JP2011166592A - Image encoding device, and image decoding device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a decoded image which does not generate visual noise like a striped pattern with a small code amount. <P>SOLUTION: An image encoding device includes: a transform unit 6 which performs discrete cosine transform of a predicted difference signal output from a subtractor 5 to output a DCT coefficient representing a discrete cosine transform result thereof, and also performs slant transform of the predicted difference signal output from the subtractor 5 to output a slant transform coefficient representing a slant transform result thereof; and a quantization unit 8 which quantizes the DCT coefficient output from the transform unit 6 and the slant transform coefficient output from the transform unit 6, and selects the DCT coefficient after the quantization or the slant transform coefficient after the quantization by reference to a distribution of AC components of the slant transform coefficient after the quantization. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an image encoding device that encodes an image signal and an image decoding device that decodes an image signal encoded by the image encoding device.

FIG. 17 is a block diagram showing a conventional image encoding device disclosed in Non-Patent Document 1, and FIG. 18 is a block diagram showing a conventional image decoding device disclosed in Non-Patent Document 1.
In the image encoding device in FIG. 17 and the image decoding device in FIG. 18, AVC / H. H.264 (ISO / IEC 14496-10 | ITU-T H.264) encoding processing and decoding processing of an image signal are performed.
The processing contents of the conventional image encoding device and image decoding device will be described below.

AVC / H. In H.264 (hereinafter, referred to as “AVC”), each frame of an image signal indicating an input image is divided into 16 × 16 macroblocks and input to an image encoding device.
The image signal for each macroblock is composed of a luminance signal 16 × 16 pixels and a color difference signal 8 × 8 pixels corresponding to the luminance signal 16 × 16 pixels.

When the intra prediction unit 101 of the image encoding device receives an image signal in units of macroblocks, the local decoding signal stored in the intra prediction memory 112 is used, and the luminance signal constituting the image signal and An intra prediction process is performed on the color difference signal.
In AVC, as an intra prediction process for a luminance signal, there are 9 types with a 4 × 4 block as a unit, 9 types with a 8 × 8 block as a unit, and 16 × 16 blocks as a unit. Four types are defined as intra prediction processes for color difference signals, and the intra prediction unit 101 determines an intra prediction method to be used from these defined intra prediction methods, and the intra prediction is performed. Intra-prediction processing is performed by the method.
When the intra prediction unit 101 performs intra prediction processing on an input signal to generate a prediction image, the intra prediction unit 101 outputs an intra prediction signal indicating the prediction image to the selection switch 104, and the intra prediction method indicating the determined intra prediction method Information is output to the entropy encoding unit 115.

When an image signal in units of macroblocks is input, the motion detection unit 102 detects a motion between the image signal and the locally decoded image signal stored in the motion compensation prediction frame memory 114, and indicates the motion The vector is output to the motion compensation prediction unit 103 and the entropy encoding unit 115.
When the motion compensation prediction unit 103 receives a motion vector from the motion detection unit 102, the motion compensation prediction unit 103 generates a prediction image using the motion vector and a locally decoded image signal stored in the motion compensation prediction frame memory 114, and the prediction image Is output to the selection switch 104.

When the selection switch 104 receives the intra prediction signal from the intra prediction unit 101 and the motion compensation prediction signal from the motion compensation prediction unit 103, for example, under the instruction of the encoding control unit 107, the selection switch 104 receives the intra prediction signal or the motion compensation prediction signal. Is selected as a prediction signal, and the prediction signal is output to the subtractor 105 and the adder 111.
The selection switch 104 also outputs prediction signal selection information indicating which of the intra prediction signal or the motion compensation prediction signal has been selected to the entropy encoding unit 115.

The subtractor 105 calculates a difference between the image signal in units of macroblocks and the prediction signal output from the selection switch 104, and outputs a prediction difference signal indicating the difference to the conversion unit 106.
Upon receiving the prediction difference signal from the subtractor 105, the conversion unit 106 performs a discrete cosine transform (DCT) on the prediction difference signal, thereby transforming the prediction difference signal from the space domain to the time domain, and the discrete cosine transform result. Is output to the quantization unit 108.
When the quantization unit 108 receives the DCT coefficient from the conversion unit 106, the quantization unit 108 quantizes the DCT coefficient according to the quantization parameter output from the encoding control unit 107, and dequantizes the quantization coefficient indicating the quantized DCT coefficient. Output to the encoding unit 109 and the entropy encoding unit 115.

Upon receiving the quantization coefficient from the quantization unit 108, the inverse quantization unit 109 performs an inverse quantization process corresponding to the quantization process of the quantization unit 108 according to the quantization parameter output from the encoding control unit 107. By performing the quantization, the quantization coefficient is inversely quantized, and the DCT coefficient corresponding to the DCT coefficient output from the conversion unit 106 (dequantization result of the quantization coefficient) is output to the inverse conversion unit 110.
When the inverse transform unit 110 receives the DCT coefficient from the inverse quantization unit 109, the inverse transform unit 110 performs inverse discrete cosine transform (inverse DCT) on the DCT coefficient, thereby returning the prediction difference signal from the time domain to the space domain, and the inverse discrete thereof. A prediction error signal (a signal corresponding to the prediction difference signal output from the subtractor 105) indicating the cosine transform result is output to the adder 111.

When the adder 111 receives the prediction signal from the selection switch 104 and receives the prediction error signal from the inverse transform unit 110, the adder 111 calculates the local decoded signal by adding the prediction signal and the prediction error signal, and the local decoded signal is calculated. Store in the intra prediction memory 112.
The loop filter 113 performs a deblocking filter process on the local decoded signal stored in the intra prediction memory 112, and uses the local decoded image signal, which is the local decoded signal after the filter process, as a motion compensated prediction frame memory 114. To store.

The entropy coding unit 115 outputs the intra prediction method information output from the intra prediction unit 101, the motion vector output from the motion detection unit 102, the prediction signal selection information output from the selection switch 104, and the output from the coding control unit 107. The quantized parameters output from the quantization unit 108 and the quantized coefficient are entropy encoded, and encoded data indicating the encoding result is output.
Note that the entropy encoding unit 115 outputs information not described above in the encoded data as necessary.

Next, when the entropy decoding unit 121 of the image decoding apparatus receives the encoded data output from the image encoding apparatus of FIG. 17, the intra prediction method information, motion vector, prediction signal selection information, quantum The quantization parameter and the quantization coefficient are entropy decoded, and the quantization parameter and the quantization coefficient are output to the inverse quantization unit 122.
Also, the entropy decoding unit 121 outputs the intra prediction scheme information to the intra prediction unit 124, outputs the motion vector to the motion compensation prediction unit 125, and outputs the prediction signal selection information to the selection switch 126.

When the inverse quantization unit 122 receives the quantization coefficient from the entropy decoding unit 121, the inverse quantization unit 122 performs inverse quantization corresponding to the quantization processing of the quantization unit 108 in FIG. 17 according to the quantization parameter output from the entropy decoding unit 121. By performing the process, the quantized coefficient is inversely quantized, and the DCT coefficient (inverse quantization result of the quantized coefficient) corresponding to the DCT coefficient output from the converting unit 106 in FIG. To do.
When receiving the DCT coefficient from the inverse quantization unit 122, the inverse transform unit 123 performs an inverse discrete cosine transform (inverse DCT) on the DCT coefficient, and a prediction error signal (subtracter 105 in FIG. 17) indicating the inverse discrete cosine transform result. The signal corresponding to the prediction difference signal output from (1) is output to the adder 127.

When the intra prediction unit 124 receives the intra prediction method information from the entropy decoding unit 121, the intra prediction unit 124 uses the decoded signal stored in the intra prediction memory 128 to perform the intra prediction process using the intra prediction method indicated by the intra prediction method information. Is performed to generate a prediction image, and an intra prediction signal indicating the prediction image is output to the selection switch 126.
When the motion compensation prediction unit 125 receives the motion vector from the entropy decoding unit 121, the motion compensation prediction unit 125 generates a prediction image using the motion vector and the decoded image signal stored in the motion compensation prediction frame memory 130. The motion compensation prediction signal shown is output to the selection switch 126.

The selection switch 126 indicates that the prediction signal selection information output from the entropy decoding unit 121 indicates that the intra prediction signal is selected by the selection switch 104 in FIG. The prediction signal is selected as the prediction signal, and the prediction signal is output to the adder 127.
On the other hand, if the prediction signal selection information output from the entropy decoding unit 121 indicates that the motion compensation prediction signal is selected by the selection switch 104 in FIG. 17, the motion compensation output from the motion compensation prediction unit 125. The prediction signal is selected as the prediction signal, and the prediction signal is output to the adder 127.

When the adder 127 receives the prediction signal from the selection switch 126 and receives the prediction error signal from the inverse transform unit 123, the adder 127 calculates the decoded signal by adding the prediction signal and the prediction error signal, and the decoded signal is intra-predicted. Stored in the memory 128.
The loop filter 129 performs deblocking filter processing on the decoded signal stored in the intra prediction memory 128 and stores the decoded image signal, which is the decoded signal after the filter processing, in the motion compensated prediction frame memory 130. At the same time, the decoded image signal is output to the outside.

The conventional image encoding device and the image decoding device are configured as described above, and the conversion unit 106 and the inverse conversion unit 110 of the image encoding device and the inverse conversion unit 123 of the image decoding device have MPEG-2 or When MPEG-4 is applied, discrete cosine transform (DCT) / inverse discrete cosine transform (inverse DCT) is used, and when AVC is applied, DCT is performed so that conversion / inverse transform is possible only by integer calculation. An integer transform that is a variant of is used.
Hereinafter, the conversion process in the case of using 8 × 8 block size DCT will be described.
FIG. 6 is an explanatory diagram showing indexes of 64 transform coefficients obtained by the transform unit.
In FIG. 6, “DC” is a DC component in the transform coefficient, “AC” is an AC component in the transform coefficient, and in particular, AC (0, 1) is the horizontal direction of all the AC components in the transform coefficient. This is the AC component located at the lowest frequency. AC (1, 0) is the AC component located at the lowest frequency in the vertical direction among all AC components in the conversion coefficient.

For example, when encoding a gradation signal (image signal) representing a sky or wall with relatively little change, encoding noise such as a change in the pixel value of the decoded image becomes a staircase and a striped pattern appears. It is known to occur (see, for example, Patent Document 1).
FIG. 7 is an explanatory diagram illustrating an example of a prediction difference signal when an image signal indicating an input image is a gradation signal that smoothly changes linearly.
As shown in FIG. 7, when DCT is performed on a prediction difference signal that smoothly changes by 2 only in the horizontal direction, a DCT coefficient that is a conversion coefficient of the prediction difference signal is a DC component as shown in FIG. And the four AC components are non-zero.
Therefore, in order to perform coding so that coding noise does not appear, it is necessary to code four AC components, which increases the amount of codes.

Here, in image coding, the component values are small, such as AC (0, 3), AC (0, 5), and AC (0, 7), and the AC components in the high frequency region are coarse. Usually, the code amount is reduced by quantization, but if these AC components are quantized to zero, the obtained decoded image is as shown in FIG.
In the original image signal, a signal that has been smoothly changing by 2 at a constant value (see FIG. 7) changes in the range of 1 to 3 (see FIG. 9). Therefore, coding noise such as a striped pattern is generated.
The reason why the coding noise such as a striped pattern is generated is that DCT is conversion based on a cosine function. When a prediction difference signal that changes linearly and smoothly is converted by DCT, FIG. As shown, many AC components are generated.
For this reason, in order to obtain a beautiful decoded image, a large amount of code is required. When the code amount is reduced, the decoded image becomes conspicuous in coding noise.

JP 2007-235377 A

MPEG-4 AVC (ISO / IEC 14496-10) / ITU-TH H.264 standard

  Since the conventional image coding apparatus is configured as described above, the prediction difference signal is always converted by DCT regardless of whether the image signal indicating the input image is a gradation signal. For this reason, when the image signal indicating the input image is a gradation signal, many AC components are generated in the DCT coefficient. As a result, in order to obtain a beautiful decoded image, a large amount of code is required. When the code amount is reduced, there is a problem that a decoded image becomes conspicuous in encoding noise.

  The present invention has been made to solve the above-described problems, and provides an image encoding device and an image decoding device capable of obtaining a decoded image that does not generate visual noise such as a striped pattern with a small code amount. The purpose is to obtain.

  The image coding apparatus according to the present invention performs a discrete cosine transform on the prediction difference signal output from the prediction difference signal calculation means, outputs a discrete cosine transform coefficient indicating the result of the discrete cosine transformation, and predictive difference signal calculation means A slant transform of the prediction differential signal output from the signal, a signal conversion means for outputting a slant conversion coefficient indicating the slant conversion result, a quantized discrete cosine transform coefficient output from the signal conversion means, and a signal conversion means Quantizing means for quantizing the slant transform coefficient output from the reference and selecting the quantized discrete cosine transform coefficient or the quantized slant transform coefficient with reference to the distribution of the AC component in the quantized slant transform coefficient The variable length encoding means variable length encodes the quantized transform coefficient selected by the quantization means, and the code It is obtained so as to output the coded data is the result.

  According to the present invention, the prediction difference signal output from the prediction difference signal calculation unit is subjected to discrete cosine transform, and the discrete cosine transform coefficient indicating the result of the discrete cosine transformation is output, and also output from the prediction difference signal calculation unit A slant transform is performed on the prediction difference signal, and a signal transform unit that outputs a slant transform coefficient indicating the result of the slant transform, and a discrete cosine transform coefficient output from the signal transform unit are quantized and output from the signal transform unit. Quantizing the slant transform coefficient, referring to the distribution of the AC component in the quantized slant transform coefficient, and providing quantization means for selecting the quantized discrete cosine transform coefficient or the quantized slant transform coefficient, The variable-length coding means performs variable-length coding on the quantized transform coefficient selected by the quantization means, and a code as a result of the coding. Since it is configured to output the data, a small amount of code, there is an effect that it is possible to obtain a decoded image which visual noise such as stripes does not occur.

It is a block diagram which shows the image coding apparatus by Embodiment 1 of this invention. It is a block diagram which shows the inside of the conversion part 6 and the quantization part 8 of the image coding apparatus by Embodiment 1 of this invention. It is a block diagram which shows the inside of the inverse transformation part 10 of the image coding apparatus by Embodiment 1 of this invention. It is a block diagram which shows the image decoding apparatus by Embodiment 1 of this invention. It is a block diagram which shows the inside of the inverse transformation part 43 of the image decoding apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the index of 64 conversion coefficients obtained by the conversion part. It is explanatory drawing which shows an example of the prediction difference signal in case the image signal which shows an input image is a gradation signal which changes smoothly linearly. It is explanatory drawing which shows an example of the conversion coefficient when the prediction difference signal of FIG. 7 is converted by DCT. It is explanatory drawing which shows the decoded image at the time of quantizing AC (0,3), AC (0,5), and AC (0,7) to zero. It is explanatory drawing which shows an example of the conversion coefficient when the prediction difference signal of FIG. 7 is slant transformed. It is a block diagram which shows the image coding apparatus by Embodiment 2 of this invention. It is a block diagram which shows the image coding apparatus by Embodiment 3 of this invention. It is a block diagram which shows the image coding apparatus by Embodiment 4 of this invention. It is a block diagram which shows the image coding apparatus by Embodiment 5 of this invention. It is explanatory drawing which shows an example of the prediction difference signal in case the image signal which shows an input image is a gentle gradation signal. It is explanatory drawing which shows an example of the conversion coefficient when the prediction difference signal of FIG. 15 is converted by DCT. 1 is a configuration diagram illustrating a conventional image encoding device disclosed in Non-Patent Document 1. FIG. It is a block diagram which shows the conventional image decoding apparatus currently disclosed by the nonpatent literature 1.

Embodiment 1 FIG.
In the first embodiment, the image encoding device and the image decoding device apply AVC, and each frame of the image signal indicating the input image is divided into 16 × 16 macroblocks and input to the image encoding device. Shall be.
The image signal for each macroblock is composed of a luminance signal 16 × 16 pixels and a color difference signal 8 × 8 pixels corresponding to the luminance signal 16 × 16 pixels.

FIG. 1 is a block diagram showing an image coding apparatus according to Embodiment 1 of the present invention.
In FIG. 1, when an intra prediction unit 1 inputs an image signal in units of macroblocks, an intra prediction for a luminance signal constituting the image signal is performed using a local decoded signal stored in the intra prediction memory 12. Implement the process.
In AVC, as intra prediction processing for luminance signals and chrominance signals, there are 9 types with 4 × 4 blocks as one unit, 9 types with 8 × 8 blocks as one unit, and 16 × 16 blocks as one unit. 4 types are defined as intra prediction processes for color difference signals, and the intra prediction unit 1 determines an intra prediction method to be used from the defined intra prediction methods, Intra prediction processing is performed using the intra prediction method.
When the intra prediction unit 1 generates a prediction image by performing an intra prediction process on the input signal, the intra prediction unit 1 outputs an intra prediction signal indicating the prediction image to the selection switch 4, and intra prediction method information indicating the determined intra prediction method Is output to the entropy encoding unit 15.

When the motion detection unit 2 inputs an image signal in units of macroblocks, the motion detection unit 2 detects a motion between the image signal and the locally decoded image signal stored in the motion compensation prediction frame memory 14, and a motion vector indicating the motion Is output to the motion compensation prediction unit 3 and the entropy encoding unit 15.
The motion compensation prediction unit 3 generates a prediction image using the motion vector output from the motion detection unit 2 and the local decoded image signal stored in the motion compensation prediction frame memory 14, and motion compensated prediction indicating the prediction image Processing for outputting a signal to the selection switch 4 is performed.

The selection switch 4 selects, for example, either an intra prediction signal output from the intra prediction unit 1 or a motion compensated prediction signal output from the motion compensated prediction unit 3 as a prediction signal under the instruction of the encoding control unit 7. Then, a process of outputting the prediction signal to the subtracter 5 and the adder 11 is performed.
Further, the selection switch 4 performs a process of outputting prediction signal selection information indicating which of the intra prediction signal or the motion compensation prediction signal is selected to the entropy encoding unit 15.
The intra prediction unit 1, the motion detection unit 2, the motion compensation prediction unit 3, and the selection switch 4 constitute a predicted image generation unit.

The subtractor 5 calculates a difference between the image signal in units of macroblocks and the prediction signal output from the selection switch 4 and performs a process of outputting a prediction difference signal indicating the difference to the conversion unit 6. The subtracter 5 constitutes a prediction difference signal calculation unit.
The transform unit 6 performs a discrete cosine transform (DCT) on the prediction difference signal output from the subtracter 5, and outputs a DCT coefficient (discrete cosine transform coefficient) indicating the result of the discrete cosine transform to the quantizer 8 and also performs subtraction. The slant transform of the prediction difference signal output from the unit 5 is performed, and the slant transform coefficient indicating the slant transform result is output to the quantization unit 8. The conversion unit 6 constitutes signal conversion means.
Here, the “slant transformation” is a kind of orthogonal transformation in which the base is composed only of diagonal linear components as described in the following non-patent document.
Non-patent literature (W. Pratt, et al., “Slant Transform Image Coding”, IEEE Transactions on Communications, vol. COM-22, No. 8, August 1974)

The encoding control unit 7 outputs a quantization parameter that is referred to when the quantization unit 8 and the inverse quantization unit 9 perform quantization / inverse quantization, and controls the code amount and the encoded image quality of the encoded data. Perform the process.
The quantization unit 8 quantizes the DCT coefficient output from the conversion unit 6 according to the quantization parameter output from the encoding control unit 7 and also quantizes the slant conversion coefficient output from the conversion unit 6. To implement.
The quantization unit 8 refers to the distribution of the AC component in the quantized slant transform coefficient, selects either the DCT coefficient after quantization or the slant transform coefficient after quantization, and the selected transform The process which outputs a coefficient to the inverse quantization part 9 and the entropy encoding part 15 as a quantization coefficient is implemented.
The encoding control unit 7 and the quantization unit 8 constitute quantization means.

  The inverse quantization unit 9 reverses the quantization coefficient (the quantized DCT coefficient or the quantized slant transform coefficient) output from the quantization unit 8 according to the quantization parameter output from the encoding control unit 7. A process of quantizing and outputting a transform coefficient (DCT coefficient or slant transform coefficient) as a result of the inverse quantization to the inverse transform unit 10 is performed.

The inverse transform unit 10 refers to the distribution of the AC component in the transform coefficient output from the inverse quantization unit 9, and selects inverse discrete cosine transform (inverse DCT) or inverse slant transform as the inverse transform method for the transform coefficient. Perform the process.
When the inverse transform unit 10 selects inverse discrete cosine transform (inverse DCT) as the inverse transform method, the inverse discrete cosine transform (inverse DCT) is performed on the transform coefficient output from the inverse quantization unit 9, and the inverse discrete cosine transform is performed. A process of outputting a prediction error signal indicating the result (a signal corresponding to the prediction difference signal output from the subtractor 5) to the adder 11 is performed. On the other hand, when inverse slant transform is selected as the inverse transform method, the transform coefficient output from the inverse quantization unit 9 is subjected to inverse slant transform, and a prediction error signal indicating the result of the inverse slant transform (prediction difference output from the subtractor 5). The signal corresponding to the signal) is output to the adder 11.

The adder 11 calculates a local decoded signal by adding the prediction signal output from the selection switch 4 and the prediction error signal output from the inverse transform unit 10, and outputs the local decoded signal to the intra prediction memory 12. Implement the process.
The intra prediction memory 12 is a recording medium such as a RAM for storing the local decoded signal output from the adder 11.

The loop filter 13 performs deblocking filter processing on the local decoded signal stored in the intra prediction memory 12, and the local decoded image signal, which is the local decoded signal after the filter processing, is stored in the motion compensated prediction frame memory 14. Perform the output process.
The motion compensation prediction frame memory 14 is a recording medium such as a RAM for storing the locally decoded image signal output from the loop filter 13.

  The entropy encoding unit 15 includes the intra prediction method information output from the intra prediction unit 1, the motion vector output from the motion detection unit 2, the prediction signal selection information output from the selection switch 4, and the encoding control unit 7 The entropy encoding is performed on the quantization parameter output from, the quantization coefficient output from the quantization unit 8 and other necessary information, and the encoded data indicating the encoding result is output. The entropy encoding unit 15 constitutes variable length encoding means.

  In FIG. 1, an intra prediction unit 1, a motion detection unit 2, a motion compensation prediction unit 3, a selection switch 4, a subtracter 5, a conversion unit 6, a coding control unit 7, and a quantization unit, which are components of the image coding device. 8, each of the inverse quantization unit 9, the inverse transform unit 10, the adder 11, the loop filter 13, and the entropy encoding unit 15 has dedicated hardware (for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, etc. However, when the image encoding device is configured by a computer, an intra prediction unit 1, a motion detection unit 2, a motion compensation prediction unit 3, a selection switch 4, a subtractor 5, a program describing processing contents of the transform unit 6, the encoding control unit 7, the quantization unit 8, the inverse quantization unit 9, the inverse transform unit 10, the adder 11, the loop filter 13 and the entropy coding unit 15. The Stored in the memory of the computer, CPU of the computer may execute a program stored in the memory.

FIG. 2 is a block diagram showing the inside of the conversion unit 6 and the quantization unit 8 of the image coding apparatus according to Embodiment 1 of the present invention.
In FIG. 2, the DCT unit 21 of the transform unit 6 performs a discrete cosine transform (DCT) on the prediction difference signal output from the subtracter 5, and quantizes DCT coefficients (discrete cosine transform coefficients) indicating the discrete cosine transform result. Processing to be output to the unit 8 is performed.
The slant transformer 22 performs a slant transform on the prediction difference signal output from the subtracter 5 and outputs a slant transform coefficient indicating the slant transform result to the quantization unit 8.

The DCT coefficient quantizer 23 of the quantization unit 8 quantizes the DCT coefficient output from the DCT unit 21 in accordance with the quantization parameter output from the encoding control unit 7, and selects the conversion method for the quantized DCT coefficient. Processing to be output to the unit 25 is performed.
The slant transform coefficient quantizer 24 quantizes the slant transform coefficient output from the slant transformer 22 according to the quantization parameter output from the encoding control unit 7, and converts the quantized slant transform coefficient to a conversion method selection unit. The process which outputs to 25 is implemented.
The conversion method selection unit 25 is the AC component in the slant transform coefficient after quantization output from the slant transform coefficient quantizer 24 and the AC component located at the lowest frequency in the horizontal direction and the lowest frequency in the vertical direction. If the non-zero AC component is not included in the AC components other than the AC component located at, the slant transform coefficient after quantization is selected, and if the non-zero AC component is included, the quantization is performed. A process of selecting a later DCT coefficient is performed.

FIG. 3 is a block diagram showing the inside of the inverse transform unit 10 of the image coding apparatus according to Embodiment 1 of the present invention.
In FIG. 3, the inverse transformation method selection unit 31 includes the AC component located at the lowest frequency in the horizontal direction among the AC components in the transformation coefficient (DCT coefficient or slant transformation coefficient) output from the inverse quantization unit 9. If the AC component other than the AC component located at the lowest frequency in the vertical direction does not contain a non-zero AC component, reverse slant transformation is selected as the inverse transformation method, and a non-zero AC component is included. If so, a process of selecting an inverse discrete cosine transform (inverse DCT) as an inverse transform method is performed.

In the inverse DCT unit 32, when the inverse discrete cosine transform (inverse DCT) is selected as the inverse transform method by the inverse transform method selection unit 31, the inverse discrete cosine transform (inverse DCT) is performed on the transform coefficient output from the inverse quantization unit 9. Then, a process of outputting a prediction error signal (a signal corresponding to the prediction difference signal output from the subtractor 5) indicating the inverse discrete cosine transform result to the adder 11 is performed.
The inverse slant transformer 33 performs inverse slant transformation on the transform coefficient output from the inverse quantization unit 9 when the inverse slant transformation is selected as the inverse transformation method by the inverse transformation method selection unit 31, and shows the result of the inverse slant transformation. A process of outputting a prediction error signal (a signal corresponding to the prediction difference signal output from the subtractor 5) to the adder 11 is performed.

FIG. 4 is a block diagram showing an image decoding apparatus according to Embodiment 1 of the present invention.
In FIG. 4, when the encoded data output from the image encoding apparatus in FIG. 1 is input to the entropy decoding unit 41, intra prediction method information, motion vectors, prediction signal selection information, quantization parameters, and quantum are determined from the encoded data. A process of entropy decoding the quantization coefficient and outputting the quantization parameter and the quantization coefficient to the inverse quantization unit 42 is performed.
In addition, the entropy decoding unit 41 outputs the intra prediction scheme information to the intra prediction unit 44, outputs the motion vector to the motion compensation prediction unit 45, and outputs the prediction signal selection information to the selection switch 46. The entropy decoding unit 41 constitutes variable length decoding means.

  The inverse quantization unit 42 inversely quantizes the quantization coefficient output from the entropy decoding unit 41 according to the quantization parameter output from the entropy decoding unit 41, and a transform coefficient (DCT coefficient or slant) as a result of the inverse quantization. The conversion coefficient is output to the inverse conversion unit 43. The inverse quantization unit 42 constitutes inverse quantization means.

The inverse transform unit 43 refers to the distribution of the AC component in the transform coefficient output from the inverse quantization unit 42 and selects inverse discrete cosine transform (inverse DCT) or inverse slant transform as an inverse transform method for the transform coefficient. Perform the process.
When the inverse transform unit 43 selects inverse discrete cosine transform (inverse DCT) as the inverse transform method, the transform coefficient output from the inverse quantization unit 42 is subjected to inverse discrete cosine transform (inverse DCT), and the inverse discrete cosine transform is performed. A process of outputting a prediction error signal indicating the result (a signal corresponding to the prediction difference signal output from the subtracter 5 in FIG. 1) to the adder 47 is performed. On the other hand, when inverse slant transformation is selected as the inverse transformation method, the transform coefficient output from the inverse quantization unit 42 is subjected to inverse slant transformation, and a prediction error signal indicating the inverse slant transformation result (output from the subtracter 5 in FIG. 1). The signal corresponding to the predicted difference signal) is output to the adder 47.
The inverse conversion unit 43 constitutes an inverse conversion method selection unit and an inverse conversion unit.

When the intra prediction unit 44 receives the intra prediction method information from the entropy decoding unit 41, the intra prediction unit 44 uses the decoded signal stored in the intra prediction memory 48 to perform the intra prediction process using the intra prediction method indicated by the intra prediction method information. By performing this, a prediction image is generated, and an intra prediction signal indicating the prediction image is output to the selection switch 46.
The motion compensation prediction unit 45 generates a prediction image using the motion vector output from the entropy decoding unit 41 and the decoded image signal stored in the frame memory 50 for motion compensation prediction, and a motion compensation prediction signal indicating the prediction image Is output to the selection switch 46.

If the prediction signal selection information output from the entropy decoding unit 41 indicates that the selection switch 4 in FIG. 1 has selected the intra prediction signal, the selection switch 46 outputs the intra prediction output from the intra prediction unit 44. A signal is selected as a prediction signal, and processing for outputting the prediction signal to the adder 47 is performed.
On the other hand, if the prediction signal selection information output from the entropy decoding unit 41 indicates that the motion compensation prediction signal is selected by the selection switch 4 in FIG. 1, the motion compensation output from the motion compensation prediction unit 45. A process of selecting the prediction signal as the prediction signal and outputting the prediction signal to the adder 47 is performed.

The adder 47 calculates a decoded signal by adding the prediction signal output from the selection switch 46 and the prediction error signal output from the inverse transform unit 43, and outputs the decoded signal to the intra prediction memory 48. carry out.
The intra prediction memory 48 is a recording medium such as a RAM for storing the decoded signal output from the adder 47.

The loop filter 49 performs a deblocking filter process on the decoded signal stored in the intra prediction memory 48 and outputs a decoded image signal, which is a decoded signal after the filter process, to the motion compensation prediction frame memory 50. To implement.
The motion compensation prediction frame memory 50 is a recording medium such as a RAM for storing the decoded image signal output from the loop filter 49.
The intra prediction unit 44, motion compensation prediction unit 45, selection switch 46, adder 47, intra prediction memory 48, loop filter 49, and motion compensation prediction frame memory 50 constitute a decoded image generating means.

  In FIG. 4, an entropy decoding unit 41, an inverse quantization unit 42, an inverse transformation unit 43, an intra prediction unit 44, a motion compensation prediction unit 45, a selection switch 46, an adder 47, and a loop filter 49, which are components of the image decoding apparatus. Each of which is configured with dedicated hardware (for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like), but the image decoding device is configured with a computer , A program describing the processing contents of the entropy decoding unit 41, the inverse quantization unit 42, the inverse transform unit 43, the intra prediction unit 44, the motion compensation prediction unit 45, the selection switch 46, the adder 47, and the loop filter 49 Stored in the memory of a computer so that the CPU of the computer executes the program stored in the memory Good.

FIG. 5 is a block diagram showing the inside of the inverse transform unit 43 of the image decoding apparatus according to Embodiment 1 of the present invention.
In FIG. 5, the inverse transform method selection unit 61 includes the AC component located at the lowest frequency in the horizontal direction among the AC components in the transform coefficient (DCT coefficient or slant transform coefficient) output from the inverse quantization unit 42. If the AC component other than the AC component located at the lowest frequency in the vertical direction does not contain a non-zero AC component, reverse slant transformation is selected as the inverse transformation method, and a non-zero AC component is included. If so, a process of selecting an inverse discrete cosine transform (inverse DCT) as an inverse transform method is performed.

When the inverse discrete cosine transform (inverse DCT) is selected as the inverse transform method by the inverse transform method selection unit 61, the inverse DCT unit 62 converts the transform coefficient output from the inverse quantization unit 42 into the inverse discrete cosine transform (inverse DCT). Then, a process of outputting a prediction error signal indicating the inverse discrete cosine transform result (a signal corresponding to the prediction difference signal output from the subtracter 5 in FIG. 1) to the adder 47 is performed.
The inverse slant transformer 63 performs inverse slant transformation on the transform coefficient output from the inverse quantization unit 42 when the inverse slant transformation is selected as the inverse transformation method by the inverse transformation method selection unit 61, and shows the result of the inverse slant transformation. A process of outputting a prediction error signal (a signal corresponding to the prediction difference signal output from the subtracter 5 in FIG. 1) to the adder 47 is performed.

Next, the operation will be described.
First, the processing content of the image coding apparatus will be described.
When the intra prediction unit 1 receives an image signal in units of macro blocks, the intra prediction unit 1 uses the local decoded signal stored in the intra prediction memory 12 to perform intra prediction on the luminance signal and the color difference signal constituting the image signal. Implement the process.
In AVC, as an intra prediction process for a luminance signal, there are 9 types with a 4 × 4 block as a unit, 9 types with a 8 × 8 block as a unit, and 16 × 16 blocks as a unit. Four types are defined as intra prediction processes for color difference signals, and the intra prediction unit 1 determines an intra prediction method to be used from the defined intra prediction methods, and the intra prediction method. Intra prediction processing is performed.
When the intra prediction unit 1 generates a prediction image by performing an intra prediction process on the input signal, the intra prediction unit 1 outputs an intra prediction signal indicating the prediction image to the selection switch 4 and indicates the determined intra prediction method. Information is output to the entropy encoding unit 15.
Since the prediction image generation process itself in the intra prediction unit 1 is a known technique, a detailed description thereof will be omitted.

When an image signal in units of macroblocks is input, the motion detection unit 2 detects a motion between the image signal and the locally decoded image signal stored in the motion compensation prediction frame memory 14, and a motion indicating the motion The vector is output to the motion compensation prediction unit 3 and the entropy encoding unit 15.
Since the motion vector detection process itself in the motion detector 2 is a known technique, a detailed description thereof will be omitted.

When the motion compensation prediction unit 3 receives the motion vector from the motion detection unit 2, the motion compensation prediction unit 3 generates a prediction image using the motion vector and the locally decoded image signal stored in the motion compensation prediction frame memory 14, and the prediction image Is output to the selection switch 4.
Since the prediction image generation process itself in the motion compensation prediction unit 3 is a known technique, a detailed description thereof is omitted.

When the selection switch 4 receives the intra prediction signal from the intra prediction unit 1 and the motion compensation prediction signal from the motion compensation prediction unit 3, for example, under the instruction of the encoding control unit 7, the selection switch 4 receives the intra prediction signal or the motion compensation prediction signal. Is selected as a prediction signal, and the prediction signal is output to the subtracter 5 and the adder 11.
Further, the selection switch 4 outputs prediction signal selection information indicating which of the intra prediction signal or the motion compensation prediction signal is selected to the entropy encoding unit 15.
When the image signal in units of macroblocks is input, the subtracter 5 calculates a difference between the image signal and the prediction signal output from the selection switch 4, and outputs a prediction difference signal indicating the difference to the conversion unit 6.

Upon receiving the prediction difference signal from the subtractor 5, the DCT unit 21 of the conversion unit 6 performs a discrete cosine transform (DCT) on the prediction difference signal, and converts the DCT coefficient indicating the result of the discrete cosine transformation into the DCT of the quantization unit 8. It outputs to the coefficient quantizer 23.
In addition, when the slant converter 22 of the conversion unit 6 receives the prediction difference signal from the subtractor 5, the slant conversion of the prediction difference signal is performed by the slant conversion of the prediction difference signal, and the slant conversion coefficient indicating the slant conversion result is converted by the slant conversion of the quantization unit 8. Output to the coefficient quantizer 24.

Here, FIG. 6 is an explanatory diagram showing indexes of 64 transform coefficients obtained by the transform unit 6.
In FIG. 6, “DC” is a DC component in the conversion coefficient, and “AC” is an AC component in the conversion coefficient.
In particular, AC (0, 1) is the AC component located at the lowest frequency in the horizontal direction among all AC components in the conversion coefficient.
AC (1, 0) is the AC component located at the lowest frequency in the vertical direction among all AC components in the conversion coefficient.

At this time, a prediction difference signal as shown in FIG. 7 (a prediction difference signal when the image signal indicating the input image is a gradation signal that linearly changes smoothly) is input to the conversion unit 6. When the DCT unit 21 performs discrete cosine transform (DCT) on the prediction difference signal, four non-zero AC components are obtained as shown in FIG.
Therefore, when a prediction difference signal as shown in FIG. 7 is input to the transform unit 6, the quantizing unit 8 quantizes the DCT coefficient output from the transform unit 6, and the quantized coefficient is dequantized. 9 and the entropy encoding unit 15, as described in the background art section, when the code amount is small, a decoded image in which encoding noise is conspicuous is obtained.

On the other hand, when the slant converter 22 of the conversion unit 6 performs slant conversion on the prediction difference signal as shown in FIG. 7, the slant conversion coefficient is the highest in the horizontal direction among all AC components as shown in FIG. Only AC (0, 1), which is an AC component located at a low frequency, becomes non-zero.
Therefore, when a prediction difference signal as shown in FIG. 7 is input to the transforming unit 6, the quantizing unit 8 quantizes the slant transform coefficient output from the transforming unit 6, and dequantizes the quantized coefficient. If it outputs to the part 9 and the entropy encoding part 15, the decoded image which does not generate | occur | produce visual noise like a striped pattern can be obtained with a small code amount.

Therefore, in the first embodiment, among the AC components in the quantized slant transform coefficient, AC (0, 1) which is the AC component located at the lowest frequency in the horizontal direction and the lowest in the vertical direction. If the non-zero AC component is not included in the AC components other than AC (1, 0) that is the AC component located at the frequency, it is determined that the image signal indicating the input image is a gradation signal, and In order to realize efficient encoding processing and decoding processing with less code amount without distorting the gradation signal, the quantized slant transform coefficient is selected and the inverse quantization unit 9 and the entropy encoding unit 15 are selected. Output to.
On the other hand, if the AC component other than AC (0, 1) and AC (1, 0) includes a non-zero AC component, it is determined that the image signal indicating the input image is not a gradation signal, and quantization is performed. The subsequent DCT coefficients are selected and output to the inverse quantization unit 9 and the entropy coding unit 15.
Specifically, it is as follows.

When the DCT coefficient quantizer 23 of the quantization unit 8 receives the DCT coefficient from the DCT unit 21 of the conversion unit 6, the DCT coefficient quantizer 23 quantizes the DCT coefficient according to the quantization parameter output from the encoding control unit 7, The converted DCT coefficients are output to the conversion method selection unit 25.
When the slant transform coefficient quantizer 24 of the quantization unit 8 receives the slant transform coefficient from the slant transformer 22 of the transform unit 6, the slant transform coefficient quantizer 24 converts the slant transform coefficient according to the quantization parameter output from the encoding control unit 7. It quantizes and outputs the quantized slant transform coefficient to the transform method selection unit 25.

When the conversion method selection unit 25 of the quantization unit 8 receives the quantized slant conversion coefficient from the slant conversion coefficient quantizer 24, it is positioned at the lowest frequency in the horizontal direction among the AC components in the slant conversion coefficient. It is determined whether or not non-zero AC components are included in AC components other than AC (0, 1) and AC (1, 0) positioned at the lowest frequency in the vertical direction.
If the non-zero AC component is not included, the conversion method selection unit 25 selects the quantized slant conversion coefficient output from the slant conversion coefficient quantizer 24 and quantizes the quantized slant conversion coefficient. The quantization coefficient is output to the inverse quantization unit 9 and the entropy coding unit 15.
On the other hand, if a non-zero AC component is included, the DCT coefficient after quantization output from the DCT coefficient quantizer 23 is selected, and the inverse quantization unit 9 uses the DCT coefficient after quantization as a quantization coefficient. And output to the entropy encoding unit 15.

Here, although the quantizing unit 8 has implemented the DCT coefficient quantizer 23 and the slant transform coefficient quantizer 24, the quantizing unit 8 has only one quantizer mounted, The quantizer may perform DCT and slant transformation in a time division manner.
When the inverse quantization unit 9 receives the quantization coefficient (quantized DCT coefficient or quantized slant transform coefficient) from the quantization unit 8, according to the quantization parameter output from the coding control unit 7, The quantization coefficient is inversely quantized, and the transform coefficient (DCT coefficient or slant transform coefficient) that is the inverse quantization result is output to the inverse transform unit 10.

When receiving the transform coefficient (DCT coefficient or slant transform coefficient) from the inverse quantization unit 9, the inverse transform method selection unit 31 of the inverse transform unit 10 is positioned at the lowest frequency in the horizontal direction among the AC components in the transform coefficient. It is determined whether or not a non-zero AC component is included in the AC components other than AC (0, 1) and AC (1, 0) positioned at the lowest frequency in the vertical direction.
If the non-zero AC component is not included, the inverse transform method selection unit 31 selects the inverse slant transform as the inverse transform method, and outputs the transform coefficient output from the inverse quantization unit 9 to the inverse slant converter 33. To do.
On the other hand, if a non-zero AC component is included, the inverse discrete cosine transform (inverse DCT) is selected as the inverse transform method, and the transform coefficient output from the inverse quantization unit 9 is output to the inverse DCT unit 32.

The inverse DCT unit 32 of the inverse transform unit 10 reverses the transform coefficient output from the inverse transform method selection unit 31 when the inverse discrete cosine transform (inverse DCT) is selected as the inverse transform method by the inverse transform method selection unit 31. Discrete cosine transform (inverse DCT) is performed, and a prediction error signal (a signal corresponding to the prediction difference signal output from the subtractor 5) indicating the inverse discrete cosine transform result is output to the adder 11.
The inverse slant converter 33 of the inverse conversion unit 10 performs inverse slant conversion on the conversion coefficient output from the inverse conversion method selection unit 31 when the inverse slant conversion is selected as the inverse conversion method by the inverse conversion method selection unit 31. A prediction error signal (a signal corresponding to the prediction difference signal output from the subtractor 5) indicating the result of the inverse slant conversion is output to the adder 11.

When the adder 11 receives the prediction signal from the selection switch 4 and receives the prediction error signal from the inverse transform unit 10, the adder 11 adds the prediction signal and the prediction error signal in order to prepare for the next encoding process. The signal is calculated, and the locally decoded signal is stored in the intra prediction memory 12.
When the adder 11 stores the local decoded signal in the intra prediction memory 12, the loop filter 13 performs a deblocking filter process on the local decoded signal, and a local decoded image signal which is a local decoded signal after the filter process Are stored in the motion compensation prediction frame memory 14.

The entropy encoding unit 15 includes intra prediction scheme information output from the intra prediction unit 1, a motion vector output from the motion detection unit 2, prediction signal selection information output from the selection switch 4, and an encoding control unit. 7 is entropy-coded with the quantization parameter output from 7, the quantization coefficient output from the quantization unit 8, and other necessary information, and outputs encoded data indicating the encoding result.
As will be described later, the image decoding apparatus can select the inverse transform method for the transform coefficient by referring to the distribution of the AC component in the transform coefficient output from the inverse quantization unit 42. It is not necessary to transmit the information indicating the conversion method in the encoded data.

Next, processing contents of the image decoding apparatus will be described.
When the encoded data output from the image encoding device in FIG. 1 is input, the entropy decoding unit 41 receives intra prediction method information, motion vectors, prediction signal selection information, quantization parameters, and quantization coefficients from the encoded data. Entropy decoding.
Further, the entropy decoding unit 41 outputs the quantization parameter and the quantization coefficient to the inverse quantization unit 42, outputs the intra prediction scheme information to the intra prediction unit 44, and outputs the motion vector to the motion compensation prediction unit 45. The prediction signal selection information is output to the selection switch 46.

  When the inverse quantization unit 42 receives the quantization coefficient from the entropy decoding unit 41, the inverse quantization unit 42 inversely quantizes the quantization coefficient in accordance with the quantization parameter output from the entropy decoding unit 41, and a transform that is a result of the inverse quantization The coefficient (DCT coefficient or slant transform coefficient) is output to the inverse transform unit 43.

When receiving the transform coefficient (DCT coefficient or slant transform coefficient) from the inverse quantization unit 42, the inverse transform method selection unit 61 of the inverse transform unit 43 is positioned at the lowest frequency in the horizontal direction among the AC components in the transform coefficient. It is determined whether or not a non-zero AC component is included in the AC components other than AC (0, 1) and AC (1, 0) positioned at the lowest frequency in the vertical direction.
If the non-zero AC component is not included, the inverse transform method selection unit 61 selects the inverse slant transform as the inverse transform method, and outputs the transform coefficient output from the inverse quantization unit 42 to the inverse slant converter 63. To do.
On the other hand, if a non-zero AC component is included, the inverse discrete cosine transform (inverse DCT) is selected as the inverse transform method, and the transform coefficient output from the inverse quantization unit 42 is output to the inverse DCT unit 62.

The inverse DCT unit 62 of the inverse transform unit 43 reverses the transform coefficient output from the inverse transform method selection unit 61 when the inverse transform method selection unit 61 selects inverse discrete cosine transform (inverse DCT) as the inverse transform method. Discrete cosine transform (inverse DCT) is performed, and a prediction error signal (a signal corresponding to the prediction difference signal output from the subtracter 5 in FIG. 1) indicating the inverse discrete cosine transform result is output to the adder 47.
The inverse slant converter 63 of the inverse conversion unit 43 performs inverse slant conversion on the conversion coefficient output from the inverse conversion method selection unit 61 when the inverse conversion method selection unit 61 selects the inverse slant conversion as the inverse conversion method. A prediction error signal indicating the result of the inverse slant conversion (a signal corresponding to the prediction difference signal output from the subtracter 5 in FIG. 1) is output to the adder 47.

When the intra prediction unit 44 receives the intra prediction method information from the entropy decoding unit 41, the intra prediction unit 44 uses the decoded signal stored in the intra prediction memory 48 to perform the intra prediction process using the intra prediction method indicated by the intra prediction method information. Is performed to generate a prediction image, and an intra prediction signal indicating the prediction image is output to the selection switch 46.
When the motion compensation prediction unit 45 receives the motion vector from the entropy decoding unit 41, the motion compensation prediction unit 45 generates a prediction image using the motion vector and the decoded image signal stored in the motion compensation prediction frame memory 50. The motion compensation prediction signal shown is output to the selection switch 46.

When the selection switch 46 receives the prediction signal selection information from the entropy decoding unit 41, the selection switch 46 refers to the prediction signal selection information and determines whether the selection switch 4 in FIG. 1 has selected the intra prediction signal or not. Determine if it is selected.
When the selection switch 46 determines that the selection switch 4 in FIG. 1 has selected the intra prediction signal, the selection switch 46 selects the intra prediction signal output from the intra prediction unit 44 as the prediction signal, and sends the prediction signal to the adder 47. Output.
On the other hand, if the selection switch 4 in FIG. 1 determines that the motion compensation prediction signal is selected, the motion compensation prediction signal output from the motion compensation prediction unit 45 is selected as the prediction signal, and the prediction signal is sent to the adder 47. Output.

When the adder 47 receives the prediction signal from the selection switch 46 and receives the prediction error signal from the inverse transform unit 43, the adder 47 calculates the decoded signal by adding the prediction signal and the prediction error signal, and the decoded signal is intra-predicted. Stored in the memory 48.
When the adder 47 stores the decoded signal in the intra prediction memory 48, the loop filter 49 performs deblocking filter processing on the decoded signal, and performs motion compensation prediction on the decoded image signal that is the decoded signal after the filter processing. And the decoded image signal is output to the outside.

  As is clear from the above, according to the first embodiment, the prediction difference signal output from the subtracter 5 is subjected to discrete cosine transform (DCT), and a DCT coefficient indicating the discrete cosine transform result is output. A slant transform is performed on the prediction difference signal output from the subtractor 5 and a slant transform coefficient indicating the slant transform result is output. A DCT coefficient output from the transform unit 6 is quantized and a transform unit. The quantizing unit 8 quantizes the slant transform coefficient output from 6 and selects the quantized DCT coefficient or the quantized slant transform coefficient by referring to the distribution of the AC component in the quantized slant transform coefficient. The entropy encoding unit 15 performs variable-length encoding on the transformed transform coefficient selected by the quantization unit 8, and outputs encoded data that is the encoding result. Since it is configured as, with a small amount of code, an effect that the picture coding apparatus for outputting coded data capable of visual noise to obtain a decoded image does not occur, such as a striped pattern is obtained.

  That is, according to the first embodiment, when it is determined that the image signal indicating the input image is a gradation signal with reference to the distribution of the AC component in the slant transform coefficient after quantization, the DCT after quantization Since a quantized slant transform coefficient can be selected instead of a coefficient, a decoded image that does not generate visual noise such as a striped pattern can be obtained even when a gradation signal is input.

  Further, according to the first embodiment, the inverse quantization unit 42 that inversely quantizes the transform coefficient entropy decoded by the entropy decoding unit 41, and the AC component in the transform coefficient after the inverse quantization by the inverse quantization unit 42 , The inverse transform method selection unit 61 that selects inverse discrete cosine transform (inverse DCT) or inverse slant transform as the inverse transform method for the transform coefficient, and the inverse transform method selection unit 61 performs the inverse transform method as the inverse transform method. When discrete cosine transform (inverse DCT) is selected, an inverse DCT unit 62 that calculates a prediction difference signal by performing inverse discrete cosine transform (inverse DCT) on the transform coefficient, and an inverse transform method selection unit 61 as an inverse transform method. When the inverse slant transform is selected, the inverse slant transform 63 that computes a prediction difference signal by performing the inverse slant transform on the conversion coefficient, and the inverse DCT converter 62 or the inverse slant transform. Since the decoded image signal indicating the decoded image is generated from the prediction difference signal calculated by the point converter 63, an image decoding apparatus capable of obtaining a decoded image that does not generate visual noise such as a striped pattern The effect is obtained.

Embodiment 2. FIG.
FIG. 11 is a block diagram showing an image encoding apparatus according to Embodiment 2 of the present invention. In the figure, the same reference numerals as those in FIG.
The image decoding apparatus in the second embodiment uses the image decoding apparatus in FIG. 4 as in the first embodiment.
The conversion unit 71 of the image encoding device performs a process of performing a discrete cosine transform (DCT) on the prediction difference signal output from the subtractor 5 and outputting a DCT coefficient indicating the discrete cosine transform result to the quantization unit 73. . The conversion unit 71 constitutes a signal conversion unit.

The encoding control unit 72 outputs the quantization parameter referred to when the quantization unit 73 and the inverse quantization unit 9 perform quantization / inverse quantization similarly to the encoding control unit 7 of FIG. A process for controlling the amount of encoded data and the encoded image quality is performed.
However, the encoding control unit 72 refers to the distribution of the DCT coefficients output from the conversion unit 71 to determine whether or not the image signal indicating the input image is a gradation signal, and the image signal is the gradation signal. In this case, AC components other than AC (0, 1), which is the AC component located at the lowest frequency in the horizontal direction, and AC components (1, 0), which are the AC components located at the lowest frequency in the vertical direction. An instruction to quantize the DCT coefficient is output to the quantization unit 73 so that becomes zero.
The quantization unit 73 quantizes the DCT coefficient output from the conversion unit 71 under the control of the encoding control unit 72, and uses the quantized DCT coefficient as the quantization coefficient to perform the inverse quantization unit 9 and the entropy coding unit. The process which outputs to 15 is implemented.
The encoding control unit 72 and the quantization unit 73 constitute a quantization unit.

  In FIG. 11, the intra prediction unit 1, the motion detection unit 2, the motion compensation prediction unit 3, the selection switch 4, the subtracter 5, the conversion unit 71, the coding control unit 72, and the quantization unit, which are components of the image coding device. 73, the inverse quantization unit 9, the inverse transform unit 10, the adder 11, the loop filter 13 and the entropy encoding unit 15 are each dedicated hardware (for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, etc. However, when the image encoding device is configured by a computer, an intra prediction unit 1, a motion detection unit 2, a motion compensation prediction unit 3, a selection switch 4, a subtractor 5, the processing contents of the transform unit 71, the encoding control unit 72, the quantization unit 73, the inverse quantization unit 9, the inverse transform unit 10, the adder 11, the loop filter 13 and the entropy coding unit 15 are described. The program stored in the memory of the computer, may execute a program that the CPU of the computer is stored in the memory.

Next, the operation will be described.
In the image encoding apparatus, except for the conversion unit 71, the encoding control unit 72, and the quantization unit 73, the configuration is the same as that of the first embodiment. Therefore, the conversion unit 71, the encoding control unit 72, and the quantization unit are mainly used. The processing contents of 73 will be described.
Since the processing contents of the image decoding apparatus are the same as those in the first embodiment, description thereof is omitted.

  Unlike the conversion unit 6 in FIG. 1, the conversion unit 71 includes only the DCT unit 21. When the prediction difference signal is received from the subtracter 5, the conversion unit 71 performs a discrete cosine transform (DCT) on the prediction difference signal. DCT coefficients indicating the result of the discrete cosine transform are output to the encoding control unit 72 and the quantization unit 73.

The encoding control unit 72 outputs quantization parameter information to be referred to when the quantization unit 73 and the inverse quantization unit 9 perform quantization / inverse quantization similarly to the encoding control unit 7 of FIG. Thus, the amount of encoded data and the encoded image quality are controlled.
However, the encoding control unit 72 refers to the DCT coefficient distribution output from the conversion unit 71 and determines whether or not the image signal indicating the input image is a gradation signal.
For example, when the image signal indicating the input image is a gradation signal and a prediction difference signal as shown in FIG. 7 is input to the conversion unit 71, the DCT coefficient output from the conversion unit 71 is as shown in FIG. The DC component and the four AC components are non-zero.
For this reason, the encoding control unit 72 has four AC components AC (0, 1), AC (0, 3), AC (0, 5), and AC (0, 7) being non-zero. The image signal indicating the input image is estimated to be a gradation signal.

  When the encoding control unit 72 estimates that the image signal indicating the input image is a gradation signal, the encoding component 72 is AC (0, 1), which is the AC component located at the lowest frequency in the horizontal direction, and the lowest in the vertical direction. An instruction to quantize the DCT coefficient is output to the quantization unit 73 so that AC components other than AC (1, 0), which is an AC component located at a frequency, become zero.

When the quantization unit 73 receives the DCT coefficient from the conversion unit 71, the quantization unit 73 quantizes the DCT coefficient in accordance with the quantization parameter output from the encoding control unit 72, and uses the quantized DCT coefficient as a quantization coefficient. The result is output to the inverse quantization unit 9 and the entropy encoding unit 15.
However, when the quantization unit 73 receives an instruction from the encoding control unit 72 to quantize the DCT coefficient so that AC components other than AC (0, 1) and AC (1, 0) become zero, The DCT coefficients are quantized so that AC components other than AC (0, 1) and AC (1, 0) become zero.
The quantization coefficient in this case (the DCT coefficient after quantization) is equivalent to the slant transform coefficient quantized by the slant transform coefficient quantizer 24 of FIG.

  When receiving the quantization coefficient from the quantization unit 73, the inverse quantization unit 9 converts the quantization coefficient to the inverse quantization according to the quantization parameter output from the encoding control unit 72, as in the first embodiment. The transform coefficient (DCT coefficient) that is the result of the inverse quantization is output to the inverse transform unit 10.

When receiving the transform coefficient (DCT coefficient) from the inverse quantization unit 9, the inverse transform method selection unit 31 of the inverse transform unit 10 in the horizontal direction out of the AC components in the transform coefficient as in the first embodiment. Whether or not non-zero AC components are included in AC components other than AC (0, 1) located at the lowest frequency and AC (1, 0) located at the lowest frequency in the vertical direction Determine whether.
If the non-zero AC component is not included, the inverse transform method selection unit 31 selects the inverse slant transform as the inverse transform method, and outputs the transform coefficient output from the inverse quantization unit 9 to the inverse slant converter 33. To do.
On the other hand, if a non-zero AC component is included, the inverse discrete cosine transform (inverse DCT) is selected as the inverse transform method, and the transform coefficient output from the inverse quantization unit 9 is output to the inverse DCT unit 32.

As in the first embodiment, the inverse DCT unit 32 of the inverse transform unit 10 uses the inverse transform method selection unit when the inverse discrete cosine transform (inverse DCT) is selected as the inverse transform method by the inverse transform method selection unit 31. The transform coefficient output from 31 is subjected to inverse discrete cosine transform (inverse DCT), and a prediction error signal (a signal corresponding to the prediction difference signal output from the subtractor 5) indicating the inverse discrete cosine transform result is supplied to the adder 11. Output.
The inverse slant converter 33 of the inverse conversion unit 10 performs inverse slant conversion on the conversion coefficient output from the inverse conversion method selection unit 31 when the inverse slant conversion is selected as the inverse conversion method by the inverse conversion method selection unit 31. A prediction error signal (a signal corresponding to the prediction difference signal output from the subtractor 5) indicating the result of the inverse slant conversion is output to the adder 11.

As apparent from the above, according to the second embodiment, the encoding control unit 72 refers to the distribution of the DCT coefficients output from the conversion unit 71, and the image signal indicating the input image is a gradation signal. If the image signal is a gradation signal, AC (0, 1), which is the AC component located at the lowest frequency in the horizontal direction, and the lowest frequency in the vertical direction Since the instruction to quantize the DCT coefficient is output to the quantization unit 73 so that the AC component other than the AC component (AC (1, 0)) that is the current AC component becomes zero, Similarly, it is possible to obtain an image encoding apparatus that outputs encoded data that can obtain a decoded image that does not generate visual noise such as a striped pattern with a small code amount.
However, in this second embodiment, it is sufficient that the transforming unit 71 mounts only the DCT device 21 and the quantizing unit 73 only mounts the DCT coefficient quantizer 23. There is an effect that the configuration of the quantifying device can be simplified.

In the second embodiment, when the image signal indicating the input image is estimated to be a gradation signal, the conversion unit 71 performs a discrete cosine transform (DCT) on the prediction difference signal, and the inverse conversion unit 10 performs the DCT coefficient. Strictly speaking, correction processing is required for the conversion coefficient.
Although the correction value depends on the characteristics of the gradation signal, it takes a value of approximately 0.96 to 1.04. Therefore, the quantization error caused by the quantization process becomes a larger value, which is substantially corrected. No processing is necessary.

Embodiment 3 FIG.
12 is a block diagram showing an image coding apparatus according to Embodiment 3 of the present invention. In the figure, the same reference numerals as those in FIGS.
As in the first embodiment, the image decoding device in the third embodiment uses the image decoding device in FIG.

In the third embodiment, as in the second embodiment, the conversion unit 71 mounts only the DCT device 21 and the quantization unit 73 mounts only the DCT coefficient quantizer 23.
However, in the third embodiment, the encoding control unit 7 of FIG. 1 is implemented, and whether or not the image signal indicating the input image is a gradation signal as in the encoding control unit 72 of FIG. There is no need to perform such processing.
Hereinafter, processing contents of the image encoding device will be described.

Unlike the conversion unit 6 in FIG. 1, the conversion unit 71 includes only the DCT unit 21. When the prediction difference signal is received from the subtracter 5, the conversion unit 71 performs a discrete cosine transform (DCT) on the prediction difference signal. The DCT coefficient indicating the result of the discrete cosine transform is output to the quantization unit 73.
As in the first embodiment, the encoding control unit 7 outputs a quantization parameter that is referred to when the quantization unit 73 and the inverse quantization unit 9 perform quantization / inverse quantization, and performs encoding. Controls the code amount of data and the encoded image quality.

When the quantizing unit 73 receives the DCT coefficient from the converting unit 71, the quantizing unit 73 quantizes the DCT coefficient according to the quantization parameter output from the encoding control unit 7, and uses the quantized DCT coefficient as a quantized coefficient. The result is output to the inverse quantization unit 9 and the entropy encoding unit 15.
However, the quantization unit 73 receives an instruction from the encoding control unit 7 to quantize the DCT coefficient so that AC components other than AC (0, 1) and AC (1, 0) become zero. Therefore, the DCT coefficient is not quantized so that AC components other than AC (0, 1) and AC (1, 0) become zero.

  When receiving the quantization coefficient from the quantization unit 73, the inverse quantization unit 9 converts the quantization coefficient into an inverse quantum according to the quantization parameter output from the encoding control unit 7 as in the first embodiment. The transform coefficient (DCT coefficient) that is the result of the inverse quantization is output to the inverse transform unit 10.

When receiving the transform coefficient (DCT coefficient) from the inverse quantization unit 9, the inverse transform method selection unit 31 of the inverse transform unit 10 in the horizontal direction out of the AC components in the transform coefficient as in the first embodiment. Whether or not non-zero AC components are included in AC components other than AC (0, 1) located at the lowest frequency and AC (1, 0) located at the lowest frequency in the vertical direction Determine whether.
If the non-zero AC component is not included, the inverse transform method selection unit 31 selects the inverse slant transform as the inverse transform method, and outputs the transform coefficient output from the inverse quantization unit 9 to the inverse slant converter 33. To do.
On the other hand, if a non-zero AC component is included, the inverse discrete cosine transform (inverse DCT) is selected as the inverse transform method, and the transform coefficient output from the inverse quantization unit 9 is output to the inverse DCT unit 32.

As in the first embodiment, the inverse DCT unit 32 of the inverse transform unit 10 uses the inverse transform method selection unit when the inverse discrete cosine transform (inverse DCT) is selected as the inverse transform method by the inverse transform method selection unit 31. The transform coefficient output from 31 is subjected to inverse discrete cosine transform (inverse DCT), and a prediction error signal (a signal corresponding to the prediction difference signal output from the subtractor 5) indicating the inverse discrete cosine transform result is supplied to the adder 11. Output.
The inverse slant converter 33 of the inverse conversion unit 10 performs inverse slant conversion on the conversion coefficient output from the inverse conversion method selection unit 31 when the inverse slant conversion is selected as the inverse conversion method by the inverse conversion method selection unit 31. A prediction error signal (a signal corresponding to the prediction difference signal output from the subtractor 5) indicating the result of the inverse slant conversion is output to the adder 11.

In the image coding apparatus according to the third embodiment, when coding a prediction difference signal having a change like a cosine function, the inverse transformation method selection unit 31 of the inverse transformation unit 10 selects the inverse slant transformation as the inverse transformation method. As a result, errors occur and the performance deteriorates. However, normal image signals have overwhelmingly more signal components that change diagonally and linearly, and errors may occur. Is extremely small.
In the third embodiment, unlike the second embodiment, since it is not necessary to mount the encoding control unit 72 having a control function, the configuration of the image encoding device can be simplified.

Embodiment 4 FIG.
13 is a block diagram showing an image coding apparatus according to Embodiment 4 of the present invention. In the figure, the same reference numerals as those in FIG.
As in the first embodiment, the image decoding device in the fourth embodiment uses the image decoding device in FIG.

Similar to the encoding control unit 72 in FIG. 11, the encoding control unit 74 refers to the distribution of DCT coefficients output from the conversion unit 71 and determines whether or not the image signal indicating the input image is a gradation signal. When the image signal is a gradation signal, AC (0, 1), which is the AC component located at the lowest frequency in the horizontal direction, and the AC component located at the lowest frequency in the vertical direction. An instruction to quantize the DCT coefficient is output to the quantization unit 73 so that AC components other than AC (1,0) become zero.
In addition, when the image signal is a gradation signal, the encoding control unit 74 outputs an instruction to quantize the DCT coefficient with higher accuracy to the quantization unit 73 than when the image signal is not a gradation signal. .
The encoding control unit 74 and the quantization unit 73 constitute a quantization unit.

  In FIG. 13, the intra prediction unit 1, the motion detection unit 2, the motion compensation prediction unit 3, the selection switch 4, the subtracter 5, the conversion unit 71, the coding control unit 74, and the quantization unit, which are components of the image coding device. 73, the inverse quantization unit 9, the inverse transform unit 10, the adder 11, the loop filter 13 and the entropy encoding unit 15 are each dedicated hardware (for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, etc. However, when the image encoding device is configured by a computer, an intra prediction unit 1, a motion detection unit 2, a motion compensation prediction unit 3, a selection switch 4, a subtractor 5, the processing contents of the transform unit 71, the coding control unit 74, the quantization unit 73, the inverse quantization unit 9, the inverse transform unit 10, the adder 11, the loop filter 13 and the entropy coding unit 15 are described. The program stored in the memory of the computer, may execute a program that the CPU of the computer is stored in the memory.

Next, the operation will be described.
Similar to the encoding control unit 72 in FIG. 11, the encoding control unit 74 refers to the DCT coefficient distribution output from the conversion unit 71 to determine whether or not the image signal indicating the input image is a gradation signal. judge.
When the encoding control unit 74 determines that the image signal is a gradation signal, AC (0, 1), which is the AC component located at the lowest frequency in the horizontal direction, and the lowest frequency in the vertical direction. An instruction to quantize the DCT coefficient is output to the quantization unit 73 so that AC components other than AC (1, 0), which is the AC component being performed, become zero.
In addition, when the image signal is a gradation signal, the encoding control unit 74 outputs an instruction to quantize the DCT coefficient with higher accuracy to the quantization unit 73 than when the image signal is not a gradation signal. .

When the quantization unit 73 receives the DCT coefficient from the conversion unit 71, the quantization unit 73 quantizes the DCT coefficient in accordance with the quantization parameter output from the encoding control unit 72, and uses the quantized DCT coefficient as a quantization coefficient. The result is output to the inverse quantization unit 9 and the entropy encoding unit 15.
However, when the quantization unit 73 receives an instruction from the encoding control unit 74 to quantize the DCT coefficient so that AC components other than AC (0, 1) and AC (1, 0) become zero, The DCT coefficients are quantized so that AC components other than AC (0, 1) and AC (1, 0) become zero.
The quantization coefficient in this case (the DCT coefficient after quantization) is equivalent to the slant transform coefficient quantized by the slant transform coefficient quantizer 24 of FIG.

Further, when the quantizing unit 73 receives an instruction to quantize the DCT coefficient with high accuracy from the encoding control unit 74, the quantizing unit 73 quantizes the DCT coefficient with high accuracy.
That is, if the image signal indicating the input image is not a gradation signal and the encoding control unit 74 does not instruct to quantize the DCT coefficient with high accuracy, the quantization unit 73 increases the quantization step size, for example. Then, the DCT coefficient is roughly quantized.
On the other hand, if the image signal is a gradation signal and the encoding control unit 74 instructs to quantize the DCT coefficient with high accuracy, for example, the quantization step size is reduced to finely quantize the DCT coefficient. To do.

In the fourth embodiment, when the image signal is a gradation signal, the DCT coefficient is quantized with higher accuracy than when the image signal is not the gradation signal. The accuracy of gradation is determined by the conversion coefficients AC (0,1) and AC (1,0), and the accuracy of these two AC components is very important to improve the reproducibility of the gradation of the decoded image. That's why.
When quantization is performed with high accuracy, the amount of code usually increases. However, when the signal is determined to be a gradation signal, the increase in the amount of code is small because there are only two non-zero AC components.

  As is apparent from the above, according to the fourth embodiment, when the image signal is a gradation signal, the encoding control unit 74 can quantize the DCT coefficient with higher accuracy than when the image signal is not a gradation signal. Since the instruction to perform the conversion is output to the quantization unit 73, there is an effect that the gradation reproducibility of the decoded image can be improved.

Embodiment 5 FIG.
FIG. 14 is a block diagram showing an image encoding apparatus according to Embodiment 5 of the present invention. In the figure, the same reference numerals as those in FIG.
In the same way as the intra prediction unit 1 in FIGS. 1 and 11 to 15, the intra prediction unit 81 inputs an image signal in units of macroblocks, and uses the local decoded signal stored in the intra prediction memory 12 to generate the image. A process for generating a predicted image is performed by performing an intra prediction process on the luminance signal constituting the signal.
However, unlike the intra prediction unit 1 of FIGS. 1 and 11 to 15, the intra prediction unit 81 determines whether or not the image signal indicating the input image is a gradation signal, and if the image signal is a gradation signal, the direction is determined. A predictive image is generated using a DC intra prediction encoding method (mode 2 encoding method defined in the AVC standard).
The intra prediction unit 81 constitutes a predicted image generation unit.

Next, the operation will be described.
In AVC, as described in the first embodiment, as the intra prediction process for the luminance signal, nine types with a unit of 4 × 4 blocks, nine types with a unit of 8 × 8 blocks, Four types are defined by a method in which 16 × 16 blocks are used as one unit.
When the image signal indicating the input image is a gradation signal, if the gradation signal has only an AC component in the horizontal direction or only an AC component in the vertical direction, the AVC intra prediction processing is also performed. Prediction is possible.
However, since it is impossible to predict a gradation signal including an oblique gradation component, the prediction difference signal includes a complex component. As a result, non-zero AC components increase, leading to a decrease in encoding efficiency.

Therefore, in the fifth embodiment, the intra prediction unit 81 determines whether or not the image signal indicating the input image is a gradation signal, and if the image signal is the gradation signal, the DC having no directivity is determined. A prediction image is generated using an intra prediction encoding method (mode 2 encoding method defined in the AVC standard).
When a prediction image is generated using a coding method of DC intra prediction, since the intra prediction signal indicating the prediction image includes only a DC component, the AC component of the prediction difference signal output from the subtracter 5 is used. Is stored as is.
For this reason, if the slant transform coefficient after quantization is selected as a quantization coefficient and the quantization coefficient is encoded, AC components other than AC (0, 1) and AC (1, 0) are zero. Therefore, a large amount of code is not generated.

Here, when the image signal indicating the input image is a gradation signal, the intra prediction unit 81 has been shown to generate a prediction image using a DC intra prediction encoding method having no directionality. The intra prediction unit 81 may not perform the intra prediction process.
In this case, the intra prediction signal output from the intra prediction unit 81 is all zero, and the selection switch 4 selects the motion compensated prediction signal output from the motion compensation prediction unit 3, and the motion compensated prediction signal is Prediction signal selection information indicating the selection is output to the entropy encoding unit 15.
Thus, when the intra prediction unit 81 is configured not to perform the intra prediction process, there is an effect that the processing amount of the image encoding device can be reduced.

Further, when the image signal indicating the input image is a gradation signal, the intra prediction unit 81 does not perform intra prediction in the pixel region adopted in AVC, but performs DC in the transform coefficient region adopted in MPEG-2. You may make it use prediction and DC / AC prediction in the conversion coefficient area | region employ | adopted by MPEG-4.
For example, when a gradation signal as shown in FIG. 7 is input and the image continues vertically, the obtained DCT coefficients and slant transform coefficients have the same value in the upper and lower blocks.
Therefore, if the process of predicting the component of AC (0, x) (x = 1 to 7) from the upper block adopted in MPEG-4 is performed, the AC components are all zero, so that the encoding efficiency is improved. Can be further enhanced.

In the case of slant transformation, only AC (0, 1) becomes a non-zero AC component. Therefore, not all seven AC components of x = 1 to 7 are predicted, but AC (0, Even if only one of 1) is predicted, the same effect can be obtained.
Here, the gradation signal changing in the horizontal direction has been described as an example. However, in the case of the gradation signal changing in the vertical direction, a component of AC (y, 0) (y = 1 to 7) is obtained from the left block. What is necessary is just to perform the process to predict.

Embodiment 6 FIG.
In the first and fifth embodiments, AC (0, 1) which is the AC component located at the lowest frequency in the horizontal direction and AC (1) which is the AC component located at the lowest frequency in the vertical direction. , 0), if the non-zero AC component is not included, the slant transformation and the inverse slant transformation are shown. However, AC (0,3), AC (0,5), Even when AC (0,7), AC (3,0), AC (5,0), and AC (7,0) are non-zero, the slant transformation and the inverse slant transformation may be selected. The same effects as those of the first and fifth embodiments can be obtained.

  The reason why the slant transformation and the inverse slant transformation are selected when these six AC components are non-zero is that the gradation signal always has a difference of “2” between the left and right pixels as shown in FIG. For example, when the gradation is more gradual and is increased by 1 every 2 pixels as shown in FIG. 15, for example, the DCT coefficient is AC (1, 0), AC as shown in FIG. This is because the four AC components of (0, 3), AC (0, 5), and AC (0, 7) are non-zero.

In addition, slant conversion and inverse slant conversion are also performed when all AC components represented by AC (0, x) and AC (y, 0) (x = 1 to 7, y = 1 to 7) are non-zero. May be selected.
In the case where the gradation signal is not uniform as generated by computer graphics but is a gradation with respect to a natural object including noise and fluctuation, AC (0, 1), AC (1,0) This is because many AC components other than are non-zero.

In Embodiments 1 to 6, it is assumed that the block size for the conversion process is 8 × 8, but the present invention is not limited to this. For example, even if the block size is 4 × 4 or 16 × 16 Good.
Also, the block size may be a rectangular block such as 8 × 16 or 16 × 8, or may be a conversion process using a three-dimensional block extending in the time direction instead of two-dimensional. Often, similar effects can be obtained by similar processing.

  In the international standardized MPEG-2 and MPEG-4, the intra prediction unit 1, the intra prediction memory 12, and the loop filter 13 are not provided, and a part of the processing is different from that of the AVC. By applying it to the unit 6 / inverse transform unit 10 and the inverse transform unit 43 of the image decoding apparatus, it is possible to obtain the same effect as in the case of AVC.

In the first to sixth embodiments, it has been described that either the discrete cosine transform or the slant transform is selected in the same manner without distinguishing between the luminance signal and the color difference signal. However, the present invention is not limited to this. Different processing may be performed depending on the color difference signal.
For example, a luminance signal containing many fine signal components is always subjected to discrete cosine transform, and a method of selecting either discrete cosine transform or slant transform only for a color difference signal having a rough signal component is conceivable. Alternatively, as described in Embodiment 6, for the luminance signal, slant conversion is selected when a plurality of AC coefficients are non-zero, and for the color difference signal, other than AC (0, 1) and AC (1, 0). A method of selecting a slant transformation when the coefficient of is non-zero is conceivable. By utilizing the difference in signal characteristics between the luminance signal and the color difference signal, it is possible to further increase the encoding efficiency.
It is also possible to easily change the criterion for selecting either the discrete cosine transform or the slant transform according to the characteristics of the input image signal. For example, in the case of the configuration as in the sixth embodiment, the slant transformation is selected when only the coefficient at any position is non-zero, or the slant transformation is selected when the number of coefficients is less than zero. This can be realized by allowing the position or number of non-zero coefficients to be changed by both the encoding device and the decoding device, and including data indicating the position or number of non-zero coefficients in a part of the encoded data. .

  DESCRIPTION OF SYMBOLS 1,81 Intra prediction part (prediction image generation means), 2 motion detection part (prediction image generation means), 3 motion compensation prediction part (prediction image generation means), 4 selection switch (prediction image generation means), 5 subtractor ( Prediction difference signal calculation means), 6 conversion section (signal conversion means), 7 encoding control section (quantization means), 8 quantization section (quantization means), 9 inverse quantization section, 10 inverse conversion section, 11 addition , 12 intra prediction memory, 13 loop filter, 14 motion compensation prediction frame memory, 15 entropy coding unit (variable length coding means), 21 DCT device, 22 slant transformer, 23 DCT coefficient quantizer, 24 Slant transform coefficient quantizer, 25 transform method selector, 31 inverse transform method selector, 32 inverse DCT device, 33 inverse slant transformer, 41 entropy decoder (variable length recovery) Means), 42 inverse quantization unit (inverse quantization unit), 43 inverse transform unit (inverse transform method selection unit, inverse transform unit), 44 intra prediction unit (decoded image generation unit), 45 motion compensation prediction unit (decoded image) Generating means), 46 selection switch (decoded image generating means), 47 adder (decoded image generating means), 48 intra prediction memory (decoded image generating means), 49 loop filter (decoded image generating means), 50 motion compensated prediction Frame memory (decoded image generating means), 61 inverse transform method selecting section, 62 inverse DCT device, 63 inverse slant transformer, 71 transform section (signal transform means), 72, 74 encoding control section (quantization means), 73 Quantization unit (quantization means), 101 intra prediction unit, 102 motion detection unit, 103 motion compensation prediction unit, 104 selection switch, 105 subtractor, 106 variable , 107 coding control unit, 108 quantization unit, 109 inverse quantization unit, 110 inverse transformation unit, 111 adder, 112 intra prediction memory, 113 loop filter, 114 motion compensation prediction frame memory, 115 entropy coding 121, entropy decoding unit, 122 inverse quantization unit, 123 inverse transform unit, 124 intra prediction unit, 125 motion compensation prediction unit, 126 selection switch, 127 adder, 128 intra prediction memory, 129 loop filter, 130 motion compensation Predictive frame memory.

Claims (7)

  A prediction image generating unit that generates a prediction image from an image signal indicating an input image and outputs a prediction signal indicating the prediction image; a difference between the prediction signal output from the prediction image generation unit and the image signal indicating the input image And a prediction difference signal calculation unit that outputs a prediction difference signal indicating the difference, and a discrete cosine transform that outputs the prediction difference signal output from the prediction difference signal calculation unit and indicates the result of the discrete cosine transformation A conversion coefficient is output, and a slant conversion is performed on the prediction difference signal output from the prediction difference signal calculation unit, and a slant conversion coefficient indicating a result of the slant conversion is output, and a signal conversion unit outputs the conversion coefficient. Quantize the discrete cosine transform coefficient and quantize the slant transform coefficient output from the signal transforming means. Quantization means that selects a discrete cosine transform coefficient after quantization or a slant transform coefficient after quantization with reference to the distribution of AC components in the slant transform coefficient, and a quantized transform selected by the quantization means An image encoding apparatus comprising: a variable-length encoding unit that performs variable-length encoding on a coefficient and outputs encoded data that is a result of the encoding.   The quantizing means includes an AC component other than the AC component located at the lowest frequency in the horizontal direction and the AC component located at the lowest frequency in the vertical direction among the AC components in the slant transform coefficient after quantization. If the non-zero AC component is not included, the quantized slant transform coefficient is selected. If the non-zero AC component is included, the quantized discrete cosine transform coefficient is selected. The image encoding apparatus according to claim 1.   A prediction image generating unit that generates a prediction image from an image signal indicating an input image and outputs a prediction signal indicating the prediction image; a difference between the prediction signal output from the prediction image generation unit and the image signal indicating the input image And a prediction difference signal calculation unit that outputs a prediction difference signal indicating the difference, and a discrete cosine transform that outputs the prediction difference signal output from the prediction difference signal calculation unit and indicates the result of the discrete cosine transformation A signal transforming means for outputting a transform coefficient; a quantizing means for quantizing the discrete cosine transform coefficient output from the signal transforming means; and a variable length coding of the discrete cosine transform coefficient quantized by the quantizing means, In the image coding apparatus comprising: variable length coding means for outputting coded data that is the coding result, the quantization means comprises the signal conversion unit. To determine whether the image signal indicating the input image is a gradation signal or not, and if the image signal is a gradation signal, the lowest frequency in the horizontal direction is determined. An image coding apparatus characterized in that the discrete cosine transform coefficient is quantized so that an AC component other than an AC component located at a position and an AC component located at the lowest frequency in the vertical direction becomes zero.   When the image signal indicating the input image is a gradation signal, the quantization means performs quantization of the discrete cosine transform coefficient output from the signal conversion means with higher accuracy than when the image signal is not a gradation signal. 4. The image encoding device according to claim 3, wherein   The predicted image generation means determines whether or not the image signal is a gradation signal when generating the predicted image from the image signal indicating the input image. If the image signal is the gradation signal, the predicted image generation unit has directionality. 5. The image encoding device according to claim 1, wherein a prediction image is generated using a coding scheme of non-DC intra prediction.   Variable length decoding means for variable length decoding transform coefficients from encoded data, inverse quantization means for inverse quantizing the transform coefficients variable length decoded by the variable length decoding means, and inverse quantization by the inverse quantization means With reference to the distribution of the AC component in the later transform coefficient, the inverse transform method selecting means for selecting inverse discrete cosine transform or inverse slant transform as the inverse transform method for the transform coefficient, and the inverse transform method selecting means selected by the inverse transform method selecting means If the inverse transform method is an inverse discrete cosine transform, a prediction error signal is calculated by inverse discrete cosine transform of the transform coefficient. If the inverse transform method is an inverse slant transform, the transform coefficient is inversely slant transformed. Inverse conversion means for calculating a prediction error signal; and decoded image generation means for generating a decoded image signal indicating a decoded image from the prediction error signal calculated by the inverse conversion means. The image decoding apparatus.   The inverse transformation method selection means is located at the lowest AC frequency component in the horizontal direction and the lowest frequency in the vertical direction among the AC components in the transform coefficient after inverse quantization by the inverse quantization means. If the AC component other than the AC component does not include a non-zero AC component, the inverse slant transform is selected as the inverse conversion method, and if the non-zero AC component is included, the inverse discrete cosine as the inverse conversion method. 7. The image decoding apparatus according to claim 6, wherein conversion is selected.

Images