JPH04180486A - Moving image data compressing device - Google Patents

Abstract

PURPOSE:To reduce the degradation of picture quality and to prevent the increase of processing time by finding the average value of the absolute values of the value of non-zero among the whole NXN coefficients, making it constant times, finding a threshold value and comparing it with the absolute value of the NXN coefficients. CONSTITUTION:Input picture signals from an input terminal 101 are divided into 8X8 blocks, luminous signal 4 blocks, color difference 2 blocks and is inputted to a movement compensation predicting means 112 and the signal of a macro block of which predicted error becomes minimum is called from the signal 1 frame before already stored. For the signal, a difference with the input picture signal is taken in a differential means 103 and it is converted by a discrete cosine converting means 105 every the block. Then, it is quantized by a quantizing means 106 and is outputted to a threshold value generating means 107. A threshold value and a quantizing coefficient are outputted from the means 107, is Huffman-encoded in a variable length encoding means 109 after code quantity control processing is executed to the quantizing coefficient in a threshold value discriminating means 108 and is outputted to an output terminal 116.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a moving image data compression device that orthogonally transforms and quantizes an image signal.

[Prior Art] Since the amount of image information is generally enormous, when storing or transmitting it, some kind of information compression is performed. This information compression means notebook city, C
There is a video encoding system (H, 261) for px 64 kbit/s audiovisual services that is being standardized in CITT and 5GXV (transmission systems and equipment). In H,261, the input image signal is divided into two-dimensional blocks of 8 x 8 pixels, subjected to orthogonal transformation (IIi scattered cotecosine transformation and quantization), and then variable-length coded and output.This variable-length code Huffman encoding is used for the conversion, and a code word is determined by a continuous zero (run) and a subsequent non-zero value (level).

Conventionally, the amount of generated codes was determined by variable-length coding the d code of the quantization means, and the amount of generated codes was controlled by changing the quantization step according to the result. The control method is
As stated in D-45 of the 1989 Institute of Electronics, Information and Communication Engineers Autumn National Conference, the output of the quantization means is variable-length coded to obtain the generated code amount several times by changing the quantization step, and the result is The quantization step for quantization with the set code amount was determined from

[Problems and Objectives to be Solved by the Invention] However, in such conventional technology, when the set code amount is set to a low rate of about 64 kbps, the quantization step size becomes considerably large, and coefficients close to DC coefficients become rough. Since the decoded image is quantized, noise and block distortion appear prominently, making the image unsightly, and processing time increases as feedback control is performed to achieve the set code amount. There is a problem.

Therefore, the present invention is intended to solve these problems, and its purpose is to reduce the deterioration in image quality between the original image and the decoded image, and to reduce the amount of generated code.
Furthermore, a method for preventing an increase in processing time is provided.

[Means for Solving the Problems] The moving image data compression device of the present invention converts image signals into N×N
(N: integer) @blocking means for dividing into elementary two-dimensional blocks; difference means for taking the difference between the block signal of N×N pixels which is the output of the blocking means and the image signal of one frame before; orthogonal transform means that orthogonally transforms the output N×Ni element difference block signal and outputs N×N transform coefficients; and quantization that quantizes the transform coefficients and outputs N×N quantized coefficients. and variable length encoding means for variable length encoding the quantization coefficients, the absolute value of a non-zero value among all N×N coefficients output from the quantization means. Threshold generation means calculates the average value of and multiplies it by a constant, and compares the threshold value which is the output of the threshold generation means with the absolute value of the N×N coefficients, and if the absolute value of the coefficient is equal to or greater than the threshold value, outputs it as is. , a threshold value determining means for outputting a value of zero when the threshold value is not reached.

[Example] DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below based on drawings showing examples thereof.

FIG. 1 is a block diagram when the present invention is applied to the intra-frame mode of an H.261 video encoder.

In FIG. 1, an input image signal from an input terminal 101 is divided by a blocking means 102 into macro blocks consisting of 8×8 blocks, 4 blocks of luminance signals, and 2 blocks of color difference signals, a total of 6 blocks. Since prediction is performed between frames, the encoding control 104 connects the switches 114 and 115 to the lower side as shown in FIG. The image signal output in macroblock units from the blocking means 102 is first input to the motion compensation prediction means 112, which calls out the signal of the macroblock whose prediction error is smaller than the already stored signal one frame before. (motion vector detection), and the signal of that macroblock is output.

After this signal passes through a loop filter 113, a difference between it and the input image signal is taken by a difference means 103. The difference signal between this input image signal and the macroblock indicated by the motion vector of one frame before is l111 for each block.
N! It is converted by i-cosine conversion means 105. and,
The signal is quantized by the quantization means 106 and output to the threshold generation means 107. A threshold value and a quantization coefficient are outputted from the threshold generation means 107, and after a code amount control process is performed on the quantization coefficient by the threshold value determination means 108, the variable length encoding means 109 encodes the threshold value and the quantization coefficient, and then Huffman encoding is performed by the variable length encoding means 109, and the output terminal 116 It will be 8 times. Threshold generating means 107 and threshold determining means 1
The processing at step 08 will be described in detail in the explanation of FIG.

The output of the threshold determination means 108 is dequantized by the dequantization means 110 and inverse cosine transformed by the inverse discrete cosine transformation means 111, and then the signal of the block indicated by the motion vector is converted into a signal passed through the loop filter 113. The sum is added and recorded in the motion compensation prediction means 112.

Here, the threshold value generating means 107 and the threshold value determining means 108
The processing will be explained according to the flowchart in FIG.

First, the threshold generation means 107 calculates the average value of the absolute values of non-zero values among the 8×8 quantization coefficients that are the pressure of the quantization means 106. Furthermore, the obtained average value is multiplied by n (n:
The threshold value is determined by (constant). Next, the threshold value determining means 108 compares the threshold value which is the output of the threshold value generating means 107 with the absolute values of all 64 quantized coefficients, and if the value is equal to or higher than the threshold value, the pressure is applied as it is to reach the threshold value. If it is not satisfied, it is output as zero.

[Effects of the Invention] As described above, in the present invention, the amount of generated codes can be reduced without increasing the quantization step size too much. In other words, when performing two-dimensional encoding of consecutive zeros (runs) and subsequent non-zero values (levels), it is effective to increase the number of zeros, and in the present invention, coefficients with low levels are prioritized. By setting it to zero, the amount of generated code can be reduced without increasing the quantization step. Furthermore, among the relatively finely quantized coefficients, some high-level coefficients remain, so although the decoded image has somewhat blurred outlines, it is smooth overall and is free from quantization noise. It is possible to reproduce images that are easy to view and have little block distortion. Furthermore, since the amount of generated code can be reduced, the number of dropped frames is also reduced, and a natural decoding interval can be obtained.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the video/moving image encoding device of the present invention. FIG. 2 is a flowchart of an embodiment of the present invention. 101... Input terminal 102... Block division means 103... Differential means 104... Encoding control 105... Discrete cosine transformation means 106... Quantization means 107... Threshold generation means 108. ...Threshold determination means 109...Variable length encoding means 110...Dequantization means 111...Inverse discrete cosine transformation means 112...Motion compensation prediction means 113...Loop filter 114...Switch 115...Switch 116...Output terminal or more Applicant: Seiko Epson Corporation

Claims (1)

[Scope of Claims] Blocking means for dividing an image signal into two-dimensional blocks of N×N (N: an integer) pixels, and a block signal of N×N pixels which is an output of the blocking means and one frame before. and a difference block signal of N×N pixels, which is the output of the difference means, is orthogonally transformed,
orthogonal transform means for outputting N×N transform coefficients, quantization means for quantizing the transform coefficients and outputting N×N quantized coefficients, and variable length code for variable length encoding the quantized coefficients. In a moving image data compression device comprising a quantizing means, (a) a threshold generating means for calculating the average value of the absolute values of non-zero values among all N×N coefficients output from the quantizing means and multiplying it by a constant; and (b) the threshold value that is the output of the threshold value generation means, and the N×
A threshold value determining means that compares the absolute values of the N coefficients and outputs the coefficient as is if it is equal to or greater than the threshold value, and outputs it as zero if it is less than the threshold value. Video data compression device.