KR20080018690A - Apparatus for improving resolution of image and method thereof - Google Patents

Abstract

An apparatus and a method for improving sharpness of an image are provided to improve sharpness in various images by differently applying a weight value according to statistical characteristics of the images and prevent an artifact by preventing excessive improvement of sharpness of an image. The first color space converting unit(100) receives red, green and blue color signals, converts them into a 'Y' signal of a luminance signal and Cb and Cr signals of color information, and outputs the same. An image characteristics extracting unit(110) extracts a local variance of the 'Y' signal outputted from the first color space converting unit and a global variance of a corresponding frame. A sharpness improving unit(120) improves sharpness of the 'Y' signal by using the extracted local variance and the global variance. The second color space converting unit(130) receives the Y signal with improved sharpness and the Cb and Cr signal outputted from the first color space converting unit, converts them into red, green and blue color signals, and outputs the same.

Description

Apparatus for improving resolution of image and method

1 is a block diagram illustrating an apparatus for improving image sharpness by a conventional unsharp masking technique.

Figure 2 is a block diagram showing an embodiment of a device for improving the sharpness of the image of the present invention.

3 is a block diagram showing the configuration of an image characteristic extraction unit according to the present invention;

Figure 4 is a block diagram showing the configuration of the sharpness improvement unit of the present invention.

5 is a graph showing weight values according to local deviations when the entire standard deviation is fixed at 128;

6 is a flowchart illustrating an embodiment of a method for improving image sharpness of the present invention.

* Explanation of symbols for the main parts of the drawings *

100: first color space conversion unit 110: image characteristic extraction unit

120: sharpness improvement unit 130: second color space conversion unit

The present invention relates to an apparatus and method for improving image sharpness.

Cathode ray tubes (CRTs), which have been the mainstay of display devices, must have a certain distance between the electron gun and the fluorescent surface in order to deflect the electron beam with respect to the display area.

Due to these characteristics, CRTs are relatively bulky and take up a lot of weight compared to the screen size, and there is a limit to the planarization and enlargement of the screen.

On the other hand, due to the rapid development of science and the abundance of life, modern people are inclined to pursue high-quality display, high-quality and large-scale display devices.

In order to overcome the demands of the modern man and the problems of the CRT, the newly emerged are liquid crystal display (LCD), plasma display panel (PDP), and projection TV. Such a flat panel display device is thin, lightweight, and can realize a large screen.

However, display devices such as LCDs and PDPs frequently reproduce output video signals having reduced resolution due to noise signals, blurring, and limitation of signal bands of input video signals as the screen size increases. do.

Conventionally, before the image data is displayed on the screen in order to improve the image clarity of the flat panel display device, the image data is applied by using an Unsharp Masking technique, which is a digital filter, so that the image can be seen more clearly. .

The unsharp masking technique emphasizes the details of the image and increases the contrast effect of the boundary part. Therefore, the unsharp masking technique is used to remove the unnecessary noise of the image image to make the image clear when the acquired image is ambiguous.

1 is a block diagram illustrating an apparatus for improving image sharpness by a conventional unsharp masking technique.

As shown in the drawing, a conventional image sharpness improving apparatus using an unsharp masking technique includes a high pass filter (HPF) 10, a multiplier 20, a signal delay unit 30, and an addition. It consists of a part (40).

The high pass filter 10 filters an input signal and outputs a high frequency signal.

That is, the high pass filter 10 removes the soft portion from the input image and passes only the sharp portion, that is, high frequency signals, to increase the sharpness of the image.

The high pass filter 10 sums the coefficients of a mask having a size of 3x3 pixels, 5x5 pixels, and 7x7 pixels by using positive coefficients at the center and negative coefficients at the periphery thereof. If this value is set to 0 and applied to the image data, a clear image can be obtained.

The multiplier 20 multiplies and outputs a fixed weight value to a signal output from the high pass filter 10.

The signal delay unit 30 delays the input signal Input for a predetermined time and then outputs the signal.

That is, the signal delay unit 30 multiplies the signal output from the high pass filter 10 by the multiplier 20 to a fixed weight value W and simultaneously outputs the delayed input signal Input-d. Output

Here, the signal delayed by the signal delay unit 30 for a predetermined time (Input-d) is the same as the signal (Input) input to the high pass filter (10).

The adder 150 adds and outputs the signal output from the multiplier 20 and the signal Input-d output from the signal delay unit 30.

A conventional method for improving the sharpness of an image using unsharp masking is expressed as a formula.

Where (m, n) represents the pixel position, y _i (m, n) represents the input brightness, W represents the weight, h _HP (m, n) represents the high pass filtered signal, y _o (m, n) represents the output luminance.

As can be seen in Equation 1, the output luminance is equal to the magnitude of the input luminance plus the weighted high pass filtered value. This method has the effect of enhancing the sharpness by strengthening the edge of the image.

However, such a classical method has a disadvantage of causing artificial distortion when applied to various images because the weight is fixed without considering the characteristics of the image.

That is, since the same weight is given to each image to the high pass filtered value, there is a problem that an unnatural image is output on the screen without properly utilizing the characteristics of the image.

Accordingly, an object of the present invention is to extract the local variance of the input luminance signal and the total variance of the corresponding frame, and adjust the weights using the extracted local variance and the total variance, thereby applying different weights according to statistical characteristics of the image. An object of the present invention is to provide an apparatus and method for improving image sharpness capable of improving natural sharpness in various images.

According to a preferred embodiment of the apparatus for improving image sharpness of the present invention, an R (red), G (green), and B (blue) color signal is input and converted into a Y signal of a luminance signal and a Cb and Cr signal of color information. A first color space converter to extract a local variance of the Y signal output from the first color space converter and a global variance of the corresponding frame; A sharpness improving unit that improves the sharpness of the Y signal by using local and total dispersions extracted by an extracting unit, and a Cb and Cr output by the Y signal and the first color space converting unit that have improved the sharpness by the sharpness improving unit And a second color space converter configured to receive a signal and convert the signal into an R (red), G (green), and B (blue) color signal.

The image characteristic extractor may include a local variance extractor configured to extract a local variance in a mask at a (m, n) pixel position of the Y signal output from the first color space converter. The local by using a global variance extracting unit for extracting a global variance of a corresponding frame of the Y signal, a memory in which the luminance information of the corresponding frame of the Y signal is stored, and the luminance information stored in the memory And a control unit configured to control the dispersion extracting unit and the total dispersion extracting unit to extract local dispersion and total dispersion of the Y signal.

The sharpness improving unit may include a high pass filter for filtering a Y signal output from the first color space converter and outputting a high frequency signal, and a local dispersion and total dispersion extracted from the image characteristic extraction unit. A weight adjusting unit for adjusting a weight to be applied to a corresponding image by using a weight adjusting unit, and a sharpness adjusting unit for receiving a high frequency signal output from the high pass filter and a weight value determined by the weight adjusting unit and outputting a luminance signal having improved sharpness Characterized in that made.

According to a preferred embodiment of the method for improving image sharpness of the present invention, an R (red), G (green), and B (blue) color signal is input and converted into a Y signal of a luminance signal and a Cb and Cr signal of color information. Extracting the local variance of the output Y signal and the global variance of the corresponding frame; and using the extracted local variance and the overall variance, the sharpness of the output Y signal is determined. And receiving the Y signal and the Cb and Cr signals having improved clarity, and converting the signals into R (red), G (green), and B (blue) color signals. do.

Hereinafter, the apparatus and method for improving image sharpness of the present invention will be described in detail with reference to FIGS. 2 to 6. In describing the present invention, when it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

2 is a block diagram illustrating an embodiment of an apparatus for improving sharpness of an image of the present invention.

As shown in the drawing, a first color space conversion unit which receives an R (red), G (green), and B (blue) color signal, converts the Y signal of the luminance signal and the Cb and Cr signals of color information, and outputs the converted color signal. 100), an image characteristic extraction unit 110 for extracting a local variance of the Y signal output from the first color space converter 100, and a global variance of the corresponding frame, and the image The sharpness improving unit 120 improves the sharpness of the Y signal by using the local dispersion and the total dispersion extracted by the feature extraction unit 110, and the Y signal and the sharpness improved by the sharpness improving unit 120. The second color space conversion unit 130 receives the Cb and Cr signals output from the one color space conversion unit 100 and converts the Cb and Cr signals into R (red), G (green), and B (blue) color signals. .

The first color space converter 100 receives luminance signals of R _i (m, n) (red), G _i (m, n) (green), and B _i (m, n) (blue). The signals are converted into Y _i (m, n) signals and Cb _i (m, n) and Cr _i (m, n) signals of color information and output.

The image characteristic extractor 110 extracts the statistical characteristic of the input luminance signal Y _i (m, n) and outputs the statistical characteristic to the sharpness improvement unit 120.

That is, the image characteristic extracting unit 110 obtains the variance of the input luminance signal Y _i (m, n) and outputs the variance to the sharpness improving unit 120.

Here, the image characteristic extraction unit 110 performs local variance in the mask and global variance of the frame at the (m, n) pixel position of the input luminance signal Y _i . Extract.

The sharpness improving unit 120 improves the sharpness of the image by using a local variance and a global variance of the input luminance signal Y _i (m, n) extracted by the image characteristic extraction unit 110. Outputs a luminance signal of Y _o (m, n).

That is, the sharpness improving unit 120 determines a weight value to be applied to the input image by using the local variance and the total variance of the input luminance signal Y _i (m, n) extracted by the image characteristic extraction unit 110. The sharpness of the image is improved according to the determined weight value.

According to this, different weight values may be applied according to statistical characteristics of an image, thereby improving natural sharpness in various images.

The second color space converter 130 may include a Y _o (m, n) signal having improved sharpness by the sharpness improver 120, and Cb _i (the output of the first color space converter 100). m, n) and Cr _i (m, n) color signals are inputted to output R _o (m, n), G _o (m, n), and B _o (m, n) color signals.

3 is a block diagram showing the configuration of an image characteristic extraction unit according to the present invention.

As shown therein, the local dispersion extracting unit 111 extracts a local variance in the mask at the (m, n) pixel position of the input luminance signal Y _i , and the input luminance signal Y _a global variance extractor 113 for extracting a global variance of the corresponding frame of _i , a memory 115 for storing luminance information of the input frame, and a frame stored in the memory 115 The local variance extractor 111 and the total variance extractor 113 are configured to control the local variance and the total variance of the input image by using the luminance information of the control unit 117.

Here, the local variance may be represented by the following Equation 2, and the local variance extractor 111 extracts the local variance in the mask at the (m, n) pixel position through Equation 2.

Where σ ² _l (m, n) represents the local variance in the mask at the (m, n) pixel position, and Y _i (k, l) represents the input luminance at the (k, l) pixel position in the mask. , mean (m, n) represents the average luminance in the mask centered on (m, n) pixels.

The total variance may be represented by Equation 3 below. The full variance extractor 113 extracts the entire variance of the corresponding frame of the input luminance signal Y _i through Equation 3.

Where σ ² _g represents the total variance of the frame, Y _i (k, l) represents the input luminance at the (k, l) pixel location within the mask, mean _g represents the average luminance of the frame, MN represents the total size of the frame. That is, M means the number of horizontal pixels of the frame, and N means the number of vertical pixels of the frame.

4 is a block diagram showing the configuration of the sharpness improving unit of the present invention.

As shown therein, a high pass filter 121 for filtering the input luminance signal Y _i (m, n) and outputting a high frequency signal h _HP (m, n), and an image characteristic extraction unit (not shown). A weight adjusting unit 124 for adjusting a weight to be applied to an input image by using the local variance and the total variance extracted from the C) and a high frequency signal h _HP (m, n) output from the high pass filter 121. And a sharpness controller 127 that receives the weight value determined by the weight controller 124 and outputs a luminance signal Y _o (m, n) having improved sharpness.

The high pass filter 121 removes the soft portion from the input luminance signal Y _i (m, n) and increases the sharpness of the image by passing only a sharp portion, that is, high frequency signals.

The high pass filter 121 sums the total coefficients by using positive coefficients in the center and negative coefficients in the center of the masks having a size of 3x3 pixels, 5x5 pixels, and 7x7 pixels. If this value is set to 0 and applied to the image data, a clear image can be obtained.

The weight adjusting unit 124 is a weight value to be applied to the input image using the local variance (σ ² _l (m, n)) and the total variance (σ ² _g ) output from the image characteristic extractor (not shown). Adjust

Here, the weight value w (m, n) using the local variance σ ² _l (m, n) and the total variance σ ² _g is determined by the following equation (4).

According to Equation 4, it can be seen that the maximum value of the weights w (m, n) is close to 1 and the minimum value is close to zero.

As the local variance σ ² _l (m, n) becomes larger than the total variance σ ² _g , the weight w (m, n) decreases, whereas the local variance σ ² _l (m, n) It can be seen that the weight w (m, n) becomes larger as it is smaller than the total variance σ ² _g .

According to Equation 4, by applying a weight value in consideration of the local variance (σ ² _l (m, n)) and the total variance (σ ² _g ) of the input image, it is possible to prevent excessive sharpness of the image, thereby preventing artifacts. ) Is effective to prevent.

5 is a graph showing weight values according to local deviations when the standard deviation is fixed to 128. FIG. As shown in the figure, it can be seen that the larger the local deviation, the smaller the weight.

As described above, according to the present invention, weights may be differently applied according to statistical characteristics of the image, and by applying weights according to Equation 4, excessive sharpness of the image may be prevented.

The sharpness controller 127 receives the high frequency signal h _HP (m, n) output from the high pass filter 121 and the weight value w (m, n) determined by the weight controller 124. The improved luminance signal Y _o (m, n) is output, and the output luminance signal Y _o (m, n) can be represented by the following equation (5).

Where (m, n) represents the pixel position, Y _i (m, n) represents the input brightness, w (m, n) represents the weight value, and h _HP (m, n) represents the high pass filtered Y _o (m, n) represents the output luminance.

6 is a flowchart illustrating an embodiment of a method for improving image sharpness of the present invention.

As shown in the drawing, first, in the image characteristic extraction unit 110, the local variance in the mask and the total variance of the corresponding frame at the (m, n) pixel position of the input luminance signal Y _i ( Global Variance) is extracted (step S100).

Next, the input luminance signal Y _i is filtered by the high pass filter 121 to output the high frequency signal h _HP (m, n), and weighted by using the local variance and the total variance extracted by the image characteristic extraction unit 110. The value w (m, n) is adjusted (step S110).

Subsequently, the sharpness adjusting unit 127 receives the high frequency signal h _HP (m, n) output from the high pass filter 121 and the weight value w (m, n) output from the weight adjusting unit 124. Outputs the improved luminance signal _Yo (m, n) (step S120).

Although the present invention has been described in detail with reference to exemplary embodiments above, those skilled in the art to which the present invention pertains can make various modifications to the above-described embodiments without departing from the scope of the present invention. I will understand.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by the claims below and equivalents thereof.

As described above, according to the present invention, the local variance of the input luminance signal and the total variance of the corresponding frame are extracted, and the weight is adjusted using the extracted local variance and the total variance, thereby weighting the weight according to statistical characteristics of the image. By applying it differently, it is possible to improve natural sharpness in various images and to prevent artifacts by preventing excessive sharpness of images.

Claims (8)

A first color space conversion unit which receives an R (red), G (green), and B (blue) color signal and converts the Y signal of the luminance signal and the Cb and Cr signals of the color information to output the color signal; An image characteristic extraction unit which extracts a local variance of the Y signal output from the first color space converter and a global variance of the corresponding frame; A sharpness improving unit which improves the sharpness of the Y signal by using local dispersion and total dispersion extracted by the image feature extraction unit; And The sharpness improving unit receives the Y signal with improved clarity and the Cb and Cr signals output by the first color space converter and converts them into R (red), G (green), and B (blue) color signals. An apparatus for improving sharpness of an image comprising a second color space converter. The method of claim 1, The image characteristic extraction unit, A local variance extractor extracting a local variance in a mask at a (m, n) pixel position of the Y signal output from the first color space converter; A global variance extracting unit which extracts a global variance of a corresponding frame of the Y signal; A memory for storing luminance information of a corresponding frame of the Y signal; And And a controller configured to control the local variance extracting unit and the total variance extracting unit to extract the local variance and the total variance of the Y signal by using the luminance information stored in the memory. The method of claim 2, The local variance extractor extracts the local variance through the following equation. (Mathematical formula)

Where σ ² _l (m, n) is the local variance in the mask at the (m, n) pixel position, Y _i (k, l) is the input luminance at the (k, l) pixel position in the mask, mean ( m, n) means the average luminance in the mask centered on the (m, n) pixels. The method of claim 2, The total variance extractor extracts the entire variance through the following equation. (Mathematical formula)

Where σ ² _g is the total variance of the frame, Y _i (k, l) is the input luminance at the (k, l) pixel position, mean _g is the average luminance of the frame, and M is the number of horizontal pixels in the frame. , N means the number of vertical pixels of the frame. The method of claim 1, The sharpness improvement unit, A high pass filter for filtering a Y signal output from the first color space converter to output a high frequency signal; A weight adjusting unit for adjusting a weight to be applied to the corresponding image by using the local variance and the total variance extracted by the image feature extractor; And And a sharpness adjuster configured to receive a high frequency signal output from the high pass filter and a weight value determined by the weight adjuster, and output a luminance signal having improved sharpness. The method of claim 5, The weight controller adjusts the weight through the following equation. (Mathematical formula)

Where w (m, n) is the weight at the (m, n) pixel position, σ ² _l (m, n) is the local variance, and σ ² _g is the total variance. Receiving the R (red), G (green), and B (blue) color signals and converting the Y signals of the luminance signal and the Cb and Cr signals of the color information to output them; Extracting a local variance of the output Y signal and a global variance of a corresponding frame; Improving sharpness of the output Y signal by using the extracted local variance and total variance; And And receiving the Y signal and the Cb and Cr signals having improved clarity, and converting the signals into R (red), G (green), and B (blue) color signals. The method of claim 7, wherein The step of improving the sharpness, Outputting a high frequency signal by filtering the Y signal; Adjusting a weight to be applied to a corresponding image using local variance and total variance of the Y signal; And And improving the clarity of the Y signal by receiving the output high frequency signal and the adjusted weight value.

Images