JPH0884342A - Video signal decoder - Google Patents

Abstract

PURPOSE: To reduce coded noise by giving a decoded image to a characteristic variable low-pass filter whose filter characteristic is decided based on the quantity of a quantization scale code. CONSTITUTION: A filter characteristic decision circuit 9a controls the pass band of a characteristic variable low-pass filter 8 to be narrow when the quantization scale code 109 is high and wide when the quantization scale code 109 is small. Thus, the coded noise of a decoded image 106 outputted from a decode image generating circuit 7 is reduced through the characteristic variable low-pass filter 8. A coefficient of a multiplier of the characteristic variable low-pass filter 8 is revised based on a filter characteristic designation signal 112 from the filter characteristic decision circuit 9a. Thus, the characteristic of the low- pass filter is changed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video signal decoding device, and more particularly to a video signal decoding device having a low pass filter as a constituent element for the purpose of reducing coding noise.

[0002]

2. Description of the Related Art FIG. 9 shows, for example, ISO / IEC 13818-2 Draft.
It is a block circuit diagram which shows the conventional video signal decoding apparatus shown by International Standard. In the figure, encoded data 101 is input from an input terminal 1, a variable length decoding circuit 3, an inverse scan circuit 4, an inverse quantization circuit 5, and an inverse D.
Output terminal 2 via CT circuit 6 and decoded image generation circuit 7
The decoded image 106 is output to the output terminal 2. Also,
The variable length decoding circuit 3 supplies control signals 108, 109, 110 and 111 to the inverse scan circuit 4, the inverse quantization circuit 5, the inverse DCT circuit 6 and the decoded image generation circuit 7, respectively.

Next, the operation will be described. Input terminal 1
The encoded data 101 input from is variable-length decoded in the variable-length decoding circuit 3. At this time, the DCT coefficient 102 and the control signal 10 necessary for reverse scanning
8. Control signal 109 necessary for inverse quantization, inverse DC
The control signal 110 necessary to generate the reproduced image and the control signal 111 necessary to generate the reproduced image are decoded. The DCT coefficient 102 decoded by the variable length decoding circuit 3 is inversely scanned by the inverse scanning circuit 4, inversely quantized by the inverse quantization circuit 5, and inversely DCT by the inverse DCT circuit 6 to obtain a decoded image. The generation circuit 7 generates the decoded image 106 and outputs it to the output terminal 2.

ISO / IEC 13818-2 shown in the conventional example is an international standard for moving picture coding, in which a video bit stream and its decoding method are specified. At this time, if the bit streams are the same, the decoding method is specified, so that the same decoded image can be obtained even if the decoding is performed by different decoding devices. That is, the quality of the decoded image depends on the bitstream.

By the way, generally, in a high-efficiency coding system such as ISO / IEC 13818-2, the image quality of the decoded image depends on the coding rate and the property of the coded image, and it depends on each frame or locally. It differs greatly for each image area. For example, in a system with a low coding rate, coding noise called block distortion occurs when an image causes a scene change.

At this time, in the video signal decoding apparatus equipped with only the decoding circuit defined by ISO / IEC 13818-2,
In a system having a low coding rate as described above, when a scene change occurs, coding noise such as block distortion is reproduced as it is.

[0007]

[Problems to be Solved by the Invention] In this way, ISO / IEC
In a conventional video signal decoding device equipped only with the decoding circuit specified in 13818-2, the quality of the decoded image depends on the coding rate and the nature of the coded image, and therefore the image quality deterioration due to coding noise occurs. There was a problem that it was noticeable.

The present invention has been made to solve the above problems, and an object thereof is to obtain a video signal decoding apparatus capable of reducing coding noise.

[0009]

A video signal decoding apparatus according to the present invention comprises means for determining the characteristic of a variable characteristic low-pass filter through which a decoded image is passed, based on the size of a quantization scale code. Is.

Further, there is provided means for determining the characteristic of the characteristic variable low-pass filter based on the magnitude of the motion vector.

Further, there is provided means for determining the characteristic of the characteristic variable low-pass filter based on the region where the DCT coefficient exists.

Further, there is provided means for calculating the activity of the decoded image and determining the characteristics of the variable-characteristic low-pass filter based on the activity.

Further, there is provided means for detecting the scene change of the decoded image and determining the characteristic of the variable characteristic low-pass filter based on this.

Further, there is provided means for determining the filter characteristic of the characteristic variable low-pass for each prescribed region.

The characteristic variable low-pass filter characteristic is based on at least two of the size of the quantization scale code, the size of the motion vector, the region where the DCT coefficient exists, the activity of the decoded image, and the scene change. It is equipped with means to determine.

Further, a means for determining the filter characteristic of the characteristic variable low-pass for each pixel is provided.

[0017]

The video signal decoding apparatus according to the present invention removes coding noise by passing the decoded image through a variable characteristic low-pass filter whose filter characteristic is determined based on the size of the quantization scale code. There is.

Coding noise is removed by passing the decoded image through a variable characteristic low-pass filter whose filter characteristic is determined based on the magnitude of the motion vector.

Coding noise is removed by passing the decoded image through a variable characteristic low-pass filter whose filter characteristic is determined based on the area where the DCT coefficient exists.

Coding noise is removed by calculating the activity of the decoded image and passing the decoded image through a variable characteristic low-pass filter whose filter characteristic is determined based on the activity.

Further, the coding noise is removed by detecting a scene change of the decoded image and passing the decoded image through a variable characteristic low-pass filter whose filter characteristic is determined based on the detected scene change.

Coding noise is removed by passing the decoded image through a variable-characteristic low-pass filter whose filter characteristic is determined based on two or more parameters.

Coding noise is removed by passing the decoded image through a variable-characteristic low-pass filter in which a filter characteristic is defined for each defined region of the decoded image.

Coding noise is removed by passing the decoded image through a variable characteristic low-pass filter having a filter characteristic defined for each pixel of the decoded image. Then, the reproduced image is obtained.

[0025]

【Example】

Example 1. The first embodiment of the present invention will be described below. FIG. 1 is a block circuit diagram showing a video signal decoding device according to the first embodiment. In the figure, the same reference numerals as those in FIG. 9 indicate the same or corresponding portions, and the third output 109 of the variable length decoding circuit 3 is the filter characteristic determination circuit 9a.
, And the output 112 of the filter characteristic determining circuit 9a is supplied to the second input of the characteristic variable low-pass filter 8. Output 107 of variable characteristic low pass filter 8
Is output as a reproduced image from the output terminal 2.

FIG. 2 is a block circuit diagram showing an example of the characteristic variable low pass filter of the first embodiment. In the figure, the decoded image 106 input from the input terminal 1b is 1
It is provided to the first inputs of the line delay circuit 10a and the multiplier 12a. Output 202 of 1-line delay circuit 10a
Is the first of the 1-line delay circuit 10b and the multiplier 12b.
Given to the input of. Output 2 of 1-line delay circuit 10b
03 is given to the 1st line delay circuit 10c and the 1st input of the multiplier 12c. The output 204 of the 1-line delay circuit 10c is the 1-line delay circuit 10d and the multiplier 12
applied to the first input of d. 1 line delay circuit 10d
Output 205 is provided to the first input of multiplier 12e. The output 206 of the multiplier 12a is given to the first input of the adder 13a, and the output 207 of the multiplier 12b is the adder 1a.
3a is supplied to the second input, and the output 20 of the multiplier 12c
8 is given to the third input of the adder 13a, and the multiplier 12
The output 209 of b is provided to the fourth input of the adder 13a, and the output 210 of the multiplier 12e is provided to the fifth input of the adder 13a.

The output 211 of the adder 13a is given to the 1-sample delay circuit 11a and the first input of the multiplier 12f. The output 212 of the 1-sample delay circuit 11a is 1
It is given to the first inputs of the sample delay circuit 11b and the multiplier 12g. Output 21 of 1-sample delay circuit 11b
3 is provided to the 1-sample delay circuit 11c and the first input of the multiplier 12h. The output 214 of the 1-sample delay circuit 11c is given to the 1-sample delay circuit 11d and the first input of the multiplier 12i. The output 215 of the 1-sample delay circuit 11d is provided to the first input of the multiplier 12j. The output 107 of the adder 13b is output from the output terminal 2 as a reproduced image.

On the other hand, the filter characteristic designating signal 112 inputted from the input terminal 1c is given to the multiplication coefficient generating circuit 14. The first output 223 of the multiplication coefficient generation circuit 14 is
The second output 224 of the multiplication coefficient generating circuit 14 provided to the second input of the multiplier 12a is provided to the second input of the multiplier 12b, and the third output 22 of the multiplication coefficient generating circuit 14 is provided.
5 is provided to the second input of the multiplier 12c, the fourth output 226 of the multiplication coefficient generation circuit 14 is provided to the second input of the multiplier 12d, and the fifth output of the multiplication coefficient generation circuit 14 is provided. 227 is applied to the second input of the multiplier 12e, and the sixth output 228 of the multiplication coefficient generation circuit 14 is applied to the multiplier 12e.
The seventh output 229 of the multiplication coefficient generating circuit 14 is applied to the second input of f, and the eighth output 230 of the multiplication coefficient generating circuit 14 is applied to the second input of the multiplier 12g. 12h of the second input, multiplication coefficient generation circuit 1
The ninth output 231 of No. 4 is given to the second input of the multiplier 12i, and the tenth output 232 of the multiplication coefficient generation circuit 14 is provided.
Is applied to the second input of the multiplier 12j.

The operation of the first embodiment will be described below.
As already described in the conventional example, in the high-efficiency coding method such as ISO / IEC 13818-2, the image quality of the decoded image depends on the coding rate and the property of the coded image, and It greatly differs for each local image area. For example,
In a system with a low coding rate, coding noise called block distortion occurs when an image causes a scene change.

Generally, as a method of reducing such coding noise, a method of passing a decoded image through a low-pass filter can be considered. However, the low pass filtering is
There is an adverse effect that the resolution of the reproduced image deteriorates.
Therefore, it is necessary to reduce coding noise by adaptively changing the pass band of the low-pass filter.

In the first embodiment, the pass band of the low pass filter is changed based on the quantization scale code.
The quantization scale code means an interval between quantization levels. Generally, when the quantization scale code is large, the generated code amount decreases, but the coding noise tends to become remarkable. For example, when a scene change occurs in an image, the generated code amount sharply increases, and the quantization scale code becomes large in order to suppress the increase of the generated code amount. As a result, the coding noise will increase. In the first embodiment, since only the decoded image having such remarkable coding noise is passed through the low-pass filter, the coding noise can be reduced without causing a significant deterioration in the resolution of the reproduced image.

In the first embodiment, the variable length decoding circuit 3
On the basis of the quantized scale code 109 decoded in, the filter characteristic determination circuit 9a narrows the pass band of the characteristic variable low-pass filter when the quantized scale code 109 is large, and the quantized scale code 109 If it is small, control is performed to widen the pass band. In this way, the decoded image 106 output from the decoded image generation circuit 7 passes through the characteristic variable low-pass filter 8 to reduce coding noise.

Variable characteristic low-pass filter 8 shown in FIG.
The configuration is a two-dimensional filter configuration in which a vertical low-pass filter and a horizontal low-pass filter on the screen are vertically connected. In the characteristic variable low pass filter 8, the coefficients of the multipliers 12a to 12j are changed based on the filter characteristic designation signal 112 from the filter characteristic determination circuit 9a. This changes the characteristics of the low pass filter. Of course, it is possible to use the variable characteristic low-pass filter 8 only as a horizontal low-pass filter or as a vertical low-pass filter.

FIG. 3 is a diagram for explaining the operation of the filter characteristic determining circuit 9a. In the characteristic variable low-pass filter 8, for example, several kinds of filter characteristics having different pass bands as shown in FIG. 3A are switched (in FIG. 3, a filter having a one-dimensional configuration is used for simplification of description. However, the same idea can be applied to a filter having a two-dimensional structure). At this time, the filter characteristic determination circuit 9a causes the filter characteristic designation signal 11 to instruct which of the above several types of filter characteristics should be used.
2 is given to the characteristic variable low pass filter 8.

Further, FIG. 3 (b) shows a screen, and the filter characteristic designating signals for the line shown by the broken line are shown on this screen in FIG. 3 (c) and FIG. 3 (d). Is.

For example, in ISO / IEC 13818-2, the quantization scale code is defined for each defined image area. Therefore, the low-pass filter 8 is provided for each defined area.
The pass band of can be changed. This is shown in FIG.

On the other hand, when the characteristics of the pass bands in two adjacent image areas are extremely different, discontinuity at the boundary of the areas may be noticeable. Therefore, the pass band of the low-pass filter 8 is changed for each pixel, not for each defined image region. This is shown in FIG. As shown in the figure, the pass band is changed gently. Such a gradual change can be realized, for example, by passing the filter characteristic designation signal 112 through a low pass filter.

In the first embodiment, the video signal decoding device has been described, but the present invention is not limited to this.
It goes without saying that the present invention can also be applied to a video signal decoding system based on the same concept.

Embodiment 2 FIG. The second embodiment of the present invention will be described below. FIG. 4 is a block circuit diagram showing a video signal decoding apparatus according to the second embodiment. The same reference numerals as those in FIG. 1 denote the same or corresponding portions, and the fifth output 111 of the variable length decoding circuit 3 is It is also given to the input of the filter characteristic determination circuit 9b and output 11 of the filter characteristic determination circuit 9b.
2 is provided to the second input of the characteristic variable low pass filter 8.

In the second embodiment, the band of the characteristic variable low pass filter 8 is changed based on the magnitude of the motion vector. The motion vector means the speed of local motion of the image. Generally, since the human visual resolution for a fast-moving image is deteriorated, it is possible to reduce coding noise without causing a large deterioration in the resolution of a reproduced image.

In the second embodiment, on the basis of the size of the motion vector 111 decoded by the variable length decoding circuit 3, the filter characteristic determining circuit 9b causes the characteristic variable low-pass filter when the motion vector 111 is large. 8 is narrowed, and when the motion vector 111 is small, the pass band is controlled to be wide. In this way, the decoded image 106 output from the decoded image generation circuit 7 passes through the characteristic variable low-pass filter 8 to reduce coding noise.

Generally, a plurality of motion vectors may exist for each local area. In such a case, the maximum value, minimum value, average value, etc. of the plurality of motion vectors can be used as the magnitude of the motion vector.

Further, the vertical motion vector and the horizontal motion vector may be used independently as the magnitude of the above motion vector. That is, when the vertical motion vector is large, the vertical pass band of the characteristic variable low-pass filter 8 is narrowed, and when the vertical motion vector is small, the vertical characteristic of the characteristic variable low-pass filter 8 is small. The pass band in the direction is controlled to be wide. When the horizontal motion vector is large, the horizontal pass band of the characteristic variable low-pass filter 8 is narrowed, and when the horizontal motion vector is small, the horizontal characteristic band of the characteristic variable low-pass filter 8 is small. The pass band in the direction is controlled to be wide.

The pass band may be changed for each defined image area or each pixel.

Although the video signal decoding apparatus has been described in the second embodiment, the present invention is not limited to this.
It goes without saying that the present invention can also be applied to a video signal decoding system based on the same concept.

Example 3. The third embodiment of the present invention will be described below. FIG. 5 is a block circuit diagram showing a video signal decoding apparatus according to the third embodiment. The same reference numerals as in FIG. 1 denote the same or corresponding portions, and the output 104 of the inverse quantization circuit 5 is a filter characteristic determination circuit. The output 112 of the filter characteristic determining circuit 9c is also supplied to the input of 9c, and is supplied to the second input of the characteristic variable low-pass filter 8.

In the third embodiment, the band of the characteristic variable low pass filter 8 is changed on the basis of the existence region of the DCT coefficient. The region where the DCT coefficient is present roughly corresponds to the frequency distribution in the local region of the image. Therefore, it can be assumed that the resolution of the encoded image is low when the area where the DCT coefficient exists is narrow. In the third embodiment, since the low-pass filter processing is performed only on such an image including only low-frequency components, it is possible to reduce coding noise without causing a significant deterioration in the resolution of the reproduced image.

In the third embodiment, D in the inverse quantization circuit 5
The filter characteristic determination circuit 9 is based on the existence region of the CT coefficient.
When c is such that the existence region of the DCT coefficient is narrow, the pass band of the characteristic variable low-pass filter 8 is narrowed, and when the existence region of the DCT coefficient is wide, the pass band is widened. In this way, the decoded image 106 output from the decoded image generation circuit 7 passes through the characteristic variable low-pass filter 8 to reduce coding noise.

The DCT coefficient existence region may be obtained independently in the vertical and horizontal directions. That is, when the existence region of the DCT coefficient in the vertical direction is narrow,
When the vertical pass band of the characteristic variable low-pass filter 8 is narrowed and the vertical DCT coefficient existence region is wide, control is performed to widen the vertical pass band of the characteristic variable low-pass filter 8. To do. When the horizontal DCT coefficient exists in a narrow area, the horizontal pass band of the variable-characteristic low-pass filter 8 is narrowed to obtain a horizontal DCT coefficient.
When the area where the coefficient exists is wide, control is performed so that the horizontal pass band of the variable characteristic low pass filter 8 is widened.

The pass band may be changed for each defined image area or each pixel.

Although the third embodiment has been described with respect to the video signal decoding apparatus, it is needless to say that the present invention is not limited to this and can be applied to a video signal decoding system based on a similar concept.

Example 4. The fourth embodiment of the present invention will be described below. FIG. 6 is a block circuit diagram showing a video signal decoding apparatus according to the fourth embodiment. The same reference numerals as those in FIG. 1 denote the same or corresponding portions, and the output 106 of the decoded image generation circuit 7 is an activity detection circuit. 15, the output 113 of the activity detection circuit 15 is supplied to the input of the filter characteristic determination circuit 9d, and the output 112 of the filter characteristic determination circuit 9d is supplied to the second input of the characteristic variable low pass filter 8.

In the fourth embodiment, the activity of the decoded image 106 is calculated, and the low pass filter 8 is calculated based on the activity.
The band of is changed, and the activity referred to here is an index indicating a local image characteristic. In the decoded image 106, when an area composed of 16 pixels × 16 lines is defined as a defined area, the activity in the defined area can be defined as follows, for example.
In the defined area of 16 pixels × 16 lines in the decoded image, the pixel value of the (i, j) component in the area is re
When c (i, j), act, which is an activity in the specified area, is defined as follows.

[0054]

[Equation 1]

The above activity roughly represents the power of the AC component of the image in the specified region. In general, the coding noise is less noticeable in the area of high activity, and the coding noise is more noticeable in the area of low activity. In the fourth embodiment, since the low-pass filter processing is performed only on such an image in which the coding noise is prominent, the coding noise can be reduced without causing a significant deterioration in the resolution of the reproduced image.

In the fourth embodiment, the activity detection circuit 15 calculates the activity 113 based on the decoded image 106 output from the decoded image generation circuit 7. Further, based on the activity 113, the filter characteristic determination circuit 9d narrows the pass band of the variable characteristic low pass filter 8 when the activity 113 is small, and widens the pass band when the activity 113 is large. Control. In this way, the decoded image 106 output from the decoded image generation circuit 7 passes through the characteristic variable low-pass filter 8 to reduce coding noise.

The activity is an index showing the power of a local image and does not necessarily have to follow the above definition. Further, the activity does not necessarily have to be calculated from the decoded image 106, and the power of the AC component of the image can be easily calculated from the output 104 of the inverse quantization circuit 5, for example.

Further, the pass band may be changed for each defined image area or each pixel.

Although the video signal decoding apparatus has been described in the fourth embodiment, the present invention is not limited to this.
It goes without saying that the present invention can also be applied to a video signal decoding system based on the same concept.

Example 5. The fifth embodiment of the present invention will be described below. FIG. 7 is a block circuit diagram showing a video signal decoding apparatus according to the fifth embodiment. The same reference numerals as in FIG. 1 denote the same or corresponding portions, and the output 106 of the decoded image generation circuit 7 is a scene change detection circuit. 16, the output 114 of the scene change detection circuit 16 is supplied to the input of the filter characteristic determination circuit 9e, and the output 112 of the filter characteristic determination circuit 9e is supplied to the second input of the characteristic variable low-pass filter 8. .

In the fifth embodiment, a scene change is detected from the decoded image, and the band of the characteristic variable low pass filter 8 is changed based on the detected scene change. If the pixel size of the decoded image 106 is 720 pixels × 480 lines,
The scene change can be defined as follows, for example. The pixel value of the (i, j) component in the nth decoded image is rec (n, i, j), and the pixel value of the (i, j) component in the n + 1th decoded image is rec (n + 1, i, j),
The scene representing the amount of scene change is defined as follows.

[0062]

[Equation 2]

Generally, since the visual resolution of human beings deteriorates immediately after a scene change, it is possible to reduce the coding noise without causing a large deterioration in the resolution of the reproduced image.

In the fifth embodiment, the scene change amount 114 is calculated by the scene change detection circuit 16 based on the decoded image 106 output from the decoded image generation circuit 7. Furthermore,
Based on the scene change amount 114, the filter characteristic determination circuit 9e narrows the pass band of the characteristic variable low-pass filter 8 when the scene change amount 114 is large, and narrows the pass band when the scene change amount 114 is small. Control to make it wider. In this way, the decoded image 106 output from the decoded image generation circuit 7 passes through the characteristic variable low-pass filter 8 to reduce coding noise.

The definition of the scene change amount is not limited to the above. The amount of scene change does not have to be calculated from the decoded image 106, and can be easily calculated from the output 104 of the inverse quantization circuit 5, for example.

Although the video signal decoding apparatus has been described in the fifth embodiment, it is needless to say that the present invention is not limited to this and can be applied to a video signal decoding system based on a similar concept.

Example 6. The sixth embodiment of the present invention will be described below. FIG. 8 is a block circuit diagram showing a video signal decoding apparatus according to the sixth embodiment, and the same reference numerals as those in FIGS. 1, 6 and 7 respectively denote the same or corresponding portions.
The third output 109 of the variable length decoding circuit 3 is also given to the first input of the filter characteristic determining circuit 9f, and the fifth output 111 of the variable length decoding circuit 3 is supplied to the first input of the filter characteristic determining circuit 9f. 2 is also given, the output 104 of the inverse quantization circuit 5 is also given to the third input of the filter characteristic decision circuit 9f, and the output 106 of the decoded image generation circuit 7 is given to the activity detection circuit 15 and the scene change. Detection circuit 1
Also given to 6. The output of the activity detection circuit 15 is given to the fourth input of the filter characteristic determination circuit 9f, and the output of the scene change detection circuit 16 is given to the fifth input of the filter characteristic determination circuit 9f. The output 112 of the filter characteristic determining circuit 9f is given to the second input of the variable characteristic low pass filter 8.

In the sixth embodiment, based on the quantization scale code, the magnitude of the motion vector, the area where the DCT coefficient exists, the activity of the decoded image, and the amount of scene change, the conditions of the first to fifth embodiments are combined. The filter characteristic determination circuit 9f is based on the characteristic variable low-pass filter 8
Control the pass band of. The decoded image 106 output from the decoded image generation circuit 7 in this manner passes through the variable characteristic low-pass filter 8 to reduce coding noise.

In the sixth embodiment, the filter characteristic of the characteristic variable low-pass filter is based on the size of the quantization scale code, the size of the motion vector, the area where the DCT coefficient exists, the activity of the composite image, and the scene change. However, the filter characteristic may be determined based on two or more of these.

The pass band may be changed for each defined image area or each pixel.

In the sixth embodiment, the video signal decoding apparatus has been described, but the present invention is not limited to this.
It goes without saying that the present invention can also be applied to a video signal decoding system based on the same concept.

[0072]

According to the present invention, the characteristics of the low pass filter are determined based on the size of the quantization scale code,
By passing the decoded image through this low-pass filter, it is possible to reduce coding noise without visually degrading the resolution.

Further, by determining the low-pass filter characteristic based on the magnitude of the motion vector and passing the decoded image through this low-pass filter, the coding noise can be reduced without any visual deterioration of resolution. .

Further, by determining the low-pass filter characteristic based on the area where the DCT coefficient exists and passing the decoded image through this low-pass filter, the coding noise can be reduced without any visual deterioration of resolution.

Further, the activity of the reproduced image is calculated, the low-pass filter characteristic is determined based on the activity, and the decoded image is passed through this low-pass filter, so that coding noise is eliminated without visual resolution deterioration. Can be reduced.

Further, the scene change of the reproduced image is detected, the low-pass filter characteristic is determined based on the detected scene change, and the decoded image is passed through this low-pass filter to encode without causing visual resolution deterioration. This is effective in obtaining a video signal decoding device that can reduce noise.

Further, the low-pass filter characteristic is determined based on at least two of the size of the quantization scale code of the decoded image, the size of the working vector, the region where the DCT coefficient exists, the activity and the amount of scene change, and the decoding is performed. By passing an image through this low-pass filter, there is an effect of obtaining a video signal decoding device capable of reducing coding noise without causing visual resolution degradation.

Further, by determining the low-pass filter characteristic for each specified region and passing the decoded image through this low-pass filter, it is possible to reduce coding noise without visually degrading the resolution.

Further, by determining the low-pass filter characteristic for each pixel and passing the decoded image through this low-pass filter, it is possible to reduce coding noise without visually degrading the resolution.

[Brief description of drawings]

FIG. 1 is a block circuit diagram of a video signal decoding device according to a first embodiment of the present invention.

FIG. 2 is a block circuit diagram showing a characteristic variable low-pass filter according to the first embodiment.

FIG. 3 is a diagram for explaining the operation of the filter characteristic determination circuit of the first embodiment.

FIG. 4 is a block circuit diagram of a video signal decoding device according to a second embodiment of the present invention.

FIG. 5 is a block circuit diagram of a video signal decoding device according to a third embodiment of the present invention.

FIG. 6 is a block circuit diagram of a video signal decoding device according to a fourth embodiment of the present invention.

FIG. 7 is a block circuit diagram of a video signal decoding device according to a fifth embodiment of the present invention.

FIG. 8 is a block circuit diagram of a video signal decoding device according to a sixth embodiment of the present invention.

FIG. 9 is a block circuit diagram of a conventional video signal decoding device.

[Explanation of symbols]

3 variable length decoding circuit, 4 inverse scan circuit, 5 inverse quantization circuit, 6 inverse DCT circuit, 7 decoded image generation circuit,
8 characteristic variable low pass filter, 9a to 9f filter characteristic determination circuit, 10a to 10d 1 line delay circuit, 11
a to 11d1 sample delay circuit, 12a to 12j multiplier, 13a and 13b adder, 14 multiplication coefficient generating circuit,
15 activity detection circuit, 16 scene change detection circuit, 106 decoded image, 109 quantization scale code, 111 motion vector, 112 filter characteristic designation signal. .

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI Technical indication location H03H 21/00 8842-5J H03M 7/30 A 9382-5K H04N 5/21 B

Claims (8)

[Claims] 1. A means for decoding a high-efficiency coded video signal data stream, a characteristic variable low-pass filter to which this decoded image is input, and a filter characteristic designation based on the size of a quantization scale code of the data. A video signal decoding device, comprising: a filter characteristic determination circuit that outputs a signal and switches the pass band characteristic of the variable low pass filter. 2. A means for decoding a high-efficiency coded video signal data stream, a characteristic variable low-pass filter to which this decoded image is input, and a filter characteristic designation signal based on the magnitude of the motion vector of the data. And a filter characteristic determining circuit for switching the pass band characteristic of the characteristic variable low-pass filter. 3. A means for decoding a high-efficiency coded video signal data stream, a characteristic variable low-pass filter to which this decoded image is input, and a filter characteristic designating signal based on the area where the DCT coefficient of the data exists. A video signal decoding device, comprising: a filter characteristic determination circuit that outputs and switches the pass band characteristic of a variable characteristic low pass filter. 4. A means for decoding a high-efficiency coded video signal data stream, a variable characteristic low-pass filter to which this decoded image is input, an activity detection circuit for detecting the activity of the decoded image, and A video signal decoding apparatus, comprising: a filter characteristic determination circuit that outputs a filter characteristic designation signal based on the detected activity to switch the pass band characteristic of the variable characteristic low pass filter. 5. A means for decoding a high-efficiency coded video signal data stream, a variable characteristic low-pass filter to which this decoded image is input, and a scene change detection circuit for detecting a scene change amount of the decoded image. And a filter characteristic determining circuit that outputs a filter characteristic specifying signal based on the detected scene change amount and switches the pass band characteristic of the characteristic variable low-pass filter. . 6. A means for decoding a high-efficiency coded video signal data stream, a variable characteristic low-pass filter to which this decoded image is input, the size of a quantization scale code of the decoded image, and a working vector. And a filter characteristic determining circuit that outputs a filter characteristic designating signal based on at least two of the DCT coefficient existence area, the activity, and the scene change amount to switch the pass band characteristic of the characteristic variable low-pass filter. A video signal decoding device characterized by being provided. 7. A means for decoding a high-efficiency coded video signal data stream, a variable characteristic low-pass filter to which this decoded image is input, and a filter characteristic designation signal for each defined region of the decoded image. 7. The video signal decoding apparatus according to claim 1, further comprising a filter characteristic determination circuit that switches a pass band characteristic of the characteristic variable low-pass filter. 8. A means for decoding a high-efficiency coded video signal data stream, a characteristic variable low-pass filter to which this decoded image is input, and a filter characteristic designation signal for each pixel of the decoded image. 7. A filter characteristic determination circuit for switching the pass band characteristic of the variable characteristic low pass filter according to claim 1.
The video signal decoding device according to any one of 1.