JPH05110865A - Video information quantity compression and/or information quantity expansion device - Google Patents

Abstract

PURPOSE:To output lots of information even in the presence of noise by detecting a DCT block having a flat part and a contour part from a information quantization compression signal with discrimination information added thereto and using the result as a control signal. CONSTITUTION:A DCT block 1a obtained by dividing picture element data from an input video signal is fed to the 2-dimension DCT transformation device 1 of a main coding means AA and the obtained transformation coefficient block 1b is fed to a quantization device 2. The coding means AA implements quantization based on a quantization step obtained based on a control signal bb and codes and adds information to discriminate the quantization step and outputs an information quantity compression signal 11a. Comparators 16, 13 compare the threshold levels of data forming 2-dimension HPF output blocks 11a and detect whether or not a flat part or a contour part is in existence depending on a prescribed operation. Lots of information is sent and outputted regardless of disturbance by quantization noise by using a signal having at least the flat part or the contour part as the above-mentioned control signal bb.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to video information amount compression and / or
Alternatively, the present invention relates to an information amount expansion device.

[0002]

2. Description of the Related Art FIG. 5 is a block diagram of a conventional still image compression apparatus, and FIG. 6 is a diagram for explaining the operation of a zigzag scanner.

Referring to FIG. 5, JPEG (Joint Photogra) is an organization that promotes international standardization of color still image coding.
The conventional still image compression device proposed by the Phic Expert Group (refer to P158 Vol44 No2 1990, The Institute of Television Engineers of Japan).
First, a DCT block obtained by extracting eight vertical and eight horizontal digital pixel data from one screen is DCT.
Supply to converter 1 (8 × 8DCT), where 2D DC
The 8 × 8 transform coefficient obtained by the T transform is applied to the linear quantizer 2
Supply to. Note that here, DCT is 1 of orthogonal transform coding.
Abbreviation for Discrete Cosine Transform.

Then, the reference quantization matrix supplied from the quantization matrix generator 3 is multiplied by the scaling factor S in the multiplier 4 to obtain a DC matrix obtained by performing linear quantization using the obtained quantization matrix. The coefficient 2a and the AC conversion coefficient 2b are supplied to the one-dimensional predictor 5 and the zigzag scanner 7, respectively.

The one-dimensional predictor 5 compresses the information amount of the DC conversion coefficient 2a by obtaining the difference value between the continuous DC conversion coefficients 2a and performing one-dimensional prediction, and supplies the compressed information to the first Huffman encoder 6. To do. Then, the first Huffman encoder 6 generates a Huffman code that allocates a smaller number of bits in the descending order of occurrence probability of an event, and supplies the Huffman code to one input of the multiplexer 9. The Huffman code is a type of entropy code because it is based on the probability of occurrence of an event. Further, the occurrence probability of the event may be obtained by using the result measured in advance, and it is not necessary to obtain it at the time of encoding.

On the other hand, the zigzag scanner 7 scans the AC conversion coefficient 2b from the low frequency component to the high frequency component as shown in FIG. 6, and supplies it to the second Huffman encoder 8. The value of the second Huffman encoder 8 is "0".
Similarly, a Huffman code is generated based on the run length of the transform coefficient and the value of the transform coefficient that is not "0", and the Huffman code is supplied to the other input of the multiplexer 9.

Then, the multiplexer 9 outputs a multiplexed output signal 9a obtained by multiplexing the respective inputs to a transmission line (not shown).

Since the above-mentioned conventional technique uses the Huffman code, the data contained in the multiplexed output signal 9a varies with respect to a constant amount of input information. Therefore, the scaling factor S is used to control, for example, the data of the multiplexed output signal 9a per field to be held constant.

[0009]

However, in the above-mentioned conventional technique, since the quantization noise described later is generated in the linear quantizer 2, there is a problem that there is a DCT block in which the image quality is significantly deteriorated. Hereinafter, in order to facilitate understanding of the problem, a one-dimensional DCT will be described as an example.

FIG. 7 is a diagram for explaining image quality deterioration due to compression / expansion using DCT due to quantization noise, and FIG. 8 is a diagram for explaining a quantization matrix. Problems to be solved by the present invention will be specifically described below with reference to the drawings.

Here, FIG. 7A having a periodic change.
Consider the one-dimensional digital waveform data string shown in FIG. 7C, which is composed of the one-dimensional digital waveform data string shown in FIG. 2 and the flat part of amplitude 11 and amplitude 0. The inverse DCT transform is defined by the following Equation 1. 7 (A) to (D)
The vertical axis represents the level of the video signal.

[0012]

[Equation 1] Then, the above-mentioned data string is subjected to a one-dimensional DCT transformation defined by the following Equation 2, and only the transform coefficient D (7) is quantized in a coarse quantization step, and then inverse quantization and inverse DCT transformation are performed. The waveform data strings obtained in this way are shown in FIG. 7B and FIG.

[0013]

[Equation 2] Now, here, the portions shown by the dotted lines in FIGS. 6B and 6D are noises based on the quantization noise generated by the quantization in the coarse quantization step, but considering the visual characteristics,
In the waveform shown in FIG. 7B, small noise is added to a place where there is a large change in the video signal level, so that the differential sensitivity (blow-out, “digital signal processing of image”) by Nikkan Kogyo Shimbun
(987) P23) is low, so that noise is difficult to detect, while the waveform shown in FIG. 9D is a noise that appears in a flat part that is easily detected, which causes a problem.

Although the case of one dimension has been described above,
The same applies to the two-dimensional case. The quantization matrix shown in FIG. 8 has a value that increases in the diagonally lower right direction in which the horizontal and vertical frequencies increase. That is, the higher the transform coefficient, the coarser the quantization step in linear quantization. This is because the amount of information is compressed by utilizing the fact that the luminosity is lowered in the portion where the spatial frequency is high and the luminosity is lowered with respect to the waveform in the oblique direction.

However, even in the case of two dimensions, FIG.
A waveform composed of a flat portion and a contour portion as in (C) has a high frequency component, and such a waveform has a two-dimensional DC.
When the transform coefficient obtained by applying the T transform is quantized using the above quantization matrix, quantization noise is mixed in the quantized transform coefficient of higher order (lower right diagonal). Then, when this is inversely quantized and subjected to inverse two-dimensional DCT conversion, noise based on quantization noise is generated, which appears as a noticeable disturbance particularly in a flat portion.

Therefore, the present invention has been made in view of the above-mentioned problems.
A DCT block having a flat portion and a contour portion has a smaller quantization step size than other DCT blocks in the periphery.

[0017]

The present invention provides the following constitution in order to solve the above-mentioned problems.

The input DCT block obtained by dividing the video signal is DCT-transformed, quantized according to the quantization step obtained based on the control signal, coded, and discrimination information for discriminating the quantization step is provided. It is characterized by further comprising encoding means for additionally outputting an information amount compressed signal, and control means for outputting the control signal obtained by detecting the input DCT block having at least a flat portion and a contour portion. Video information amount compression or information amount compression / decompression device.

According to a first aspect of the present invention, there is provided a video information amount expansion device or information amount compression / expansion device which expands the information amount compressed signal by the information amount, wherein inverse quantization is performed based on the discrimination information. Video information amount expansion or information amount compression / expansion device.

A video information amount compression or information amount compression / expansion device according to claim 1, wherein the control means determines whether or not the level of the luminance signal applied to the video signal is within a predetermined range. When detecting the input DCT block having the flat portion and the contour portion, which is provided with the first discriminating means, even a pixel regarded as the flat portion is invalid for a pixel outside the predetermined range. A video information amount compression or information amount compression / expansion device characterized by being configured as follows.

The image information amount compression or information amount compression / decompression device according to claim 1, wherein the control means includes second discrimination means for discriminating the magnitude of noise based on the quantization noise accompanying the quantization. A video information amount compression or information amount compression / decompression device, comprising: control means for outputting the control signal according to the magnitude of noise based on the quantization noise.

[0022]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram for explaining a main part of a video information amount compression apparatus of the first embodiment, and FIG. 2 shows a main part of a video information amount compression apparatus or an information amount expansion apparatus of the first embodiment. The figure which showed the numerical example for explaining, FIG. 3 is the block diagram for demonstrating the principal part of the video information amount expansion apparatus of 1st Example,
FIG. 4 is a block diagram for explaining the main part of the video information amount compression apparatus of the second embodiment. The first embodiment will be described below with reference to the drawings.

(First Embodiment) In the first embodiment, when the amount of information is compressed, the contour portion is neither too bright nor too dark.
The CT block is specified, and the DCT block is quantized by reducing the scaling factor S,
Encode. Then, when the information amount expansion is performed, it is inversely quantized and decoded according to the scaling factor S. Hereinafter, an example of the main part of the video information amount compression device according to the present invention will be described, and then an example of the main part of the video information amount expansion device according to the present invention will be described. (Video Information Amount Compressing Device) In FIG. 1, main parts of the video information amount compressing device are a main encoding means AA for performing an information amount compression on an input video signal, and a control means BB for controlling a scaling factor S. It depends on. Hereinafter, the main encoding means AA will be described in order.

A DCT block 1a obtained by dividing 8 vertical and 8 horizontal digital pixel data from an input video signal is supplied to a DCT block divider (not shown) in the main encoding means AA.
It is supplied to the dimensional DCT transformer 1, and the transform coefficient block 1 b obtained by the two-dimensional DCT transform is supplied to the quantizer 2.
Here, as an example of the DCT block 1a, FIG.
Considering the DCT block shown in FIG. 1, description will be given below based on this numerical example. It should be noted that the numbers representing the digital data in FIG. 9A are quantized by 8 bits,
It can take an integer value of 8 to 127. The input video signal is a luminance (Y) signal and color difference (RY, BY).
Any signal that represents image information such as a signal may be used.
However, the Y signal, the RY signal, the BY signal, and the like are encoded separately.

Then, the multiplier 4 supplies the quantization matrix 4a obtained by multiplying the reference quantization matrix supplied from the quantization matrix generator 3 and the scaling factor S to the quantizer 2, where The quantized transform coefficient 2 c obtained by quantizing the transform coefficient block 1 b with the quantization matrix 4 a is supplied to the first encoder 10. Here, the scaling factor S is supplied from the scaling factor generator 20, and is obtained based on the control signal bb supplied from the control means BB.

Then, the first encoder 10 supplies an encoded signal 10a obtained by subjecting the quantized transform coefficient 2c to well-known encoding to the ID adder 21, where it is obtained based on the control signal bb. The information amount compression signal 21a obtained by adding discrimination information for discriminating the scaling factor S to be output is output to a transmission line (not shown). The first encoder 10 includes, for example, the one-dimensional predictor 5, the first and second Huffman encoders 6 and 8, the zigzag scanner 7 and the multiplexer 9 in the conventional technique described above. It may be However, when the one-dimensional prediction is used for the DC conversion coefficient, the scaling factor is not controlled for the DC conversion coefficient.

The control means BB will be described below.
The two-dimensional HPF output block 11 obtained by applying the DCT block 1a to the three-dimensional HPF 11 and performing spatial HPF
a is supplied to the first absolute value converter 12. Here, an example of the two-dimensional HPF 11 is shown in FIG. The coefficient “8” surrounded by a thick frame in the same figure (B) represents the coefficient of the filter which is multiplied by the target pixel, and the other coefficient “−1” is the filter which is multiplied by the pixels around the target pixel. Represents the coefficient of. That is, the two-dimensional HPF output block 11a is configured by adding nine pieces of data obtained by multiplying the digital data of the target pixel by eight and multiplying the digital data of the peripheral pixels by -1. I have data. Since the target pixel is at the edge of the DCT block, 8
If there are no such pixels, the filter coefficient of the target pixel is made equal to the number of effective peripheral pixels. The two-dimensional HPF output block 11a thus obtained is shown in FIG.

The first absolute value converter 12 takes an absolute value for each data forming the two-dimensional HPF output block 11a, and takes this absolute value as the first and second threshold values, respectively.
The signals are supplied to the second comparators 16 and 13, respectively. The first comparator 16, the first counter 17, and the third comparator 18 are
The second comparator 13, the second counter 14, and the fourth comparator 15 are the main parts that detect whether or not there is a flat portion in the DCT block 1a. It is a main part that detects whether or not there is a contour part. The first discriminating means BB1, which will be described later and is composed of the fifth and sixth comparators 51 and 52, the first NOR circuit 53 and the second AND circuit 54, has a luminance signal level outside the predetermined level range. The pixel corresponding to this is determined, and the pixel corresponding to this is determined as an invalid pixel.

The first comparator 16 compares each data forming the absolute-valued two-dimensional HPF output block 11a with a first threshold value set to a value smaller than the second threshold value, When it is smaller than the first threshold value, "1" is obtained, and in other cases, "0" is obtained for each data, and the low-frequency block 16a is supplied to one input of the second AND circuit 54. Further, a first NOR circuit output signal 53a described later is supplied to the other input of the second AND circuit 54, and when the level of the luminance signal is detected and the pixel is too bright or too dark, "1" in the low-frequency block 16a
Second AND circuit output signal 5 obtained by replacing
4a is supplied to the first counter 17. Pixels having a value of "1" in the low frequency block 16a are pixels having no high frequency component.

Then, in the first counter 17, the second A
A first cumulative addition value 17 obtained by cumulatively adding the number of "1" per low-frequency block 16a of the ND circuit output signal 54a.
a is supplied to the third comparator 18, where the first cumulative addition value 17a is compared with the third threshold value. If the first cumulative addition value 17a is large, "1" is set to the other value. In this case, "0" is supplied to one input of the first AND 19 as the third comparator output signal 18a. In this way, it is determined whether or not the number of pixels having no high frequency component per DCT block exceeds a certain value (third threshold value), and there is a flat portion in each DCT block 1a. It is detected whether or not. Further, the first and second counters 14 and 17 are naturally reset for each DCT block.

On the other hand, the second comparator 13 compares each data of the absolute valued two-dimensional HPF output block 11a with the second threshold value, and when it is larger than the second threshold value, "1". In other cases, the high frequency block 13a which has obtained “0” for each data is supplied to the second counter 14. A pixel having a value of "1" in the high frequency block 13a is a pixel having a high frequency component.

Then, the second counter 14 cumulatively adds the number of "1" s per high-frequency block 13a to obtain a second value.
Is supplied to the fourth comparator 15, where the second cumulative addition value 15a is compared with the fourth threshold value.
When the second cumulative addition value 15a is large, "1" is supplied, and otherwise "0" is supplied to the other input of the first AND19.

In this way, it is determined whether or not the number of pixels having a high frequency component per DCT block exceeds a certain value (fourth threshold value), and the contour portion is formed in each DCT block 1a. It is detected whether or not there is.

Then, the scaling factor generator 20 obtains the control signal bb obtained by specifying the DCT block having the flat portion and the contour portion at the same time by taking the logical product of both inputs in the first AND circuit 19. Supply to. That is, when the control signal bb is "1", the DCT block has the flat portion and the contour portion at the same time, and when the control signal bb is "0", the DCT block does not have the flat portion and the contour portion.

Here, the first discriminating means BB1 composed of the fifth and sixth comparators 51 and 52, the first NOR circuit 53 and the second AND circuit will be described. First, D will be described.
The CT block 1a is supplied to the fifth comparator 51 that determines when the level of the luminance signal is high and the sixth comparator 52 that determines when the level of the luminance signal is low. Then, in the fifth comparator 51, the digital luminance signal data in the DCT block 1a is changed to a sixth
When compared with a fifth threshold value larger than the threshold value of "1", if it exceeds this threshold value, "1" is supplied; otherwise, "0" is supplied to one input of the first NOR circuit 53 for each digital pixel data. To do. Here, the digital pixel data represents a luminance signal or a color difference signal of a certain pixel, and in the case of the luminance signal, the data itself is used, and in the case of the color difference signal, the luminance signal data of the pixel is used. Further, the sixth comparator 52 compares the digital luminance signal data in the DCT block 1a with the sixth threshold value, and if they do not exceed the threshold value, "1" is given; otherwise, "0". Is supplied to the other input of the first NOR circuit 53 for each digital pixel data.
The first NOR circuit 53 obtains a first NOR circuit output signal 53a obtained by taking and inverting the logical sum of both inputs.

The logical product of both inputs is obtained because the pixel having the value "1" in the low-frequency block 16a is a pixel having no high-frequency component, but the first NOR circuit output signal 53a.
Is 0, the level of the luminance signal is too high or too low, so even if it is a flat portion, it is a pixel in which noise is visually suppressed. is there. The low-frequency block 16a and the first N
Of course, the timing with the OR circuit output signal 53a is adjusted by a known timing adjustment circuit. Further, it goes without saying that the first discriminating means need not be provided.

In this way, it is possible to specify a DCT block having a flat portion and a contour portion in which noise due to quantization noise poses a problem, and further, a DCT block having a flat portion and a contour portion.
A control signal b obtained by removing a DCT block in which noise based on quantization noise is inconspicuous even in a block
The scaling factor S can be determined based on b. Then, for example, when the control signal bb is "1", the scaling factor S is "0.5", and when the control signal bb is "0", the scaling factor S is "1". Information amount compression signal 21a with reduced amount
Can be obtained. (Video Information Amount Expanding Device) To describe the video information amount expanding device with reference to FIG. 3, the same components as those in FIG. 1 are designated by the same reference numerals and the description thereof will be omitted. The video information amount expansion device is the first encoder 1 in the video information amount compression device shown in FIG.
0 corresponds to the case where the one-dimensional predictor 5, the first and second Huffman encoders 6 and 8, the zigzag scanner 7 and the multiplexer 9 described in the prior art are used. ..

First, the information amount compression output signal 21a is supplied to the first selector 22, where the discrimination information added by the ID adder 21 and the video information are separated, and the former is I
The latter is supplied to the D decoder 23 and the second selector 24, respectively.

The ID decoder 23 supplies the ID decoder output signal 23a obtained by decoding the discrimination information to the scaling factor generator 20 described later. On the other hand, the second selector 24 outputs the DC conversion coefficient signal 24a and the AC conversion coefficient signal 24b corresponding to the DC conversion coefficient of the entire DCT block described above to the first and second Huffman decoders 2.
Supply to 5 and 26 respectively.

Then, the first Huffman decoder 25 supplies the decoded DC conversion coefficient signal 25a obtained by decoding the DC conversion coefficient signal 24a to the one-dimensional predictive decoder 27. Further, the second Huffman decoder 26 supplies the decoded AC conversion coefficient signal 26a obtained by decoding the AC conversion coefficient signal 24b to the inverse zigzag scanner 28.

This one-dimensional predictive decoder 27 corresponds to the above-mentioned one-dimensional predictor 5, and obtains a DC conversion coefficient and supplies it to one input of the multiplexer 29. The inverse zigzag scanner 28 is for returning the zigzag scanning for each DCT block performed by the zigzag scanner 7 to the original, and is composed of a memory for rearranging data. Then, the restored AC conversion coefficient is supplied to the other input of the multiplexer 29.

Then, the multiplexer 29 supplies the multiplexer output signal 29a composed of 8 × 8 data obtained by multiplexing the DC conversion coefficient and the AC conversion coefficient to the dequantizer 30. The inverse quantizer 30 supplies the inverse DCT converter 32 with an inverse quantized output signal obtained by inversely quantizing the multiplexer output signal 29a using the inverse quantization matrix.
The dequantization matrix is obtained by multiplying the scaling factor S output from the scaling factor generator 20 by the multiplier 4 and the reference dequantization matrix supplied from the dequantization matrix generator 31. is there.

Then, the inverse DCT converter 32 outputs the information amount expansion signal 32a obtained by applying the inverse DCT to the inverse quantized output signal to a transmission line (not shown).

Here, the noise due to the difference in the scaling factor S will be described with reference to FIGS. 2D and 2E.
3A and 3B respectively show the results obtained by DCT transforming, quantizing, and inverse DCT transforming the DCT block in FIG. 9A with scaling factors S = 0.5 and S = 1. Comparing both figures, it is clear that noise is suppressed when S = 0.5.

As described above, in the present embodiment, by reducing the scaling factor S for the DCT block having the flat portion and the contour portion, it is possible to suppress the generation of visually noticeable noise. (Second Embodiment) In the first embodiment, when the transform coefficient block obtained by the two-dimensional DCT transform is quantized, noise generated due to the quantization noise mixed in the quantized transform coefficient is flattened. To suppress, a DCT block having a flat portion and a contour portion is specified, and in such a case, the scaling factor S is reduced to transmit a large amount of information, and in other cases, the scaling factor S is increased to reduce the amount of information. It was something to do.

However, when the quantization noise mixed in the quantized transform coefficient is sufficiently small and the sum of the absolute values thereof is sufficiently small, the noise generated by the quantization noise is also small and does not pose a problem.

Therefore, in the second embodiment, such a case is detected and the scaling factor S is increased to reduce the amount of information even in the DCT block having the flat portion and the contour portion.

The video information amount compression apparatus according to the second embodiment will be described with reference to FIG. 4. The same components as those in FIG. 1 are designated by the same reference numerals and the description thereof will be omitted. Further, the image information amount expansion device has the same configuration as that of the first embodiment, and therefore its explanation is omitted.

In the figure, the difference from the video information amount compression apparatus according to the first embodiment is that the quantization noise mixed in the quantization conversion coefficient in the control means BB is sufficiently small, and the absolute value of its absolute value is small. The point is that the second discriminating means BB2 for discriminating the DCT block 1a having a sufficiently small total sum is provided.

Therefore, to explain the second discriminating means BB2, the transform coefficient block 1b obtained by the two-dimensional DCT transform is supplied to the second quantizer 39. The second quantizer 39 inverses the second quantizer output signal 39a obtained by performing quantization using a quantization matrix obtained by a large scaling factor S (S = 1 in the first embodiment). It is supplied to the quantizer 40.

Then, after the inverse quantizer 40 performs inverse quantization corresponding to the second quantizer 39, the subtractor 4
Quantization noise, which is supplied to one input of 1 and is subtracted from the transform coefficient block 1b supplied to the other input, is supplied to the second absolute value converter 42. In this way, the auxiliary control unit BB1 determines the quantization noise when the large scaling factor S is used.

The absolute value quantization noise 42a obtained by converting the quantization noise into the absolute value by the second absolute value converter 42 is supplied to one input of the adder 43 and the seventh comparator 46. To be done. Here, the adder 43, the register 44, and the ninth comparator 45 determine when the total sum of the absolute values of the quantization noise is sufficiently small, and the seventh comparator 46, the third counter 47, and the third comparator 47 8
The comparator 48 has a role of discriminating when the quantization noise is sufficiently small.

In the adder 43, the absolute value quantization noise 42a for each pixel data and the cumulative addition quantization noise 44a supplied from the register 44 to the other input are added in the register 44. Then, the cumulative addition quantization noise 4
4a is supplied to the ninth comparator 45, where the cumulative addition quantization noise 44a is compared with the ninth threshold value for each DCT block, and if it exceeds this value, "1" is given; otherwise, "0" is given. Is used as the ninth comparator output signal 45a for the first OR
Supply to one input of circuit 49. Then register 4
Of course, 4 is cleared for each DCT block.

On the other hand, in the seventh comparator 46, the absolute value quantization noise 42a is compared with the seventh threshold value for each pixel data, and if it exceeds this value, "1" is given, otherwise. "0" is supplied to the third counter 47, where the number of "1" is accumulated, and the result is compared with the eighth threshold value for each DCT block by the eighth comparator 48, and this is calculated. If it exceeds, "1" is supplied, and otherwise, "0" is supplied to the other input of the first OR circuit 49 as the eighth comparator output signal 48a.

Then, the first OR circuit output signal 49a obtained by ORing the eighth and ninth comparator output signals 48a and 45a in the first OR circuit 49 is converted into the first AND output signal 19a. The control signal bb is supplied to the other input of the third AND circuit 50, which is supplied to one input, and the logical product of the two is supplied to the scaling factor generator 20 and the ID adder 21.

In the above-described embodiment, the output of one two-dimensional HPF 11 is taken as an absolute value and compared with the first and second threshold values, but two two-dimensional HPFs having different characteristics are used.
Are provided, the absolute values of those outputs are taken, and the first and second values are obtained, respectively.
Of course, the threshold value may be compared with the threshold value.

In the above embodiment, the reference quantization matrix is multiplied by the scaling factor S to vary the quantization step. However, plural kinds of quantization matrices are prepared and one of them is selected. May be. The same applies to inverse quantization. For example, one of a plurality of kinds of quantization matrices prepared is shown in FIG. 8, and the other one has a flat portion and a contour portion as a matrix in which the value at the lower right of the quantization matrix is smaller. In the case of a DCT block, the latter may be selected.

Although the video information amount compression device and the video information amount decompression device have been individually described in the above-mentioned embodiment,
Of course, an information amount compression / decompression device that integrates both may be used.

[0059]

As described above, according to the present invention, the input D
Output a video information amount compressed signal obtained by performing a DCT transform on a CT block and performing quantization according to a quantization step obtained based on a control signal obtained by detecting an input DCT block having at least a flat portion and a contour portion. Since it is possible to provide a video information amount compression or information amount compression / expansion device that performs the above, it is possible to transmit and output a large amount of information for an input DCT block in which noise based on quantization noise associated with quantization appears as a significant disturbance. There is an effect.

Further, since it is possible to provide a video information amount expansion or information amount compression / expansion device which performs inverse quantization based on the discrimination information for discriminating the quantization step, it is possible to quantize when the above video information amount compressed signal is decoded. There is an effect that noise based on digitized noise can be suppressed.

Further, the control means is provided with a first discriminating means for discriminating whether or not the level of the luminance signal applied to the video signal is within a predetermined range, and the input DCT having the flat portion and the contour portion. It is possible to provide a video information amount compression or information amount compression / decompression device characterized in that when detecting a block, pixels outside the above-mentioned predetermined range are invalidated. When the noise is not visually noticeable, there is an effect that the information amount of the video information amount compression signal can be reduced.

Further, the control means is provided with a second discriminating means for discriminating the magnitude of the noise based on the quantization noise associated with the quantization, and the control is performed according to the magnitude of the noise based on the quantization noise. Since it is possible to provide the information amount compression or information amount compression / expansion device configured to output a signal, it is possible to reduce the information amount of the video information amount compression signal when the noise based on the quantization noise is small. ..

[Brief description of drawings]

FIG. 1 is a block diagram illustrating a main part of a video information amount compression device according to a first embodiment.

FIG. 2 is a diagram showing a numerical example for explaining a main part of a video information amount compression device or an information amount decompression device of the first embodiment.

FIG. 3 is a block diagram for explaining the main part of the video information amount expansion device of the first embodiment.

FIG. 4 is a block diagram illustrating a main part of a video information amount compression device according to a second embodiment.

FIG. 5 is a block diagram of a conventional still image compression apparatus.

FIG. 6 is a diagram for explaining the operation of the zigzag scanner.

FIG. 7 is a diagram for explaining image quality deterioration due to compression / expansion using DCT due to quantization noise.

FIG. 8 is a diagram for explaining a quantization matrix.

[Explanation of symbols]

 1a DCT block 21a Information amount compression signal bb Control signal AA Encoding means BB Control means BB1 First discriminating means BB2 Second discriminating means

Claims (4)

[Claims] 1. A discrimination for discriminating the quantizing step by subjecting an input DCT block obtained by dividing a video signal to a DCT transform, quantizing in accordance with a quantizing step obtained on the basis of a control signal, and coding. It has an encoding means for adding information and outputting an information amount compression signal, and a control means for outputting the control signal obtained by detecting the input DCT block having at least a flat portion and a contour portion. A video information amount compression or information amount compression / expansion device. 2. A video information amount expansion or information amount compression / expansion device for expanding the information amount compressed signal according to claim 1, wherein the information amount expansion signal is inversely quantized based on the discrimination information. Video information amount expansion or information amount compression / expansion device. 3. The image information amount compression or information amount compression / expansion device according to claim 1, wherein the control means determines whether or not the level of the luminance signal applied to the image signal is within a predetermined range. When detecting the input DCT block having the flat portion and the contour portion, the first discriminating means for discriminating is applied to a pixel which is regarded as the flat portion but is out of the predetermined range. The video information amount compression or information amount compression / decompression device characterized in that is configured to be invalid. 4. The image information amount compression or information amount compression / expansion device according to claim 1, wherein the control means determines the noise magnitude based on the quantization noise associated with the quantization. Video information amount compression or information amount compression / expansion device, further comprising: means for outputting the control signal according to the magnitude of noise based on the quantization noise.