JPH10155154A - Moving image coder and moving image decoder - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize efficient coding. SOLUTION: A DC component coefficient is given to a subtractor 16 through a terminal 14. A memory 18 has already stored a DC component coefficient of one preceding block. In the case that information in an N-th frame original image from a DC component prediction on/off decision section 7 indicates '1', since a terminal 19 of a switch 8 is thrown to a terminal 20, a difference obtained by the subtractor 16 between the DC component coefficient of the processing block and the DC component coefficient of one preceding block is fed to an output terminal 31 through terminals 19, 35. On the other hand, in the case that information in an N-th frame original image from the DC component prediction on/off decision section 7 indicates '0', since the terminal 19 of the switch 8 is thrown to a terminal 21, the DC component coefficient of the processing block itself is fed to the output terminal 31.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The present invention relates to an inter-frame difference signal between an encoding target frame of a moving image and a reference frame.
The present invention relates to a moving image encoding device that encodes for each predetermined block and a moving image decoding device that decodes encoded data encoded by the moving image encoding device.

[0002]

2. Description of the Related Art An apparatus for encoding / decoding a moving image includes an apparatus for encoding / decoding an inter-frame difference signal between a frame to be encoded / decoded and a reference frame, and an apparatus for encoding / decoding a frame to be encoded / decoded. There is an apparatus that performs encoding / decoding using only a signal.

In an apparatus that performs encoding / decoding using only a signal of a frame to be encoded / decoded, the signal itself contains a lot of redundancy. In many cases, a component value is predicted from a DC component value in a block of a neighboring block that has already been encoded / decoded, and the prediction error is used for encoding / decoding.

On the other hand, in an apparatus for encoding / decoding an inter-frame difference signal between a frame to be encoded / decoded and a reference frame, the difference signal already contains little redundancy. Therefore, if the DC component value of the inter-frame difference signal is predicted for each predetermined block from a block that has already been encoded / decoded, the coding efficiency often decreases instead. Therefore, the apparatus that encodes / decodes the inter-frame difference signal between the encoding / decoding target frame and the reference frame has not performed this processing.

[0005]

However, an image in which flicker occurs or a strobe illuminates between an image frame to be encoded / decoded and a reference image frame, or an entire screen such as a fade-in or fade-out image, is used. Even if a method of encoding / decoding an inter-frame difference signal of a video in which a change has occurred is used, the inter-frame difference signal has redundancy included in the signal of the encoding / decoding target frame itself. A lot of redundancy of almost the same nature as above is still included. Therefore, unless the intra-block DC component prediction processing is performed on the inter-frame difference signal,
The coding efficiency is greatly reduced. On the other hand,
As described above, if the intra-block DC component value prediction processing is performed on an inter-frame difference signal for a video in which a substantially uniform luminance change does not occur globally, the coding efficiency is reduced.

It is an object of the present invention to provide a moving picture coding apparatus and a moving picture decoding apparatus with high coding efficiency.

[0007]

A moving picture coding apparatus according to the present invention converts a DC component value of an inter-frame difference signal in a block to be coded in a frame to be coded into a frame which has already been coded in the frame to be coded. An inter-frame difference signal DC component prediction means for predicting the inter-frame difference signal DC component value of the block and outputting the prediction error value, and a prediction error value output from the inter-frame difference signal DC component prediction means for encoding. Or whether to use the inter-frame difference signal DC component value in the encoding target block for encoding is determined for each defined area, and the inter-frame difference signal DC component prediction that outputs the determination information is determined. The decision information output from the on / off determination means and the inter-frame difference signal DC component prediction on / off If was component prediction on ",
The prediction error value output from the inter-frame difference signal DC component prediction means is used for encoding, and the decision information output from the inter-frame difference signal DC component prediction on / off determination means is “inter-frame difference signal”. Signal DC component prediction off "
In the case of, there is provided an inter-frame difference signal DC component prediction on / off switching means for using the inter-frame difference signal DC component value in the encoding target block for encoding.

Further, the moving picture decoding apparatus of the present invention provides an inter-frame difference signal DC component prediction on / off input from an input terminal.
When the OFF determination information is “inter-frame difference signal DC component prediction ON”, the already decoded inter-frame difference signal DC component value is added to the inter-frame difference signal DC component prediction error value separately input from the input terminal. The added value is used for decoding, and if the inter-frame difference signal DC component prediction on / off determination information input from the input terminal is “inter-frame difference signal DC component prediction off”, the input is separately input from the input terminal. And an inter-frame difference signal DC component prediction on / off switching means used for decoding the inter-frame difference signal DC component value itself.

As described above, the inter-frame difference signal DC component prediction processing is performed on a frame in which a luminance change occurs in the entire screen, and the inter-frame difference signal DC component prediction processing is not performed on other frames. Thus, efficient encoding is possible in both cases.

[0010]

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of a moving picture coding apparatus according to a first embodiment of the present invention, FIG. 2 is a view showing values of discrete cosine transform coefficients before quantization, and FIG. 3 is a view after quantization. FIG. 6 is a diagram illustrating values of discrete cosine transform coefficients.

The moving picture coding apparatus according to the first embodiment comprises an input terminal 1, frame memories 2 and 3, a motion detecting section 4, a motion compensating section 5, an intra-block difference average calculating section 6, a DC component prediction ON / OFF OFF determining unit 7, switch 8, discrete cosine transform unit 9, quantizing unit 10, switches 11, 15 and subtractor 16
, An output terminal 17, a memory 18, a switch 22, an inverse quantization unit 26, an inverse discrete cosine transform unit 27, an adder 28, a frame memory 29, a subtractor 30, output terminals 31, 32, a frame memory 36, and a motion compensation unit 37. It is composed of

An N-th frame original image is input from an input terminal 1 to a frame memory 2, a frame memory 3, a motion detecting section 4, and an intra-block difference average calculating section 6. In the frame memory 3, the N-th frame original image is stored, and the (N-1) -th frame original image stored so far is stored in the motion detector 4,
It is sent to the motion compensator 5. The motion detector 4 calculates a motion vector between the Nth frame original image and the (N-1) th frame original image for each macroblock of 16 × 16 pixels by a block matching method that has been widely used in the past. I do. The motion vector for each macroblock is calculated by the motion compensator 5.
, And stored in the frame memory 36 for one frame, and then input to the motion compensator 37 and the output terminal 17. The motion compensator 5 calculates a pixel shifted by the motion vector for each macroblock input from the motion detector 4 with respect to the (N-1) -th original image input from the frame memory 3, Send to 6. In the intra-block difference average value calculation unit 6, for each block of 8 × 8 pixels,
An intra-block average value of a difference between the N-th frame original image from the input terminal 1 and the motion-compensated (N-1) -th frame original image from the motion compensation unit 5 is calculated. The calculated difference average value for each block is input to the DC component prediction on / off determination unit 7. In the DC component prediction on / off determining unit 7, the average value M of the difference average value for each block over one frame is calculated.
And the variance V are calculated. When the absolute value of the average value M is equal to or larger than the predetermined threshold T1 and the variance V is equal to or smaller than the predetermined threshold T2, the DC component prediction is turned on for the Nth frame original image. Is determined, and data “1” is transmitted to the switch 8 as the information, and is transmitted from the output terminal 32 to the outside. In other cases, the DC component prediction is determined to be OFF for the N-th frame original image, and data "0" is transmitted to the switch 8 as the information and the data is transmitted from the output terminal 32 to the outside. . The above processing is performed while the N-th original image is stored in the frame memory 2.

Thereafter, the (N + 1) th original image is input to the input terminal 1
While the above processing is being performed on the (N + 1) th frame original image, the Nth frame original image is output from the frame memory 2 and input to the subtractor 30.
The (N-1) th frame reproduced image has already been stored in the frame memory 29, and the motion compensator 37 calculates, for each macroblock, a pixel at a position shifted by the motion vector input from the frame memory 36, The data is sent to the adder 28. The subtracter 30 calculates the difference between the frames of the two images, and inputs the difference signal between the frames to the discrete cosine transform unit 9. The discrete cosine transform unit 9 subjects the inter-frame difference signal to discrete cosine transform for each block of 8 × 8 pixels. Now, it is assumed that the discrete cosine transform coefficients of a certain block are obtained as shown in FIG. Here, one coefficient at the upper left corner represents a DC component in the block, and the other 63 coefficients are equivalent to representing AC components in the block. Each of these coefficients is quantized by the quantization unit 10, that is, the coefficient is divided by a quantization step value and a fraction is rounded down. For example, assuming the quantization step 10, each coefficient in FIG. 2 becomes as shown in FIG. 3 after quantization.

When each quantized coefficient is output from the quantization unit 10, if the coefficient is 63 coefficients representing an AC component (hereinafter, referred to as an AC component coefficient), the switch 11 Terminal 12 is connected to terminal 13 and switch 22
The terminal 23 of the switch 15 is connected to the terminal 24, and the terminal 33 of the switch 15 is connected to the terminal 34. Therefore, the AC component coefficient is directly input to the inverse quantization unit 26 and sent out from the output terminal 31.

On the other hand, when the coefficient is the remaining one coefficient representing a DC component (hereinafter referred to as a DC component coefficient), the terminal 12 of the switch 11 is connected to the terminal 14;
The terminal 23 of the switch 22 is connected to the terminal 25, and the terminal 33 of the switch 15 is connected to the terminal 35. Therefore, the DC component coefficient is directly input to the inverse quantization unit 26. The DC component coefficient is subtracted by the subtractor 16 and the memory 1
8. Input to the terminal 21 of the switch 8. At the same time, in the memory 18, the DC component coefficients of one previous block already stored are sent to the subtractor 16. The subtractor 16 calculates the difference between the DC component coefficient of the current block and the DC component coefficient of the immediately preceding block, and sends the DC component coefficient prediction error value to the terminal 20 of the switch 8. If the information data is “1” from the DC component prediction on / off determining unit 7 in the N-th frame original image, the terminal 19 of the switch 8 is connected to the terminal 20, so that the DC component coefficient prediction error value 19, 35, and 33 are sequentially sent to the output terminal 31. On the other hand, when the information from the DC component prediction on / off determining unit 7 in the N-th frame original image is data “0”, the terminal 19 of the switch 8 is connected to the terminal 21, so that the DC component coefficient itself is The values are terminals 19, 35, 33
Are sequentially transmitted to the output terminal 31.

The inverse quantization unit 26 inversely quantizes the DC component coefficients and the 63 AC component coefficients of the processing block, performs inverse discrete cosine transform in the inverse discrete cosine transform unit 27, and outputs the (N− 1) An N-th frame reproduced image is added to the frame reproduced image, and is newly stored in the frame memory 29. The N-th frame reproduced image stored in the frame memory 29 is used as a reference image when encoding the (N + 1) image.

In the moving picture coding apparatus according to the first embodiment, the DC component prediction on / off determining unit 7 uses the average value M and the variance V of the difference average value of each block for one entire frame. However, the present invention is not limited to these statistics, and for example, an autocorrelation coefficient which is also widely used in signal processing technology may be used. That is,
The DC component prediction on / off determining unit 7 calculates an autocorrelation coefficient for the entire frame of the difference average value for each block, and when the autocorrelation coefficient is equal to or larger than a predetermined threshold T3, It may be determined that the DC component prediction is ON for the N-frame original image, and that the DC component prediction is OFF for the N-th frame original image if the DC component prediction is less than the threshold value T3. It is possible.

In the moving picture coding apparatus according to the first embodiment, when a change in brightness occurs on the entire screen by adjusting a fade-in / fade-out or a diaphragm of a camera which is frequently used at the time of video editing, a direct-current (DC) The change in brightness that has occurred in the entire screen can be compensated for using information as small as one bit per frame obtained from the component prediction on / off determining unit 7.

FIG. 4 is a block diagram showing a moving picture coding apparatus according to a second embodiment of the present invention.

The moving picture coding apparatus according to the second embodiment comprises an input terminal 51, a subtractor 52, a discrete cosine transform section 53, a quantizing section 54, switches 55, 59, 63, and a DC component prediction on / off determining section. 67, a subtractor 68, a memory 69, a switch 70, an inverse quantization unit 74, and an inverse discrete cosine transform unit 75
, An adder 76, a frame memory 77, a motion detecting section 78, a motion compensating section 79, and output terminals 80, 81, 82.

The N-th frame original image is input from an input terminal 51 to a subtracter 52 and a motion detector 78. The (N-1) th frame reproduced image has already been stored in the frame memory 77, and the motion detecting section 78 has the Nth frame original image from the input terminal 51 and the (N-1) th frame from the frame memory 77.
A motion vector between a frame playback image and a frame reproduction image is calculated as 16 × 1 by a block matching method which has been widely used.
The calculation is performed for each macroblock of 6 pixels. The motion vector for each macroblock is input to the motion compensator 79 and the output terminal 82. In the motion compensating unit 79, the pixel at the position shifted by the motion vector input from the motion detecting unit 78 is transmitted to the subtractor 52 and the adder 76 for each macro block. The subtracter 52 calculates the difference between the frames of the two images, and inputs the difference signal between the frames to the discrete cosine transform unit 53. In the discrete cosine transform unit 53, the 8 ×
A discrete cosine transform is performed for each block of 8 pixels.
The variance cosine coefficients output by this and the process of quantizing those coefficients are the same as in the example described in the moving picture coding apparatus of the first embodiment.

When each quantized coefficient is output from the quantization section 54, if the coefficient is 63 coefficients representing an AC component (hereinafter referred to as an AC component coefficient), the switch 55 Terminal 56 is connected to terminal 57 and switch 63
The terminal 64 of the switch 70 is connected to the terminal 65, and the terminal 71 of the switch 70 is connected to the terminal 72. Therefore, the AC component coefficient is input to the inverse quantization unit 74 as it is and transmitted from the output terminal 81.

On the other hand, when the coefficient is the remaining one coefficient representing a DC component (hereinafter referred to as a DC component coefficient), the terminal 56 of the switch 55 is connected to the terminal 58 and the terminal of the switch 63 is connected. 64 is connected to the terminal 66, and the terminal 71 of the switch 70 is connected to the terminal 73. Therefore,
The DC component coefficient is directly input to the inverse quantization unit 74.
The DC component coefficient is input to a subtractor 68, a memory 69, a terminal 60 of a switch 59, and a DC component prediction on / off determining unit 67. At the same time, the DC component coefficient of the immediately preceding block which has been stored in the memory 69 is sent to the subtractor 68. The subtracter 68 calculates the difference between the DC component coefficient of the processing block and the DC component coefficient of the immediately preceding block.
1. It is sent to the DC component prediction on / off determining unit 67.
In the DC component prediction on / off determining unit 67, the switch 55
, And the DC component coefficient prediction error value from the subtractor 68 are input, and the magnitudes of the respective absolute values are compared. If the absolute value of the DC component coefficient prediction error from the subtractor 68 is smaller, it is determined that the DC component prediction is on for the encoding target block, and data “1” is transmitted to the switch 59 as the information. At the same time, the signal is transmitted from the output terminal 80 to the outside. Conversely, the DC component prediction on / off determination unit 67
If the absolute value of the DC component coefficient from 8 is smaller,
It is determined that the DC component prediction is off for the encoding target block, and the data “0” is used as the information as the switch 59.
, And at the output terminal 80 to the outside. When the information from the DC component prediction on / off determining unit 67 for the block to be encoded is data “1”, the terminal 62 of the switch 59 is connected to the terminal 61, so that the DC component coefficient prediction error It is sent to the output terminal 81 through 62, 66 and 64 in sequence. On the other hand, if the information from the DC component prediction on / off determining unit 67 for the encoding target block is data “0”, the switch 5
Since the terminal 62 of No. 9 is connected to the terminal 60, the value of the DC component coefficient itself is sent to the output terminal 81 through the terminals 62, 66 and 64 in sequence.

The inverse quantization unit 74 inversely quantizes the DC component coefficients and the 63 AC component coefficients of the processing block, performs inverse discrete cosine transform in the inverse discrete cosine transform unit 75, and performs motion compensation in the adder 76. The image is added to the (N-1) th frame reproduced image to form an Nth frame reproduced image, which is newly stored in the frame memory 77. The N-th frame reproduced image stored in the frame memory 77 is the (N +
1) Used as a reference image when encoding a frame image.

In the moving picture coding apparatus according to the second embodiment, the DC component prediction on / off determining unit 67 generates 1-bit information per block, and the moving picture coding apparatus according to the first embodiment. Although more information is available, for example, for an image in which the brightness of a spotlight illuminating a part of the screen changes, DC component prediction is turned on in the part illuminated by the spotlight, and DC Since the component prediction can be turned off, a change in brightness can be compensated for more narrowly than in the moving picture coding device according to the first embodiment.

FIG. 5 is a block diagram of a moving picture decoding apparatus according to one embodiment of the present invention.

The moving picture decoding apparatus according to the present embodiment comprises input terminals 101 and 102, switches 103 and 107, and an adder 11
1, memory 112, switch 113, and frame memory 1
17, an inverse quantization unit 118, and an inverse discrete cosine transform unit 119
, An adder 120, an output terminal 121, an input terminal 122, and a motion compensation unit 123.

In FIG. 5, the data transmitted from the output terminal 32 of the moving picture coding apparatus in FIG. 1 or the data transmitted from the output terminal 80 of the moving picture coding apparatus in FIG. I do. Input terminal 1
From 02, the data transmitted from the output terminal 31 of the moving picture coding apparatus in FIG. 1 or the data transmitted from the output terminal 81 of the moving picture coding apparatus in FIG. 4 is input. From the input terminal 122, the data transmitted from the output terminal 17 of the video encoding device in FIG. 1 or the data transmitted from the output terminal 82 of the video encoding device in FIG. 4 is input.

When the data of the AC component coefficient value among the data input from the input terminal 102 is input, the terminal 104 of the switch 103 is connected to the terminal 105 and the terminal 108 of the switch 107 is connected to the terminal 109. Therefore, the data is directly input to the inverse quantization unit 118. On the other hand, when the data of the DC component coefficient value or the DC coefficient predicted value is input, the terminal 104 of the switch 103 is connected to the terminal 1
06, and the terminal 108 of the switch 107 is connected to the terminal 11
Connect to 0. Therefore, the data is first added to adder 1
11. Input to terminal 115 of switch 113. The DC component coefficient value of the previous block is already stored in the memory 112, and the DC component coefficient value or the DC component coefficient prediction error value of the block to be decoded is added to the adder 111 by the adder 111. The DC component coefficient values are added and sent to the terminal 116 of the switch 113.

The information from the input terminal 101 is data "1".
In this case, the terminal 114 of the switch 113 is connected to the terminal 116. In this case, since the output from the terminal 106 of the switch 103 is the DC component coefficient prediction error value of the decoding target block, the output of the terminal 116 of the switch 113 is the DC component coefficient value of the decoding target block.
The value is input to the inverse quantization unit 118 via the terminals 114, 110, and 108 sequentially, and is input to the memory 112. On the other hand, if the information from the input terminal 101 is data “0”, the terminal 114 of the switch 113
Is connected to the terminal 115. In this case, the switch 103
Since the output from the terminal 106 is the DC component coefficient value of the decoding target block, the output of the terminal 115 of the switch 113 is the DC component coefficient value of the decoding target block, and the value is output to the terminals 114, 110, and 108. The signals are sequentially input to the inverse quantization unit 118 and input to the memory 112.

Each coefficient is inversely quantized by the inverse quantization unit 118, and is inversely discrete cosine transformed by the inverse discrete cosine transform unit 119, thereby obtaining an inter-frame difference signal. The (N-1) th frame reproduced image has already been stored in the frame memory 117, and the motion compensation unit 123 has the input terminal 1
The value of the pixel shifted by the motion vector input from 22 is input to the adder 120. In the adder 120, the two images are added and output from the output terminal 121 as an N-th frame reproduced image and stored in the frame memory 117.

When the moving picture decoding apparatus of the present embodiment decodes the data coded by the moving picture coding apparatus of the first embodiment of the present invention, the data is inputted from the input terminal 101 in frame units. , And the switch 113 is also switched in frame units. On the other hand, when the video decoding device of the present embodiment decodes data encoded by the video encoding device of the second embodiment of the present invention, data is input from the input terminal 101 in block units. Therefore, the switch 113 is also switched in block units.

[0034]

As described above, in the present invention, the intra-block DC component value of the inter-frame difference signal is already encoded /
Means for predicting from the DC component value in the decoded neighboring block and using the prediction error for encoding / decoding are provided, and the above means is applied to a frame in which a luminance change has occurred in an entire area. Efficient coding can be realized by using the above means for the other areas.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a video encoding device according to a first embodiment of the present invention.

FIG. 2 is a diagram illustrating values of discrete cosine transform coefficients before being quantized.

FIG. 3 is a diagram illustrating values of discrete cosine transform coefficients after quantization.

FIG. 4 is a configuration diagram of a video encoding device according to a second embodiment of the present invention.

FIG. 5 is a configuration diagram of a video decoding device according to an embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 1 input terminal 2, 3 frame memory 4 motion detecting unit 5 motion compensating unit 6 intra-block difference average calculating unit 7 DC component prediction on / off determining unit 8 switch 9 discrete cosine transform unit 10 quantizing unit 11 switch 12 to 14 terminal Reference Signs List 15 switch 16 subtractor 17 output terminal 18 memory 19-21 terminal 22 switch 23-25 terminal 26 inverse quantization unit 27 inverse discrete cosine transform unit 28 adder 29 frame memory 30 subtractor 31, 32 output terminal 33-35 terminal 51 Input terminal 52 Subtractor 53 Discrete cosine transform unit 54 Quantization unit 55 Switch 56 to 58 Terminal 59 Switch 60 to 62 Terminal 63 Switch 64 to 66 Terminal 67 DC component prediction on / off determination unit 68 Subtractor 69 Memory 70 Switch 71 to 71 73 terminal 74 inverse quantization unit 75 inverse discrete In-conversion unit 76 Adder 77 Frame memory 78 Motion detection unit 79 Motion compensation unit 80-82 Output terminal 101, 102 Input terminal 103 Switch 104-106 terminal 107 Switch 108-110 terminal 111 Adder 112 Memory 113 Switch 114-116 terminal 117 Frame memory 118 Inverse quantization unit 119 Inverse discrete cosine transform unit 120 Adder 121 Output terminal 122 Input terminal 123 Motion compensation unit

Claims (8)

[Claims] 1. A moving image encoding apparatus that encodes an inter-frame difference signal between an encoding target frame of a moving image and a reference frame for each predetermined block, wherein the encoding target in the encoding target frame is Inter-frame difference signal DC component prediction means for predicting an inter-frame difference signal DC component value in a block from an inter-frame difference signal DC component value of an already coded block in a frame to be encoded and outputting a prediction error value thereof Whether to use the prediction error value output from the inter-frame difference signal DC component prediction means for encoding, or whether to use the inter-frame difference signal DC component value in the encoding target block for encoding. An inter-frame difference signal DC component prediction on / off determining means for determining in a predetermined area unit and outputting the determined information; When the determination information output from the inter-frame difference signal DC component prediction on / off determination means is “inter-frame difference signal DC component prediction on”, the information is output from the inter-frame difference signal DC component prediction means. When the prediction error value is used for encoding, and conversely, when the decision information output from the inter-frame difference signal DC component prediction on / off determining means is “inter-frame difference signal DC component prediction off”, A moving image encoding apparatus comprising: an inter-frame difference signal DC component prediction on / off switching unit that uses an inter-frame difference signal DC component value in the encoding target block for encoding. 2. The method according to claim 1, wherein said inter-frame difference signal DC component prediction on / off determining means uses a prediction error value output from said inter-frame difference signal DC component prediction means for encoding, or said encoding means. The moving picture coding apparatus according to claim 1, wherein whether to use the DC component value of the inter-frame difference signal in the target block for coding is determined on a frame basis. 3. An inter-frame difference signal DC component prediction on / off determining means, comprising: a motion detecting unit that detects a motion between the encoding target frame and the reference frame; A motion compensating unit that shifts, an intra-block difference average value calculating unit that calculates an intra-block average value of a difference between the encoding target frame and the reference frame shifted by the motion for each of the predetermined blocks, The average value of one frame of the average difference value within a block,
Calculate the variance value, or correlation coefficient value, based on the value,
The moving image according to claim 2, further comprising a DC component prediction on / off determining unit that determines whether to turn on the inter-frame difference signal DC component prediction or to turn off the inter-frame difference signal DC component prediction for the encoding target frame. Image coding device. 4. An inter-frame difference DC component prediction on / off
Off determining means for using the prediction error value output from the inter-frame difference signal DC component prediction means for encoding, or for encoding the inter-frame difference signal DC component value in the encoding target block for encoding The moving picture coding apparatus according to claim 1, wherein whether to use the moving picture coding unit is determined in units of the predetermined block. 5. The inter-frame difference signal DC component prediction on / off determining means includes an absolute value of an inter-frame difference DC component of an encoding target block, and an inter-frame difference DC component value of an encoding target block. The absolute value of the difference between the inter-frame difference DC component value of the block subjected to encoding and the absolute value of the inter-frame difference DC component of the encoding target block is smaller, and it is determined that the inter-frame difference signal DC component prediction is off. If the absolute value of the difference between the inter-frame difference DC component value of the encoding target block and the inter-frame difference DC component value of the already encoded block is smaller, it is determined that the inter-frame difference signal DC component prediction is ON. The moving picture coding apparatus according to claim 4, further comprising a DC component prediction on / off determining unit that performs the prediction. 6. The inter-frame difference signal DC component prediction means and the inter-frame difference signal DC component prediction on / off switching means output an AC component value and a DC component value quantized by a quantization unit, respectively. First and second signal paths, and a first switch for outputting the AC component value and the DC component value quantized by the quantization unit to the first and second signal paths, respectively;
A memory that receives the output of the second signal path and stores it as a DC component value of a block that has already been encoded, and calculates the difference between the DC component value output to the second signal path and the output of the memory, A subtractor that outputs a DC component prediction error value that is an output to a third signal path; and a subtractor that outputs “interframe difference signal DC component prediction on” information from the interframe difference signal DC component prediction determination means. The output on the second signal path is selected when the information on the signal path No. 3 is selected and the information of “DC component prediction off between frames is output” is output.
And an output on the first signal path when an AC component value is output from the quantization unit, and an output of the second switch when a DC component value is output from the quantization unit. And an output on the first signal path when an AC component value is output from the quantization unit, and a DC component value is output from the quantization unit when the AC component value is output from the quantization unit. 2. The moving picture coding apparatus according to claim 1, further comprising a fourth switch that selects an output on the second signal path when output, and outputs the selected signal to an inverse quantization unit. 7. A moving picture decoding apparatus for decoding coded data coded by the moving picture coding apparatus according to claim 1, comprising: an inter-frame difference signal DC component prediction on / off determination information input from an input terminal. Is "inter-frame difference signal DC component prediction ON", the value obtained by adding the already decoded inter-frame difference signal DC component value to the inter-frame difference signal DC component prediction error value separately input from the input terminal. Is used for decoding, and if the inter-frame difference signal DC component prediction on / off determination information input from the input terminal is “inter-frame difference signal DC component prediction off”, the inter-frame difference signal DC component prediction Moving picture decoding characterized by comprising inter-frame difference signal DC component prediction on / off switching means used for decoding the difference signal DC component value itself. Location. 8. A first signal path through which the inter-frame difference signal DC component prediction on / off switching means transmits the AC component value output from an output terminal of the moving picture coding apparatus according to claim 1. And a second signal path through which the DC component value or the DC component prediction error value output from the output terminal of the video encoding device according to claim 1 is transmitted, and wherein the AC component value is a first signal. And the DC component value or the DC component prediction error value is added to the first switch for outputting to the second signal path, and the output of the second signal path and the DC component predicted value are added to the third switch. An adder that outputs the signal on the third signal path, and selects the output on the third signal path when the inter-frame difference signal DC component prediction on / off determination information is “inter-frame difference signal DC component prediction on”. The inter-frame difference signal DC component prediction ON When the / off determination information is “inter-frame difference signal DC component prediction off”, a second switch for selecting an output on the second signal path, and storing the output of the second switch, the DC component prediction value 8. A memory for outputting to said adder as a memory, and a third switch for outputting an AC component value on a first signal path and a DC component value output from a second switch to an inverse quantization unit, respectively. Video decoding device.