JPH11205802A - Data encoding device - Google Patents

Abstract

(57) [Summary] Even if the amount of motion of input image data is large, encoded image data with high image quality with little deterioration in image quality is generated. In a data encoding apparatus for encoding input data by a predetermined encoding method, a motion amount of image data is detected, and when the detection result exceeds a predetermined threshold level, the image data is encoded. A pixel number converting means for reducing the number of pixels, and a code amount of a motion vector which is reduced according to the image data having the reduced pixel number by the pixel number converting means. Encoding means for encoding the coefficient data so that the generated code amount can be obtained by allocating to the code amount of the coefficient data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[Table of Contents] The present invention will be described in the following order.

[0002] 2. Description of the Related Art [0004] Prior Art (FIGS. 3 and 4) Problems to be Solved by the Invention Means for Solving the Problems

[0003]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data encoding device, and is suitably applied to, for example, a data encoding device used in a digital broadcasting system.

[0004]

2. Description of the Related Art In recent years, various compression / encoding methods have been proposed as methods for reducing the amount of video and audio information, and a typical one is MPEG2 (Moving Picture Experts Gr.).
There is a method called oup Phase 2). MPEG2
2. Description of the Related Art A digital broadcasting system has been started in which video and audio broadcast data are compression-encoded using a system and broadcast using terrestrial waves or satellite waves.

In FIG. 3, reference numeral 50 denotes a data encoding device as a whole, and image data D1 is supplied from a video signal supply device (not shown) such as a video tape recorder to a pre-filter 5.
1 is input. The image data D1 is stored in the pre-filter 51.
After undergoing band limiting processing according to the frequency characteristic control signal D55 supplied from the quantization rate control unit 55 in
The high-frequency component of the image data D1 is reduced, and is output to the pixel number conversion unit 52 as band-limited video data D51. Here, the high-frequency component of the image data D1 expresses a fine portion of the image, and the pre-filter 51 omits the fine portion of the image by reducing the high-frequency component to reduce the influence on the entire image. While reducing the bandwidth of the image data D1.

The pixel number conversion section 52 performs a pixel number conversion process on the input band-limited video data D51. That is, assuming that the number of pixels in the horizontal direction of the input image data D1 is the number of horizontal pixels M, the pixel number conversion unit 52 performs band-limited video data D51 obtained by band-limiting the image data D1.
Is reduced to the number N of reduced horizontal pixels that satisfies the relationship of N <M, and is sent to the encoding unit 53 as pixel number converted image data D52. Here, the number N of reduced horizontal pixels is set to a large value for image contents requiring high image quality and a small value for image contents not requiring high image quality, according to the image contents of the image data D1.

[0007] The pixel number converted image data D52 is subjected to motion compensation processing and DCT (Discrete Cosine) in an encoding unit 53.
Transform, discrete cosine transform) processing, quantization processing, and VLC (Variable Length Coding) processing, compression-encoded, and variable-length encoded data D53.
Is sent to the buffer 54. At this time, the encoding unit 5
3 adjusts the quantization rate in the quantization processing based on the quantization control signal D56 supplied from the quantization rate control unit 55. Further, the coding unit 53 is supplied with coding unit control information D57 from the coding unit control unit 57, and the coding unit 53 performs coding timing and motion compensation processing based on the coding unit control information D57. Set the motion vector search range in.

Here, the generated code amount of each frame composed of the variable length coded data D53 varies according to the image content of the image data D1. For this reason, the encoding unit 53 determines the quantization rate of the quantization process and the pre-filter 5 based on the occupation amount of the variable-length encoded data D53 in the buffer 54.
1 to control the variable-length coded data D5
3 are stored in the buffer 54, and then, for example, G
The generated code amount per OP (Group Of Pictures) is output as fixed-length coded data D50 controlled to be constant.

That is, the quantization rate control unit 55 constantly monitors the accumulation state of the variable-length encoded data D53 in the buffer 54, and monitors the accumulation state with the occupancy rate information D5.
As a result, the quantization control signal D56 and the frequency characteristic control signal D55 are generated based on the occupation ratio information D54, and are supplied to the encoding unit 53 and the pre-filter 51. Thus, the encoding unit 53 generates the fixed-length encoded data D50 in which the generated code amount per fixed period is controlled to be constant.

[0010]

In the data encoding device 50, the larger the amount of motion of the image of the image data D1, the greater the amount of generated motion vectors per frame. That is, FIG.
And (B) show the ratio between the code amount of the DCT coefficient and the code amount of the motion vector per frame before performing the quantization process, and the motion vector MV2 obtained by coding an image having a large motion amount is: The code amount is larger than that of the motion vector MV1 obtained by coding an image with a small motion amount.

DCT coefficient DCT1 and DCT coefficient DCT
The amount of generated code 2 hardly changes regardless of the amount of motion of the image. For this reason, the data encoding device 50 determines that the DCT coefficient DC
Since the generated code amount of T2 and the entire motion vector MV2 is increased, the quantization is coarsely performed by increasing the quantization step when quantizing the DCT coefficient. As a result, the data encoding device 50 can generate the fixed-length encoded data D50 in which the amount of generated code per fixed period is controlled to be constant. However, since the quantization step is increased, the image quality is deteriorated accordingly. There was a problem that occurs.

The present invention has been made in view of the above points, and has been made in consideration of the above points. Even when the amount of motion of input image data is large, data encoding that can generate high-quality encoded data with little deterioration in image quality is performed. It is intended to propose a device.

[0013]

According to the present invention, there is provided a data encoding apparatus for encoding input data by a predetermined encoding method, wherein a motion amount of image data is detected. When the detection result exceeds a predetermined threshold level, the pixel number conversion means for reducing the number of pixels of the image data, and the number of pixels is reduced according to the reduced image data by the pixel number conversion means. Encoding means for encoding the coefficient data so that the generated code amount is obtained by allocating the code amount of the motion vector to the code amount of the coefficient data obtained by converting the image data by a predetermined method. By assigning the reduced code amount of the motion vector to the code amount of the coefficient data, the generated code amount of the coefficient data is increased, and the quantization step size is set smaller by that amount. It can be encoded.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings.

In FIG. 1, reference numeral 10 denotes a data encoding device as a whole, which is supplied from a video signal supply device (not shown) such as a video tape recorder or the like to an HDTV (High Definition Television).
vision) standard image data D1 is input to the pre-filter 12.

The pre-filter 12 performs a band limiting process on the image data D1 based on the frequency characteristic control signal D18 supplied from the quantization rate control unit 18, and the image data D1
1, the bandwidth of the image data D1 is reduced and the system delay 13 is used as the band-limited image data D2, while omitting small parts of the image and reducing the influence on the entire image. And the motion amount detection unit 15.

The motion amount detector 15 detects the motion amount of each frame image of the band-limited image data D2. That is, the motion amount detection unit 15 stores each frame image of the band-limited image data D2 as a reference image in an internal frame memory (not shown), and calculates the sum of absolute differences of pixels between frames. This is sent to the pixel number conversion determination unit 17 as the motion amount data D15.

The pixel number conversion judging section 17 calculates the motion amount data D1.
5 is compared with a predetermined threshold value. When the motion amount data D15 is larger than the threshold value, the value of the pixel number conversion control signal D17 for converting the number of pixels is set to “1”. When the motion amount data D15 is smaller than the threshold value, The value of the pixel number conversion control signal D17 that does not directly convert the pixel number is set to “0” and the pixel number conversion unit 14
And to the encoding unit control unit 16.

The band-limited image data D2 sent to the system delay 13 is temporarily written into a memory (not shown) of the system delay 13 for each frame, and then read out for each frame after a predetermined standby time. , To the pixel number conversion unit 14. Here is system day 1
3 is the band-limited image data D in the motion amount detection unit 15.
2 after the elapse of the same standby time as the time required for the motion amount detection processing per frame of the pixel number conversion control signal D1.
7 is input to the pixel number conversion unit 14 in synchronization with the timing at which the number 7 is input to the pixel number conversion unit 14.

The pixel number conversion unit 14 is a pixel number conversion determination unit 1
7 converts the number of pixels in the horizontal direction of the band-limited image data D2 based on the pixel number control signal D17 supplied from. That is, the pixel number conversion unit 14 sets the number of pixels in the horizontal direction of the band-limited image data D2 obtained by band-limiting the image data D1 to M
Assuming that the number of horizontal pixels after pixel number conversion is N, if the value of the pixel number conversion control signal D17 is “1”, the number of horizontal pixels of the band-limited image data D2 is reduced to N pixels,
This is sent to the encoding unit 20 as the pixel number conversion image data D3, and if the value of the pixel number conversion control signal D17 is “0”, the band-limited image data D2 is directly used as the pixel number conversion image data D3. 20.

The encoding unit control unit 16 supplies the encoding unit control signal D16 to the encoding unit 20 when the pixel number control signal D17 is input from the pixel number conversion determining unit 17. Encoding unit 2
0 performs an encoding process on the pixel number converted image data D3 of N pixels by operating based on the encoding unit control signal D16.

That is, when the value of the pixel number control signal D17 is "1", the band-limited image data D2 of the horizontal pixel number M shown in FIG. The data is transmitted to the encoding unit 20 as the pixel number converted image data D3 having the horizontal pixel number N shown in FIG.

Here, the pixel number-converted image data D3 is not transmitted continuously, but after N pixels are continuously transmitted as shown in FIG. Transmission is stopped for a corresponding period. The encoding unit 20 performs the encoding process in accordance with the encoding unit control signal D16 during a period corresponding to N pixels, and stops the encoding process in a period corresponding to (MN) pixels, thereby operating clock. While maintaining the constant, the encoding process is performed on the pixel number converted image data D3 of N pixels.

On the other hand, when the value of the pixel number conversion value D17 is “0”, the pixel number conversion unit 14 converts the band-limited image data D2 of the horizontal pixel number M shown in FIG. The image data is sent to the encoding unit 20 as the pixel number converted image data D3 without changing the horizontal pixel number M. The encoding unit 20 continuously performs the encoding process based on the encoding unit control signal D16. Further, the encoding unit 20 sets a motion vector search range in the motion compensation processing according to the encoding unit control signal D16.

In the encoding section 20, the pixel number converted image data D3 is input to the preprocessing section 21. Pre-processing unit 21
Specifies, for each frame image of the pixel number converted image data D3, which of three image types of I, P, and B pictures is to be processed according to the set sequence, According to the image type of the frame image, the frame image is rearranged in the encoding order.
It is divided into macroblocks composed of a luminance signal of pixels × 16 lines and a color difference signal corresponding to the luminance signal, and this is divided as macroblock data D4 by the arithmetic unit 22.
And the motion vector detecting unit 31. Incidentally, the preprocessing unit 21 converts the raster scan into the block scan, thereby changing the encoding order of the images.

The motion vector detecting section 31 calculates the motion vector of each macro block of the macro block data D4 based on the macro block data D4 and the reference image data D28 stored in the frame memory 29, and calculates the motion vector data. The motion compensator 30 and VL as D31
Each of them is sent to a variable length coding (C) unit 25.

The arithmetic unit 22 determines whether the macroblock data D4 supplied from the preprocessing unit 21 is an intra mode, a forward prediction mode, or a backward prediction mode based on the image type of each macroblock of the macroblock data D4. Alternatively, motion compensation is performed in one of the bidirectional prediction modes. Here, the intra mode is a method in which a frame image to be encoded is directly used as transmission data, and the forward prediction mode is a method in which a prediction residual between the frame image to be encoded and a past reference image is transmitted data. It is a method. The backward prediction mode is a method in which a prediction residual between a frame image to be encoded and a future reference image is used as transmission data, and the bidirectional prediction mode is a method in which a frame image to be encoded is This is a method in which a prediction residual between an average value of two prediction images, a reference image and a future reference image, is used as transmission data.

First, the case where the macroblock data D4 is an I-picture will be described. In this case, the macro block data D4 is processed in the intra mode. That is, the arithmetic unit 22 converts the macroblock of the macroblock data D4 into the DCT (Di
screte Cosine Transform, discrete cosine transform) unit 23
To send to.

The DCT unit 23 performs a DCT conversion process on the operation data D5 to convert the operation data D5 into DCT coefficients, and sends the result to the quantization unit 24 as DCT coefficient data D6. Quantizer 24
Performs a quantization process on the DCT coefficient data D6 and sends out the quantized DCT coefficient data D7 to the VLC unit 25 and the inverse quantization unit 26, respectively. At this time, the quantization unit 24
Adjusts the quantization rate in the quantization processing according to the quantization control signal D20 supplied from the quantization rate control unit 18, and controls the code generation amount of the quantized DCT coefficient data D7.

The quantized DCT sent to the inverse quantization unit 26
The coefficient data D7 undergoes an inverse quantization process, and is sent to the inverse DCT unit 27 as DCT coefficient data D26. The DCT coefficient data D26 undergoes an inverse DCT process in an inverse DCT unit 27, and becomes an arithmetic unit 28 as arithmetic data D27.
And stored in the frame memory 29 as reference image data D28.

Next, a case where the macroblock data D4 is a P-picture will be described. In this case, arithmetic unit 2
Numeral 2 performs a motion compensation process on the macroblock data D4 in either the intra mode or the forward prediction mode.

When the prediction mode is the intra mode, the computing unit 22 sends the macroblock of the macroblock data D2 as it is to the DCT unit 23 as the computation data D5, as in the case of the I-picture. Calculation data D5
Is subjected to a DCT transformation process in the DCT unit 23 and a quantization process in the quantization unit 24 in the same manner as in the case of the above-mentioned I-picture, and to the VLC unit 25 and the inverse quantization unit 26 as quantized DCT coefficient data D7. Sent out. At this time, the quantization unit 24 adjusts the quantization rate in the quantization process in accordance with the quantization control signal D20 supplied from the quantization rate control unit 18, and generates the quantized DCT coefficient data D
7 is controlled.

The quantized DCT sent to the inverse quantization unit 26
The coefficient data D7 undergoes an inverse quantization process, and is sent to the inverse DCT unit 27 as DCT coefficient data D26. DCT
The coefficient data D26 undergoes inverse DCT processing in the inverse DCT unit 27, is sent to the arithmetic unit 28 as operation data D27, and is stored in the frame memory 29 as reference image data D28.

In the case where the prediction mode is the forward prediction mode, the arithmetic unit 22 calculates the forward prediction image data D30F supplied from the motion compensator 31 for the macroblock data D4.
Is subtracted using.

Here, the forward prediction image data D30F is
Reference image data D stored in frame memory 29
28 is calculated by performing motion compensation according to the motion vector data D31. That is, in the forward prediction mode, the motion compensator 30 reads out the reference image data D28 by shifting the read address of the frame memory 29 in accordance with the motion vector data D31, and uses the reference image data D28 as the forward prediction image data D30F. 28. The arithmetic unit 22 subtracts the forward prediction image data D30F from the macroblock data D4 to obtain difference data as a prediction residual, and sends it to the DCT unit 23 as operation data D5.

The operation data signal D5 undergoes DCT conversion processing in the DCT section 23, further undergoes quantization processing in the quantization section 24, and is sent to the VLC section 25 and the inverse quantization section 26 as quantized DCT coefficient data D7. . At this time, the quantization unit 24 adjusts the quantization rate in the quantization process according to the quantization control signal D20 supplied from the quantization rate control unit 18, and generates the quantized DCT coefficient data D7.
Is controlled.

The quantized DCT sent to the inverse quantization unit 26
The coefficient data D7 undergoes an inverse quantization process, and is sent to the inverse DCT unit 27 as DCT coefficient data D26. The DCT coefficient data D26 undergoes inverse DCT processing in the inverse DCT unit 27, and is sent to the arithmetic unit 28 as operation data D27. Here, the forward predictive image data D30F is supplied from the motion compensation unit 30 to the arithmetic unit 28, and the arithmetic unit 2
8 indicates that the forward prediction image data D3
By adding 0F, the reference image data D28 (P picture) is locally reproduced and stored in the frame memory 29.

Next, the case where the macroblock data D4 of B picture is supplied from the preprocessing unit 21 to the arithmetic unit 22 will be described. In this case, the arithmetic unit 22 performs any one of the motion compensation processing of the intra mode, the forward prediction mode, the backward prediction mode, and the bidirectional prediction mode on the macroblock data D4.

When the prediction mode is the intra mode or the forward prediction mode, the macroblock data D2 is subjected to the same processing as in the case of the above-mentioned P picture, but the B picture is not used as a reference picture, so that the frame memory 2
9 is not stored.

When the prediction mode is the backward prediction mode, the arithmetic unit 22 performs the backward prediction image data D30B supplied from the motion compensator 30 on the macroblock data D4.
Is subtracted using.

Here, the backward prediction image data D30B is
Reference image data D stored in frame memory 29
28 is calculated by performing motion compensation according to the motion vector data D31. That is, in the backward prediction mode, the motion compensating unit 31 reads the reference image data D28 by shifting the read address of the frame memory 29 according to the motion vector data D31, and uses the reference image data D28 as the backward prediction image data D30B. 28. The calculator 22 subtracts the backward prediction image data D30B from the macroblock data D4 to obtain difference data as a prediction residual, and obtains the DCT unit 23 as calculation data D5.
To send to.

The operation data D5 is stored in the DCT section 23 as D
After undergoing a CT conversion process, the quantization unit 24 further performs a quantization process, and performs VL as the quantized DCT coefficient data D7.
It is sent to the C section 25 and the inverse quantization section 26. At this time, the quantization unit 24 adjusts the quantization rate in the quantization process according to the quantization control signal D20 supplied from the quantization rate control unit 18 and controls the code generation amount of the quantized DCT coefficient data D7. I do.

The quantized DCT sent to the inverse quantization unit 26
The coefficient data D7 undergoes an inverse quantization process, and is sent to the inverse DCT unit 27 as DCT coefficient data D26. DCT
The coefficient data D26 undergoes inverse DCT processing in the inverse DCT unit 27, and is sent to the arithmetic unit 28 as operation data D27. Here, the backward prediction image data D30B is supplied from the motion compensation unit 30 to the computing unit 28, and the computing unit 2
8 is the backward prediction image data D3
The reference image data D28 (B-picture) is locally reproduced by adding 0B. However, since the B-picture is not used as another predicted reference image, the reference image data D28 is not stored in the frame memory 29.

When the prediction mode is the bidirectional mode, the arithmetic unit 22 calculates the motion compensation unit 30 from the macroblock data D4.
The difference data as a prediction residual is obtained by subtracting the average value of the forward prediction image data D30F and the backward prediction image data D30B supplied thereto, and the difference data is transmitted to the DCT unit 23 as operation data D5.

The operation data signal D5 undergoes DCT conversion processing in the DCT section 23, further undergoes quantization processing in the quantization section 24, and is sent to the VLC section 25 and the inverse quantization section 26 as quantized DCT coefficient data D7. You. At this time, the quantization unit 24 adjusts the quantization rate in the quantization process in accordance with the quantization control signal D20 supplied from the quantization rate control unit 18, and generates the quantized DCT coefficient data D
7 is controlled.

The quantized DCT sent to the inverse quantization unit 26
The coefficient data D7 undergoes an inverse quantization process, and is sent to the inverse DCT unit 27 as DCT coefficient data D26. DCT
The coefficient data D26 is converted by the inverse DCT section 27
After receiving the processing, it is sent to the arithmetic unit 28 as arithmetic data S27. Here, the forward predictive image data D30F and the backward predictive image data D30B are supplied from the motion compensation unit 30 to the arithmetic unit 28, and the arithmetic unit 28
The reference image data D28 (B picture) is generated by adding the average value of the forward prediction image data D30F and the backward prediction image data D30B to 27, but the B picture is not used as another prediction reference image. , The reference image data D28 is not stored in the frame memory 29.

Thus, the pixel number converted image data D3 input to the data encoding device 10 is subjected to motion compensation processing, DC
After undergoing the T processing and the quantization processing, it is supplied to the VLC unit 25 as quantized DCT coefficient data D7.

The VLC unit 25 performs a variable length coding process on the quantized DCT coefficient data D7 and the motion vector data D31, and sends the result to the buffer 19 as variable length coded data D8. In the buffer 19, the variable-length coded data D8 is temporarily stored (GOP
) Is read out as fixed-length coded data D10).

In this case, the pixel number converted image data D3 is
The number of pixels of the image data D1 is N by the pixel number conversion unit 14.
Since the number of pixels is reduced, the number of macroblocks of the macroblock data D4 obtained by dividing the image of the pixel number converted image data D3 into 16 pixels × 16 lines is naturally reduced.

Here, the generated code amount of the motion vector in the motion compensation processing is almost proportional to the number of macroblocks. Therefore, when the number of macroblocks is reduced, the generated code amount of the motion vector is also reduced.

Therefore, since the generated code amount of the fixed-length coded data D10 is constant, the decrease in the code amount of the motion vector due to the decrease in the number of macroblocks is calculated by the DCT coefficient data D10.
6 as a result, the quantization rate control unit 18 obtains D in a finer quantization step.
Since the CT coefficient data D6 can be quantized, high-quality fixed-length coded data D10 in which deterioration in image quality is prevented is generated.

In the above configuration, the data encoding device 1
0 is the motion data D supplied from the motion detector 15
When 15 is larger than a predetermined threshold value, it is determined that the motion vector of the image is large, and a pixel number conversion control signal D17 having a value of “1” is supplied to the pixel number conversion unit. The number-of-pixels conversion unit 14 obtains the number-of-pixels converted image data D3 by reducing the number of pixels in the horizontal direction of the band-limited image data D2 to N pixels based on the number-of-pixels conversion control signal D17. Is converted to 16 pixels ×
It is divided into macroblock data D4 of 16 lines.

As a result, the number of macroblocks in the macroblock data D4 decreases as much as the number of pixels of the frame image is reduced, and as a result, the amount of generated motion vector codes decreases. That is, since the number of macroblocks decreases in accordance with the amount of reduction in the number of pixels, the size of each macroblock increases, thereby reducing the number of pixels required to represent a motion vector. Therefore, the larger the motion vector of an image is, the greater the rate at which the amount of generated codes of the motion vector is reduced. Therefore, the data encoding device 1
0 indicates that the number of macroblocks decreases with a reduction in the number of pixels when the motion vector of the image is large.
It can be assigned to the generated code amount of the coefficient data D6.

Thus, the data encoding device 10 can assign the amount of generated code of the DCT coefficient data D6 to the amount of reduced code of the generated code of the motion vector, so that the quantization step size of the quantization unit 24 can be assigned to that amount. Thus, even when the image moves rapidly and the generated code amount increases, the coefficient value of the DCT coefficient data D6 can be increased to prevent a decrease in resolution in the DCT space.

According to the above configuration, the data encoding device 1
0 indicates that the number of pixels is reduced when the motion vector of the frame image is large, and the generated code amount of the motion vector is reduced in accordance with the reduced number of macroblocks. By allocating to the code amount, the quantization step size can be reduced by that amount and quantization can be performed, and high-quality coded data with improved image quality can be generated.

In the above-described embodiment, the case where the number of pixels is reduced based on the result of comparison between the motion amount data D15 and a predetermined threshold has been described. However, the present invention is not limited to this. , A plurality of types of threshold values, and a plurality of reduced horizontal pixel numbers corresponding to the plurality of types of threshold values, for example, 1440 pixels, 1280 pixels, 1152 pixels, and 960
Pixels may be set, and the number of horizontal pixels may be reduced to a certain number of pixels based on a comparison result between the motion amount data D15 and a predetermined threshold. In this case, the pixel number control signal D
Reference numeral 17 denotes not a 1-bit signal of "1" or "0" but a k-bit (j≤2 ^k ) corresponding to a plurality of types of thresholds j
Is used, and the number of pixels is reduced at a predetermined integer ratio.

Further, in the above-described embodiment, a case has been described where the HDTV standard image data D1 is input to the data encoding device 10. However, the present invention is not limited to this, and the present invention is not limited to this. Data may be input.

Further, in the above-described embodiment, a case has been described where the number of pixels in the horizontal direction is reduced in accordance with the degree of compression difficulty of an image having a sharp movement. However, the present invention is not limited to this. The vertical direction, ie, the number of scanning lines, may be reduced, or the vertical direction and the horizontal direction, ie, the number of scanning lines, and the number of pixels in the horizontal direction may be simultaneously reduced.

Furthermore, in the above-described embodiment, a case has been described in which the present invention is applied to the data encoding device 10 of the MPEG2 system. However, the present invention is not limited to this, and various other encoding systems are used. The present invention may be applied to a data encoding device according to

[0060]

As described above, according to the present invention, in a data encoding apparatus for encoding input data by a predetermined encoding method, a motion amount of image data is detected, and the detection result is determined by a predetermined value. A pixel number conversion unit for reducing the number of pixels of the image data when the threshold level is exceeded, and a code amount of a motion vector reduced according to the image data having the reduced number of pixels by the pixel number conversion unit And coding means for coding the coefficient data so that the generated code amount is obtained by allocating the image data to the code amount of the coefficient data obtained by converting the image data by a predetermined method. By assigning the code amount of the vector to the code amount of the coefficient data to increase the generated code amount in the coefficient data, it is possible to set the quantization step size to be smaller by that amount and perform coding. In case the motion vector of the image data input Te is larger, it is possible to produce a high quality coded data with less degradation in image quality.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a data encoding device according to the present invention.

FIG. 2 is a schematic diagram used for explaining pixel number conversion.

FIG. 3 is a schematic diagram illustrating an increase in a motion vector amount.

FIG. 4 is a block diagram showing a conventional data encoding device.

[Explanation of symbols]

10, 50 ... data encoding device, 12, 51 ... pre-filter, 13 ... system delay, 14, 52 ...
Pixel number conversion unit, 15 ... Motion amount detection unit, 16, 57 ...
Encoding unit control unit 17, pixel number conversion determination unit 18, 5,
5: Quantization rate control section, 19, 54 ... Buffer,
20, 53... Encoding unit.

 ────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor: Adult, Funato 6-7-35 Kita-Shinagawa, Shinagawa-ku, Tokyo Sony Corporation (72) Inventor: Takeshi Uehara 6-35, Kita-Shinagawa, Shinagawa-ku, Tokyo 7-35 Sony Inside the company

Claims (4)

[Claims] 1. A data encoding apparatus for encoding input image data by a predetermined encoding method, wherein a motion amount of the image data is detected, and the detection result exceeds a predetermined threshold level. In this case, the pixel number conversion means for reducing the number of pixels of the image data; and the code amount of the motion vector reduced according to the image data whose number of pixels has been reduced by the pixel number conversion means, Coding means for coding the coefficient data so that the generated code amount is obtained by allocating the coefficient data to the code amount of the coefficient data converted by a predetermined method. apparatus. 2. The data encoding apparatus according to claim 1, wherein said pixel number conversion means reduces the number of pixels in the horizontal direction of said image data. 3. The data according to claim 1, wherein said pixel number conversion means does not reduce the number of pixels of said image data unless said detection result exceeds a predetermined threshold level. Encoding device. 4. The data encoding apparatus according to claim 1, wherein said pixel number conversion means reduces the number of pixels of said image data at a predetermined integer ratio.