JP2001218227A - Moving image compressor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a moving image compressor that can easily relive a load of its CPU and integrate compression functions to an application program. SOLUTION: A 1st compression control section 33 conducts a part of compression processing in the case of receiving input image data and stores the image data, after the processing to a buffer 34. A 2nd compression control section 35 compresses image data stored in the buffer 43. The 1st and 2nd compression control sections 33, 35 execute the processing in parallel.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving picture compression apparatus for compressing an input picture stored in a memory at a fixed address in real time at fixed time intervals. The present invention relates to a moving image compression apparatus capable of real-time compression with a reduced CPU load.

[0002]

2. Description of the Related Art Conventionally, an image has been digitized and a CD-R
When recording on a recording medium such as an OM or a hard disk,
Since the data amount is huge, it is usually recorded by compression encoding. This compression encoding method includes various compression encoding methods. In particular, a relatively large number of DCT-based encoding methods for performing compression by using the property that the spatial frequency of an image is concentrated at a low frequency are relatively common. It is used. This is JPEG (Joint Photographic Ex
part group), MPEG (Moving)
This is adopted as an international standard encoding system such as Pictures Expert Group 1 or MPEG2.

[0003] In the following, moving images are recorded in real time by MPEG.
The compression operation in the case of compression will be described. FIG. 11 is a block diagram showing a configuration of a conventional moving picture compression apparatus. The moving image compression apparatus shown in FIG. 11 includes a capture unit 11 for converting a TV signal or the like into an original image of digital data, and a color signal format (YUV) capable of compressing the input image data captured by the capture unit 11 by MPEG. (YUV conversion unit 12) that converts the image size into an input image size and an image size to be compressed when the image size to be compressed is different from each other. A motion search unit 14 for searching for each area of 16 pixels × 16 pixels in MPEG), a DCT unit 15 for converting a block image into a spatial frequency, and a quantization unit 16 for quantizing frequency-converted DCT data. ,
Inverse quantization unit 19 that performs inverse quantization to restore the quantized data
An inverse DCT unit 18 for performing an inverse frequency transform for restoring the frequency-converted DCT data, a motion compensating unit 17 for compensating the motion of the image in block units to generate a new reference frame, And a variable length coding unit 20 for performing variable length coding for performing variable length coding.

A moving image compression apparatus converts an original image captured by a capture unit 11 into YUV data by a YUV conversion unit 12 and converts it into an image size to be compressed by a size conversion unit 13 as shown in FIG. Compression according to each picture type is performed.

In the case of an I picture, the YUV data is divided into blocks, converted into spatial frequencies by a DCT unit 15 for each block, quantized by a quantization unit 16, and subjected to variable length coding by a variable length coding unit 20. To output the compression code. Further, the quantized block is inversely quantized by the inverse quantization unit 19 and inverse frequency transformed by the inverse DCT unit 18 to create a reference frame.

In the case of a P picture, the YUV data is divided into blocks, and the motion search unit 14 finds a block having the highest correlation among the blocks of the previous frame stored as a reference frame for each block, and obtains the correlation. Is converted to a spatial frequency by the DCT unit 15, quantized by the quantization unit 16, variable-length coded by the variable-length coding unit 20, and output as a compressed code. In addition, the quantized block is inversely quantized by the inverse quantization unit 19, inversely frequency-transformed by the inverse DCT unit 18, and added to the motion-compensated block by the motion compensation unit 17 to create a reference frame.

In the case of a B picture, YUV data is divided into blocks, and the motion search unit 14 finds a block having the highest correlation among the blocks of the previous / next frame stored as a reference frame for each block. The DCT unit 15 converts the difference between the block of the previous / next frame having the highest correlation into a spatial frequency, quantizes it by the quantization unit 16, and performs variable-length coding by the variable-length coding unit 20, and outputs a compressed code. . In the case of a B picture, there is no need to create a reference frame.

As an example of performing such real-time compression, there is a moving image compression / decompression device disclosed in Japanese Patent Application Laid-Open No. Hei 9-116891. This moving picture compression / decompression device performs compression and decompression in real time by having both compression and decompression means, and performs bidirectional communication. Also, Japanese Unexamined Patent Application Publication No.
Japanese Patent Publication No. 177,893 proposes an encoder that analyzes video characteristic information in real time and dynamically changes and compresses control parameters.

Further, as an example of performing processing in case that compression cannot be performed in time for real-time MPEG compression, there is a technique disclosed in Japanese Patent Application No. 10-197500. In this method, the frame configuration is changed according to the frame rate that can be actually compressed, and the B picture is regarded as a frame having no difference and compressed, so that the compression is not delayed.

[0010]

SUMMARY OF THE INVENTION However, MPE
In G compression, images of a plurality of frames must be left in order to perform backward prediction with a B picture. However, usually, image data captured at a fixed address is stored.
It is necessary to copy the captured image data to another memory area so as not to be overwritten, and there is a problem that the CPU load increases. In addition, it is necessary to prepare a memory area capable of storing image data of a continuous number of B pictures or more, and to search for an area in which the previous compression is completed and new image data can be stored in order to store sequentially captured image data. However, there is a problem that it takes time to perform the processing.

Also, in any of the above-mentioned prior arts, little consideration is given to incorporating the MPEG compression function into another application having no MPEG compression function. For this reason, there is a problem that it is difficult to incorporate the MPEG compression function into another application. SUMMARY An advantage of some aspects of the invention is to provide a moving image compression apparatus that can reduce a CPU load and easily incorporate a compression function into an application.

[0012]

A moving image compression apparatus according to the present invention comprises a first compression control means (3) for receiving input image data.
3) and a second compression control means (35) for compressing the image data received by the first compression control means. The first compression control means and the second compression control means Is executed. As described above, by providing the first compression control unit and the second compression control unit that execute the processing in parallel, the captured image data can be transferred to the moving image without the application knowing the details of the moving image compression method. A system for compressing a moving image can be constructed simply by passing it to the image compression device.

One example of the configuration of the moving picture compression apparatus of the present invention is to convert input image data in R (red), G (green), B (blue) format into Y (luminance) U (R-Y color difference) V ( It has YUV conversion means (36a) for converting into image data of the (B-Y color difference) format. As described above, by converting the input image data in the RGB format into the image data in the YUV format,
Data capacity can be reduced. Further, as one configuration example of the moving image compression device of the present invention, the first compression control unit includes:
When the input image data is received, a part of the compression processing is performed. As described above, by performing a part of the compression process when the first compression control unit receives the input image data, the process of copying the image data can be reduced. In addition, a part of the compression processing includes:
R (red) G (green) B (blue) input image data is Y
(Luminance) UUV (R-Y color difference) V (B-Y color difference) YUV conversion for converting image data, size conversion for converting the size of input image data, frequency conversion for converting the frequency of input image data It is.

[0014] Further, one configuration example of the moving picture compression apparatus of the present invention has a buffer (34) having an area for storing a plurality of frames of image data processed by the first compression control means, The first compression control means stores the processed image data in a buffer, and the second compression control means compresses the image data stored in the buffer. Further, as one configuration example of the moving picture compression apparatus of the present invention, the second compression control means does not use the input image data or the image data processed by the first compression control means when the compression processing is delayed. In addition, the B picture is compressed assuming that the B picture has no difference. As described above, by compressing the B picture as a frame having no difference, the compression process can be prevented from being delayed. Further, as one configuration example of the moving picture compression apparatus of the present invention, the second compression control means determines that the compression processing is delayed when the buffer is not empty. Also, as one configuration example of the moving image compression apparatus of the present invention, the second compression control unit may be configured to perform the operation when the number of buffer areas is insufficient for the number of input image data received by the first compression control unit. , The compression process is determined to be delayed.

[0015]

BEST MODE FOR CARRYING OUT THE INVENTION [First Embodiment]
G (Moving Picture Experts
Embodiments of the present invention for the (Group) compression method will be described in detail with reference to the drawings. FIG. 1 is a configuration diagram showing a code format conforming to MPEG video.

The video data consists of one or more video sequences VSC and ends with a video sequence end code VSE. Each video sequence VSC
Is a video sequence header VSH and one or more GOs
P (Group Of Picture).

Each GOP is composed of one or more pictures, and one picture indicates one picture. At the head of each picture, a picture header including information such as a picture type is arranged. There are three types of pictures: an I picture composed of only intra-frame codes, a P picture composed of only forward inter-frame codes, and a B picture composed of forward and backward bi-directional inter-frame codes. .

A picture is composed of a plurality of slices divided into arbitrary areas. Each slice is composed of a plurality of macroblocks arranged in left-to-right or top-to-bottom order.

Macroblocks are roughly classified into two types: intrablocks, which are intra-frame codes, and interblocks, which are forward and bidirectional interframe codes. An I picture is composed of only intra blocks, but a P picture or a B picture may include not only inter blocks but also intra blocks.

Each macro block is composed of luminance components Y1, Y2, Y3, and Y4 obtained by further dividing a 16 × 16 dot block into 8 × 8 dot blocks, and an 8 × 8 dot block of an area corresponding to the luminance component. It is composed of six blocks of color difference components Cb and Cr. The block of 8 × 8 dots is the minimum unit of encoding.

FIG. 2 is a block diagram showing the configuration of the moving picture compression apparatus according to the first embodiment of the present invention. The moving image compression apparatus according to the present embodiment transfers a capture unit 31 for converting a TV signal or the like into an original image of digital data and image data captured by the capture unit 31 to a first compression control unit described later. A capture compression application 32, a first compression control unit 33 that receives image data from the capture compression application 32, performs a part of the compression process, and stores the data that has been partially compressed by the first compression control unit 33. It comprises a buffer 34 for storing a plurality of frames, and a second compression control unit 35 for executing a compression process on the data stored in the buffer 34.

A first compression control unit 33 controls the entire system, and a YUV for converting the input image data captured by the capture unit 31 into a color signal format (YUV format) compressible by MPEG. Conversion unit 36a
And a size conversion unit 37 that converts the image size when the image size of the input image data is different from the image size to be compressed.
a and a DCT (Discr) that converts an input image for each block (16 pixels × 16 pixels in MPEG) into a spatial frequency.
et Cosine Transform) section 39a
It is composed of

The buffer 34 has a plurality of areas 34-1, 3-3.
4-2, 34-3,... The second compression control unit 35 includes a size conversion unit 37b that converts the image size when the image size of the data stored in the buffer 34 is different from the image size to be compressed, and the motion of the image of the previous / next frame and the current frame. , A DCT unit 39b that converts a block image into a spatial frequency, a quantization unit 40b that quantizes the frequency-converted DCT data, and a An inverse quantization unit 43b for performing inverse quantization to restore the original reference frame; an inverse DCT unit 42b for performing inverse frequency transformation for restoring the frequency-converted DCT data to the original; A motion compensating unit 41b to generate;
Variable length coding unit 44b for performing variable length coding on the quantized data
It is composed of

In the moving picture compression apparatus of FIG. 2, the capture compression application 32 passes the original image captured by the capture section 31 to the first compression control section 33, and the first compression control section 33 performs a part of the compression processing. Execute buffer 3
4 and the second compression control unit 35 performs a compression process on the data stored in the buffer 34, and the capture compression application 32 compresses the compression code output from the second compression control unit 35. Write to file.

The YUV converter 3 in the first compression controller 33
6a, when the image data captured by the capture unit 31 is R (red), G (green), and B (blue) data, the RGB image data is converted into Y (luminance) U (RY-color difference) V
(B-Y color difference) data.

The size converter 3 in the first compression controller 33
If the image size of the image data captured by the capture unit 31 and the image size to be compressed are different, 7a converts the YUV data into an image size to be compressed.

The DCT unit 39a in the first compression control unit 33
If the frame to be compressed is an I picture, frequency conversion is performed on this frame. The data processed by the first compression control unit 33 is stored in a buffer 34 by the control unit 33a.
Is stored in an empty area.

The second compression control unit 35 executes compression according to each picture type. In the case of an I-picture, the control unit 35b in the second compression control unit 35 extracts data to be compressed from the buffer 34 and, if size conversion is necessary, passes the data to the size conversion unit 37b. The size converter 37b converts the data into an image size to be compressed.

Further, if the image data to be compressed stored in the buffer 34 or the image data whose size has been converted by the size converter 37b has not been subjected to frequency conversion (DCT), the controller 35b converts the data to the DCT unit 3b.
9b. The DCT unit 39b receives the input YUV
Frequency conversion of data.

The quantizing unit 40b performs frequency-converted DCT.
The data is quantized, and the variable length coding unit 44b performs variable length coding on the quantized data and outputs a compressed code. On the other hand, the inverse quantization unit 43b inversely quantizes the quantized block, and the inverse DCT unit 42b inversely transforms the inversely quantized data to create a reference frame.

In the case of a P picture, the second compression control unit 35
The control unit 35b stores the data to be compressed in the buffer 34.
If size conversion is necessary, the data is passed to the size conversion unit 37b. The size converter 37b converts the data into an image size to be compressed.

The motion retrieving unit 38b is a unit for storing the image data to be compressed stored in the buffer 34 or the size converting unit 37.
The block having the highest correlation among the blocks of the previous frame stored as the reference frame for each block of the image data subjected to the size conversion by b is calculated, and the difference data with the block of the previous frame having the highest correlation is calculated. I do.

The DCT unit 39b converts the difference data into a spatial frequency, and the quantization unit 40b quantizes the frequency-converted DCT data. The variable length coding unit 44b performs variable length coding on the quantized data and outputs a compressed code.

On the other hand, the inverse quantization unit 43b inversely quantizes the quantized block, and the inverse DCT unit 42b inversely transforms the inversely quantized data. The motion compensation unit 41b adds the inverse DCT unit 42b
To create a reference frame by adding the difference data output from the.

In the case of a B picture, the second compression control unit 35
The control unit 35b stores the data to be compressed in the buffer 34.
If size conversion is necessary, the data is passed to the size conversion unit 37b. The size converter 37b converts the data into an image size to be compressed.

The motion search section 38b is a section for storing image data to be compressed stored in the buffer 34 or a size conversion section 37.
b, the block having the highest correlation among the blocks of the previous / next frame stored as the reference frame for each block of the image data subjected to the size conversion by b is determined. Calculate difference data.

The DCT unit 39b converts the difference data into a spatial frequency, and the quantization unit 40b quantizes the frequency-converted DCT data. The variable length coding unit 44b performs variable length coding on the quantized data and outputs a compressed code. In the case of a B picture, there is no need to create a reference frame.

Next, the overall operation of the present embodiment will be described in detail with reference to FIGS. FIG. 3 is a flowchart of the capture compression. First, the capture compression application 32
Determines whether there is capture data in the capture unit 31 (step S1), and if there is no capture data, returns to step S1.

If there is capture data in the capture section 31, the capture compression application 32 passes the capture data to the first compression control section 33 (step S2). The first compression control unit 33 executes a process described later. Next, capture compression application 3
2 judges whether or not the first compression control unit 33 has the processed compression code (step S3). If there is no compression code, the process proceeds to step S5.

The capture compression application 32
If there is a compression code in the first compression control unit 33, this compression code is written in a compressed file (step S4). First
If there is no processed compression code in the compression control unit 33, or after the end of step S4, the capture compression application 32 determines whether or not capture has ended (step S5), and if not, returns to step S1. If the capture has ended, the capture compression processing ends.

FIG. 4 is a flowchart showing the operation of the first compression control unit 33, and FIG. 5 is a flowchart showing the operation of the second compression control unit 35. First compression control unit 33
And the operation of the second compression control unit 35 are executed in parallel.

First, the control unit 33 of the first compression control unit 33
a determines whether or not the buffer 34 is empty (step S11), and when the buffer 34 is not empty, notifies the second compression control unit 35 that processing is skipped (step S12). When the buffer 34 is free, the control unit 33a of the first compression control unit 33 executes preprocessing of the input image described later (step S13).

Subsequently, the control section 33a increments the input counter by 1 (step S14), and the second compression control section 35
Is notified that the input data has been received (step S
15). On the other hand, the control unit 35b of the second compression control unit 35
Determines whether there has been a notification from the first compression control unit 33 (step S21), and if not, returns to step S21.

The control unit 35b of the second compression control unit 35
When there is a notification from the first compression control unit 33, it is determined whether to skip (step S22). When a notification of a skip is received from the first compression control unit 33, the control unit 35b executes a skip process described later (step S23).

Then, the control unit 35b determines whether or not the capture is completed (step S24). If the capture is not completed, the process returns to step S21, and if the capture is completed, the process is completed.

If not skipped in step S22, the control section 35b executes image compression, which will be described later (step S25). Then, the control unit 35b decrements the input counter by 1 (step S26), determines whether or not the capture is completed (step S27). If the capture is not completed, the process returns to step S21, and if the capture is completed, the process is completed. .

FIG. 6 is a flowchart of the input image preprocessing (step S13) described above. The control unit 33a in the first compression control unit 33 determines whether the input image data captured by the capture unit 31 is RGB data (Step S51).

When the input image data is RGB data, the control section 33a outputs the input image data to the YUV conversion section 36a in the first compression control section 33. The YUV conversion unit 36a converts the input image data into YUV data and passes it to the control unit 33a. The control unit 33a stores the YUV data in an empty area of the buffer 34 (Step S52).

The control unit 33a includes the capture unit 31
If the input image data captured is not RGB data but YUV data, it is determined whether the image size of the input image data is different from the image size to be compressed (step S53).

If the image sizes are different, the controller 33a
Outputs the input image data (YUV data) to the size conversion unit 37a in the first compression control unit 33. The size conversion unit 37a converts the size of the YUV data and
Hand over to 3a. The control unit 33a stores the converted YUV data in an empty area of the buffer 34 (Step S54). In MPEG compression, the image size is 1
Since the width or height of the input image data is not a multiple of 16, since the width or height of the input image data is not a multiple of 16, the input image data is enlarged or reduced to be a multiple of 16.

When the size conversion is not performed, the control unit 33a
It is checked whether the frame to be compressed is an I picture (step S55). If the picture is an I picture, the control unit 3
3a outputs the input image data (YUV data) to the DCT unit 39a in the first compression control unit 33. DCT part 3
9a frequency-converts the YUV data and passes it to the control unit 33a. The control unit 33a stores the frequency-converted YUV data in an empty area of the buffer 34 (Step S).
56).

When the frame to be compressed is not an I picture, the control unit 33a does not convert the input image data (YUV data) into a frequency,
Is copied to an empty area (step S57).

FIG. 7 shows the above-described image compression (step S2).
It is a flowchart figure of 5). Second compression control unit 35
The control unit 35b determines whether the data to be compressed stored in the buffer 34 has been subjected to the size conversion in the preprocessing (step S13) and whether the size conversion is not necessary (step S61). ).

When the size conversion has been completed or when the size conversion is not necessary, the control unit 35b does not execute the size conversion. When it is necessary to convert the size of the data to be compressed stored in the buffer 34, the control unit 35 b extracts the data to be currently compressed from the buffer 34, and stores the data in the second compression control unit 35. Output to the size conversion unit 37b. The size conversion unit 37b converts the size of the input data and passes it to the control unit 35b (step S62).

Next, the control unit 3 in the second compression control unit 35
5b judges the picture type of the data to be compressed stored in the buffer 34 or the data whose size has been converted by the size conversion unit 37b (step S63).
Compression is performed for each picture.

In the case of an I picture, the control unit 35b determines that the data has been subjected to the frequency conversion (D
It is determined whether or not (CT) has been performed (step S64). If the frequency has been converted, the frequency-converted DCT data is output to the quantization unit 40b in the second compression control unit 35.

If the data has not been frequency-converted, the control section 35b outputs the data to the DCT section 39b in the second compression control section 35. The DCT unit 39b frequency-converts the YUV data input from the control unit 35b and passes the YUV data to the quantization unit 40b (step S65).

Next, the quantizing unit 40b quantizes the frequency-converted DCT data input from the control unit 35b or the DCT unit 39b, and transmits the quantized DCT data to the variable-length encoding unit 44b in the second compression control unit 35. The data is passed to the inverse quantization unit 43b (step S66). The variable length coding unit 44b includes the quantization unit 40
Variable-length encoding is performed on the quantized data output from b (step S67). The compressed code output from the variable length coding unit 44b is passed to the capture compression application 32 via the control unit 35b.

On the other hand, the inverse quantization section 43b is
The quantized data output from b is inversely quantized and passed to the inverse DCT unit 42b in the second compression control unit 35 (step S68). The inverse DCT unit 42b performs inverse frequency transform on the inversely quantized data output from the inverse quantization unit 43b to create a reference frame (Step S69).

Then, the control unit 35b releases the area of the buffer 34 in which the data to be processed this time has been stored, so that new data can be stored (step S81).

In the case of a P picture, the second compression control unit 35
The motion search unit 38b searches for the motion of the macroblock between the reference frame and the data to be compressed stored in the buffer 34 or the data whose size has been converted by the size conversion unit 37b, and calculates the difference from the previous frame. Is calculated, and the calculated difference data is stored in the second compression control unit 35 as D
Output to the CT unit 39b (step S70).

The DCT unit 39b frequency-converts the difference data input from the motion search unit 38b, and
(Step S71). The quantization unit 40b uses the DCT
The frequency-converted DCT data input from the unit 39b is quantized and passed to the variable-length coding unit 44b and the inverse quantization unit 43b in the second compression control unit 35 (step S7).
2).

Next, the variable length coding unit 44b performs variable length coding on the quantized data output from the quantization unit 40b (step S73). The compressed code output from the variable length coding unit 44b is passed to the capture compression application 32 via the control unit 35b. On the other hand, the inverse quantization unit 4
3b inversely quantizes the quantized data output from the quantization unit 40b, and outputs the inverse DCT unit 4 in the second compression control unit 35.
2b (step S74).

The inverse DCT unit 42b performs an inverse frequency transform on the inversely quantized data output from the inverse quantization unit 43b,
Is passed to the motion compensator 41b in the compression controller 35 (step S75). The motion compensation unit 41b adds the difference data output from the inverse DCT unit 42b to the macroblock of the previous frame to create a reference frame, and passes the reference frame to the motion search unit 38b (Step S76).

Then, the control unit 35b releases the area of the buffer 34 in which the data to be processed this time has been stored so that new data can be stored (step S81).

In the case of a B picture, the second compression control unit 35
The motion search unit 38b searches for the motion of the macroblock between the reference frame and the data to be compressed stored in the buffer 34 or the data whose size has been converted by the size conversion unit 37b. Is calculated, and the calculated difference data is output to the DCT unit 39b in the second compression control unit 35 (step S77).

The DCT section 39b frequency-converts the difference data input from the motion search section 38b, and
(Step S78). The quantization unit 40b uses the DCT
The frequency-converted DCT data input from the unit 39b is quantized and passed to the variable length coding unit 44b in the second compression control unit 35 (step S79).

The variable-length coding unit 44b includes a quantization unit 40b
Is subjected to variable-length coding (step S80). The compressed code output from the variable length coding unit 44b is passed to the capture compression application 32 via the control unit 35b. Then, the control unit 35b
The area of the buffer 34 in which the data to be processed this time is stored is released so that new data can be stored (step S81).

FIG. 8 shows the skip processing (step S
It is a flowchart figure of 23). Second compression control unit 3
The control unit 35b in 5 judges whether or not the frame to be compressed is a B picture (step S91). If the frame is not a B picture, the control unit 35b regards the previously processed data as the data to be currently processed and described with reference to FIG. Image compression is performed (step S92).

If the frame to be compressed is a B picture, the control unit 35b inserts a code indicating that it is the same as the previous frame into the frame to be compressed (step S9).
3). Simply inserting a code indicating that the frame is the same as the previous frame does not actually perform compression, so that the compression operation can be accelerated.

As described above, in the present invention, the first compression control unit 33 and the second compression control unit 35 perform image compression processing so that the capture compression application 32 does not need to directly participate in compression processing. Aggregate and process.

Conventionally, when compressing to MPEG video,
Since bidirectional prediction is performed, images of a plurality of frames must be left.However, since captured image data is stored at a fixed address, it is necessary to copy the image data to another memory area so that the image data is not overwritten. Increase. On the other hand, in the present invention, the processing of copying image data is omitted by performing a part of the compression processing in the first compression control unit 33, so that the CPU load can be reduced.

[Second Embodiment] FIG. 9 is a flowchart for explaining the operation of a first compression control unit according to a second embodiment of the present invention, and FIG. 10 is a second embodiment of the present invention. FIG. 13 is a flowchart for explaining the operation of the second compression control unit according to the first embodiment. Also in the present embodiment, the configuration as a moving image compression apparatus is the same as that of the first embodiment, and therefore, the description will be made using the same reference numerals as those of the first embodiment.
As in the first embodiment, the operation of the first compression control unit 33 and the operation of the second compression control unit 35 are executed in parallel.

First, the control unit 33 of the first compression control unit 33
a executes the preprocessing of the input image described in FIG. 6 (step S31). Subsequently, the control unit 33a increments the input counter by 1 (step S32), and notifies the second compression control unit 35 that the input data has been received (step S33).

On the other hand, the control unit 35 of the second compression control unit 35
b determines whether or not there has been a notification from the first compression control unit 33 (step S41), and if not, returns to step S41.

When the control unit 35b receives the notification from the first compression control unit 33, the control unit 35b stores the value of the input counter in the buffer 3
It is determined whether the number is larger than a value obtained by multiplying the number of regions by 4 by a constant α (α is an arbitrary decimal value less than 1.0 and is set to an optimum value for each device) (step S42).

When the value of the input counter> the number of areas of the buffer 34 × α is satisfied, the control unit 35b determines that the number of areas of the buffer 34 is not larger than the number of input image data received by the first compression control unit 33. When it is determined that it is sufficient, the skip processing described in FIG. 8 is executed (step S43). Then, the control unit 35b determines whether or not the capture has ended (step S44). If the capture has not ended, the process returns to step S41. If the capture has ended, the process ends.

If the value of the input counter> the number of areas in the buffer 34 × α is not established, the control unit 35b executes the image compression described with reference to FIG. 7 (step S45). Then, the control unit 35b subtracts 1 from the input counter (step S4).
6), and determine whether or not capture is completed (step S4)
7) If the capture has not ended, the process returns to step S41, and if the capture has ended, the process ends.

In the present embodiment, whether or not to skip is determined by comparing the value of the input counter with the number of buffers. However, it may be determined from the time required for compression and the number of compressed frames. In that case, skip when the following expression holds. Compression time × frame rate of image code × β> number of compressed frames (1) In equation (1), β is an arbitrary decimal number of 1.0 or more, and is set to an optimum value for each device. . Alternatively, the number of frames to be skipped may be determined based on the time required for compression, the number of compressed frames, the value of the input counter, and the number of areas in the buffer 34, as in the following calculation example. Skip2 = (Rate− (Frame−Skip1) / Time) × γ1 + (Input / Buffer) × γ2 (2) Here, if the calculation result is a decimal, the result is rounded down, rounded up or rounded off, and an integer is calculated. And Rate is the compression frame rate to be originally processed described in the header of the video code, Frame is the number of frames for obtaining the average compression frame rate (for example, the number of frames of one GOP),
Skip 1 is the number of frames skipped while compressing the number of frames, Time is the time required to compress the number of frames, Input is the current input counter value, Buffer is the number of areas in the buffer 34, γ 1 , Γ2 is an arbitrary decimal number of 1.0 or more, and is set to an optimum value for each device. Skip2 is the next Fr
This is the number of frames to be skipped when compressing a number of frames. In this way, skipping is performed evenly by the number of Skip2.

If the voice is also compressed at the same time,
The determination may be made from the size of the voice code. In that case, skip when the following expression holds. Compression time = size of audio code / bit rate of audio code compression time × frame rate of image code × γ> number of compressed frames (3) In equation (3), γ is an arbitrary decimal number of 1.0 or more. And is set to an optimum value for each device.

[0081]

According to the present invention, by providing the first compression control means and the second compression control means for executing the processing in parallel, the processing can be centralized in the moving picture compression apparatus and the application can be directly executed. A system that compresses moving images by passing captured image data to a moving image compression device without the application having to know the details of the moving image compression method is constructed because the system does not need to be involved in the compression processing. can do. As a result, it becomes easy to incorporate the compression function into the application.

Also, a YUV for converting input image data captured in RGB format into YUV format image data
By providing the conversion means, the data capacity can be reduced as compared with the case of copying in the RGB format, so that the required memory capacity can be reduced.

Further, by performing a part of the compression processing when the first compression control means receives the input image data, the processing of copying the image data can be reduced, and the load on the CPU can be reduced. it can.

Also, without using the input image data or the image data processed by the first compression control means,
By regarding the picture as a frame having no difference and inserting a code indicating that the difference is 0 from the previous frame without actually compressing the picture, the compression process can be prevented from being delayed.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a code format conforming to MPEG video.

FIG. 2 is a block diagram illustrating a configuration of a moving image compression device according to the first embodiment of the present invention.

FIG. 3 is a flowchart illustrating an operation of capture compression by the moving image compression apparatus of FIG. 2;

FIG. 4 is a flowchart for explaining the operation of a first compression control unit in FIG. 2;

FIG. 5 is a flowchart for explaining the operation of a second compression control unit in FIG. 2;

FIG. 6 is a flowchart of preprocessing of an input image by a first compression control unit in FIG. 2;

FIG. 7 is a flowchart of image compression by a second compression control unit in FIG. 2;

FIG. 8 is a flowchart of a skip process by a second compression control unit in FIG. 2;

FIG. 9 is a flowchart illustrating an operation of a first compression control unit according to the second embodiment of the present invention.

FIG. 10 is a flowchart illustrating an operation of a second compression control unit according to the second embodiment of the present invention.

FIG. 11 is a block diagram illustrating a configuration of a conventional moving image compression device.

[Explanation of symbols]

31 capture unit, 32 capture compression application, 33 first compression control unit, 34 buffer, 3
5: second compression control unit, 33a: control unit, 36a: YU
V converter, 37a: size converter, 39a: DCT unit,
35b: control unit, 37b: size conversion unit, 38b: motion search unit, 39b: DCT unit, 40b: quantization unit, 41b
... Motion compensator, 42b inverse DCT, 43b inverse quantizer, 44b variable length encoder.

 ──────────────────────────────────────────────────続 き Continued on front page F term (reference) 5C057 AA06 DA04 EA01 EA02 EA07 EA11 EG08 EL01 EM09 EM13 EM16 GL02 5C059 KK13 MA00 MA23 MC11 MC38 ME01 NN01 PP05 PP06 PP07 PP15 PP16 SS11 SS20 SS30 TA25 TA60 TB04 TC15 UA06 AA03 CA23 EE03 GA01 GA02 GA05 HA01 KE07 5C078 AA09 BA21 CA25 CA31 CA35 DA01 9A001 EE04 HH15 HH30 HH31

Claims (8)

[Claims] A first compression control unit that receives input image data; and a second compression control unit that compresses the image data received by the first compression control unit. The moving image compression apparatus, wherein the means and the second compression control means execute processing in parallel. 2. The moving picture compression apparatus according to claim 1, wherein the input image data in R (red), G (green), and B (blue) format is converted into Y (luminance) U (R-Y color difference) V (B A moving image compression apparatus, comprising: a YUV conversion unit for converting image data into (−Y color difference) format image data. 3. The moving image compression apparatus according to claim 1, wherein said first compression control means performs a part of a compression process when receiving said input image data. apparatus. 4. The moving picture compression apparatus according to claim 3, wherein a part of the compression processing is R (red) G (green) B
(Blue) format of the input image data is represented by Y (luminance) U (R-
(Y color difference) YUV conversion for converting the image data into V (BY color difference) format image data, size conversion for converting the size of the input image data, and frequency conversion for converting the frequency of the input image data. Moving image compression apparatus. 5. The moving picture compression apparatus according to claim 3, further comprising: a buffer having an area for storing a plurality of frames of image data processed by the first compression control means, A moving image compression apparatus, wherein the compression control means stores the processed image data in the buffer, and the second compression control means compresses the image data stored in the buffer. 6. The moving picture compression apparatus according to claim 5, wherein the second compression control means, if the compression processing is delayed, the input image data or the image processed by the first compression control means. A moving picture compression apparatus characterized in that a B picture is regarded as a frame having no difference and compressed without using data. 7. The moving picture compression apparatus according to claim 6, wherein said second compression control means determines that said compression processing is delayed when said buffer is not empty. Image compression device. 8. The moving picture compression apparatus according to claim 6, wherein the second compression control means has a buffer area number that is smaller than the number of input image data received by the first compression control means. A moving image compression apparatus, which determines that the compression processing is delayed when sufficient.