JP2565096B2 - Compressed video / audio signal expansion device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a decompression device for decompressing a compressed video / audio signal into a video reproduction and an audio reproduction which represent an original video signal and an original audio signal and are synchronized with each other according to a compression information table. This compressed video / audio signal is a signal generated by an opposite compression device that compresses an original video signal and an original audio signal that are synchronized with each other according to a compression information table, and includes used compression information as set compression information. There is.

[0002]

2. Description of the Related Art In a compression device of this kind, either such a compressed signal is transmitted to a transmission medium or is stored in a recording medium such as a compact disc read only memory (CD-ROM) or a hard disc. To
It is desirable to compress the original video signal and the original audio signal into a highly compressed video / audio signal according to the compression information table. The compressed signal comprises a compressed video signal component and a compressed signal component which are synchronized with each other. The original video signal comprises the original image and thus represents the original image.

The decompressor establishes synchronization between the video reproduction and the audio reproduction from the compressed signal according to the compressed data table and reproduces the video reproduction signal and the audio reproduction signal representing the original video signal and the original audio signal, thereby recording and reproducing. It is essential when using compressed signals in the device. In the compressed signal, the compressed video signal component is represented by a video code. Video playback is played from video code.

As will be described in more detail later, such a compression apparatus compresses an original image into a number-compressed image according to a table holding arrangement for holding a compression information table indicating image compression information and the image compression information. An image compression arrangement for encoding an input image into a video code, and an audio encoding arrangement for encoding an original audio signal into an audio code representing a compressed audio signal component. There is. In a conventional decompressor, the number compressed image is used to represent the compressed video signal in the video code and is used as the input image. That is, in this case, the feed arrangement feeds the number compressed image to the video coding arrangement as an input image.

In the conventional compression apparatus, the supply arrangement is a discrete cosine transform processing arrangement for subjecting a Nieber compressed image to a discrete cosine transform in order to generate a discrete cosine transform processed image, and a discrete cosine transform processed image is a quantized image. And a quantization arrangement for quantizing. In this way, the feed arrangement feeds as a quantized image to the video coding arrangement as an input image.

The decompressor decompresses the video code representing the compressed video signal component and the audio code representing the compressed audio signal into video reproduction and audio reproduction, respectively. The conventional decompression device has a table holding arrangement for holding the above-mentioned compression information table for instructing the image compression information, a video decoding arrangement for decoding a video code into a number-compressed image, and a video reproduction. An image interpolation arrangement for interpolating the additional image into the number compressed image according to the image compression information and an audio reproduction arrangement for reproducing the audio reproduction from the audio code are provided.

In each of the compression and decompression devices,
Random access memory is used to record various signals that are processed during compression of the original video signal and during decompression of the video code.

[0008]

However, in general,
Random access memory does not process signals fast enough. Moreover, the original video signal and video code are processed at a slower rate than the original audio signal and audio code. This results in a loss of synchronization between the compressed video signal component and the compressed audio signal component and between the video and audio reproductions. For example, suppose the random access memory processes each image in 5 milliseconds. When processing various images at a frame rate of 30 images per second, the random access memory must process each image in about 28 milliseconds. If it takes 50 milliseconds to process each image, the compressed video signal component is generated about 22 milliseconds behind the compressed audio signal component. Moreover, the discrete cosine transform processing arrangement and the quantized array cannot be operated fast enough.

Therefore, the technical problem of the present invention is to provide a compression device capable of processing an original video signal into a compressed video signal component quickly, and to quickly process a video code representing the compressed video signal component into a video reproduction representing the original video signal. It is to provide a stretching device capable of

Another object of the present invention is to represent an original audio signal between a video code and an audio code generated from the original audio signal for representing a compressed audio signal component, and yet to represent the original audio signal. Another object of the present invention is to provide a compressed video / audio signal decompressing device capable of maintaining synchronization between video reproduction decompressed from the audio code and audio reproduction.

[0011]

According to the present invention, the compressed video / audio signal is separately expanded into a video reproduction signal representing the original image of the original video signal and an audio reproduction signal representing the original audio signal. The compressed video / audio signal is a series of punctured frames, and the frame rate compression information and the setting high-frequency component punctuation information are specified, and each of these punctured frames corresponds to the quiescent video code and the quiescent audio signal. In the compressed video / audio signal decompressor when the high frequency components of the original image are compressed, the compression information table showing a plurality of high frequency component spacing information is held. Table holding means and the thinned frames according to the frame rate compression information. An interpolated additional frame to decompress the interlaced video code and the interleaved audio code into a replay video code and a replay audio code, producing a interleaved frame with these replay video code and replay audio code. Interpolation means, audio decoding means for decoding the reproduced audio code into the audio reproduced signal, video decoding means for decoding the reproduced video code into a high frequency component excluded image, and the compression according to the set high frequency component gap information. High frequency component recovery means for recovering the high frequency components of the high frequency component excluded image and outputting a reproduced image according to one selected from the high frequency component spacing information of the information table; and video reproducing means for decoding these reproduced images into the video reproduced signal. A compressed video / audio signal decompressor with .

In compressing the synchronized original video signal and the original audio signal into a compressed video / audio signal including the synchronized compressed video signal component and compressed audio signal component, the original video signal converts the original image. Table holding means for holding a compression information table, video encoding means for encoding the input image into video codes with interspersed, and encoding the original audio signal into an audio code to be the compressed audio signal component In a video / audio signal compression device including audio encoding means, high frequency component compression means for compressing high frequency components of the original image according to high frequency component spacing information and outputting high frequency component compressed images, and the high frequency component compressed images Supply to the video encoding means as the input image Means and a spacing means for spacing the video code according to the frame rate compression information and outputting a spacing video code which becomes the compressed video signal component, wherein the compression information table includes the high frequency component spacing information. A video and audio signal compression device that specifies the frame rate compression information and the frame rate compression information is obtained.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT Referring to FIG. 1, ISO 10918-1
Prior to describing a compressed video / audio signal decompression device according to a preferred embodiment of the present invention for use in a video coding system conforming to (Joint Photographic Coding Experts Group JEPG), this video coding system will be described. Used this compressed video
A video and audio signal compression device, which is a device opposite to the audio signal expansion device, will be described.

Referring to FIG. 2, the compression device has a video signal input terminal 31, an audio signal input terminal 33, a compression signal output terminal 35, and a place for installing a compression device hard disk 37. Although the original video signal is input to the video signal input terminal 31, the original video signal may be an analog signal or a color video signal or a monochrome video signal. In the following, the original video signal is a color analog signal unless otherwise specified. The original audio signal is input to the audio signal input terminal 33, and the original video signal is synchronized with this. A compressed video / audio signal is sent from the compressed signal output terminal 35 to a propagation medium (not shown). The compressed video / audio signal comprises a compressed video signal or a compressed video signal component and a compressed audio signal or a compressed audio signal component which are synchronized with each other.

The compressor comprises a compressor central processing unit (CPU) 39 which controls the video and audio compression operations of the device, and a compressor keyboard 41 which specifies this operation. Compressor Read-only memory (R
The OM) 43 stores in advance a compression information table for designating a plurality of pieces of compression information including high-frequency component gap information and frame rate compression information so as to become clear later.

A variety of video and audio signals are stored in the compressor random access memory (RAM) 4 so that they can be read and stored by the central processing unit 39.
5 is retained as described in detail below. Controlled by the central processing unit 39 and specified by the keyboard 41, the random access memory 45 sends the compressed video / audio signal to the hard disk 37 or the compressed signal output terminal 35 via the transmitter 47. A part of the random access memory 45 thus operates as a frame memory.

Analog-to-digital converter (A / D) 4
9 converts the analog video signal into a digital video signal and stores it in the random access memory 45 as a storage original video signal. The digital video signal, and thus the stored original video signal, represents a series of original images or video continuations during a common, pre-determined duration, such as a frame or field in a television signal. As known to those skilled in the art, each original image of this predetermined duration can be divided into a series of original blocks, each of which typically consists of 8x8 pixels or dots.

Controlled by the central processing unit 39 and the high frequency component spacing information read from the compression device read-only memory (ROM) 43, the DCT (discrete cosine transform processing) unit 51 performs DCT conversion on the original image. Output DCT transformed image. In these DCT transformed images, the high frequency components of the original image are compressed according to the high frequency component spacing information. More specifically, the DCT unit 51 refers to the high frequency component spacing information and compresses the high frequency components of the original image, and outputs the component compressed image. The DCT unit 51 then performs DCT conversion on these component compressed images and outputs a DCT converted image. Therefore, the DCT conversion is performed in a short time.

Similarly controlled, the quantizer 53 quantizes the DCT transformed image and outputs the quantized image. In these quantized images, the high frequency components of the original image are compressed again according to the high frequency component spacing information.
In particular, all high frequency components of the original image may not be compressed or may remain in the DCT transformed image as residual high frequency components. The quantizer 53 first compresses the residual high frequency components by referring to the high frequency component spacing information and outputs a residual component compressed image. The quantizer 53 then quantizes the residual component compressed image into a quantized image. In this way, the quantizer 53 can quantize the DCT transformed image in a short time without widening the quantization step.

The combination of the DCT transform unit 51 and the quantizer 53 operates as a high frequency component compression means for compressing the high frequency component of the original image according to the high frequency component spacing information and outputting the quantized image as a high frequency component compressed image. I understood. For better understanding of the compression device, the high frequency component compression means is shown as a high frequency component (HF) compression unit 55 by a broken line block.

When the analog video signal is a color video signal, the stored original video signal is, for example, luminance Y and red color difference C.
It represents a series of color images of pixels characterized by color components such as r and blue-side color difference Cb, which are divided into color blocks consisting of a luminance block, a red-side color difference block and a blue-side color difference block. In this case, controlled by the central processing unit 39 and specified by the keyboard 41, the DCT unit 51 divides the color image into a luminance image, a red side color difference image, and a blue side color difference image in advance.

Next, the DCT unit 51 processes the color image into a color processed image of a selected color component of the color components selected so as to always have luminance according to the color component compression information read from the read-only memory 43. To do.
Therefore, the color image is processed into the color processed image. Therefore, the color processed image always comprises a luminance image. Depending on what the color component compression information looks like, the color processed image is a red side color difference image. Alternatively, the color processed image is a blue side color difference image.

Then, the DCT unit 51 treats the color processed image as an original image. As a result, the compression device is provided as a part of the DCT unit 51 or a part of the high frequency component compression unit 55, and a luminance / color difference (Y / C) separator 57 is separately drawn as a block of a broken line for convenience of illustration.
You can see that it has. Further, the luminance / color difference separator 57 separates the color image into only the luminance image, into the luminance image and the red side or blue side image, or “separates” into all of the luminance image and the red side and blue side color difference images.
I know what to do. In such a case, it can be seen that the DCT unit 51 temporarily operates as a component DCT processing means for converting a color processed image into a DCT processed image.

The video encoder 59 encodes the quantized image, that is, the high frequency component compressed image into a video code and stores it in the random access memory 45. These video codes are referred to herein as interspersed video codes.

The conventional compression device does not perform high frequency component compression. In this case, the DCT unit 51
And the quantizer 53 operate as a supply means for outputting the high frequency component-containing images and supplying these high frequency component-containing images to the video encoder 59 as input images. In such a conventional compression apparatus, the video encoder 59 encodes the input image as it is into a video code to obtain a compressed video signal.

The audio encoder 61 encodes the original audio signal into an encoded audio signal and stores it in the random access memory 45. This encoded audio signal is
Used as a compressed audio signal, it is represented by a series of spaced audio symbols.

Depending on the high frequency component spacing information, the spacing audio code and the spacing audio code are output as a one-to-one correspondence as a pair. The interleaved video code and the corresponding interleaved audio code are synchronized. Each of the interleaved video codes and one of the corresponding interleaved audio codes will be referred to as an interleaved frame.

Controlled by the central processing unit 39 and the compression information table stored in the read-only memory 43,
The frame skipping unit 63 skips the skipping frames according to the frame rate compression information indicated by the compression information table to make the skipping frames. Each of these gap frames comprises a gap video code and a gap audio code as will be described shortly. Therefore, the intermittent continuity of such interleaved video codes represents a compressed video signal. A similar sequence of thin audio codes represents a compressed audio signal.

Finally, referring to FIG. 1, the stretching device described above will be described. The decompressor has a compressed signal input terminal 65, a video signal output terminal 67, an audio signal output terminal 69 and a place for a decompressor hard disk 71. The compressed video / audio signal is supplied to the compressed signal input terminal 65 via the propagation medium, and is further received by the receiving unit 73.
Is supplied to. The hard disk 71 may be an unused hard disk or the compression device hard disk 37 shown in FIG. 1 in which compressed video / audio signals are recorded.

The decompressor comprises a decompressor central processing unit 75 which utilizes video and audio decompression operations, and a decompressor keyboard 77 which specifies this operation. The decompressor read-only memory 79 stores the above compression information table in advance. The usage of this table is different from that of the compression information table described for the compression device.

Various video and audio signals can be read and stored by the central processing unit 75.
It will be stored in the decompressor random access memory 81 as will be described later in detail. The random access memory 81 is controlled by the central processing unit 75 and specified by the keyboard 77.
The compressed video / audio signal from 3 is stored as a stored video / audio signal. The random access memory 81 partially functions as a frame memory.

Controlled by the central processing unit 75 and the compression information table stored in the read-only memory 79, the frame interpolator 83 interpolates additional frames into the interleaved frames in accordance with the frame rate compression information and reproduces the reproduced video code. The reproduced audio signal alternates with the subsequent sequence. These codes correspond to the interleaved video code and interleaved audio code and are stored in the random access memory 79. The central processing unit 75 reads the reproduced video code and the reproduced audio code separately.

Under such control by the central processing unit 75, the audio decoder 85 decodes the reproduced audio code into an audio reproduced signal and sends it to the audio signal output terminal 69. The audio playback signal represents the original audio signal. Under the same control, the video decoder 87 decodes the reproduced video code into a high frequency component excluded image corresponding to the high frequency component compressed image.

The inverse quantizer 89, which is controlled by the central processing unit 75 and the compression information table, recovers the high frequency components of the high frequency component excluded image according to the high frequency component spacing information and creates a high frequency component containing image. Next, the inverse quantizer 89 inversely quantizes the high frequency component-containing image into an inverse quantized image.

The high frequency components of the original image are generally not yet fully recovered in the high frequency component containing image and thus in the dequantized image. In such a situation, the inverse DCT unit 91 recovers the high frequency component of the inverse quantized image according to the high frequency component spacing information to create a high frequency component recovery image. Next, the inverse DCT unit 91 performs inverse DCT conversion on the high frequency component restored image and outputs a series of reproduced images.

The reproduced image corresponds to the original image, the luminance image, the red color difference image, or the blue color difference image. However, it should be reiterated that the high-frequency components of the original image do not have to be completely restored. In any case, the processing brightness of the inverse quantizer 89 and the inverse DCT unit 91 can be increased.

The combination of the inverse quantizer 89 and the inverse DCT unit 91 operates as high frequency component recovery means for recovering the high frequency component of the high frequency component excluded image according to the high frequency component spacing information and outputting the high frequency component recovery image. I found out that The high frequency component recovery means is shown as a high frequency component recovery unit 93 separately from this combination by a broken line block.

The conventional decompression device also has an inverse quantizer 89 and an inverse DC.
And a T unit 91. The inverse quantizer 89 and the inverse DCT unit 91 do not process the above-mentioned high frequency components, either alone or in combination.

When processing a color video signal, the central processing unit 75 combines the luminance image with the red side color difference image, the blue side color difference image, or the red side and blue side color difference images to correspond to the original image of the color video signal. A composite image. Therefore, it can be understood that the decompression device includes the luminance / color difference combiner 95 shown by a broken line block.

The series of reproduced images represent a digital video signal. Digital-to-analog converter (D / A) 97
Converts this digital video signal into an analog video signal and sends it to the video signal output terminal 67.

Referring again to FIGS. 1 and 2, gap information is carried from the compression information table according to the processing speeds of the decompressor and compressor as will be described next. Some of the device elements of the decompressor and compression device are preferably implemented in software. For example, such high-frequency component compression means and high-frequency component recovery means will be described later.

Now referring to FIG. 3 again and referring again to FIGS. 1 and 2, the compression information table is stored in advance in each of the ROMs 43 and 79 of the compression / expansion device. The compression information table has a plurality of storage units.
In the illustrated example, each storage unit includes first to fourth storage areas each having a length of 1 byte.

The first storage area of each storage unit is for displaying the processing rate PR of the related one of the compression / expansion devices. The processing rate is the rate or speed at which each original image or video period is compressed. In the illustrated example, such processing rates are specified by 1, 2, 3, etc. The processing rate of 1 represents the processing rate of one original image in less than 30 milliseconds. The processing rate of 2 represents the processing rate of one original image between 30 milliseconds and 60 milliseconds. The processing rate of 3 is 6 for each original image.
The processing rate between 0 ms and 90 ms is displayed. The processing rate of 4 (not shown) is 9 for one original image.
The processing rate between 0 and 120 milliseconds is displayed.

The second storage area of the storage unit represents the high frequency component compression information Ch according to the processing rate of each storage unit, and is specified by 1, 2 or the like. The information of 1 represents one-to-one, that is, not omitted. The information of 2 represents a 2-to-1 space. The 4 (not shown) information represents a 4-to-1 space.

The third storage area of the storage unit represents the frame rate compression information Cf according to the processing rate of each storage unit, and is specified and displayed by 0.1 or the like. Information of 0 represents 30.0 frames per second. The information of 1 represents 15.0 intermittent frames per second. The second piece of information (not shown) represents 10.0 skip frames per second. Information 3 (not shown) represents 7.5 frames per second. 4 information (not shown) is 6.0 per second.
Represents an individual frame.

The fourth storage area of the storage unit represents the color compression information Cc according to the processing rate of each storage unit, and the color element of each original pixel is specified by 7 and others. As described above, the color elements are displayed by the luminance Y, the red color difference Cr, and the blue color difference Cb.

The information 7 indicates that the luminance, the red color difference, and the blue color difference are used without omission. Information 5 (not shown) indicates that only the luminance and the blue side color difference are used. The third item of information (not shown) indicates the use of luminance and blue side color components. Information 1 (not shown) indicates the use of luminance only, i.e. the processing of the color video signal as a monochrome video signal.

If the image ponder is used as the high-frequency component compression unit 55 together with the facing section (not shown) used in the decompressor, the compression information table includes one storage area in each storage section. Should be. These areas of the storage unit display the image compression information according to the processing rate of each storage unit, and are specified by 1, 2, 4, etc.

The image compression information of 1 displays the original pixel or original block without omission. The second information indicates the omission of every other original pixel, that is, the original block, in the horizontal direction. The information of No. 4 indicates the omission of every other original pixel, that is, the original block, in each of the horizontal and vertical line directions.

Various compression information is selected from the compression information table, as previously described and later described with respect to the decompression device. In the compressor, such compression information is selected to allow each original image to be processed with a time interval defined by the processing rate. This keeps the synchronization between the equivalent of the skip video code and the skip audio code. Similarly, synchronization is established between video playback and audio playback.

For example, suppose it takes 5 milliseconds to process each original image for the compressor. In this case, each original image must be processed within 28 milliseconds to select the compressed information representing 30.0 interlaced frames per second. If each original image is processed within 28 milliseconds, then compressing the original image is unnecessary. If each original image is processed between 28 and 61 ms, the high frequency component information is 2
Display the interval for one-to-one. As a result, the frame rate compression information displays 15.0 intermittent frames per second. If each original image is processed between 62 and 94 ms, the interstitial frames must be further compressed. The frame rate compression information of 15.0 intermittent frames per second is selected together with the high-frequency component compression information for 2 to 1 intermittent frames.

In the example shown in FIG. 3, the compression information is selected for the compression device as described below. When the processing rate is slower than 30 milliseconds, the high frequency compression information displays one-to-one gaps such as displaying 30 gap frames per second and one-to-one gaps of high frequency components. . For color video signals, all of the luminance and red and blue color differences are processed. Processing rate is 30 milliseconds and 6
When it is between 0 ms, the high frequency component information is a 2: 1 ratio so that 15.0 interlaced frames are displayed every second, and luminance, red side color difference and blue side color difference are all processed. Display the space.

Referring to FIG. 4, the compressed video / audio signal is provided with a standardized format according to ISO 10918-1 as described below. The compressed video and audio signal results from a series of frames, referred to herein as pre-interlaced frames, as described above. Each interspacing frame is a succession of start code SOI (start of image), compressed video code Vf, compressed audio code Af, and end code EOI.
The start code indicates the start of each frame before each interval, and it is 5 digits 1 together with ending H as 0FFD8H.
It is specified by a hexadecimal number. The compressed video / audio code is
It contains compression parameters and will be discussed shortly. The end code indicates the end of the previous frame and the other 5 digit 1
Specified by the hexadecimal number 0FFD9H.

Turning to FIG. 5, the compressed video code Vf is given a modified format modified according to an example of the present invention. In the example shown, the compressed video code consists of a series of information areas or blocks INF, a video start code SOF, and a compressed video area VCOM. The information area will now be described. The video start code indicates the start of the compressed video area and is specified by another 5-digit hexadecimal number 0FF00H. The compressed video area represents the pre-interleaved video code and includes its compression parameters in a standardized format.

In FIG. 5, the information area is a series of application data sub area APP and byte length sub area L.
p, reserved sub-areas Cp and Mp, video compressed data start sub-area Tp and video parameter sub-area P
It consists of ARAM. The application data sub area indicates the start of the information area and is specified by another 5-digit hexadecimal number 0FFE0H. The byte length sub area indicates the byte length of a part of the information area following the application data sub area. In the illustrated example, the byte length sub-area displays 16 bytes. The compressed data start sub-area indicates the start of the video compressed data sub-area and is specified by the 4-digit number 0002H. The parameter sub area includes another reserved sub area RES and displays the high frequency component compression information Qh, the frame rate Ft, and the color component compression information Co.

Still referring to FIG. 6, the compressed audio code Af is provided in another modified format. The compressed audio code consists of a series of application data areas (some identical reference symbols are thus used) APP, a byte length area Lp, reserved areas Cp and Mp, a video / audio compressed data start area T.
p and a parameter area PARAM described below. The application data area is similar to the application data sub area and has the above-mentioned 5-digit hexadecimal number 0FF.
Specified by E0H. The byte length area represents the byte length of a part of the compressed audio code that follows the application data area. In the example shown, the byte length field represents at least 30 bytes. The video / audio compressed data start area represents the start of the parameter area and is specified by another four-digit number 0001H.

In FIG. 6, the parameter area includes a reserved sub area TIME, a frame number sub area Fn, a video / audio information sub area Ft (the same reference symbol is used), another reserved sub area RES1, an audio length sub area. It comprises a region Fs, a compression mode sub-region M, a sampling frequency sub-region Fr, a channel sub-region Ch, yet another reserved sub-region RE2, and a compressed audio region ACOM of at least 1 byte. The number-of-frames sub-region represents the ordinal number given to one of the pre-interleaved frames shown and containing the compressed audio code Af. This ordinal number is serial number 0 (zero) to 7
It is represented by one of the digits 0FFFFFFH.
The serial number is successively given to the pre-spacing frames which have not yet been skipped in the slack frame. The bytes of each sub-region consist of the zeroth (least significant digit) bit to the fifteenth (maximum digit) bit.

When the compressed video code Vf displays the frame rate in its parameter sub-area, the video / audio information sub-area does not need to display the frame rate (0th to 10th bits) but the 14th and the 14th. 1
Link data is displayed by 5 bits. According to Fig. 1, the link data is preceded by only the pre-skipping audio code.
10 "and by" 11 "that both pre-interleaved video and audio codes are linked. The audio length sub-area indicates the byte length of the compressed audio area as a two digit number 00H to FFH (decimal). Of 1
To 256)). 1
The byte length in bytes is the default. The compressed mode sub-region is, for example, an adaptive differential pulse code modulation (ADPCM) of a pre-interlaced audio code by a two-digit number 00H.
Display the compression mode like. The sampling frequency sub-region displays the sampling frequency of the original audio signal. The sampling frequency sub-region is a 2-digit number 05H,
When displaying 04H, 03H and 02H, the sampling frequency is 8.27kHz (default), 1
1.03 kHz, 16.54 kHz, and 22.5 kHz
Hz. The channel sub-regions are represented by 00H for the left monoaural channel, 01H for the right monoaural channel, and 02H for every two stereo channels (default). The compressed audio code represents a pre-interlaced audio code.

Referring now to FIG. 7 and additionally to FIG. 1, the compressor performs the video and audio compression operation as follows. Suppose the compressor is supplied with an analog color video signal and an original audio signal which are synchronized with each other.

In the first compression step CS1, the compressor keyboard 41 displays transmission on or off. When the transmission ON is displayed, the compressed video / audio signal is sent from the transmitter 47 to the compressed signal output terminal 35. When the transmission off is displayed, the compressed video / audio signal is stored in the compressed hard disk 37.

In the second compression step CS2, the keyboard 41 displays the number of compressed frames and the file name. In the third compression step CS3, the compression central processing unit 39 reads the compression information table from the compression device ROM 43.

Meanwhile, the analog color video signal is converted into a digital color video signal in a fourth compression step CS4. In the fifth compression step CS5, the digital color video signal is separated into a luminance signal and a color difference signal by the luminance and color difference separator 51. These signals are stored in the compression device RAM 45 as the original image described above.

In a sixth compression step CS6, the original image is subjected to a video compression step, which will be described in due course. In a little more detail, the compression central processing unit 39
The original image is read next from the RAM 45, and the forward discrete cosine processing unit 51 is made to generate the forward discrete cosine transform processed image, the quantizer 53 is made to generate the quantized image, and the video encoder 59 is made to generate the pre-interlaced video code. . A series of discrete cosine transform processed images include a final block as one final image. It contains the final block as the final image in the series of quantized images. Each such sequence comprises a portion encoded in a pre-interleaved video code, referred to as a block. The compressed video / audio signal contains one final space frame as the final frame.

In the seventh compression step CS7, the central processing unit 39 determines whether the pre-interlaced video code has been finally processed up to the last block. If the final block has not yet been processed, the sixth compression step is repeated.

Meanwhile, the original audio signal is encoded by the audio encoder 61 into pre-interleaved audio codes for storage in the RAM 45, as shown in the eighth compression step CS8. The central processing unit 39 then in a ninth compression step CS9, the frame ponder 63.
To generate a skip frame for storage in the RAM 45.

In the tenth compression step CS10,
Central processing unit 39 checks keyboard 41 to determine if transmission is on or off. If the transfer is on, the central processing unit 39 causes the transmitter 47 to send the compressed video / audio signal from the RAM 45 in the eleventh compression step CS11. At the same time or if transmission is off instead of on,
The central processing unit 39 stores the compressed video / audio signal in the hard disk 37 in the twelfth compression step 12.

If the central processing unit 39 finds in the thirteenth compression step CS13 that the last frame has already been processed, the operation in the video / audio compression mode ends. Otherwise, the fourth compression step and the subsequent compression steps are performed by the central processing unit 39 first.
Iterate until it is determined that the final block has been processed in the compression step of 3.

Referring again to FIGS. 3 to 6, the frame rate compression information Ft of the compressed video / audio signal is
Used in stretching equipment. The high frequency component compression information Qh and the color component compression information Co of the compressed video / audio signal are
Here, it is called the set high-frequency component compression information and the set color component compression information, and one of the color component information in the compression information table for use in the decompressor is the color component set as described below. Selected according to compression information. Similarly, the high frequency component compression information C of the compression information table
One of h is selected according to the set high frequency component information.

Turning to FIG. 8 and with additional reference to FIGS. 1, 3 and 5, the implementation of the sixth compression step CS6 will be described with reference to FIG. As previously mentioned, the central processing unit 39 reads from the RAM 45 one after another the original image or video period of the selected one of the luminance and chrominance signals. Let us assume that the luminance signal has been processed. Color difference signals are processed similarly.

In the first video compression step VC1, the central processing unit 39 checks each original image as will be described later. Second video compression step VC
At 2, the central processing unit 39 determines whether the original image should be maintained as one of the maintenance images for use in the pre-interleaved video code or omitted or skipped as the omitted image. If the original image in question should be omitted, the central processing unit 49
Do not check the pixels of the image that were omitted in the video compression step VC3 of. Subsequently, the first video compression step is repeated to check the next original image.

If the original image is to be maintained as a preserved image, the CPU 39 in turn checks the original block of the preserved image in the fourth video compression step VC4. In the fifth video VC5, the CPU
39 determines whether the original block should be maintained as a maintenance block or omitted or skipped as an omission block. If the original block in question should be skipped, the CPU 39 returns to the fourth video compression step without checking the pixels of the omitted block in the sixth video compression step VC6.

If the original block should be maintained, the CPU 39 processes the sustain block. More specifically, such a maintenance block is compressed into a number compressed image in a seventh video compression step VC7. In the eighth video compression step VC8, the number-compressed image is converted into a forward discrete cosine transform by the discrete cosine transform processor 55. The forward quantizer 57 quantizes such a discrete cosine transform processed signal into a quantized video signal in the ninth video compression step VC9. The video encoder 59 encodes the pre-interleaving video code in the tenth video compression step VC10 and the eleventh video compression step VC11.
In RAM 45. The video compression operation ends.

With reference to FIG. 9 and with additional reference to FIG. 2, the decompressor operates in the video / audio decompression mode as follows. It is assumed that the compressed signal input terminal 65 is supplied with a compressed video / audio signal from the above-mentioned compression device, or the decompression hard disk 71 is a compression device hard disk 37 (FIG. 1) in which a compressed video / audio signal is stored. .

In the first expansion step ES1, the expansion device keyboard 77 displays reception ON or OFF. The reception ON is displayed when the reception unit 73 is to be used. In this case, a new hard disk can be used as the expansion device hard disk 71. When reception is off, the compressed video / audio signal of the hard disk 71 is expanded.

In the second decompression device ES2, the keyboard 77 is operated to display the number of compressed frames of the compressed video / audio signal and the file name by which the compressed frames should be read from the hard disk. The CPU 75 of the decompression device reads the compressed data table from the ROM 79 of the decompression device in the third decompression device ES3,
In the fourth expansion step ES4, it is determined whether or not the keyboard 77 displays reception on.

If reception is on, the CPU 75 stores the compressed video / audio signal in the expansion device RAM 81 as the received video / audio signal from the receiving section 73 in the fifth expansion step ES5. If reception is not on or off, the compressed video / audio signal is the sixth
In a decompression step ES6, the received video / audio signals are stored in the RAM 81 from the hard disk 71.
If desired, the compressed video and audio signals are
In a decompression step ES7 of, the data is further stored in a new hard disk used as the hard disk 71. Suppose the compressed video signal of the compressed video audio signal originates from the luminance signal.

The CPU 75 separates the skipped video code and the skipped audio code of the compressed video / audio signal from the received video / audio signal and reads them. The frame interpolator 83 generates a reproduced video code and a reproduced audio code. Thus controlled by the CPU 75, the audio decoder 85 decodes the reproduced audio code into audio reproduction for sending to the audio signal output terminal 69 in the eighth expansion step ES8.

In the ninth expansion step ES9, the reproduced video code uses a video decoder 87, an inverse quantizer 89, an inverse discrete cosine transform processing unit (inverse DCT) 91 and an image interpolator (HF recovery) 93. To be decrypted. This step will be detailed later.

In the tenth expansion step ES10,
The CPU 75 determines whether or not the pre-interlaced video code has been expanded to one of the pre-interlaced video codes included in the compressed video signal at the end as the final block. If the final block has not yet been processed, the ninth decompression step is repeated.

If the last block has already been processed, the compressed video signal of the luminance signal is expanded into the luminance signal. The compressed video signal of the color difference signal is generated similarly to the color difference reproduction.

In the eleventh expansion step ES11,
In the CPU 75, the luminance and color difference signal combiner 95 couples the luminance reproduction and the color difference reproduction to the digital signal reproduction. If the compressed video signal originates from a monochrome video signal, the tenth decompression step directly provides digital signal reproduction.

In the twelfth expansion step ES12,
A digital-to-analog converter (D / A) 97 converts digital signal reproduction into video reproduction synchronized with audio reproduction. The video reproduction is transmitted to the video output terminal 67.

The CPU 75 uses the thirteenth expansion step ES.
If, at 13, it senses that the final frame has already been processed, the operation of the video audio decompression mode ends. Otherwise, the fourth to thirteenth expansion steps are repeated.

Turning to FIG. 10 and with additional reference to FIGS. 2, 3 and 5, the ninth expansion step ES9 described with reference to FIG. 9 will be described. As mentioned earlier, CP
The U75 successively supplies the video decoder 87 with the reproduced video code corresponding to the pre-interlaced video code in the inter-precessed frame.

In the first video decompression step VE1, the CPU 75 checks whether or not the reproduced image of each preceding frame should be processed. In a second video decompression step VE2, the CPU 75 determines whether the pre-intermittent frame in question contains one of the video blocks to be processed. If there is no preceding frame in question, the CPU 75 does not check this preceding frame in the third video decompression step VE3 and repeatedly executes the first video decompression / expansion step.

If the previous frame is to be held as a holding frame, the CPU 75 checks each playback block in the fourth video expansion step VE4.
In the fifth video expansion step VE5, the CPU 75
Determines whether each playback block should be decompressed. If the play block does not need to be decompressed, this play block is skipped to the sixth video decompression step VE6. The sixth video decompression step VE6 turns to the fourth decompression step.

If the playback block in question is to be retained as a decompressed retained image, the CPU 75 processes the retained image. More specifically, such a retained image is decoded by the video decoder 87 into a high frequency component excluded image in a seventh video decompression step VE7.

The inverse quantizer 89 first converts the high frequency component excluded image into an inverse quantized input image with reference to the high frequency component compression information. Then, the inverse quantizer 89
In the video decompression step VE8, the dequantized input image is converted into the dequantized image.

Compared to the original image used in the compressor, each dequantized image does not transform high frequency components sufficiently. In other words, some of the high frequency components are excluded or eliminated as high frequency components.

Therefore, the inverse DCT 91 first adds some or all of the rejected high frequency components to produce the processor input image. Therefore, the inverse DCT 91 applies the inverse discrete cosine transform to the processor input image to obtain the ninth
In the video decompression step VE9 of FIG.

In this way, the high frequency component recovery arrangement is
The high-frequency component excluding the image is restored with respect to the high-frequency component compression information, and a processor output image as a high-frequency component restoration image is generated. It should be noted that the high frequency components need not be completely recovered. This increases the speed of stretching due to the high frequency recovery arrangement. Ninth
The video decompression step VE9 of step 1 returns to the first video decompression step as long as the preceding frame remains unreproduced in the compressed video signal of the received video / audio signal.

With reference to FIG. 11 and with repeated reference to FIGS. 2, 3 and 5, the decompressor read-only read out various compression information used for the video decompression operation and similar operations described with reference to FIG. To extract and decide from the memory 79, the procedure is as follows. Various variables in FIG. 11 are variables in the decompressor random access memory 81.

The variable r represents the processing rate described with reference to FIG. Another variable Cut ([5], [2], [3])
Is called a Cut variable and represents high frequency component spacing information Ch, frame rate Cf, and color component compression information Co in the compression information table. The first of this cut variable
Represents the processing rate in the compression information table. The second argument, which represents the index j of the compression information table, is not used in the present case. The third argument, in decimal, represents the next step of determining the compression information.

In the first determining step DS1, the decompressor central processing unit 75 uses the received video / audio signal stored in the digital access memory 81 as a header, that is, the parameter of the information area of the compressed video code which is the interleaved video code. Read the sub area. In the second decision step DS2, the high frequency component spacing information C of the header
h is stored as the variable ML. Third decision step DS
In 3, the frame rate Ft of the header is stored as another variable FL. Fourth decision step DS
4, the color component compression information Co of the header is stored as another variable CL.

In the fifth determining step DS5, a processing rate is selected according to the program and stored as the variable r. In the sixth determination step DS6, the central processing unit 75 causes the 0th cut variable Cut ([r],
[J], [0]) determines whether or not the variable ML is exceeded. If the 0th Cut variable is larger,
This variable is newly stored as the variable ML in the seventh determination step DS7.

In the eighth decision step DS8, the central processing unit 75 causes the first cut variable Cut ([r],
[J], [1]) is determined to be larger than the variable FL. If the first cut variable is larger, this variable is newly stored as the variable FL in the ninth determination step DS9. If the 0th cut variable becomes larger in the sixth decision step, the 0th cut variable is changed to the first cut variable and the eighth decision step is immediately performed.

In the tenth determination step DS10, the central processing unit 75 causes the second cut variable Cut.
If ([r], [j], [2]) is smaller than the variable CL, this variable is newly stored as the variable CL in the eleventh determination step DS11. If the second cut variable is not small in the eighth decision step, the first cu
The t variable is changed to the second cut and the tenth determination step is immediately performed.

When the second cut variable is stored as the variable CL in the eleventh determination step, the extraction of the compressed information from the read-only memory 79 ends. That is, the compression information is given by the variables ML, FL and CL.

With reference to FIG. 12 and with additional reference to FIGS. 2, 6 and 11, frame determination, ie FIG.
The first video decompression step VE1 described with reference to 0 will be described. In FIG. 12, the variable n represents the ordinal number given to the frame number sub-region Fn in the parameter region of the compressed audio code Af which primarily represents the interleaved audio code in the received video / audio signal. The other variable i will become clear as the description progresses.

As described with reference to FIG. 10, the interlaced frames of the received video / audio signal are processed one after another. Let's assume that one of these gap frames is being processed as the current frame.

In the first sub-step SS1, the ordinal number of the current frame is stored in the variable n from the decompressor random access memory 81. In preparation for the second substep SS2, the central processing unit 75 divides the variable n by the sum of the variables FL and 1 and calculates the remainder. In the second substep, this remainder is stored in the variable i.

In the third sub-step SS3, the central processing unit 75 determines whether the variable i is equal to 0. If the variable i is not equal to 0, then in the fourth substep SS4 it is decided to process the current frame. If i is equal to 0, the fifth substep SS
In 5, it is decided not to process the current frame. Each of the fourth and fifth substeps returns to the first substep for processing the next frame.

With reference to FIG. 13 again and with reference to FIG. 2 and others, the block determination, that is, the fourth video decompression step VE4 described with reference to FIG. 10, will be described. FIG. 13 is applied to the color components of pixels forming the original image in the video compression step, and is embodied by software that can be called color processing steps.

In the video decompression operation, each pre-spacing frame consists of a series of pre-spacing blocks consisting of compressed pixels of different color components corresponding to the original image. These pre-interception blocks are realized one after another. One of these blocks currently being processed is called the current block.

In the first color processing step CP1, the decompressor central processing unit 75 specifies which of the luminance Y, the red side color difference Cr and the blue side color difference Cb the color component compression information Co of the current block specifies. To judge. Variables c and i
As described below.

In the second color processing step CP2, if the color component is luminance, a decimal number 1 is given to the variable c.
In the third color processing step CP3, if the color component is the red side color difference, a decimal number 4 is given to the variable c.

In the fifth color processing step CP5, the central processing unit 75 stores, in the variable i, the logical sum of the variable c and the variable CL determined during the determination described with reference to FIG. In the sixth color processing step CP6, the central processing unit 75 determines whether the variable is equal to 0.

If the variable i is not equal to 0, it is decided to process the current block in the seventh color processing step CP7. If the variable i is equal to 0, it is determined that the current block is not processed in the eighth color processing step CP8. In either case, the first and subsequent color processing steps are repeated for the next block.

Next, referring to FIG. 14 and additionally referring to FIG. 1 and others, the inverse quantizer 89 performs the inverse quantization processing operation described as the eighth video expansion step VE8 in FIG. 10 as follows. To do. In FIG. 14, the variable ML represents the high frequency component spacing information that is determined next while the determination described with reference to FIG. 11 is performed on the high frequency component spacing information. Now, let us assume that the number of high-frequency components in each original image is constant.

In the first dequantization step DQ1, the decompressor central processing unit 75 determines which of the values 1, 2 and 4 the variable ML represents. According to this value, the high frequency component at the time of inverse quantization is recovered.

In the second dequantization step DQ2,
The value of the high-frequency component gap information is 1. In this case, the inverse quantizer 89 restores a high frequency component up to a fixed number. In the third dequantization step DQ3, the value of the high frequency component spacing information is 2. In this case, the inverse quantizer 89 restores the number of high frequency components to half of the fixed number. In the fourth dequantization step DQ4, the value of the high frequency component spacing information is 4. In this case, the inverse quantizer 89 restores the number of high frequency components to a fixed number of quarters. The first dequantization step and one of the second to fourth dequantization steps are repeated for the next block.

Referring back to FIG. 15 and additionally referring to FIGS. 1, 14 and others, the second dequantization step DQ2 described with reference to FIG. 14 is performed as follows. Variable x,
y, and dd, buffer Buffer ([x],
[Y]) and the coefficient Q ([x], [y]) are used as described below. The buffer here is for inverse quantization. Buffer arguments x and y
Are the variables x and y, which represent the pixels in each block and can vary from 0 to 7.

In the first quantization step DQ1.1,
Use 0 for the variable y. In the second quantization step DQ1.2, 0 is also used as the variable x. Third quantization step DQ
In 1.3, the value of the buffer is used as the variable dd.
In the fourth quantization step DQ1.4, the value of the buffer used for the variable dd is multiplied by the coefficient Q to create a product variable.

In the fifth quantization step DQ1.5,
Store this product variable in a buffer. In the sixth quantization step DQ1.6, 1 is added to the variable x to create a sum x variable.

In the seventh quantization step DQ1.7,
The decompressor central processing unit 75 determines if the sum x variable is less than eight. If the sum x variable is less than 8, the third dequantization step DQ1.3 and below is repeated. If it is not smaller than the sum x variable 8, the sum y variable is created by adding 1 to the variable y in the eighth quantization step DQ1.8. Ninth quantization step DQ
In 1.9, compare the sum y variable with 8. If the sum y variable is less than 8, then the second dequantization step D
Q1.2 and below are repeated. If the sum y variable is not less than 8, the second dequantization step DQ2 ends.

Turning further to FIG. 16, the third dequantization step DQ3 of FIG. 14 will be described in detail. Use variables, buffers and coefficients Q as described above.

First to sixth inverse quantization steps DQ2.
1 to DQ2.6 are the same as the first to sixth dequantization steps DQ1.1 to DQ1.5 in FIG.

In the seventh quantization step DQ2.7,
In the central processing unit 75, the sum variable x has a maximum value x (max
2 [8]). If the sum x variable is not smaller than this maximum value, the sum x variable and 8 are compared in the eighth dequantization step DQ2.8.

If the sum x variable is smaller than 8, 0 is stored in the buffer in the ninth dequantization step DQ2.9. Next, in the tenth quantization step DQ2.10, 1 is added again to the variable x to create a resuming variable. The eighth dequantization step DQ2.8 is performed again using this re-sum variable.

In the eighth inverse quantization step DQ2.8, if the sum x variable or the re-sum variable is not smaller than 8,
The eleventh and twelfth quantization steps DQ2.11 and DQ2.12 are performed similarly to the eighth and ninth dequantization steps DQ1.8 and DQ1.9 in FIG.

Turning further to FIG. 17, the fourth dequantization step DQ4 of FIG. 14 will be described in detail. Variables, buffers and coefficients Q are used as above.

First to sixth inverse quantization steps DQ3.1
To DQ3.6 are the same as the first to sixth dequantization steps DQ1.1 to DQ1.6 in FIG.

In the seventh quantization step DQ3.7,
In the central processing unit 75, the sum x variable is the maximum value x (max
4 [8]). When the sum x variable is smaller or smaller than this maximum value, the eighth to twelfth dequantization steps DQ of FIG.
The same steps from 2.8 to DQ2.12 are the same as the eighth to twelfth dequantization steps DQ3.8 to DQ3.12.
Until done.

Referring again to FIGS. 15 to 17, when the high-frequency component gap information indicates 1: 2, x (max2 [8]) is used as the maximum value, and when the high-frequency component gap information indicates 1: 4. Uses x (max4 [8]) as the maximum value. In the case of FIG. 15, all calculations are 6 without omission.
Do 4 times. In the case of Fig. 16, the number of calculations is (7 + 6 + 5 + 4 + 4 + 3 + 2 + 1) times or 32 times by using the maximum value. In the case of FIG. 17, the maximum value is used to calculate the number of calculations by (5 + 4 + 3 + 2 + 1 + 1 + 0
+0) or 16 times.

Referring now to FIG. 18 and additionally to FIG. 1, the inverse DCT unit 91 is shown in FIG.
The ninth video decompression step VE9 described for 0 is performed as follows. The variable ML is used as in FIG.
It is assumed that the number of high frequency components in each original image is constant.

In the first inverse DCT variable step TS1, the decompressor central processing unit 75 determines which of the values 1, 2 and 4 the variable ML represents. According to this value, the high frequency component in the case of the inverse DCT variable is recovered.

The second inverse DCT conversion step TS2 is when the value of the high frequency component spacing information is 1. in this case,
The inverse DCT unit 91 restores the high frequency component until it becomes a fixed number. The third inverse DCT conversion step TS3 is when the value of the high frequency component gap information is 2. In this case reverse D
The CT conversion unit 91 recovers high frequency components until the number becomes half of a fixed number. Fourth inverse DCT conversion step TS4
Is when the value of the high-frequency component gap information is 4. In this case, the inverse DCT conversion unit 91 recovers until the number of high frequency components becomes a quarter of a fixed number. The first inverse DCT transform step and one of the second to fourth inverse DCT transform steps are repeated for the next block.

The second inverse DCT conversion step TS2 of FIG. 18 will be described with reference to FIG. 19 and FIG. Variable x,
The y, dd and buffer are used as in FIGS. However, the buffer is for inverse DCT conversion. Furthermore, the variables u and v and the area variable t ([v], [u]) are used. The area variable is an area where the variable changes. Also, the coefficient Coeff ([v],
[U]) is used. Depending on the case, the domain variables and Huffers and coefficients are t ([v], [y]) or t ([y],
[U]), Buffer ([y], [u]), Coeff ([v], [y]) or Coeff
([X], [u]) are shown.

In the first substep TS1.1, 0 is used for the variable y. 0 in the second substep TS1.2
Is also used for the variable v. Further, 0 is used for the variable dd in the third sub-step TS1.3. Fourth sub-step TS
In 1.4, 0 is also used for the variable u. In the fifth sub-step TS1.5, the variable dd and the buffer Buf
fer ([y], [u]) and coefficient Coeff
Change the variable dd to the sum of the product of ([v] and [u]).

In the sixth sub-step TS1.6, 1 is added to the variable u to create a sum u variable. 7th sub-step TS
Comparing the sum u variable and 8 in 1.7. What if u
If the variable is less than 8, then from the 7th substep to the 5th
Return to sub-step. If the sum u variable is not smaller than 8, the variable dd is stored in the area variable t ([v], [y]) in the eighth sub-step TS1.8.

In the ninth sub-step ST1.9, 1 is added to the variable v to create a sum v variable. Tenth substep T
Compare the sum v variable and 8 in S1.10. If the sum v variable is less than 8, then the tenth substep returns to the third substep. If the sum v variable is not smaller than 8, the sum y variable is created by adding 1 to the variable y in the eleventh substep ST1.11.

In the 12th substep TS1.12, the sum y variable and 8 are compared. If the sum y variable is less than 8, return to the second substep. If the sum y variable is not less than 8, then the thirteenth substep TS
In 1.13, 0 is used for the variable y. In the 14th sub-step TS1.14, 0 is used for the variable x. In the 14th sub-step T1.14, 0 is also used for the variable x. First
In the 5th sub-step TS1.15, 0 is further used for the variable dd. In the 16th substep TS1.16, 0 is also used for the variable u. 17th sub-step TS
In 1.17, the variable dd is changed to the variable dd and the area variable t.
([Y], [u]) and the coefficient Coeff ([x],
Change to the sum of the product of [u]).

In the 18th sub-step TS1.18, 1 is added to the variable u to create a sum u variable. In the 19th substep TS1.19, the sum u variable is compared with 8. If the sum u variable is smaller than 8, the process returns to the 17th substep.

If the sum u variable is not smaller than 8, the variable dd is added in the 20th sub-step TS1.20.
Is stored in the buffer Buffer ([x], [y]). In the 21st substep TS1.21, the variable x is incremented by 1 to create a sum x variable. Twenty-second substep 1.2
Comparing the sum x variable and 8 in 2. If the sum x variable is less than 8, return to the fifteenth substep. If the sum x variable is not smaller than 8, the sum y is obtained by adding 1 to the variable y in the 23rd sub-step TS1.23.
Create a variable.

In the 24th substep TS1.24, the sum y variable and 8 are compared. If the sum y variable is less than 8, return to the 14th substep. If the sum y variable is not less than 8, then the second inverse DCT transform step ends.

Referring to FIG. 20 and other drawings as well as FIG. 20 (a) and FIG. 20 (b), FIG.
The inverse DCT conversion step TS3 will be described.

The first to sixth sub-steps TS2.1 to TS2.6 are the same as the first to sixth sub-steps TS1.1 to TS1.6 in FIG. 19A.

In the seventh sub-step TS2.7,
Compare the sum u variable and the maximum value u (max2 [y]).
If the sum u variable is smaller than this maximum value, the process returns to the fifth sub-step TS2.5. If the sum u variable is not smaller than the maximum value, the same 8th to 19th substeps TS2.8 to TS1.8 to TS1.19 in the 8th to 19th substeps TS1.8 to TS1.19 of FIG.
2.19 is performed.

In the twentieth substep TS2.20, the variable dd is stored in the buffer Buffer ([x],
[Y]) and Buffer ([x + 1], [y]). Twenty-first substep TS2.21
In, the variable y is added by 2 to make the sum y variable. Then, the same 22nd to 24th substeps TS2.22 to TS2.24 are performed in the 22nd to 24th substeps TS1.22 to TS1.24 of FIG. 19A and FIG. 19B.

The fourth inverse DCT conversion step TS4 of FIG. 18 will be described with reference to FIG. 21A and FIG. 21B and with reference to FIG. 1 and others.

The first to sixth sub-steps TS3.1 to TS3.6 are the same as the first to sixth sub-steps TS1.1 to TS1.6 in FIG. 19A.

In the seventh sub-step TS3.7,
The sum u variable is compared with the maximum value u (max4 [y]). If the sum u variable is smaller than this maximum value, the process returns to the fifth sub-step TS3.5. If the sum u variable is not smaller than the maximum value, the eighth to nineteenth values in FIG.
The same 8th to 19th substeps TS3.8 to TS3.19 are carried out in substeps TS1.8 to TS1.19.

In the twentieth substep TS3.20, the variable dd is stored in the buffer Buffer ([x],
[Y]) and Buffer ([x + 1], [y]) and Bu
offerer ([x], [y + 1]) and Buffer ([x
+1] and [y + 1]). 21st
In sub-step TS3.21, the variable x is incremented by 2 to create the sum x variable. In the 22nd substep TS3.22, the sum x variable and 8 are compared. If sum x variable is 8
If it is smaller than the above, the process returns to the 15th sub-step TS3.15. If the sum x variable is not less than 8, then the second
In 3 substep TS3.23, 2 is added to the variable y to create a sum y variable. Then, the same 24th sub-step TS3.
24.

Referring again to FIGS. 19A to 21B, when the high frequency component spacing information indicates 1: 1, the inverse DCT calculation is 8 × 8 × 8 × 2 times, that is, 10 times.
24 times. When the high-frequency component spacing information indicates 1: 2, the calculation is 8 × (7 + 6 + 5 + 4 + 4 + 3 + 2 +
1) + 8 × 8 × 4 times, that is 512 times. When the high frequency component gap information indicates 1: 4, the calculation is 8 × (5+
4 + 3 + 2 + 1 + 1) + 8 × 4 × 4 times, that is, 256 times.

10 to 21B, selection of color component compression information, frame interpolator 83 and inverse quantizer 8
9, inverse DCT unit 91 and high frequency component recovery unit 9
The software that realizes 3 and 3 is described. Similarly, and partly more easily, selection of other compression information and the frame spacing unit 63, the DCT unit 51, the quantization unit 53, and the high frequency component compression unit 55 can be realized by software.

[0146]

As described above, according to the present invention, the high frequency component spacing information is stored in the compression information table and is selected according to the processing speed of the compression device. Therefore, the video code representing the compressed video signal component is selected. An advantage is that it is possible to provide a decompression device that can process the video reproduction representing the original video signal quickly.

[Brief description of drawings]

FIG. 1 is a block diagram of a compressed video / audio signal expansion device according to an embodiment of the present invention.

FIG. 2 is a block diagram of a compression device opposite the compression video / audio signal expansion device described with reference to FIG.

3 illustrates a compression information table for use with the decompression device shown in FIG. 1 and the compression device shown in FIG.

4 shows the format of a compressed video / audio signal supplied to the decompression device shown in FIG. 1 and produced by the compression device shown in FIG.

5 shows a format of a compressed video signal included in a compressed video / audio signal generated by the compression apparatus shown in FIG.

6 shows a format of a compressed audio signal included in the compressed video / audio signal shown in FIG.

7 shows a flow chart for use in describing the operation of the compression device shown in FIG.

FIG. 8 shows a flowchart for use in detailing a portion of the flowchart shown in FIG.

FIG. 9 shows a flow chart for use in describing the operation of the decompression device shown in FIG.

FIG. 10 shows a flowchart for use in detailing a portion of the flowchart shown in FIG.

11 shows a flow chart for use in describing the determination of various compression information from the compressed data table shown in FIG. 3 for use in the decompressor shown in FIG.

FIG. 12 shows a flowchart for use in detailing a portion of the flowchart shown in FIG.

FIG. 13 shows a flow chart for use in detailing different parts of the flow chart shown in FIG.

FIG. 14 shows a flowchart for use in detailing yet another portion of the flowchart shown in FIG.

15 shows a flow chart for use in detailing one of the steps of the flow chart shown in FIG.

16 shows another flowchart for use in describing the steps shown in FIG.

FIG. 17 shows yet another flowchart for use in describing the steps shown in FIG.

FIG. 18 shows a flowchart for use in detailing yet another portion of the flowchart shown in FIG.

FIG. 19 shows a flow chart for use in detailing one of the steps of the flow chart shown in FIG.

FIG. 20 shows a flow chart for use in detailing one of the steps of the flow chart shown in FIG.

FIG. 21 is another flow chart for use in describing the steps shown in FIGS. 19 and 20.

22 is another flow chart for use in describing the steps shown in FIGS. 19 and 20. FIG.

FIG. 23 is yet another flowchart for use in describing the steps shown in FIGS. 19 and 20.

FIG. 24 is yet another flowchart for use in describing the steps shown in FIGS. 19 and 20.

[Explanation of symbols]

 31 Video Signal Input Terminal 33 Audio Signal Input Terminal 35 Compressed Signal Output Terminal 37 Compressor Hard Disk 39 Central Processing Unit (CPU) 41 Compressor Keyboard 43 Compressor Read Only Memory (ROM) 45 Random Access Memory (RAM) 47 Transmitter 49 Analog / digital converter (A / D) 51 DCT (discrete cosine transform) unit 53 Quantizer 55 High frequency component (HF) compression unit 57 Luminance / color difference (Y / C) separator 59 Video encoder 61 Audio code Optimizer 63 Frame spacing unit 65 Compressed signal input terminal 67 Video signal output terminal 69 Audio signal output terminal 71 Expander hard disk 73 Receiver 75 Expander central processing unit 77 Expander keyboard 79 Expander read-only Memory 81 Decompression device Random access memory 83 Frame interpolator 85 Audio decoder 87 Video decoder 89 Inverse quantizer 91 Inverse DCT unit 93 High frequency component recovery unit 95 Luminance / color difference synthesizer 97 Digital / analog converter (D / A)

Claims (4)

(57) [Claims] 1. When the compressed video / audio signal is separately expanded into a video reproduction signal representing an original image of the original video signal and an audio reproduction signal representing an original audio signal, the compressed video / audio signal is interleaved. It is a series of frames and specifies frame rate compression information and set high-frequency component spacing information, and each of the spacing frames is provided with a spacing video code and a spacing audio signal in synchronization with each other. In the compressed video / audio signal decompressor when the high frequency component of the original image is compressed, table holding means for holding a compression information table showing a plurality of high frequency component gap information, and the frame rate compression information Interpolate additional frames into the skipped frames according to Interpolating means for expanding the skip video code and the skip audio code into a playback video code and a playback audio code to generate a skip frame having the playback video code and the playback audio code, and the playback audio code. Audio decoding means for decoding the reproduced audio signal, video decoding means for decoding the reproduced video code into a high-frequency component exclusion image, and high-frequency component spacing information of the compression information table according to the set high-frequency component spacing information Decompressing a compressed video / audio signal including high frequency component recovery means for recovering high frequency components of the high frequency component excluded image and outputting a reproduced image according to one selected, and video reproducing means for decoding these reproduced images into the video reproduced signal. apparatus. 2. The high frequency component recovery means dequantizes the high frequency component exclusion image according to a first one selected from the high frequency component spacing information of the compression information table according to the set high frequency component spacing information, and dequantizes the high frequency component exclusion image. An inverse quantization means for outputting a digitized image, and an inverse DCT transform of the inverse quantized image according to a second one selected from the high frequency component spacing information of the compression information table according to the set high frequency component spacing information, The compressed video / audio signal expansion apparatus according to claim 1, further comprising an inverse DCT conversion unit that outputs an image subjected to inverse DCT conversion as the reproduction image. 3. When compressing a synchronized original video signal and an original audio signal into a compressed video / audio signal including a synchronized compressed video signal component and compressed audio signal component, the original video signal is used to compress an original image. Table holding means for holding the compression information table,
A video / audio signal compression apparatus comprising: a video encoding means for encoding an input image into video codes in a narrowly spaced manner; and an audio encoding means for encoding the original audio signal into an audio code to be the compressed audio signal component. High frequency component compression means for compressing high frequency components of the original image in accordance with high frequency component spacing information and outputting high frequency component compressed images; and supply means for supplying these high frequency component compressed images to the video encoding means as the input images. And a gap unit that gaps the gap video code according to frame rate compression information and outputs a gap video code that becomes the compressed video signal component, wherein the compression information table includes the high frequency component gap information and the gap information. Video and audio that specify frame rate compression information and Dio signal compression device. 4. High frequency component means, DCT transform means for DCT transforming the original image and outputting a DCT transformed image according to the high frequency component spacing information, and quantizing these DCT transformed images, and according to the high frequency component spacing information. 4. Quantization means for outputting a quantized image that is a high-frequency component compressed image.
Video and audio signal compression device according to.