JP3334255B2 - High efficiency coding device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used in a digital signal recording / reproducing apparatus for recording digital video information data, such as a digital video tape recorder (hereinafter, referred to as a digital VTR). The present invention relates to a high-efficiency encoding apparatus for compressing and encoding the data amount.

[0002]

2. Description of the Related Art FIG. 17 is a block diagram showing a configuration of a conventional high efficiency coding apparatus. In the figure, reference numeral 1 denotes input digital image data, which is a luminance signal (hereinafter, referred to as Y data or simply Y) and two types of color difference signals (hereinafter, two types of color difference signals are referred to as U data and V data, or simply U, V
). 2, the input digital image data 1, that is, Y, U, and V are divided into blocks of 8 pixels × 8 lines (hereinafter, this block is referred to as a DCT block). The shuffling processing circuit 3 for performing the shuffling process in units of the DCT blocks is used to perform discrete cosine transform (hereinafter, referred to as DCT transform or simply DC
DCT that performs operations and outputs DCT coefficients
A transformer 4 is a quantizer that performs a quantization process on DCT coefficients at a predetermined value, 5 is a variable length encoder that performs variable length encoding of quantized data, and 6 is a variable length encoded data. Buffer for temporarily storing, 8 is encoded output data,
Reference numeral 7 denotes a rate control circuit that measures the amount of data input to the buffer 6 and controls the quantization value of the quantizer 4 based on the amount of data.

Next, the operation will be described. The input image digital data 1 is formed by a shuffling unit 2 into DCT blocks each having 8 (lines) × 8 (pixels) as one block in the vertical and horizontal directions.
The data value at the i-th row and the j-th column in the block is represented by f (i, j). Also, i and j are called spatial coordinates.) This DC
Shuffling processing is performed in units of T blocks.
6 in the DCT block subjected to shuffling processing
DCT conversion is performed on the data of the four pixels by the DCT converter 3.

Here, the discrete cosine transform (DCT transform) will be briefly described. This conversion is performed by calculating the following equation.

[0005]

(Equation 1)

The conversion result is obtained as 8 × 8 data blocks as shown in FIG. 18 or FIG.
The data in the u-th row and the v-th column is represented as F (u, v). The data of sequence 0 in the upper left corner of the data block is
Sequences 1 to 63 are referred to as AC coefficient values, and sequences 1 to 63 are referred to as AC coefficient values. Among them, the one with a small sequence number is called a low-frequency coefficient, and the one with a large sequence number is called a high-frequency coefficient. (See FIG. 19) Generally, in the data subjected to DCT conversion, the power spectrum is concentrated on the above-mentioned low-frequency component. Each transform coefficient after the DCT transform is output from the DCT converter 3 in the order of the sequence number shown in FIG. 19 and in order from the lower band side data where the power spectrum is concentrated by a scanning method called zigzag scanning. In FIG. 18, the output order is indicated by arrows. As shown in the drawing, the output is sequentially performed from the sequences 0 to 63.

[0007] The data after the orthogonal transform is quantized by a quantizer 4.
And quantized. This quantization operation is shown in FIG.
Description will be made using 0. In the figure, reference numeral 41 denotes an input terminal of a DCT transform coefficient F (u, v) output from the DCT converter 3, 42 denotes an input terminal to which a quantization table value is input,
44 is a multiplier, 45 is a divider, 46 is a rounding circuit for rounding the output of the division circuit 45 to a predetermined bit width, and 47 is an output terminal.

First, the quantization table value T (u, v) input from the input terminal 42 is multiplied by the rate control variable K output from the rate control circuit 7 by the multiplier 44. As the quantization table values, appropriate numerical values are determined in advance as 8 × 8 numerical value tables as shown in FIG. As in the case of DCT, the quantization value of the u-th row and the v-th column is represented by T
(U, v). The rate control variable K is given by the rate control circuit 7, and the above-described quantization table value T
Multiplication is performed by the multiplier 44 with (u, v).

The DCT transform coefficient F (u, v) input to the input terminal 41 is divided by the division circuit 45 by the above multiplication result. The result of the division is rounded to an appropriate significant figure by a rounding circuit 46, and the resulting numerical value is output from an output terminal 47. This operation can be expressed by the following equation.

[0010]

(Equation 2)

Next, the variable length encoder 5 (hereinafter abbreviated as VLC) will be described. FIG. 22 is a block diagram showing a general configuration of a conventional variable length encoder. In the figure, 51
Is input digital data to the encoding means, and 52 is output digital data. The variable-length encoder 5 itself includes an encoding table 5b and an encoder 5a, and is connected to each other by connection lines 5c and 5d.

Next, the operation of the variable length encoder 5 will be briefly described. The input digital data 51 is transmitted to the encoder 5a.
The encoding table 5b is read via the signal line 5d according to the input data value, and the code is read by the signal line 5c.
Then, an output operation is performed as the output digital 52.

In this case, the code table 5b is set so that the bit length of the code data is variable according to the occurrence probability of the input data 51. That is, assuming that the occurrence probability of the data Di is Pi and the encoded code length at that time is Li, if the occurrence probability in each case is Pj> Pk for the input data Dj and Dk, If the code length is set so that Lj ≦ Lk, the total amount of post-code data can be made smaller than the input data as a whole.

If the input data is a fixed length of 6 bits 0, 1,
0, 20, and 30 with an occurrence probability of 50
%, 30%, 10%, and 10%.
It is assumed that codes are assigned to the respective data in ascending order of probability of occurrence as shown in FIG.

In this case, if the input data string is, for example, 0, 2,
Assuming that 0, 0, 0, 30, 10, 10, 0, 10, 0, and 10, the required total number of bits of input data is 6 bits × 1
Although 0 words = 60 bits, if encoding is performed using the table shown in FIG. 23, encoding can be performed with all 17 bits, and as a result, the total data amount can be reduced.

[0016] The coded data that has undergone the information compression is input to the buffer 6. Here, the encoded data is temporarily stored. On the other hand, the rate control circuit 7 measures the coded data amount stored in the buffer 6 up to that point in time, and outputs the rate control variable K so that the code amount finally becomes an appropriate value. The rate control variable K is connected to the above-described quantizer 4 and performs the above-described operation.

A description will be given of an example in which this situation is used for a digital VTR. For example, consider a case where input video data for one screen is recorded on ten tracks on a tape. As a matter of course, the amount of data contained in one track is equal. Here, when the number of input data is encoded to half of one screen, if the code amount is the amount of five tracks, ideal encoding has been performed so far. However, if there is an amount of data for seven tracks at that time, the remaining half of the image data must be stored in three tracks, and it is impossible to continue the encoding performed so far. Further information compression has to be done. For that purpose, it is necessary to increase the rate control variable K and reduce the data value output from the quantizer 4.

When the rate control variable K is increased, the output value of the quantizer 4 decreases and the number of possible values decreases.
This means that the output numerical value is limited, and the probability of taking a specific numerical value increases. Therefore, in the variable-length encoder 5, a short code is assigned.
As a result, the code amount can be reduced. Similarly, if the code amount at a certain point in time is too small with respect to the target value, conversely, if the rate control variable K is reduced, the code amount will increase. By appropriately operating the rate control variable K, data for one screen can be accurately written on ten tracks.

The data compressed and encoded as described above are sequentially recorded on a tape. By the rate control, video data for one screen (hereinafter abbreviated as frame data or simply a frame) can be accurately written on 10 tracks in the above example.

As described above, in a digital VTR equipped with a conventional high-efficiency coding apparatus, video data coded in DCT block units is read out sequentially from the buffer 6 in DCT block units, and the video data is read on a magnetic tape. Was recorded above. On the other hand, a storage medium represented by a digital VTR requires a special reproduction function such as a high-speed reproduction in addition to a normal recording and reproduction function. In a home magnetic recording / reproducing apparatus (hereinafter, referred to as VTR), several types of variable speed reproduction speeds are required. In particular, double speed reproduction is an indispensable product concept. When high-speed reproduction is performed by a digital VTR having a recording format as described above, video data is intermittently reproduced from the magnetic tape. When a reproduced image is synthesized by using intermittently obtained reproduction data, if the reproduction speed is an integer multiple speed such as 2 ×, only a specific portion (specific position) on the magnetic tape is reproduced, and the reproduction screen is displayed. There is a problem that data of a DCT block at a specific position cannot be read, and appears as noise of a fixed pattern on a reproduced image.

[0021]

Since the conventional high-efficiency coding apparatus is configured as described above, there is a problem that noise of a fixed picture remains on a reproduction screen during high-speed reproduction.
Hereinafter, the cause of this problem will be briefly described.

First, a description will be given of a reproduction state at the time of high-speed reproduction. FIG. 24 shows the relationship between the track pattern and the scanning trajectory of the rotary head at the time of double speed reproduction. As the head configuration, a two-channel combination head arranged 180 ° facing each other as shown in FIG. 25 was used. As a recording format, a case was described in which one frame was recorded on ten tracks. At the time of high-speed reproduction, the rotary head scans across a plurality of tracks, so that a reproduction signal can be obtained only intermittently. This is because only the signal of the same azimuth-recorded portion as that of the scanning head can be obtained at the time of reproduction, and the hatched portion in the drawing is the portion to be reproduced. Therefore, the information recorded at both ends of the track is not reproduced at all, and the video information recorded at that portion is not reproduced at all. As a result, fixed picture noise remains on the screen. This has a fatal effect on the reproduced image. In order to solve such a problem, for example, a method using a movable head or an additional head dedicated to special reproduction is conceivable, but all of these methods cause a significant increase in cost and are not realistic.

FIG. 26 shows a simulation result of double speed reproduction in this example. FIG. 26 is a reproduced image when the drum rotation speed is the same for both recording and reproduction. The portion where video information cannot be obtained because the head scanning locus is fixed is noise of a fixed picture on the screen (gray block). There is a problem that it appears.

The present invention has been made to solve the above problems, and has as its object to provide a high-efficiency encoding apparatus capable of obtaining a good reproduced image.

[0025]

Means for Solving the Problems] According to the invention of claim 1 high-efficiency encoding apparatus, in the high efficiency encoding device for encoding to compress the digital image data, block the image data for each of a plurality of pixels And a shuffling means for outputting the blocks in a predetermined order. An orthogonal transformation means is connected to a stage subsequent to the shuffling means, and performs orthogonal transformation on the block data. .
A quantizing means is connected to a stage subsequent to the orthogonal transform means, and quantizes the transform coefficients obtained by the orthogonal transform means. The quantizing means has a rate control coefficient, which will be described later, as an input, and operates so as to make the quantization step variable according to the input value. A variable length encoding unit is connected to a stage subsequent to the quantization unit, and performs variable length encoding on the quantized data obtained by the quantization unit. The subsequent stage of the variable length coding means is connected to the buffer means temporarily Ru stored variable length coded data by the variable length coding means.
Is connected to the output sequence replaced means downstream of said buffer means, in the order in which the buffer means predetermined perform divided into those belonging to what the high frequency band belonging to the lower band of the orthogonal transform coefficient to the output Perform output . Then, the battery
Code amount and high band in data stored in
The code amount of each code amount is measured and determined according to each measured value.
The rate control coefficient is output to the quantization means .

According to a second aspect of the present invention, there is provided a high-efficiency encoding apparatus according to the first aspect of the present invention, wherein the quantization means in the high-efficiency encoding apparatus according to the first aspect comprises a plurality of predetermined quantization tables. Data stored in the buffer means
The amount of low-band code and the amount of high-band code are measured.
The quantization table type of the quantization means is switched based on a rate control coefficient output according to the measured value .

According to a third aspect of the present invention, there is provided a high-efficiency encoding apparatus, comprising: a shuffling means for dividing image data into blocks for each of a plurality of pixels and outputting the blocks in a predetermined order. It is connected to orthogonal transform means.
The orthogonal transformation means performs orthogonal transformation on the block data. The orthogonal transform unit is connected to a subsequent first quantizer, and the first quantizer performs quantization on the transform coefficient obtained by the first orthogonal transform unit. Further, the first quantizing means has a subsequent first rate control coefficient as an input, and operates such that the quantization step is variable according to the value. A first variable length coding unit is connected to a stage subsequent to the first quantization unit, and performs variable length coding on the quantized data obtained by the first quantization unit. A first buffer means is connected to a stage subsequent to the first variable-length coding means, and temporarily stores the data variably encoded by the first variable-length coding means, and stores the stored data amount. And outputs a first rate control coefficient to the first quantization means according to the measured value. Further, an inverse quantization means is connected to the output of the first quantization means, and performs inverse quantization on the output of the first quantization means. At the subsequent stage of the inverse quantization means, a difference calculation means is connected,
The difference calculation means is also connected to the output of the orthogonal transform means, and calculates and outputs a difference between the output of the orthogonal transform coefficient and the output value of the inverse quantization means. A second quantization means is connected to a stage subsequent to the difference calculation means, and performs quantization on the difference calculation output. The second quantizing means has a second rate control coefficient as an input, and operates so that the quantization step is variable according to the value. A second variable length encoding unit is connected to a stage subsequent to the second quantization unit, and performs variable length encoding on the quantized data obtained by the second quantization unit. A second buffer means is connected to a stage subsequent to the second variable length coding means, temporarily stores the data variably coded by the second variable length coding means, and measures the amount of the stored data. Then, a second rate control coefficient is output to the second quantization means in accordance with the measured value.

The high-efficiency coding apparatus according to a fourth aspect of the present invention is the high-efficiency coding apparatus according to the third aspect of the present invention, particularly corresponding to the order of the orthogonal transform coefficients input to the first quantization means. A plurality of predetermined quantization tables are prepared, and the quantization table type of the first quantization unit is determined in accordance with the code amount value stored in both the first buffer unit and the second buffer unit. This is connected to first quantization table switching means for switching.

A high efficiency coding apparatus according to a fifth aspect of the present invention is the high efficiency coding apparatus according to the third aspect of the present invention, in which the predetermined high efficiency coding apparatus corresponds to the order of orthogonal transform coefficients input to the quantization means. A plurality of quantization tables are prepared and connected to a quantization table switching means for switching the quantization table type of the quantization means according to the type of input data.

A high efficiency coding apparatus according to a sixth aspect of the present invention is the high efficiency coding apparatus according to the third aspect of the present invention, wherein the coding table in which the coded data is described in the second variable length coding means is provided. A plurality of types are provided, and the encoding table is switched according to the rate control coefficient value.

[0031]

According to the first aspect of the present invention, each DCT block is divided into low-frequency data and high-frequency data at a predetermined ratio, and encoded data is output in a predetermined order. The low-frequency side data can be placed at a position on the track where it can be obtained, and stored in the buffer means.
Each code of low band code amount and high band code amount in the obtained data
Signal and outputs it to the quantization means according to each measurement.
The quantization step is switched according to the rate control coefficient
In, it becomes possible to good high-speed playback, the image quality
Can be planned .

According to the second aspect of the present invention, in the first aspect of the present invention, as the operation of the quantization means, the quantization table is adaptively switched in accordance with the amount of data stored in the buffer means, so that a high-quality, high-speed reproduced image can be obtained. realizable.

According to the third aspect of the present invention, the encoding is performed by dividing into an encoding unit that performs encoding equal to the reproduction data acquisition rate at a predetermined high reproduction speed and a residual encoding unit. In addition, it is possible to improve the quality of a high-speed playback image without deteriorating the normal playback image quality.

According to a fourth aspect of the present invention, in the third aspect of the invention, the first quantization operation is particularly performed by adaptively storing the quantization table in accordance with the amount of data stored in the first and second buffer memories. Since the switching is performed, it is possible to improve the quality of the high-speed reproduced image.

According to the fifth aspect of the present invention, the operation configuration is such that the quantization operation (quantization table) is switched according to the type of input data in the third aspect of the invention, so that higher image quality can be achieved.

According to the sixth aspect of the present invention, since the operation of the second variable length encoding means is switched particularly in conjunction with the rate control coefficient in the third aspect of the invention, the encoding efficiency is increased, Since more image information can be recorded by that amount, higher image quality can be achieved.

[0037]

[Embodiment 1] FIG. 1 is a block diagram illustrating a configuration of a high-efficiency encoding device according to the first embodiment of the present invention. In the figure, the same parts as those of the conventional example are denoted by the same reference numerals, and description thereof will be omitted. 9 is an output order changing circuit.

First, the relationship between a video signal that has been subjected to image compression processing using DCT conversion and VLC and the reproduced image quality will be described. As described above, the input video is subjected to DCT transform to an 8 (line) × 8 (pixel) DCT block, the DCT coefficient is quantized, and Huffman encoding is performed through a variable length encoder. Normally, the code lengths of DCT blocks are widely distributed over the entire screen. On the other hand, adjacent DCT block code lengths have similar sizes. This means that the allocation of the code amount is small for a uniform image such as the sky, but the allocation of the code amount is large for an image with a fine pattern.

FIG. 2 shows the distribution of the code amount of the luminance signal in one DCT block when the entire sample image is subjected to data compression processing to 1/5. It can be seen that the code amount after the compression process is widely distributed from 2 bytes to 36 bytes. Also, the recording code amount in one frame must be kept within a certain value, and it is necessary to control the code amount. However, from the viewpoint of reproduction image quality, these image compression processes can control the code amount almost uniformly over the entire screen. It is desired to be done.
For this reason, shuffling processing is required to perform image compression processing uniformly over the entire screen. This operation is also an important factor in determining the high-speed playback image quality because the position of the acquired data on the tape and the image output position on the screen during high-speed playback are related at the same time. A ring was used to remove the effects of the shuffling itself.

Next, the data acquisition situation will be considered again with the example of double speed reproduction. As shown in FIG. 24, at the time of double speed playback,
We can always get the track center. In addition,
The acquisition area is indicated by oblique lines in the figure. The data amount corresponds to 1/2 of the total data amount obtained during normal reproduction.
In such cases, using partially acquired data,
It is important how far the best picture can be displayed.

FIG. 3 shows the relationship between the acquisition rate and the S / N ratio with respect to the total data amount after encoding. This means that the number of transmission sequences of DCT sequence data is one (only DC is transmitted)
Is a plot of the relationship between the code amount and the S / N ratio when the number is changed to 64 (all DCT data is transmitted). In the figure, the horizontal axis represents the data transmission rate of the DCT sequence data (in this embodiment, it substantially matches the data acquisition rate of the reproduced data), and the vertical axis represents the video S
/ N ratio. Note that the data acquisition rate here is 1 using the data actually reproduced during high-speed reproduction with respect to the total data amount of one frame of image data after high-efficiency encoding.
It is the ratio of the reproduction data amount that can be used to compose a frame image. In this embodiment, the output of the reproduction signal is repeated at a period of 20 tracks on the magnetic tape. At this time, the amount of data that can be obtained is about 50% of the total amount of data necessary to compose one frame. (3/10 of one frame is recorded in the first three scans, and 2/10 of the ten tracks on which the next frame is recorded in the remaining two scans. The 3/10 portion and the 2/10 portion are reproduced. Since the portion has a different data position within one frame, the data amount that can ultimately contribute to image restoration is 50% of the sum of the two.)

The curve shown by a square plot in the same figure shows the case where the code amount control is performed at 100% of the recording capacity.
The diamond plot shows a case where the code amount control is performed in advance so that the code amount control becomes 50% of the storage capacity at the time of encoding. In the double speed reproduction in this example, the data acquisition rate is 50
%, In principle, it is possible to reproduce with the image quality up to the portion indicated by the vertical line. (A detailed description of the present invention using this drawing will be described later.)

The operation will be described below with reference to FIG. The digital video data input to the shuffling circuit 2 is 8
After blocking is performed on the (line) × 8 (pixel) DCT block, shuffling processing is performed on a DCT block basis. The DCT converter 3 performs a two-dimensional DCT transform on the DCT block as in the conventional example. The digital data subjected to the DCT conversion is converted into a rate control variable K output from the rate control circuit 7 by the quantizer 4.
Is quantized based on The quantized data is subjected to variable-length encoding by a variable-length encoder 5,
Output to The rate control circuit 7 sequentially measures the amount of data stored in the buffer 6 and outputs a rate control coefficient K to the quantizer 4.

The output order changing circuit 9 outputs all the D data constituting one screen out of the encoded data output from the buffer 6.
In the CT block, an operation of separating the portion from the low frequency side to 50% and the remaining high frequency side 50% is performed.

FIG. 4 is a block diagram showing the configuration of the output order changing circuit according to the first embodiment of the present invention. In the figure, 9
1 is a coded data input, 92 is a block start signal indicating that the input data is the head of the DCT block,
94 is a memory for temporarily storing the high-frequency side data of the encoded data, 95 is a memory for temporarily storing the low-frequency side, and 93 is a memory for storing the input encoded data in the high-frequency side memory 94 or in the low-frequency side memory 95. Input selection switch for selecting whether to store, 9
Reference numeral 7 denotes an output selection switch for selecting whether to output the data output from the high-frequency side memory 94 or the low-frequency side memory 95, and the code is output from the output terminal 98. A control unit 96 controls the input selection switch 93 and the output selection switch 97.

When the control section 96 detects that the input data 91 is the head data of the DCT block by the block start signal 92, the control section 96 first switches the switch 93 to the low-frequency side memory 95 side to transfer the data to the low-frequency side memory. Start storing. When the reception data 91 in one DCT block reaches a predetermined value, the switch 93 is switched to the high-frequency side memory 94 this time. In this way, it is possible to sort the high-frequency data and the low-frequency data in one DCT block data. Note that the block start signal is a signal indicating the head data of the DCT block as described above, but in this example, the block start signal is a signal supplied simultaneously with data from the buffers 6 in all stages.

Thereafter, by performing the same operation for all the DCT blocks constituting one screen, it is possible to divide the data for one screen into high-frequency data and low-frequency data.

Next, the output operation of the output order changing circuit 9 will be described. First, it is assumed that the rotary head for actually recording data on the magnetic tape is at the start end of one track. At this time, the recording data output 98 is first read from the high frequency side memory. Thereafter, when the rotary head continues scanning and enters the center of the track, the switch 97 is connected to the low-frequency memory 95 side and low-frequency data is output. Therefore, low-frequency data is recorded at the center of the track. Next, when the rotary head enters the end portion, the output data is switched again to the high frequency side and recording is performed. These series of operations are controlled by the control unit 96. In order to switch the output switch, a signal for detecting the rotating head scanning state (generally, the drum PG and drum FG signals are used as signals for detecting the scanning state of the rotating head) is naturally required. Is not shown because it is not the essence of the present invention. Similarly, other signals necessary for controlling the memories 94 and 95 include other necessary signals such as a write / read control signal.

When the above operation is performed on all the data constituting one screen, all the low-frequency data constituting one screen are finally recorded at the center of each track on the magnetic tape. The high-frequency data is recorded in the section. FIG. 6 is a recording format diagram of a digital VTR using the high-efficiency encoding device according to the first embodiment of the present invention. FIG. 6 shows data of a low-frequency component and a high-frequency component of the DCT coefficient on a magnetic tape in one frame. Show the arrangement.

The operation in the case where double speed reproduction is performed on a digital VTR having the above recording format (the recording format shown in FIG. 6) will be described. As in the conventional example, the relationship between the track pattern and the scanning trajectory of the rotary head is as shown in FIG. In this case, 20 tracks or 2 tracks
It is possible to obtain all the data of the central portion corresponding to 10 tracks necessary to form one frame from the tracks corresponding to the frames. In the case of a high-efficiency coding method based on orthogonal transform represented by DCT, even if only the low-frequency component is reproduced, the original image can be restored to some extent by setting the high-frequency component to 0. It becomes possible. (However, since high frequency components are not reproduced, the resolution,
Alternatively, the S / N is deteriorated. Therefore, by using the low-frequency components of the DCT coefficients arranged in the center portion of the track, it is possible to substantially restore the data of all the DCT blocks in one frame. Therefore, in the case of the double speed reproduction, the data of all the DCT blocks in one screen is rewritten every scan of the rotary head 5 and the noise pattern of the fixed picture seen in the conventional example can be removed. (However, the S / N of the composite image is about 29 dB as shown in FIG. 3)

FIG. 5 shows the result of a computer simulation of the image output situation when double-speed playback is actually performed with the configuration of this embodiment. The noise pattern of the fixed picture, which is a problem in the conventional example, is completely removed.

Embodiment 2 FIG. In the first embodiment, the method in which the data after the high-efficiency coding based on the DCT transform is divided into a low-frequency component and a high-frequency component at the time of recording and recorded is described.
In this case, the S / N of the composite image is approximately 29 dB,
Although the noise pattern of the fixed picture does not remain, deterioration of the S / N is visually noticeable. In particular, in high-speed reproduction on a relatively low-speed side such as double-speed reproduction, the deterioration of the S / N is particularly anxious. The purpose of this embodiment is to improve the reproduction quality at the time of high-speed reproduction. Hereinafter, the basic concept of the present invention will be briefly described.

FIG. 3 shows the relationship between the data transmission rate and the reproduced image S / N ratio. As shown in this figure, the code amount control is 100% coded with respect to the storage capacity, and 5%
Compared to reconstructing an image using up to 0%,
If the method of restoring an image using data coded so that the code amount is controlled to 50% of the storage capacity from the beginning has the same data transmission rate, good reproduction image quality can be obtained.

Therefore, in order to improve the reproduction quality of a high-speed reproduction image in which the yield of data reproduction data decreases, it is necessary to control the code amount from the beginning with 100% of the storage capacity as shown in the first embodiment. Also, if the code amount is controlled in advance to a small amount with respect to the storage capacity, and this data is arranged in an area where data can be acquired at the time of high-speed reproduction, the reproduction image quality can be greatly improved. In this embodiment, an improvement in the image quality at the time of 2 × speed reproduction, in which the deterioration of the reproduction image quality is particularly noticeable visually, will be described. If it is desired to improve the image quality during double-speed playback, high-efficiency encoding is performed in advance with a code amount of 50%, and if the encoded data is arranged in an obtainable area on the tape, the playback image quality can be greatly improved. become. In particular, in this example, the frame update period is 1 frame and the original image frame distribution is 2 frames, and the visual sensation is also good.

FIG. 7 is a block diagram showing a configuration of a high-efficiency encoding apparatus according to the second embodiment of the present invention. In the figure,
The same parts as those in the conventional example are denoted by the same reference numerals, and description thereof will be omitted. 400 is an inverse quantizer; 410 is a subtractor for subtracting the output of the inverse quantizer 400 from the output of the DCT transformer 3;
Is a second quantizer, 50 is a second variable length encoder, 60
Is a second buffer, and 190 is a second output rearranging circuit.

First, the operation of the high efficiency coding apparatus shown in FIG. 7 will be briefly described. In the figure, a first rate control block in the upper part serves as a processing unit for creating compressed image data used when restoring an image at the time of high-speed reproduction, and the code amount is controlled so as to be within the data acquisition rate. On the other hand, the second rate processing block at the bottom is a residual data encoding unit, and the code amount is controlled so that the data amount falls within the remaining recordable capacity. Then, as described above, the data output from the first rate control block and used for decoding at the time of high-speed reproduction is recorded in the data obtainable area on the magnetic tape, and the residual data is recorded in other free areas. I do. By arranging the respective data in this manner, good high-speed reproduction can be performed.

The operation will be described below with reference to FIG. The digital video data input to the shuffling circuit 2 is 8
After blocking is performed on the (line) × 8 (pixel) DCT block, shuffling processing is performed on a DCT block basis. The DCT converter 3 performs a two-dimensional DCT transform on the DCT block as in the first embodiment. The digital data subjected to the DCT transform is quantized by a quantizer 4 based on a rate control variable K output from a rate control circuit 7. The quantized data is subjected to variable-length encoding by a variable-length encoder 5 and output to a buffer 6. The rate control circuit 7 sequentially measures the amount of data stored in the buffer 6 and outputs a rate control coefficient K to the quantizer 4. Note that the quantizer 4 in FIG.
The first rate control block composed of the variable length encoder 5, the buffer 6, and the rate control circuit 7 is functionally the same as that of the first embodiment except that the code amount control is performed with respect to the storage capacity. Are identical. In the first embodiment, the code amount of one frame is controlled so as to be equal to the storage capacity, whereas in the second embodiment, the code amount of one frame is controlled in order to improve the reproduction quality at double speed. Control 50% of recording capacity
The description will be continued below.

The output from the first rate control block is such that the coded data thus obtained becomes data used at the time of high-speed reproduction, and is output from the output terminal 8. Output terminal 8
Is output to a track position that can be obtained during high-speed playback. For example, at the double speed, the center of the track can be reproduced, so that the track is recorded in that part.

On the other hand, the inverse quantizer 400 inversely quantizes the output of the quantizer 4. That is, the quantized data is input, and is multiplied by the quantization step table value used for quantization and the rate control value to return to the original DCT coefficient value. However, since it has been once quantized by the first quantizer 4, it does not completely return to the original DCT coefficient value.
This occurs because the division result is rounded by the rounding circuit 46 in the quantizer 4.

The subtracter 410 subtracts the output of the inverse quantizer 400 from the output of the DCT transformer 3 to calculate the above-mentioned error component (hereinafter referred to as a residual component or simply a residual). In the second rate control block for the residual signal output from the subtractor 410, the second quantizer 40,
The second variable length encoder 50, the second buffer 60, the second
Compression coding is performed by the rate control circuit 70 of FIG. The series of processing itself is exactly the same as the above-described first rate control block, and thus description thereof is omitted. However, the code amount control of the residual component is performed by subtracting the code capacity for high-speed reproduction from the recordable capacity. Is controlled such that the amount of In the second embodiment, the data acquisition rate at the time of double speed reproduction is 50% because the improvement of the reproduction image quality at the time of double speed reproduction is considered, so that the rate control in the second rate control block is 50% of the recording capacity. %.

The second output order permutation circuit 190 includes a compressed image data used for restoring a high-speed reproduction image output from the first rate control block, and a residual image output from the second rate control block. When the rotary head is at the start end of the track, the residual signal is read out. After that, when the rotary head continues scanning and enters the center of the track, the compressed image data is read out. Therefore, the compressed image data is recorded at the center of the track. Next, when the rotary head enters the end portion, the output data is switched to a residual signal and recording is performed.

When the above operation is performed for one frame period, finally, the compressed image data constituting one frame is recorded in the center portion of the track on the magnetic tape (reproduction data acquisition area at the time of double speed reproduction), and the residual The encoded data of the signal is at both ends on the track (data non-acquisition area during double-speed playback)
Will be recorded.

The encoding is performed as described above, and the compressed image data and the residual signal are divided into a data acquisition area and a non-acquisition area during high-speed reproduction, and recording is performed. The noise pattern of the picture can be completely removed. In particular, when the compressed image data is arranged in the center of the track as shown in the second embodiment, it is necessary to form an image for one frame within the scanning period of the rotary head 5 at the time of double speed reproduction as in the first embodiment. Since all the compressed image data recorded at the center of the track can be reproduced, the noise pattern of the fixed picture can be removed.

On the other hand, in the normal reproduction, all information can be obtained. Therefore, by decoding the high-speed reproduction data and the residual difference data together, a higher-quality reproduced image can be obtained. Note that the decoding process is the reverse of the encoding process, and a detailed description thereof will be omitted.

When the processing is performed by the method of the first embodiment, as shown in FIG. 5, the double-speed reproduced image has blurs and block-like distortions in the video outline. The S / N ratio in this case is about 29 dB. FIG. 8 is a diagram showing an image by simulation for explaining the operation of the second embodiment of the present invention, and shows a case where encoding is performed with a code amount of 50% and double-speed reproduction is similarly performed. It can be seen that the blur and the block-like distortion of the outline in FIG. 5 are greatly improved. At this time, the S / N ratio was about 35 dB, and an improvement effect of about 6 dB was obtained. In the second embodiment, the circuit scale is increased as compared with the first embodiment, but a significant improvement in image quality during high-speed reproduction is achieved. These are the code amount shown in FIG.
This is consistent with the contents shown in the graph of the N ratio.

Embodiment 3 FIG. FIG. 12 is a block diagram illustrating a configuration of a quantizer according to the third embodiment of the present invention. In the drawings, the same reference numerals are assigned to the same parts in the conventional example and other drawings,
Description is omitted. 420 is the two quantization tables 42
421, a switch for switching the output of 423 and 423
Is an input terminal of a control signal for controlling the switch 420;
2 and 423 are quantization tables.

First, the concept of the third embodiment will be described. FIGS. 9 and 10 are diagrams for explaining the first quantization and the second quantization of the third embodiment. FIG. 3 (using the quantization table in FIG. 9) and FIG. 11 (using the quantization table in FIG. 10) respectively show the relationship between the code amount and the reproduction image quality when the encoding of the first embodiment is performed using both tables. ). As is clear from the figure, changing the quantization table value changes the tendency of the curve in the graph. (At this time, the S / N during normal reproduction also slightly changes.) In particular, by changing the quantization table,
Comparing the point that the acquisition rate of the reproduced data is 50%, the S / N differs by about 2 dB. (It goes without saying that this curve differs depending on the input image.) This corresponds to the image quality difference at the time of the double speed reproduction (at the time of the 50% acquisition rate). To be extreme, this means that the image quality at the time of double speed reproduction is different even if the image quality at the time of normal reproduction is the same.

The tendency of the curve in the above graph slightly changes depending on the contents of the input video. However, the ratio of the high-frequency data to the low-frequency data is controlled so that the normal reproduction image quality is kept substantially equal. This indicates that the image quality of the reproduced image at the time of double speed reproduction can be further improved.

Therefore, in order to improve the reproduction image quality at the time of high-speed reproduction in which the data acquisition rate is reduced, it is necessary to select a quantization table that can increase the S / N in a portion where the data acquisition rate is small. Therefore, in order to improve the reproduction image quality at the time of high-speed reproduction, a plurality of quantization tables are prepared on the encoder side, and the plurality of quantization tables may be switched according to the characteristics of image data. This is because the code amount control uses a value obtained by multiplying the quantization table value by a constant coefficient value (rate control value) as the quantization step as described above, and the step ratio between the low band and the high band is one type of quantization. This is because there is no change when a table is used. If encoding is performed by changing the ratio between the low band and the high band of the quantization step, the high-speed reproduction image quality can be increased even if the total code amount and the normal reproduction image quality are equal.

Therefore, in this embodiment, for example, a plurality of types of quantization tables are prepared in advance in the quantization means, and the rate control circuit 7 detects the amount of high-band encoded data and the amount of low-band encoded data. Then, a quantization table switching signal may be output to the quantizer 4 so as to switch the quantization table appropriately according to the code amounts of both signals.

FIG. 12 shows an embodiment of the quantization table switching operation in the quantizer 4. In the drawing, the same reference numerals are given to the same portions in the conventional example and other drawings, and the description is omitted. Hereinafter, only the difference between the configuration and operation of the quantizer 4 will be described. As for the quantization table, the quantization table T1
It has two types, (u, v) and T2 (u, v), and uses the changeover switch 420 to select which to use for quantization. The changeover switch 420 operates according to the quantization table changeover signal 421. In this example, as described above, the transmission sequence amount of the low-frequency data obtained from the rate control circuit 7 and the Is generated according to the transmission sequence amount.

Even if there is only one type of quantization table, for example, a high-frequency rate control variable K1 and a low-frequency rate control variable K2
Is given to the quantizer 4 individually, but the same effect is obtained.
Needless to say, this operation is basically equivalent to switching and using a plurality of quantization tables.

Embodiment 4 FIG. Although the third embodiment has been described by taking the improvement of the first embodiment as an example, it goes without saying that completely the same improvement is possible in the second embodiment. That is, the quantizer 4
Where the quantization table is used, the total code amount is the same, and even if the normal reproduction image quality is the same, the code amount and the reproduction S /
The image quality differs during special reproduction in which the reproduction operation is partially performed because the curve shape of the N ratio is different. Therefore, if a plurality of quantization tables are prepared in advance and adaptively switched according to the encoding situation, special reproduction with higher image quality can be performed.

In this case, the switching signal of the quantization table can be obtained by measuring the respective code amounts in the first buffer 6 or the second buffer 60 and adaptively switching according to the value.

Embodiment 5 FIG. As a modification of the fourth embodiment, for example, when the quantization table type is further switched and quantized for each of the Y, U, and V components, higher image quality reproduction is possible.

FIG. 13 shows the residual code amount for each of the Y, U, and V components when the rate control value is changed, and the normal reproduction image quality at that time. In this case, it can be seen that the change in the quantization step and the change in the code amount and the quality of the reproduced image are different for each of the Y, U, and V components.

This means that, even when the total code amount of Y, U, and V is the same when compressing to a predetermined code amount, a higher quality, high-speed reproduction image can be output if the allocation amount is properly selected. It is shown that. Image quality is Y, U, V
This is because it is not necessarily a good image even if the S / N ratio is high only for the Y component since it is determined by the balance of each component.

In this example, for example, the rate control value is set to 6 for Y.
By quantizing 0, U, and V as 64, respectively, high-speed reproduction with higher image quality can be realized while the code amount as a whole is kept constant.

As an embodiment, as an example, Y,
The U and V identification signals may be input to the quantizers 4 and 40, and the other components may be configured so that the tables can be switched in the same manner as the quantization table switching means of the fourth embodiment. is there.

Embodiment 6 FIG. In the third and fourth embodiments, it is possible to further improve the image quality by making the encoding table used by the variable length encoding means variable.

FIGS. 14 and 15 are diagrams for explaining the operation of the sixth embodiment of the present invention, and FIG. 14 shows the residual code amount when a sample image is coded. FIG. 15 shows the code amount when variable length coding is performed using different coding tables. The difference between the two code amounts indicates that the code amount can be reduced by appropriately changing the coding table. That is, if the total code amount is constant, the recording information can be further packed as much as the code amount can be compressed, and as a result, the reproduction image quality including the special reproduction image quality can be improved.

In particular, the residual component greatly depends on the operation of the first and second quantizers 4 and 40 on which the calculation is based. That is, if the quantization step of the first quantizer 4 is large, the residual component naturally has a numerically wide value. The same can be said for the quantization step of the second quantizer 40.

Therefore, if this encoding table is switched according to the quantization state of the first or second quantizer, a higher quality reproduced image can be obtained as a result.

As an actual embodiment, it can be easily realized by a method according to the method of switching the quantization table. FIG. 16 is a block diagram showing a configuration of a variable length encoder of a digital VTR using a high-efficiency encoder according to Embodiment 6 of the present invention. In the figure, 5e is the output of the encoding table 5b, 5f is the output of the encoding table 5g, 5
g is a switch for switching between the output of the first coding table 5b and the output of the second coding table, and 5h is a switch for switching the output of the second coding table.

Next, the operation will be described. As for the inputted digital data 51, the encoder 5a reads the input data value and stores the value in the first encoding table 5b and the second encoding table 5b.
Is output to the encoding table 5g. Encoding table 5
In b and 5g, a variable length code corresponding to input data is output to signal lines 5e and 5f. The switch 5h receives the quantization table switching signal 421 output from the rate control circuit 7, and switches the output of the encoding table 5b and the output of the encoding table 5g based on this signal.

Thus, the variable length encoders 5 and 5
By operating 0, image compression can be performed efficiently and the reproduction quality of a high-speed reproduction image can be improved.

Since the control of the changeover switch of the encoding table is appropriately switched by the operation of the quantizers 4 and 40, the control may be performed, for example, according to a rate control variable signal, and the same effect is obtained. Needless to say,

Embodiment 7 FIG. In the above embodiment, the buffers 6 and 60 and the output order changing circuits 9 and 19 are used.
Although 0 is provided, the present invention is not limited to this, and a similar effect can be obtained even if the memory arranged in the output order rearranging circuit 9 or 190 is also used as the buffers 6 and 60 to reduce the circuit scale.

Embodiment 8 FIG. In the above embodiment, the case where DCT transform is used as one embodiment of the high efficiency coding method has been described. However, the present invention is not limited to this, and high efficiency based on orthogonal transform such as Hadamard transform and KL transform is used. The important components of image information and residual information are also magnetically encoded by the coding method, the ADRC conversion, the high-efficiency coding method based on DPCM, or the high-efficiency coding method based on sub-band or wave red transform. The same effect can be obtained by dividing and recording on tracks on the tape.

Embodiment 9 FIG. In the present embodiment, the case where the rate control of the recording signal is performed in units of one frame period has been described, but in the present embodiment, the buffer control is performed in units of one frame. In this embodiment, a case will be described in which the buffer control is performed in units of one track. As described above, at the time of double-speed reproduction, synthesis is performed based on information of 20 tracks. At this time, track number 1 from the first frame,
The information at the center of the tracks 2, 5, 6, 9, and 10 is
Also, from the second frame, track numbers 3, 4, 7, 8
The information in the central part of is reproduced. When synthesizing a reproduced image using these pieces of information, if rate control is performed in units of one screen, the amount of code changes depending on the characteristics of the image. Therefore, when synthesizing a reproduced image using the reproduction signal, D is not reproduced.
A CT block occurs. This block is a high-frequency component (or residual component) on the recording track during code amount control.
When the DCT block arranged near the switching between the low frequency component and the low frequency component (or the compressed image component) reproduces at double speed, the information in the above portion may not be rewritten stably.
Therefore, if the rate control is performed in units of one track, the above-mentioned problem is solved and a good double-speed playback image can be obtained.

Embodiment 10 FIG. In the above, in the ninth embodiment, the case where the rate control is performed in units of one track in consideration of the double speed reproduction is described. Good. This is 2
In the case of double-speed reproduction, as shown in FIG. 22, information at the center of the track, that is, all low-frequency components (or compressed image components) of the image are reproduced. Will be replaced. That is, since the low-frequency component of the image arranged in the center portion of the track is not reproduced for one track unlike the case of 2 × speed reproduction, the area of the updated portion on the screen varies when the reproduced image is synthesized. , Resulting in an awkward composite image. Therefore,
Considering other double-speed reproduction when encoding image information, if the minimum unit of rate control is performed in units of predetermined DCT blocks, the above-described problem is solved and a good double-speed reproduction image can be obtained.

Embodiment 11 FIG. In the above embodiment, the case where one frame is recorded on ten tracks using a two-channel facing head (see FIG. 25) is described. However, the present invention is not limited to this. In a combination head, a three-channel head, or recording on a track of one frame), the same effect can be obtained by dividing and recording important components of image information and residual information on tracks on a magnetic tape.

Embodiment 12 FIG. In the above embodiment, the case of 2 × speed reproduction has been described, but the present invention is not limited to this.
It is needless to say that, even with other double speeds, the reproduction efficiency is slightly lower than in the case of double speed reproduction, but good high speed reproduction can be realized.

Embodiment 13 FIG. In this embodiment, the basic data (low-frequency signal or compressed image data) of the image information is recorded in the center of the track on the magnetic tape. However, the present invention is not limited to this. It goes without saying that the same effect can be obtained even if the data is recorded in several positions.

[0095]

As described above, according to the present invention, the digital
High-efficiency encoding device that compresses and encodes image data
Image data into blocks for each pixel.
To output the block in a predetermined order.
Ring means and orthogonal transformation for block data
Orthogonal transformation means for performing the transformation and the orthogonal transformation means
Quantization step of transform coefficient is cut by rate control coefficient.
The quantization means for performing quantization in
Variable length encoding for variable-length coding on quantized data
Encoding means, and variable-length coded by the variable-length coding means.
Buffer means for temporarily storing the stored data;
Belonging to the low-frequency portion of the orthogonal transform coefficient for the output of the
And those that belong to the high-frequency part, and
Output order changing means for performing output in the beginning.
Low-frequency code amount and high-frequency code in the data stored in the
Measures each code amount of the signal amount and controls the rate according to each measured value
Since the coefficients are output to the quantization means , a good high
High-speed reproduction with high image quality can be realized.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a high-efficiency encoding device according to a first embodiment of the present invention.

FIG. 2 is a diagram for explaining the operation of the first embodiment of the present invention.

FIG. 3 is a diagram for explaining the operation of the first embodiment of the present invention.

FIG. 4 is a block diagram illustrating a configuration of an output order permutation circuit according to the first embodiment of the present invention.

FIG. 5 is a diagram showing an image by simulation for explaining the operation of the first embodiment of the present invention.

FIG. 6 is a recording format diagram of a digital VTR using the high-efficiency encoding device according to the first embodiment of the present invention.

FIG. 7 is a block diagram illustrating a configuration of a high-efficiency encoding device according to a second embodiment of the present invention.

FIG. 8 is a diagram showing an image by simulation for explaining the operation of the second embodiment of the present invention.

FIG. 9 is a diagram for explaining a first quantization table according to the third embodiment of the present invention.

FIG. 10 is a diagram for explaining a second quantization table according to the third embodiment of the present invention.

FIG. 11 is a diagram for explaining the operation of the third embodiment of the present invention.

FIG. 12 is a block diagram illustrating a configuration of a quantizer according to a third embodiment of the present invention.

FIG. 13 is a diagram for explaining the operation of the fifth embodiment of the present invention.

FIG. 14 is a diagram for explaining the operation of the sixth embodiment of the present invention.

FIG. 15 is a diagram for explaining the operation of the sixth embodiment of the present invention.

FIG. 16 is a block diagram illustrating a configuration of a variable-length encoder of a digital VTR using a high-efficiency encoding device according to a sixth embodiment of the present invention.

FIG. 17 is a block diagram illustrating a configuration of a conventional high-efficiency encoding device.

FIG. 18 is a diagram for explaining an output order of transform coefficients after DCT.

FIG. 19 is a diagram for explaining sequence numbers of DCT coefficients.

FIG. 20 is a block diagram showing a configuration of a conventional quantizer.

FIG. 21 is a diagram for explaining a quantization table used in a conventional quantizer.

FIG. 22 is a block diagram showing a general configuration of a conventional variable-length encoder.

FIG. 23 is a diagram for explaining the concept of a variable length coding scheme.

FIG. 24 is a diagram for explaining a high-speed reproduction operation of the digital VTR.

FIG. 25 is a head layout diagram of a rotary head of a conventional digital VTR.

FIG. 26 is a diagram illustrating an image obtained by simulation when high-speed reproduction is performed by a conventional digital VTR.

[Explanation of symbols]

 2 Shuffling Circuit 3 Orthogonal Transformer 4 Quantizer 5 Variable Length Encoder 6 Buffer 7 Rate Control Circuit 9 Output Order Rearranging Circuit 400 Inverse Quantizer 410 Difference Operation Device 40 Second Quantizer 50 Second Variable Long encoder 60 second buffer

────────────────────────────────────────────────── ─── Continued on the front page Examiner Akira Suzuki (56) References JP-A-6-62371 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04N 5/91-5 / 956 H04N 7/24-7/68 G11B 20/10-20/12

Claims (6)

(57) [Claims] 1. A high efficiency encoding apparatus for compressing and encoding digital image data, wherein said image data is divided into blocks for each of a plurality of pixels, and said blocks are output in a predetermined order. And orthogonal transform means for performing orthogonal transform on the block data, and a quantization step of transform coefficients obtained by the orthogonal transform means.
Quantizing means for performing quantization Te switching Rikae by the rate control factor, and the variable length coding means for performing variable length coding on the quantized data obtained by the quantization means, the variable length coding and temporarily <br/> stored Luba Ffa means variable-length encoded data by means belonging to the lower band of the orthogonal transform coefficient to the buffer means output and to those belonging to the high-frequency portion Output order changing means for dividing the data into a predetermined order and outputting in a predetermined order, and storing the data in the buffer means.
Each code of low band code amount and high band code amount in the obtained data
The rate control coefficient according to each measurement value.
A high-efficiency coding device for outputting to said quantization means . 2. The apparatus according to claim 1, wherein said quantization means includes a plurality of predetermined quantization tables, and quantization table switching means for switching a quantization table type of said quantization means according to said rate control coefficient. 2. The high-efficiency encoding apparatus according to claim 1, wherein: 3. A high-efficiency encoding apparatus for compressing and encoding digital image data, wherein said image data is divided into blocks for each of a plurality of pixels, and shuffling means for outputting said blocks in a predetermined order. And orthogonal transform means for performing orthogonal transform on the blocked data, and quantizing the transform coefficient obtained by the orthogonal transform means with a first rate control coefficient output from a first buffer means. First quantization means for performing quantization by switching steps, first variable-length encoding means for performing variable-length encoding on the quantized data obtained by the first quantization means, The variable-length-encoded data is temporarily stored by the first variable-length encoding means, the amount of the stored data is measured, and the first rate control is performed by the first quantization means in accordance with the measured value. Output coefficient A first buffer means, an inverse quantization means for performing inverse quantization on the output of the first quantization means, and a difference between the orthogonal transform coefficient output and the output value of the inverse quantization means are output. A difference calculation unit, a second quantization unit that performs quantization by switching a quantization step based on a second rate control coefficient output from a second buffer unit with respect to the difference calculation output, and the second quantization unit. A second variable-length encoding unit that performs variable-length encoding on the quantized data obtained by the quantization unit, and temporarily stores the variable-length encoded data by the second variable-length encoding unit. And a second buffer means for measuring an amount of the stored data and outputting a second rate control coefficient to the second quantizing means in accordance with the measured value. Efficiency coding device. 4. The first quantization means comprises a plurality of predetermined quantization tables, and the first quantization means has a plurality of quantization tables in accordance with code amount values stored in both the first buffer means and the second buffer means. 4. The high-efficiency encoding apparatus according to claim 3, further comprising a first quantization table switching unit that switches a quantization table type of the first quantization unit. 5. The apparatus according to claim 1, wherein said first or second quantization means includes a quantization table switching means for switching a quantization table type of the quantization means for an image component of an input orthogonal transform coefficient. The high-efficiency coding apparatus according to claim 3, wherein 6. The apparatus according to claim 1, wherein said second variable length coding means comprises a plurality of types of coding tables describing coded data, and switches between said coding tables according to said rate control coefficient value. Item 3. A high-efficiency encoding device according to Item 3.