JPH0779434A - Reception/reproduction device for digital picture signal - Google Patents

Abstract

PURPOSE:To efficiently generate a decoded picture by using correction data when a signal subject to orthogonal transformation coding is decoded. CONSTITUTION:A coefficient decoding circuit 32 decodes coefficient data CD according to coded coefficient data DT and a threshold level TH. A quantization circuit 33 applies quantization to supplied coefficient data CD to generate quantization data QD. A correction data table 34 corrects the coefficient data CD corresponding to a pattern of the supplied quantization data QD and applies correction coefficient data ND to a data conversion circuit 35. The data conversion circuit 35 applies clipping to the correction coefficient data ND according to coding coefficient data DT and a threshold level TH. The correction coefficient data ND are fed to an IDCT circuit 38. The IDCT circuit 38 receives a DC component, the coefficient data CD and the correction coefficient data ND respectively and applies IDCT to them and outputs a decoded picture data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital image signal receiving / reproducing apparatus applied to recording / reproducing a digital image signal by, for example, a digital VTR, and more particularly to converting quantized data into a restored value. For a decoding device for.

[0002]

2. Description of the Related Art When a digital video signal is recorded on a recording medium such as a magnetic tape, the amount of information to be recorded is large. Therefore, high efficiency encoding compresses the digital video signal to obtain a transmission rate at which recording / reproduction is possible. It is usually achieved. High-efficiency coding for compressing a digital video signal is performed by dividing the digital video signal into a large number of small blocks and processing each block by ADRC or DC.
T (Discrete Cosine Transform) and the like are known.

The DCT performs cosine transform on pixels in a block and requantizes coefficient data obtained by the cosine transform. Further, variable length coding is performed on the requantized coefficient data. Entropy coding such as Huffman coding is often used for this variable length coding. Therefore, the image data is orthogonally transformed to be divided into a large number of frequency data from low frequencies to high frequencies.

When requantizing the divided frequency data, the low frequency data, which is important in consideration of human visual characteristics, is finely quantized and the human visual characteristics are taken into consideration. With respect to high-frequency data of low importance, it is possible to maintain high image quality and achieve efficient compression by performing rough quantization. Such adaptive quantization is performed in the coefficient data requantization circuit. Furthermore, the amount of encoded data can be controlled by varying the quantization step width at the time of requantization, and a buffering process that keeps the output data rate constant is possible.

In the conventional decoding using DCT, quantized data for each frequency component is converted into a representative value of the code,
Inverse DCT (IDCT: Inverse DC) for these components
The reproduction data is obtained by applying T). When converting to this representative value, the quantization step width at the time of encoding is used. When performing adaptive quantization, the decoding side needs to know the quantization step width at the time of encoding, and therefore this quantization step width is transmitted. This transmission information is called a threshold TH. 6 to 10 show DCT as an example.
The encoding and decoding of the coefficient of each frequency component of is shown. The image data is shown in FIG. DCT for this image data
By performing the processing of (1), coefficient data for each frequency component can be obtained. This coefficient data is shown in FIG. After the coefficient data is obtained, quantization is performed.

Here, for simplification, the coefficient data of each frequency component other than the DC component is divided into a low frequency region and a high frequency region as shown in FIG. 11, and the coefficient data in the low frequency region is "4". Divide by, and divide by 8 for coefficient data in the high frequency range. Quantization is performed by truncating each fractional data part of the data calculated by this division. Information "4" and "8" of these quantization steps or information indicating this is transmitted as the threshold value TH. Figure 7
FIG. 8 shows an example in which the coefficient data shown in is quantized. Generally, quantized data is subjected to entropy coding such as Huffman coding, but it is omitted here for simplicity.

Next, the decoding operation will be described. From the entropy code, the quantized data decoded into a normal code, for example, the quantized data shown in FIG. 8 is decoded into coefficient data. On the encoding side, it is known from the threshold value TH that the coefficient data is divided by `4 ', and the quantization process of cutting off the fractional data part is performed, so the quantized data in the low frequency region is multiplied by 4, Coefficient data is obtained by multiplying the quantized data in the high frequency region by eight. FIG. 9 shows the quantized data shown in FIG. 8 converted into coefficient data. The decoded image is obtained by performing IDCT on the coefficient data as shown in FIG. FIG. 10 shows the coefficient data shown in FIG. 9 subjected to IDCT. This ID
When the CT processing is completed and the reproduction code is obtained, the decoding is completed.

[0008]

As described above, the DCT
In the compression using the orthogonal transform such as the above, there is a feature that high quality can be maintained and highly efficient compression can be realized by performing encoding in consideration of human visual characteristics. However, if the quantization for each coefficient data is roughened to increase the compression rate, the error of the restored image with respect to the original image increases, and the restored image deteriorates. The distortion is the blur of the image,
It appears in the form of so-called mosquito noise at the edge portion or block distortion, which is a big problem.

By the way, image data generally has a strong correlation. In the method of dividing into small regions and performing quantization with the same step width and treating the quantized data as encoded data, the encoded data also has a correlation like the image data. Similarly, the image data after the orthogonal transformation and the coefficient data in the periphery also have a correlation.

In the conventional method, the coefficient data of the component of interest is decoded without using the coefficient data of the peripheral component. Therefore, when the compression rate is high, there is a problem that image deterioration is noticeable. By utilizing the local correlation between coefficient data, it is possible to improve the decoding accuracy of coefficient data and create a decoded image with reduced image deterioration.

Therefore, it is an object of the present invention to form finer and appropriate decoding coefficient data without increasing the amount of information in the coding using the orthogonal transform, thereby reducing the quantization error. It is to provide a possible digital image signal receiving / reproducing apparatus.

[0012]

SUMMARY OF THE INVENTION According to the present invention, in a digital image signal receiving / reproducing apparatus which encodes coefficient data of each component after orthogonal transformation and decodes transmission data, the coefficient data of interest in the same block. And the coefficient data of a plurality of components has a memory in which prediction coefficient data that minimizes the error statistically is previously stored for each pattern, and the coefficient data of interest and the coefficient data of a plurality of components are decoded. Correction coefficient data generation means for inputting to the memory and generating correction coefficient data corresponding to the coefficient data of interest from the memory; and means for generating decoded data by performing inverse conversion based on the correction coefficient data. And a digital image signal receiving / reproducing device.

[0013]

A mapping table is prepared in which the coefficient of the filter for generating an appropriate decoding coefficient is prepared by using the coded data of the coefficient of the component of interest and the coded data of the coefficient of a plurality of adjacent components. There is. The received / reproduced coefficient data is input to this mapping table, an appropriate decoding coefficient is output, and IDCT is applied to it,
Appropriate decoded data can be obtained as compared with the related art.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below. FIG. 1 shows an example of this embodiment, namely, a digital VT.
1 shows a schematic configuration of R signal processing. A video signal is supplied from an input terminal 1 and one sample is digitally processed by the A / D converter 2, for example, 8 bits. The output data of the A / D converter 2 is supplied to the blocking circuit 3. In this embodiment, in the blocking circuit 3, the effective area of one frame is (4 × 4) pixels, (8 × 4) pixels.
8) It is divided into blocks of a size such as pixels.

A digital video signal scan-converted in the order of blocks from the blocking circuit 3 is supplied to the shuffling circuit 4. The shuffling circuit 4 performs shuffling in block units, for example. The output of the shuffling circuit 4 is supplied to the block encoding circuit 5.
The block coding circuit 5 requantizes the pixel data for each block to compress the pixel data. Here, the shuffling circuit 4 may be provided after the block encoding circuit 5.

In this embodiment, the block coding circuit 5 performs DCT processing on an input signal divided into (8 × 8) blocks, for example. As a result, 63 pieces of AC component coefficient data and 1 piece of DC component coefficient data can be obtained from the input signal divided into (8 × 8) blocks. This AC component coefficient data is requantized. As a method of requantization, in this embodiment, as an example, a method of dividing by the quantization step width of each of the low frequency band and the high frequency band indicated by the threshold value TH and rounding down after the decimal point is used. . The requantized coefficient data of a plurality of AC components are subjected to variable length coding with different bit lengths, that is, entropy coding, according to the appearance probability, and are output from the block coding circuit 5. The coefficient data of the DC component is output from the block coding circuit 5 without being requantized and entropy coded.

The output data of the block coding circuit 5 is supplied to the framing circuit 6. Recording data is generated from the framing circuit 6. The framing circuit 6 generates the parity of the error correction code and also generates the record data having a structure in which sync blocks are continuous. As the error correction code, for example, a product code that performs error correction coding in each of the horizontal direction and the vertical direction of the matrix array of data can be adopted. In the sync block, a sync block sync signal and an ID signal are added to encoded data and parity. The recording data in which the sync blocks are continuous is the channel encoding circuit 7.
Then, the channel encoding circuit 7 receives the channel encoding processing for reducing the DC component of the supplied recording data.

The output data of the channel encoding circuit 7 is converted into a bit stream and further supplied to the rotary head H via the recording amplifier 8, and the recording data is recorded on the magnetic tape T.
Recorded as a diagonal track on top. Multiple rotary heads are typically used, but for simplicity only one head is shown.

The reproduced data taken out from the magnetic tape T by the rotary head H is supplied to the channel decoding circuit 12 through the reproducing amplifier 11 and subjected to channel coding decoding. The output data of the channel decoding circuit 12 is supplied to the frame decomposing circuit 13, and various data is separated from recorded data and error correction processing is performed. The output data generated from the frame disassembling circuit 13 includes an error flag indicating the presence or absence of an error after error correction in addition to the reproduced data.

The output data of the frame disassembling circuit 13 is supplied to the block decoding circuit 15. Further, the block decoding circuit 15 is configured to generate a decoded value by generating a decoding coefficient by referring to a mapping table for correction and applying IDCT to it as described later.

The decoded data of the block decoding circuit 15, that is, the restored data corresponding to each pixel is supplied to the deshuffling circuit 16. This deshuffling circuit 1
6 is complementary to the shuffling circuit 4 on the recording side,
Performs processing to return the spatial position of a block to its original position.
The output data of the deshuffling circuit 16 is supplied to the block decomposition circuit 17. In the block decomposition circuit 17, the order of data is returned to the order of raster scanning. The output data of the block decomposition circuit 17 is supplied to the error interpolation circuit 18. The error interpolation circuit 18 performs error detection on a pixel-by-pixel basis. Pixel data detected as an error is interpolated with peripheral pixel data.

As the interpolation processing, for example, a spatially, that is, a two-dimensional direction interpolation circuit and a time direction interpolation circuit which are sequentially connected can be used. The output data of the error interpolation circuit 18 is supplied to the D / A converter 19, and the output terminal 20
, The reconstruction data corresponding to each pixel is obtained in the raster scanning order.

The present invention is applied to the block decoding circuit 15 described above. FIG. 2 is an example of the block decoding circuit 15 according to the present invention. Reproduction data is supplied from the input terminal indicated by 31, and this reproduction data is supplied to the frame disassembling circuit 13. In the frame decomposing circuit 13, the encoded coefficient data DT and the threshold value TH are separated from the supplied reproduction data and taken out respectively. The data extracted from the frame decomposing circuit 13 is the coefficient decoding circuit 3
2 are respectively supplied to the coding coefficient data DT, and after the entropy code is decoded, the representative value conversion is performed by multiplying the low and high quantization step widths indicated by the threshold value TH. The coefficient data CD of each component is decoded. Of the coefficient data CD of each component, the coefficient data of the component of high importance is supplied to the quantization circuit 33, and the coefficient data of the other components is supplied to the memory 36. In this example, the data of the highly important components are five components (coefficient data CD1 to CD5) starting from the low-order components as shown in FIG. Here, the memory 36 holds the supplied coefficient data, outputs it from the memory 36 to the IDCT circuit 38 after delaying for a certain time.

The quantizing circuit 33 has a low-order 5 degree of importance.
Quantize the coefficient data of the component. This quantization is for reducing the number of patterns of coefficient data of highly important components. The coefficient obtained by performing the DCT process has a characteristic of showing a strong tendency of concentration centering on '0'. Therefore, for example, nonlinear quantization as shown in FIG. 4 is performed. As a result, when the DCT processing is performed on the coefficient data in the existence range of (−255 to +255), the quantized data is compressed in the range of (−8 to +8).
Data obtained by performing such quantization on the coefficient data CD1 to CD5 is referred to as quantized data QD1 to QD5.
The quantization circuit 33 supplies the coefficient data CD1 to CD5 and the quantized data QD1 to QD5 to the correction data table 34, respectively.

The correction data table 34 includes coefficient data C
The correction values of D1 to CD5 are generated. The correction data table 34 is composed of a memory and stores filter coefficients formed by training in advance as described later. When the patterns of the quantized data QD1 to QD5 are supplied from the quantization circuit 33 to the correction data table 34, the quantized data QD1 is displayed in the correction data table 34.
To coefficient data CD1 to CD corresponding to the QD5 pattern
The correction coefficient data ND1 to ND5 are calculated by reading out the filter coefficients for performing the correction of No. 5 and multiplying the read filter coefficient with the coefficient data CD1 to CD5, respectively. In the pattern of the quantized data QD1 to QD5,
The filter coefficient for obtaining the correction coefficient data ND1 is (a
1, a2, a3, a4, a5), the correction coefficient data ND1 is obtained by the following equation (1).

ND1 = a1 × CD1 + a2 × CD2 + a3 × CD3 + a4 × CD4 + a5 × CD5 (1)

Similarly, when the filter coefficient for obtaining the correction coefficient data ND2 is (b1, b2, b3, b4, b5), the correction coefficient data ND2 is obtained by the following equation (2).

ND2 = b1 × CD1 + b2 × CD2 + b3 × CD3 + b4 × CD4 + b5 × CD5 (2)

When the correction data is obtained for the five components of the correction coefficient data ND1 to ND5, the quantized data QD1
5 sets of filter coefficients are stored for each pattern of QD5. Correction coefficient data ND1 to ND5 are calculated from the correction data table 34. The calculated correction coefficient data ND1 to ND5 are supplied from the correction data table 34 to the data conversion circuit 35.

The data conversion circuit 35 clips the correction coefficient data ND1 to ND5. According to the encoded coefficient data DT and the threshold value TH supplied from the frame decomposition circuit 13 via the coefficient decoding circuit 32, the data conversion circuit 35 calculates the original existence range of the correction coefficient data of each component. For example, the coding coefficient data DT is
3'and the threshold value TH is "4", 3 x 4 =
Twelve. Here, since the threshold value TH is synonymous with the quantization step, when the coding coefficient data DT is “3”, the original value of the coding coefficient data DT is 12 or more and 1 or more.
It exists in the range of less than 6. This range is called the existence range.

This existence range and the correction coefficient data ND1 to ND1
The correction coefficient data ND1 to ND5 are compared by comparing the values of ND5.
If the respective values of are not included in their original range of existence,
Clipping is performed by replacing the correction coefficient data with the closest value among the coefficient data included in the existing range. For example, when the existence range of the correction coefficient data ND1 is 12 or more and less than 16, the correction coefficient data ND1 is 1
When it is 1.6 and is not included in the existing range, this correction coefficient data ND1 is clipped from 11.6 to 12. The output of the data conversion circuit 35 is supplied to the IDCT circuit 38. The IDCT circuit 38 is supplied with the output of the data conversion circuit 35 and the output of the memory 36, respectively, and the DC component is supplied from the terminal 37 to the IDCT circuit 38.

The IDCT circuit 38 performs IDCT processing on the supplied coefficient data of each component and decodes it into image data. The decoded image data is supplied to the output terminal 39.

FIG. 5 shows an example of a block diagram at the time of training for creating the correction data table 34. In FIG. 5, a digital video signal is input to the input terminal 41, and the input digital video signal is supplied to the blocking circuit 42. The digital video signal blocked in the blocking circuit 42 is supplied from the blocking circuit 42 to the DCT coding circuit 43, and DCT coding is performed in the DCT coding circuit 43. This input data is preferably a standard digital video signal for training.

In the DCT coding circuit 43, the coded code DT after quantizing the supplied digital video signal and the threshold value TH, and the coefficient data RD of each component before being quantized are respectively supplied. Output. In the coefficient decoding circuit 44, the encoded code D from the DCT encoding circuit 43.
T and threshold TH are supplied respectively. The supplied encoded code DT is decoded according to the threshold value TH to generate coefficient data CD, and the coefficient decoding circuit 44 is a block which performs the same processing as the coefficient decoding circuit 32. The coefficient data CD generated by the coefficient decoding circuit 44 is supplied to the quantization circuit 45. Quantization circuit 45
Is a block that performs the same processing as that of the quantization circuit 33, and performs quantization of coefficient data of five components of high importance.
The quantization circuit 45 reduces the number of patterns of coefficient data of highly important components.

The simplest method of classifying is to use the pattern of the decoded coefficient data of each component as it is as a class. However, with this method, the number of patterns becomes enormous, and a very large capacity ROM is required. Therefore, the number of classes is effectively reduced by selecting a highly important component and performing quantization.

Quantized data QD1 to QD5 output from the quantization circuit 45, coefficient data RD1 to RD5 output from the DCT encoding circuit 43 before quantization, and coefficients output from the coefficient decoding circuit 44. Data CD1 to CD5
Are respectively supplied to the normal equation adding circuit 46.

Here, the normal equation used in the normal equation adding circuit 46 will be described. Quantized data QD1 to QD5 using the above-described coefficient data CD1 to CD5 and coefficient data RD1 to RD5 before being quantized.
Coefficient w _1, ... for each class defined by the pattern
_An n-tap linear estimation formula based on w _n is shown as formula (3) below.

RD1 = w ₁ CD1 + w ₂ CD2 + w ₃ CD3 + w ₄ CD4 + w ₅ CD5 (3)

Before training, w _i is an undetermined coefficient. In addition, in this method, the coefficient data RD1 to RD5 before being quantized are corrected, so in practice, an equation must be set for each of the coefficient data RD1 to RD5.

Training is performed on a plurality of signal data for each class. When the number of data is m, according to the equation (3),

RD1 _j = w ₁ CD1 _j + w ₂ CD2 _j + w ₃ CD3 _j + w ₄ CD4 _j + w ₅ CD5 _j (j = 1,2, ..., m) (4)

When m> n, w _1, ... _, W _n are not uniquely determined, so the elements of the error vector E are

E_j= RD1_j-{W₁CD1_j+ W₂CD2_j+ W₃CD3_j+ W_FourCD4 _j + W_FiveCD5_j} (J = 1, 2, ..., M) (5)

The coefficient that minimizes the equation (6) shown below is determined.

[0045]

[Equation 1]

This is a so-called least squares method. Here, the partial differential coefficient by w _i of the equation (6) is obtained.

[0047]

[Equation 2]

Since each w _i may be obtained so that the expression (7) is set to `0 ',

[0049]

[Equation 3]

Using a matrix as

[0051]

[Equation 4]

It becomes This equation is generally called a normal equation. The normal equation adding circuit 46 adds the normal equations. After the input of all the training data is completed, the normal equation addition circuit 46 outputs the normal equation data to the prediction coefficient determination circuit 47. The prediction coefficient determination circuit 47 solves w _i using a general matrix solution method such as a sweeping method of the normal equation and calculates the prediction coefficient. The prediction coefficient determination circuit 47 writes the calculated prediction coefficient in the memory 48.

As a result of training as described above,
The memory 48 stores prediction coefficients statistically closest to the true value for estimating the coefficient data of interest RD1 (and RD2 to RD5) for each pattern defined by the quantized coefficient data QD1 to QD5. . The table stored in the memory 48 is the correction data table 34 used in the block decoding circuit 15 described above. Here, each obtained w _i is the filter coefficient (a1,
a2, a3, a4, a5), etc., used at the time of decoding.

In this embodiment, the data of the five components is estimated from the data of the five components of high importance, but the combination is not limited to this.

Further, in this embodiment, the prediction coefficient is stored in the memory 48, but it is also possible to use the case where the representative value obtained by the centroid method is stored in the memory 48.

[0056]

According to the present invention, since the amount of data to be transmitted can take many decoding levels at least in the coefficient of the component of high importance, quantization error and block distortion can be reduced and restored. The image can be good. Further, the present invention has an advantage that there is no increase in transmission information at all.

[Brief description of drawings]

FIG. 1 is a digital V to which the present invention can be applied.
It is a block diagram of a recording / reproducing circuit of TR.

FIG. 2 is a block diagram showing an example of a block decoding circuit to which the present invention is applied.

FIG. 3 is a schematic diagram for explaining a case of creating a correction data table.

FIG. 4 is a block diagram showing a configuration at the time of training for creating a correction data table in one embodiment of the present invention.

FIG. 5 is a schematic diagram for explaining a case of creating a correction data table.

FIG. 6 is a schematic diagram for explaining encoding by orthogonal transform.

FIG. 7 is a schematic diagram for explaining encoding by orthogonal transform.

FIG. 8 is a schematic diagram for explaining encoding by orthogonal transform.

FIG. 9 is a schematic diagram for explaining encoding by orthogonal transform.

FIG. 10 is a schematic diagram for explaining encoding by orthogonal transform.

FIG. 11 is a schematic diagram for explaining a frequency domain after performing orthogonal transformation.

[Explanation of symbols]

 32 coefficient coding circuit 33 quantization circuit 34 correction data table 35 data conversion circuit 38 IDCT circuit

Claims (4)

[Claims] 1. A digital image signal receiving / reproducing apparatus for encoding coefficient data of each component after orthogonal transformation and decoding transmission data, wherein coefficient data of interest of a same block and coefficient data of a plurality of components are included. Prediction coefficient data such that the error is statistically minimized for each pattern determined by has a memory previously stored, input the coefficient data of interest to be decoded and the coefficient data of the plurality of components to the memory, It comprises correction coefficient data generating means for generating correction coefficient data corresponding to the attention coefficient data from the memory, and means for generating decoded data by performing inverse conversion based on the correction coefficient data. A digital image signal receiving / reproducing apparatus characterized by the above. 2. A digital image signal receiving / reproducing apparatus which encodes coefficient data of each component after orthogonal transformation and decodes transmission data, in which coefficient data of interest of the same block and coefficient data of a plurality of components Prediction coefficient data such that the error is statistically minimized for each pattern determined by has a memory previously stored, input the coefficient data of interest to be decoded and the coefficient data of the plurality of components to the memory, Correction coefficient data generating means for generating correction coefficient data corresponding to the attention coefficient data from the memory, and means for generating decoded data by performing inverse conversion based on the correction coefficient data, A digital image signal receiving / reproducing apparatus characterized in that correction is performed only on low-order coefficient data of one block. 3. The digital image signal receiving / reproducing apparatus according to claim 1, wherein the prediction coefficient data stored in the memory is a result of nonlinear quantization of the coefficient data of interest and the coefficient data of a plurality of components. A device for receiving / reproducing digital image signals classified based on the data of. 4. The digital image signal receiving / reproducing apparatus according to claim 1, wherein when the correction coefficient data from the correction coefficient data generating means does not fall within the existing range, the correction coefficient data is A receiving / reproducing apparatus for a digital image signal, characterized by comprising clipping means for performing clipping.