JPH04271664A - Picture data compressor and picture data decoder - Google Patents

Abstract

PURPOSE:To obtain a coding data not including redundant information with respect to the picture data compressor and picture data decoder. CONSTITUTION:The picture data compressor and picture data decoder is featured to have an orthogonal transformation means 111 applying orthogonal transformation to a picture data representing a multi-value picture for each block comprising MXN picture elements and obtaining respectively an M-row N-column coefficient matrix comprising transformation coefficients, an extraction means 112 extracting a prescribed component from the coefficient matrix corresponding to each block respectively and generating reduced coefficients of K-row and L-column comprising only the extracted component, a reduced picture generating means 113 applying inverse orthogonal transformation to the reduced coefficients to generate a reduced picture comprising KXL picture elements corresponding to each block, and a transformation coding means 114 applying transformation coding to a set of the reduced pictures corresponding to each block as a picture data and sending the data.

Description

[Detailed description of the invention]

0001

[Field of Industrial Application] The present invention relates to an image data compression device that converts and encodes image data representing a multivalued image and compresses it, and an image data restoration device that restores image data based on the obtained encoded data. It is. An adaptive discrete cosine transform encoding method (Ad
aptive Discrete Cosine Tr
(hereinafter referred to as ADCT method) is widely used.

[0002] This ADCT method divides a multilevel image into blocks each consisting of a predetermined number of pixels (for example, 8×8 pixels), and orthogonally transforms the image data for each block to obtain transform coefficients (hereinafter referred to as DCT coefficients). The amount of data is compressed by determining a matrix consisting of the following elements (referred to as 1), quantizing each component of this matrix using a corresponding visual adaptive threshold (described later), and then variable-length encoding.

0003

2. Description of the Related Art FIG. 22 shows the configuration of an image data compression apparatus to which a conventional ADCT method is applied. Also, in Figure 23,
An example of blocks obtained by dividing a multivalued image is shown. D.C.
The T-transform unit 611 performs two-dimensional discrete cosine transform (hereinafter referred to as DCT transform) processing on the input block, and transforms the input block into 8
It is converted into a matrix with 8 rows and 8 columns (hereinafter referred to as DCT coefficient D). FIG. 24 shows an example of this DCT coefficient D.

Each component of this DCT coefficient D is sequentially inputted to a divider 622 as a dividend by a DCT coefficient input section 621 of a linear quantization section 620. Further, the linear quantization unit 620 is provided with a quantization matrix VTH consisting of visually adaptive thresholds corresponding to each frequency component in advance, and is inputted to this quantization matrix VTH and the quantization matrix VTH prior to encoding one screen. An 8-by-8 matrix consisting of the product with the quantization parameter is stored in the quantization threshold holding unit 623 as the quantization threshold QTH. The above-mentioned visual adaptation threshold is predetermined based on experimental results regarding visual sensitivity to each spatial frequency component, and the quantization control parameter is a coefficient that determines the quantization accuracy of an image.

Therefore, in accordance with the input of each component of the DCT coefficient D, the corresponding component of the above-mentioned quantization threshold QTH is input to the divider 622 as a divisor, so that each component of the DCT coefficient D is visually adapted. A quantized coefficient is obtained by quantizing using a threshold value. Generally, human vision is
It has high sensitivity to low spatial frequencies and low sensitivity to high spatial frequencies. Accordingly, as shown in FIG. 25, the value of each component of the above-mentioned quantization matrix VTH is such that the absolute value of the component corresponding to a low spatial frequency is small, and conversely, the absolute value of the component corresponding to a high spatial frequency is small. The value is set large. Therefore, 8
A matrix with 8 rows and 8 columns (hereinafter referred to as quantization coefficient DQU) is
As shown in FIG. 26, only the component in the upper left corner of the matrix indicating the DC component (hereinafter referred to as DC component) and a very small number of AC components indicating low spatial frequency components around this DC component are non-zero. Most AC components are effective coefficients that have a value, and most AC components are invalid coefficients that have a value of zero.

The quantization coefficient DQU obtained in this way
is converted into a one-dimensional array using a scanning order called zigza scan shown in FIG. 27, and then input to the encoding unit 631. In addition, this encoding unit 631
is the one-dimensional array described above with the effective coefficient (index) and the continuous length (run) of the invalid coefficients before this index.
Based on the code table 632, each combination is replaced with a code corresponding to its frequency of appearance, variable length coding is performed, and the resulting coded data is sent out.

[0007] Furthermore, the encoded data obtained in this manner is restored to image data by an image data restoration device shown in FIG. Decoding unit 711 of image data restoration device
Contrary to the code table 632 described above, the system has a decoding table 712 that shows combinations of runs and indexes corresponding to codes, and decodes sequentially input codes to find combinations of indexes and runs. , based on these, the quantization coefficient DQU is restored and input to the inverse quantization unit 720.

[0008] This inverse quantization section 720 converts the quantization coefficient DQ
By multiplying each component of U by the corresponding component of the quantization threshold QTH, each component of the quantization coefficient DQU is dequantized and the DCT coefficient D is restored. This DCT coefficient D is the inverse DC
is input to the T transform section 731, and this inverse DCT transform section 731
The image data of the corresponding block is restored by the inverse DCT transformation process.

Normally, as described above, all the effective coefficients included in the quantization coefficient DQU corresponding to each block are encoded and sent as an index, and the quantization coefficient DQU is restored from this encoded data. The image data is restored based on this quantization coefficient DQU. On the other hand, in the case of hierarchical restoration in which a detailed image is restored step by step from a rough image, the above-mentioned encoding unit 631 sequentially extracts some components corresponding to low spatial frequencies of the quantization coefficient DQU, A matrix containing only the extracted components as effective coefficients is variable-length encoded and sent as encoded data (spectral selection method). For example, in the first stage of hierarchical restoration, the quantization coefficient DQU
Only the DC component (hereinafter referred to as DC component) is extracted, and then, in the second and third stages, two AC components adjacent to the DC component (in FIG. ) and three AC components adjacent to these components (indicated by symbols 4, 5, and 6 in FIG. 27) are extracted, encoded, and sent out, respectively.

Accordingly, in the first stage of layer restoration, the above-described decoding section 711 and inverse quantization section 720 restore the DCT coefficient D in which only the DC component is an effective coefficient. Similarly, in the second stage of hierarchy restoration, the DC
The T coefficient D is restored, and in the third stage of hierarchical restoration, the DCT coefficient D whose only effective coefficients are the three AC components described above is restored.

[0011] Here, the DC component of the DCT coefficient D corresponds to the DC component of the image data, that is, the average value of the image data, and the AC components included in the first row and the first column correspond to the respective blocks. It corresponds to each spatial frequency component indicating changes in pixel data in the row direction and column direction. Figure 29
shows the basis vector representing the fundamental waveform of each spatial frequency component. In the figure, the 0th order basis vector is a basis vector corresponding to the DC component, and the 1st to 7th order basis vectors are basis vectors corresponding to each component in the first row and first column.

As shown in FIG. 29, since the zero-order basis vector corresponding to the DC component is a constant, by performing inverse DCT transformation on the DCT coefficient D of the first stage described above,
An image of one block is restored in which all pixel data has the average value of the pixel data of each block, and a rough image can be obtained from these blocks. Also,
The first-order basis vector corresponding to the AC component included in the DCT coefficient D restored in the second stage described above is
It shows a monotonous decrease (or increase) in the row or column direction. Therefore, the above-mentioned DCT coefficient D is inverse DCT
By performing the conversion, one block of image consisting of pixels whose pixel data monotonically decreases (or increases) in the row and column directions is obtained.

Similarly, by inverse DCT-transforming the DCT coefficients D restored in the third step, image data in which pixel data changes according to the second-order basis vector is obtained. In this way, by sequentially superimposing the image data restored at each stage, a more detailed image is gradually restored.

[0014]

[Problem to be Solved by the Invention] By the way, when the DCT coefficient D includes only a DC component and three AC components surrounding this DC component as effective coefficients,
One block of image obtained by inverse DCT transformation is an image in which the value of each pixel monotonically decreases (or monotonically increases) according to the first-order base vector in each direction. That is, in this case, if pixel data corresponding to the pixels at the four corners of one block is obtained, other pixel data can be obtained by using these pixels as base points and interpolating using the waveform of the first-order base vector. The value can be determined approximately. Also,
If the DCT coefficient D includes a component corresponding to the second-order basis vector as an effective coefficient, the waveform representing the change in pixel data in the restored image has one extreme value in the row and column directions. is expected. Therefore, in this case, the pixel that takes the above-mentioned extreme value is set as the inflection point, and the pixel data corresponding to other pixels is approximately determined from the pixel data of the pixel corresponding to this inflection point and the above-mentioned base point. Can be done.

[0015] Therefore, as described above, in applications where it is permissible to obtain pixel data approximately, only information regarding the base point and inflection point or spatial frequency components contributing to approximate gradation changes is required. It is sufficient to transmit the information, and there is no need to transmit information regarding all pixels of one block or information regarding all spatial frequency components as encoded data.

However, in the conventional system, the DCT coefficients D are encoded and transmitted for each of a huge number of blocks, regardless of whether the above-mentioned approximation is allowed or not. For this reason, the encoded data includes redundant information regarding unnecessary pixels and unnecessary spatial frequency components for each block, which increases the amount of encoded data as a whole and increases the time required to transmit the encoded data. was long.

On the other hand, in the lower layers of hierarchical restoration, there is a desire to restore and output a rough image as quickly as possible. There is a need for an image data compression method that encodes at a compression rate and a corresponding image data restoration method. SUMMARY OF THE INVENTION An object of the present invention is to provide an image data compression device that obtains encoded data that does not include redundant information, and also to provide an image data restoration device that restores the aforementioned encoded data.

[0018]

Means for Solving the Problems FIG. 1 is a diagram showing the configuration of an image data compression apparatus according to a first aspect of the present invention. The invention of claim 1 provides orthogonal transformation means 111 for orthogonally transforming image data representing a multivalued image for each block of M×N pixels to obtain a coefficient matrix of M rows and N columns each consisting of transform coefficients; Extraction means 112 extracts each predetermined component from a coefficient matrix corresponding to a block and generates reduction coefficients of K rows and L columns consisting only of the extracted components; A reduced image generation means 113 that generates a reduced image corresponding to each block, and a conversion encoding means 114 that converts and encodes a set of reduced images corresponding to each block as image data and sends it out as encoded data.
It is characterized by having the following.

FIG. 2 is a diagram showing the configuration of an image data compression apparatus according to a second aspect of the present invention. The invention according to claim 2 provides the image data compression apparatus according to claim 1, wherein the conversion encoding means 114
Instead, the present invention is characterized in that it includes predictive encoding means 121 that predictively encodes image data consisting of reduced images corresponding to each block. FIG. 3 is a diagram showing the configuration of an image data compression device according to a third aspect of the present invention.

[0020] The invention according to claim 3 includes orthogonal transformation means 111 for orthogonally transforming image data representing a multivalued image block by block to obtain a coefficient matrix, and extracting a plurality of components from the coefficient matrix corresponding to each block. extracting means 112 for generating reduction coefficients of K rows and L columns including only the extracted components;
The present invention is characterized by comprising a transform encoding means 115 that transforms and encodes a set of reduction coefficients corresponding to each block as image data and sends it out as encoded data.

FIG. 4 is a diagram showing the configuration of an image data compression apparatus according to a fourth aspect of the present invention. The invention according to claim 4 is the image data compression device according to claim 2, wherein the conversion encoding means 115
Instead, the present invention is characterized in that it includes predictive encoding means 122 that predictively encodes image data consisting of reduction coefficients corresponding to each block. The invention of claim 5 is the image data compression apparatus according to any one of claims 1 and 3, wherein the extraction means 112 extracts the DC component of the coefficient matrix and the surrounding AC components as predetermined components. It is characterized by a configuration that generates a reduction coefficient.

FIG. 1 shows the configuration of an image data compression apparatus according to claim 1 to which the invention according to claim 6 is applied. The invention according to claim 6 is the image data compression apparatus according to any one of claims 1 and 3, wherein the extracting means 112 analyzes the characteristics of the multivalued image based on the coefficient matrix corresponding to each block. An analysis means 131 is provided, and this feature analysis means 131
The present invention is characterized in that it is configured to determine the components to be extracted from the coefficient matrix according to the analysis result.

FIG. 1 shows the configuration of an image data compression apparatus according to claim 1 to which the invention according to claim 7 is applied. The invention of claim 7 provides the image data compression apparatus according to any one of claims 1 and 3, in which the extraction means 112 extracts coefficients of components to be extracted in response to input of hierarchical restoration information indicating a stage of hierarchical restoration. The present invention is characterized in that it includes a coefficient position information holding means 132 that outputs coefficient position information regarding a position on the matrix, and determines a component to be extracted from the coefficient matrix in response to input of the coefficient position information.

FIG. 3 shows the configuration of an image data compression apparatus according to claim 3 to which the invention according to claim 8 is applied. The invention of claim 8 provides the image data compression apparatus according to any one of claims 1 and 3, wherein the extracting means 112 quantizes the input coefficient matrix components using a corresponding quantization threshold value. It comprises a quantization means 141 for obtaining a quantization coefficient, and an inverse quantization means 142 for inverse quantizing the quantization coefficient using a corresponding quantization threshold, and consists of transform coefficients obtained by inverse quantizing the quantization coefficient. It is characterized by a configuration that generates a reduction coefficient.

FIG. 5 is a diagram showing the configuration of an image data restoring apparatus according to a ninth aspect of the invention. The invention according to claim 9 includes a reduced image restoring means 151 for restoring a set of reduced images based on input encoded data, and a coefficient matrix of M rows and N columns corresponding to each block based on each reduced image. coefficient restoring means 152 that restores each coefficient matrix, and performs inverse orthogonal transformation on each coefficient matrix,
Inverse orthogonal transformation means 15 for restoring image data of each block
3.

FIG. 6 is a diagram showing the configuration of an image data restoring apparatus according to a tenth aspect of the present invention. The invention according to claim 10 provides a reduced image restoring means 151 for restoring a set of reduced images based on input coded data, and a reduced image restoring means 151 for restoring a set of reduced images based on input coded data, and a reduced image restoring means 151 for restoring a set of reduced images based on input coded data, and a reduced image restoring means 151 for restoring a set of reduced images based on input coded data, and a reduced image restoring means 151 for restoring a set of reduced images based on input coded data, and a reduced image restoring means 151 for restoring a set of reduced images based on input coded data, and a reduced image restoring means 151 for restoring a set of reduced images based on input coded data, and a reduced image restoring means 151 for restoring a set of reduced images based on input coded data, and a The block restoration means 154 estimates the pixel data of the block and restores the corresponding block.

FIG. 7 is a diagram showing the configuration of an image data restoring apparatus according to an eleventh aspect. The invention according to claim 11 includes a reduction coefficient restoring means 161 for restoring a set of reduction coefficients corresponding to each block based on input encoded data, and a matrix corresponding to each component of the reduction coefficients corresponding to each block. Coefficient restoring means 162 that allocates the upper position and restores each M-row N-column coefficient matrix, and inverse orthogonal transform means 153 that restores image data of each block by inversely orthogonally transforming each coefficient matrix. It is characterized by:

[0028]

[Operation] According to the invention of claim 1, the extraction means 112 extracts a predetermined component from the coefficient matrix obtained by the orthogonal transformation means 111 to generate a reduction coefficient, and the reduction image generation means 113
However, by inverse orthogonal transformation of this reduction coefficient, a reduced image containing only information regarding the extracted spatial frequency components can be obtained corresponding to each block. Furthermore, by converting and encoding a set of these reduced images as image data by the conversion encoding means 114, the plurality of reduced images described above can be encoded at once.

Furthermore, by encoding the set of reduced images described above using the predictive encoding means 121 instead of the transform encoding means 114 described above, it is possible to perform encoding without losing information regarding the reduced images. can. In the invention of claim 3, the extraction means 112 obtains a reduction coefficient consisting of only predetermined components, and the transformation encoding means 115 converts and encodes the set of reduction coefficients as image data, thereby obtaining the above-mentioned predetermined image data. It is possible to obtain encoded data that includes only information regarding the spatial frequency component corresponding to the component of .

Similarly, by encoding the above-mentioned set of reduction coefficients using the predictive encoding means 121 instead of the transform encoding means 115, it is possible to encode the set of reduction coefficients without losing information regarding the reduction coefficients. Can be done. Further, the above-mentioned extraction means 112 extracts the DC component and the surrounding AC component as predetermined components, thereby obtaining a reduction coefficient that includes a spatial frequency component that contributes to a rough change in gradation in the original image. It is possible to obtain encoded data that reliably includes these spatial frequency components. Furthermore, by performing inverse orthogonal transformation on the reduction coefficients described above, a reduced image that reflects rough changes in gradation in the original image can be obtained.

Furthermore, the feature analysis means 131 analyzes the features of the multivalued image, and according to the analysis results, the extraction means 1
By determining the components to be extracted by step 12, it is possible to generate a reduction coefficient consisting of components corresponding to the features of the multivalued image. Further, based on the coefficient position information obtained by the coefficient position information holding means 132, the extraction means 112 generates a reduction coefficient consisting of the corresponding components, thereby obtaining a reduction coefficient corresponding to the hierarchical restoration information.

Furthermore, quantization means 141 and inverse quantization means 1
42, by once quantizing the transform coefficients and then inversely quantizing them, the extracting means 112 can obtain reduction coefficients consisting of the transform coefficients obtained by dequantizing the quantized coefficients. It is possible to eliminate noise components contained in the transform coefficients. In the invention of claim 9, the coefficient restoring means 152 restores a coefficient matrix corresponding to each block based on the reduced image restored by the reduced image restoring means 151, and this coefficient matrix is converted into the inverse orthogonal transform means 1.
53 performs inverse orthogonal transformation, the image of the original block can be approximately restored based on information regarding the spatial frequency components included in the reduced image.

Furthermore, when the above-mentioned reduced image reflects rough gradation changes of the original image, the block restoration means 154 performs other pixel data based on each pixel data constituting the restored reduced image. By estimating the pixel data of pixels, the original block image can be approximately restored. According to the eleventh aspect of the invention, based on the reduction coefficient restored by the reduction coefficient restoration means 161, the coefficient restoration means 16
2 restores the coefficient matrix corresponding to each block, and the inverse orthogonal transform means 153 performs inverse orthogonal transform on this coefficient matrix to transform the original block image based on the information regarding the spatial frequency components contained in the reduced image. It can be approximately restored.

[0034]

Embodiments Hereinafter, embodiments of the present invention will be described in detail based on the drawings. FIG. 8 shows the configuration of an embodiment of the image data compression device of the present invention. In FIG. 8, image data input to the image data compression device is subjected to DCT transformation for each block of 8×8 pixels by a DCT transformation unit 211 corresponding to the orthogonal transformation means 111, and sequentially transferred to a buffer 22.
It is held at 1.

In addition, in FIG. 8, the buffer 2 described above
21, the address calculation section 231, the readout circuit 232, and the buffer 233 form the extraction means 112. In this extraction means 112, the address calculation section 231
The components to be extracted from the T coefficients D may be determined in advance, the addresses of the corresponding components may be sequentially calculated for the DCT coefficients D of each block, and the reading circuit 232 may be instructed to read these components. For example, component d(i
, j) (i=1, K j=1, L). Normally, it is sufficient to extract a DC component and three surrounding AC components as predetermined components and generate reduction coefficients in 2 rows and 2 columns.

In this way, the readout circuit 232
By selectively reading out the components of the T coefficient D, only the above-mentioned predetermined components are extracted, and K consisting of these components is extracted.
The reduction coefficients in rows and columns L are sequentially stored in the buffer 233. Also, at this time, the address calculation unit 231 may sequentially calculate addresses of the buffer 221 corresponding to these components and designate them as write addresses.

The reduction coefficient generated by this extraction means 112 is calculated by an inverse DCT corresponding to the reduced image generation means 113.
The data is input to the transform unit 241, and the inverse DCT transform unit 241 performs two-dimensional inverse DCT transform processing corresponding to a matrix of K rows and L columns, thereby generating reduced images each consisting of K×L pixels. In this way, reduced images are sequentially generated based on the reduction coefficients corresponding to all blocks, and after all of these reduced images are stored in the buffer 242,
The set of reduced images stored in this buffer 242 is input to the conversion encoding means 114.

The transform encoding means 114 is formed of a DCT transform section 261, a linear quantization section 262, and an encoding section 263, and converts a set of input reduced images in the same manner as the conventional ADCT method. The image data is configured to undergo DCT transformation, quantization, and variable length encoding for each 8×8 pixel. As mentioned above, the DCT of each block
By extracting a predetermined component from the coefficient D, information regarding the gradation change of the image of each block is selected, and the obtained reduction coefficient is subjected to inverse DCT transform to obtain the floor corresponding to the necessary spatial frequency component. A reduced image containing only information regarding key changes can be obtained.

Furthermore, by converting and encoding a set of reduced images corresponding to each block as image data, information regarding a plurality of blocks can be compressed all at once. For example, if a reduced image of 2 rows and 2 columns is generated corresponding to each block, the reduced images corresponding to 16 blocks are DCT-transformed all at once and sent out as encoded data corresponding to one block. Ru.

[0040] As a result, redundant information can be eliminated and image data can be compressed and transmitted at a much higher compression rate than before, so even when using a transfer line with restrictions on transmission speed, The time required to transmit encoded data can be shortened, and image data restoration processing can be speeded up. Furthermore, since the above-mentioned reduced image is data with the same number of bits as the image data,
A configuration may be adopted in which the DCT conversion unit 211 performs DCT conversion on the original image and the set of reduced images, without separately providing a DCT conversion unit 261 that performs DCT conversion on the set of reduced images.

FIG. 9 is a block diagram of another embodiment of the image data compression device. In FIG. 9, the image data compression device first sends one screen worth of image data to the D
The configuration is such that the data is input to the CT converter 211, and the result of conversion by the DCT converter 211 is stored in the buffer 221 via the demultiplexer 272. Further, after the DCT conversion process by the DCT conversion unit 211 and the inverse DCT conversion process by the inverse DCT conversion unit 241 are completed, the selector 271 and the demultiplexer 272 described above are switched, and the reduced image stored in the buffer 242 is subjected to DCT conversion. The output of the DCT transformer 211 may be input to the linear quantizer 262 at the same time. As a result, the DCT transform section 211 and the linear quantization section 26
2 and the encoding unit 263 realize the function of the transform encoding means 114.

In this case, since it is not necessary to provide two DCT conversion sections, the circuit scale of the image data compression apparatus does not increase. Alternatively, a configuration may be adopted in which the DCT coefficients D obtained by the DCT transform unit 211 are once quantized and then inversely quantized, and a reduced coefficient matrix is generated based on the obtained DCT coefficients D.

FIG. 10 is a block diagram of an embodiment of the image data compression apparatus according to claim 8. In FIG. 10, the extraction means 112 of the image data compression device is the extraction means 11 shown in FIG.
2, a linear quantization unit 234 and a multiplier 235 are added. This linear quantization section 234 corresponds to the quantization means 141, and is a quantization threshold holding section 2 that holds quantization thresholds corresponding to each component of the DCT coefficient D.
36, and the DC obtained by the DCT conversion unit 211
Each component of the T coefficient D is divided by the corresponding quantization threshold value, and the result is stored in the buffer 221. In this case, the address calculation unit 231 determines that the reading circuit 2
32 to read out the above-described predetermined component, and at the same time instructs the above-mentioned quantization threshold holding unit 236 to output the quantization threshold corresponding to this component. Further, in accordance with this instruction, the quantization threshold holding unit 236 inputs the corresponding quantization threshold to the multiplier 235,
This multiplier 235 is configured to multiply the output of the readout circuit 232 by the above-mentioned quantization threshold. In this way, the function of the dequantization means 142 is realized by the quantization threshold holding unit 236 and the multiplier 235, and by sequentially storing the dequantization results in the buffer 233, the quantization coefficients are decoded. A reduction coefficient consisting of DCT coefficients obtained by quantization is obtained.

In general, by once quantizing the DCT coefficient D and then inversely quantizing it, in particular, noise components contained in high spatial frequency components can be removed to obtain a DCT coefficient D that does not contain noise components. Can be done. Therefore, by configuring the extraction means 112 as described above, it is possible to eliminate noise components from the reduction coefficients corresponding to each block and reduce the amount of information of the reduced image generated from these reduction coefficients.

Furthermore, as shown in FIG.
2, the address calculation unit 231 and the readout circuit 232 read out a predetermined component of the DCT coefficient D from the buffer 221 and input it to the linear quantization unit 234, and the quantization result by the linear quantization unit 234 is converted into a corresponding quantization result. It may also be configured such that it is input to the multiplier 235 together with the threshold value. In this case, only the components designated by the address calculation unit 231 described above are selectively quantized, so it is possible to omit the division process for quantizing the components that are not extracted, and the division process required for the quantization process can be omitted. It is possible to shorten the time and speed up image data compression processing.

By the way, an image whose gradation changes gradually as a whole may be inputted to the image data compression apparatus, and an image whose gradation changes drastically may be inputted into the image data compression apparatus. Furthermore, even in a single screen image, there may be a mixture of parts where the gradation changes are gradual and parts where the gradation changes are rapid. If a reduced image is generated by extracting the same number of DCT coefficients corresponding to each block that makes up such an image, the amount of information will be excessive in parts where the gradation changes are gradual; In areas where changes are rapid, the amount of information may be insufficient.

A method of compressing image data while taking such image characteristics into account will be described below. Figure 1
2 is a block diagram of an embodiment of the image data compression apparatus according to claim 6. In FIG. 12, the image data compression device shown in FIG.
It is constructed by adding a distribution determining section 281, a feature storage section 282, and an extraction instruction section 283 to the extraction means 112 shown in FIG. The distribution discriminator 281 described above includes the linear quantizer 23
The feature storage unit 282 is configured to determine the distribution pattern of effective coefficients in the quantized coefficients DQU input from 4, and the feature storage unit 282 is configured to store this determination result as a feature of each block. Further, after the determination of the distribution patterns of all blocks is completed in this way, the extraction instruction unit 283 divides the portions where blocks having the same or similar distribution patterns are consecutive into regions, and The address calculation unit 231 may be instructed on the components to be extracted according to the distribution pattern applicable to each area. As a result, reduction coefficients corresponding to the distribution pattern applicable to each area are generated, and based on these reduction coefficients, a reduced image having a size corresponding to the above-described distribution pattern is generated.

[0048] In this way, depending on the characteristics of each area,
By changing the components of the DCT coefficients D to be extracted, it becomes possible to select information regarding gradation changes according to the image characteristics of each region. This makes it possible to flexibly handle cases where images with various characteristics are input, or cases where the input image contains parts with different characteristics. It is possible to prevent information from being sent.

Furthermore, in this case, the transform encoding means 114
The set of reduced images obtained as described above is variable-length encoded for each area, and information about the components included in the reduction coefficient is written as encoding information at the beginning of the encoded data for each area. Just add it and send it. Further, as described above, the feature analysis means 131 is formed by the linear quantization section 234, the distribution discriminating section 281, the feature storage section 282, and the extraction instruction section 283, and the quantization coefficient DQU is input to the distribution discriminating section 281. Therefore, the feature analysis means 131 can analyze the features of the image without being affected by the noise components included in the DCT coefficients D.

Next, an image data restoring device for restoring image data based on encoded data obtained by the above-described image data compression device will be described. Figure 1
3 is a configuration diagram of an embodiment of the image data restoration device of the present invention. In FIG. 13, a decoding section 311 and an inverse quantization section 312
and the inverse DCT transformation section 313 are the reduced image restoration means 151
is formed. This reduced image restoration means 151 restores a set of reduced images by decoding input encoded data, inverse quantization, and inverse DCT transformation in the same manner as in the past, and stores these reduced images in the buffer 321. The configuration is such that the files are stored sequentially. In addition, the decoding unit 31 described above
1 is configured to detect encoded information added to input encoded data and send the obtained encoded information to coefficient restoring means 152.

The coefficient restoring means 152 is formed of a DCT transform section 331 and a rearrangement processing section 332. In this coefficient restoring means 152, the DCT transform section 331
is configured to decompose the above-mentioned set of reduced images into reduced images consisting of K×L pixels, and perform DCT conversion processing on each of these reduced images to obtain the reduction coefficient corresponding to each reduced image. . Further, the rearrangement processing unit 332 is configured to restore the corresponding DCT coefficient D by assigning and rearranging each component of these reduction coefficients a corresponding position on a matrix of 8 rows and 8 columns. D mentioned above
The CT conversion unit 331 normally performs DCT conversion processing on the assumption that each reduced image has a preset size, and when encoded information is input, it performs DCT conversion processing corresponding to the size indicated by this encoded information. All you have to do is perform the conversion process. Further, the rearrangement processing unit 332 normally performs rearrangement processing by assigning each component of each reduction coefficient to a preset position on a matrix, and according to the input of encoding information, All you have to do is to allocate and perform the sorting process.

In this way, the above-mentioned extraction means 112
The DCT coefficient D containing only the components extracted by can be restored. Furthermore, the above-mentioned extraction means 112 extracts the component d(i,j)(
A reduced image obtained by extracting K j = 1, L) and generating K rows and L columns of reduction coefficients reflects rough gradation changes in the original image. Therefore, in this case, based on each pixel of the reduced image corresponding to each block, it is possible to estimate the pixel data of other pixels forming the corresponding block.

In this case, as shown in FIG. 14, the pixel data input section 334 and the arithmetic processing section 335 constitute the block restoration means 154, and as will be described later, the pixel data of the corresponding block is extracted from the reduced image. All you have to do is restore it. The above-mentioned pixel data input section 334 is normally configured to read out a matrix of a preset size from the buffer 321 as a reduced image corresponding to each block, and input it to the arithmetic processing section 335. In addition, when encoded information is detected by the decoding unit 311 described above, this pixel data input unit 334 sequentially reads out reduced images of the size indicated by this encoded information, and
35. In addition, in response to the input of the reduced image described above, the arithmetic processing unit 335 sets each pixel of the reduced image as a base point and an inflection point of gradation change in the image of the corresponding block, and calculates the change in pixel data between these points. The configuration is such that other pixel data of the corresponding block is approximately obtained by interpolating with a straight line.

For example, a DC component and three surrounding AC components
If a reduced image consisting of 4 pixels is generated in response to a reduction coefficient consisting of Then, the pixel data of each other pixel can be approximately obtained. Alternatively, the base vectors corresponding to the components indicated by the encoded information may be superimposed to obtain a function representing the gradation change in the corresponding block, and the interpolation process may be performed using this function.

As mentioned above, if the reduced image reflects the gradation changes in the original image, the arithmetic processing unit 334
By performing relatively simple arithmetic processing, image data can be approximately restored. This makes it possible to reduce the number of times the inverse DCT transformation process and the DCT transformation process, which require a large amount of calculation processing, are executed, thereby making it possible to speed up the restoration process.

On the other hand, when a reduction coefficient is generated by extracting an arbitrary component from the DCT coefficient D, it is necessary for the image data restoration device to DCT transform the restored reduced image to obtain the reduction coefficient. Based on the reduction coefficient, the DCT coefficients D in 8 rows and 8 columns are restored. Therefore, in this case, although it is possible to obtain a high compression ratio, it is not possible to reduce the amount of calculation required for the restoration process.

Hereinafter, a method of encoding and transmitting the reduction coefficient itself obtained by the extraction means 112 and restoring the reduction coefficient from the encoded data will be explained. Figure 15 shows
FIG. 3 is a configuration diagram of an embodiment of an image data compression device according to claim 3; In this way, the reverse DC shown in FIGS. 8 and 10 to 12
An image data compression device is configured except for the T-transformer 241 and the buffer 242, and the set of reduction coefficients generated by the extracting means 112 is used as image data, and transform encoding is configured similarly to the transform encoding means 114 described above. Means 11
5 may be configured.

In this case, the reduction coefficients corresponding to a plurality of blocks are collectively subjected to DCT transformation by the DCT transformation unit 261.
This conversion result is quantized by the linear quantization unit 262. At this time, the linear quantization unit 262 may perform the quantization process using a quantization threshold that does not reduce accuracy, depending on the magnitude of the value of the conversion result described above. Furthermore, the obtained quantization result is processed by the encoding unit 263.
The data is variable-length encoded in the same manner as before and sent out as encoded data.

[0059] In this way, encoded data containing only information regarding the components extracted by the extracting means 112 can be obtained, so image data can be efficiently compressed and the time required to transmit encoded data can be shortened. can do. Further, by applying the invention of claim 6, the extracting means 11
When 2 is configured to extract components corresponding to features of a multivalued image (see FIG. 12), the encoding unit 263 adds information regarding the extracted components to the encoded data as encoded information. Just send it.

However, in this case, the transform encoding means 115
The DCT transformation section 261, linear quantization section 262, and encoding section 263 need to have a configuration corresponding to the number of bits of data representing each component of the reduction coefficient. Therefore, as shown in FIG. 9, the function of the DCT converter 261 cannot be performed by the DCT converter 211. The encoded data obtained as described above is restored to image data by the image data restoration device shown in FIG.

In FIG. 16, the image data restoration device is
Decoding section 311, inverse quantization section 312, and inverse DCT transformation section 31
3, a rearrangement processing unit 332 corresponding to the coefficient recovery unit 162, and an inverse DCT transformation unit 333 corresponding to the inverse orthogonal transformation unit 153. In this image data restoring device, the reduction coefficient restoring means 161 includes the reduced image restoring means 1 described above.
Similarly to 51, the set of reduction coefficients is restored by decoding encoded data, inverse quantization, and inverse DCT transformation. Further, when encoded information is added to the encoded data, the decoding section 311 sends the encoded information to the rearrangement processing section 332.

By performing the above-mentioned rearrangement processing by the rearrangement processing unit 332, a DCT coefficient D containing only information regarding the selected spatial frequency component is obtained for each block based on the reduction coefficient, This DC
By subjecting the T coefficient D to inverse DCT transformation by the inverse DCT transformation unit 333, the image data of the corresponding block can be approximately restored.

Therefore, in this case, the reduction coefficient is inverse DCT
Compared to encoding the reduced image obtained by conversion,
Since it is possible to reduce the amount of calculations during compression processing and decompression processing, it is possible to obtain a high compression ratio and to speed up the decompression processing. Furthermore, since the restored DCT coefficient D includes many invalid coefficients as described above, when performing inverse DCT transformation on this DCT coefficient D,
By executing only calculations related to effective coefficients and omitting calculations related to invalid coefficients, it is possible to further reduce the time required for the restoration process. For example, the inverse DCT transform unit 333 may be configured using the technique disclosed in Japanese Patent Application No. 2-259484 entitled "Image Data Restoration Method and Image Data Restoration Apparatus" by the present applicant.

Furthermore, when a very small number of components are extracted from the original DCT coefficient D, the restored DCT coefficient D contains many invalid coefficients, so the technique described above is used to extract the inverse DCT coefficient D.
By configuring the CT transformation unit 333, the effect of shortening the time required for inverse DCT transformation processing is significant. In particular, when performing hierarchical restoration, it is desired to shorten the time required for restoration processing at low restoration stages, so the method of encoding reduction coefficients as described above is suitable.

FIG. 17 is a block diagram of an embodiment of the image data compression apparatus according to claim 7. In FIG. 17, the image data compression device is constructed by adding coefficient position information holding means 132 to the extraction means 112 shown in FIG. In this extracting means 112, the coefficient position information holding means 132 is configured to output coefficient position information in response to the input of hierarchy restoration information.
31 instructs the readout circuit 232 to read out the corresponding component.

The coefficient position information holding means 132 described above is
Information indicating a component to be extracted corresponding to hierarchical restoration information indicating a restoration stage may be stored as coefficient position information, and may be configured using a look-up table or the like. For example, when performing hierarchical restoration using the spectral selection method, coefficient position information indicating that only the DC component is extracted is stored in correspondence with the hierarchical restoration information indicating the first stage of hierarchical restoration, and In response to the hierarchy restoration information indicating
Coefficient position information indicating two AC components adjacent to each component may be stored. Similarly, coefficient position information corresponding to each stage may be stored.

In this case, in response to the input of hierarchical restoration information indicating each stage, the extracting means 112 generates reduction coefficients consisting only of the corresponding components, and this set of reduction coefficients is then processed by the transform encoding means 115. It is encoded and sent. In this case, since the components to be extracted are determined in accordance with the layer restoration information, the transform encoding means 115 may add this layer restoration information to the encoded data as encoded information and send it out.

The encoded data at each stage obtained in this way is restored by the image data restoration device shown in FIG. In FIG. 18, the image data restoration device is shown in FIG.
The image data restoring device shown in 6 includes the above-mentioned coefficient position information holding means 132, an image data holding section 341 that holds the image data restored in the previous stage, and an inverse DCT on the output of this image data holding section 341. It is configured by adding an adder 351 that adds the output of the converter 333.

In this case, the above-mentioned coefficient position information holding means 132 sends information regarding the position of the corresponding component to the rearrangement processing unit 332 in accordance with the layer restoration information added to the encoded data as encoded information. It may be configured as follows. In response, the rearrangement processing unit 332 sets each component of the reduction coefficient corresponding to each block as the component at the corresponding position, thereby restoring the DCT coefficients D in 8 rows and 8 columns. Since this DCT coefficient D includes only information regarding the spatial frequency component newly added at the stage of the corresponding layer restoration, the above-mentioned spatial An image whose gradation changes depending on the frequency component is obtained.

In addition, this image and the image held in the image data holding section 341 described above are input to the adder 351,
By inputting this addition result to the image data holding unit 341 again, it is possible to restore image data that includes information regarding all the spatial frequencies contained in the encoded data input so far. In this way, by repeating the restoration processing at each stage one after another, detailed image data can be restored step by step.

In this case, a memory with a large capacity is required as the image data holding section 341, but as described above, the DCT coefficient D input to the inverse DCT transformation section 333
contains many invalid coefficients, so by omitting the arithmetic processing regarding the invalid coefficients, it is possible to speed up the restoration process. On the other hand, if you do not want to increase the circuit scale of the image data restoration device, as shown in FIG.
2 may be provided to hold the DCT coefficient D of the previous stage. In this case, the adder 351 adds the DCT coefficient D obtained by the coefficient restoring unit 332 and the DCT coefficient D of the previous stage described above, and converts this addition result into an inverse DC
Input to the T conversion unit 333 and coefficient holding unit 342
The configuration may be such that the information is input to

In this case, since there is no need to provide a large capacity memory, the circuit scale of the image data restoration device does not increase. By the way, transform encoding methods such as the ADCT method described above are non-information preserving type encoding methods, and in the encoding process, part of the information regarding the set of reduced images and the set of reduction coefficients is lost, resulting in poor image quality. The deterioration of For this reason, image data compression devices and image data decompression devices that use the above-mentioned transform encoding method cannot perform compression that is higher than image quality, such as when performing hierarchical restoration using a transfer line with limited transmission speed. Suitable for applications where efficiency is a priority.

On the other hand, there are also applications that require a high compression rate and do not allow deterioration in image quality. In the following, a method for obtaining a high compression rate while maintaining image quality in order to support such uses will be described using an example in which a reduction coefficient is encoded. FIG. 20 is a block diagram of an embodiment of the image data compression apparatus according to claim 4. In FIG. 20, the image data compression apparatus includes predictive encoding means 122 in place of the transform encoding means 115 of the image data compression apparatus shown in FIG. This predictive encoding means 122, for example,
For each component of the reduction coefficient stored in the buffer 233 of the extraction means 112 described above, it is configured to include a difference circuit 291 that calculates a difference between previous values, and an encoding section 292 that performs variable length encoding on the obtained difference data. Bye.

In this case, the predictive encoding means 12 described above
2, the difference data of the reduction coefficients is encoded as is, so information regarding the reduction coefficients is encoded without being lost. Therefore, by encoding the reduction coefficient using the predictive encoding method, which is an information-preserving encoding method, it becomes possible to accurately restore the reduction coefficient on the restoration side, and this reduces the possibility of image quality deterioration. can be suppressed.

Also in this case, since the encoded data includes only the information extracted by the extracting means 112, a higher compression rate than the conventional ADCT method can be obtained. The encoded data obtained in this way can be restored by the image data restoration device shown in FIG.

In FIG. 21, the image data restoration device is
In place of the inverse quantization section 312 and the inverse DCT transformation section 313 shown in FIG. 16, the reduction coefficient restoring means 161 includes an integration circuit 314, and the output of the integration circuit 314 is sequentially stored in the buffer 321. And it is sufficient. In this case, since the above-mentioned difference data itself is obtained as the decoding result by the decoding unit 311, the reduction coefficient can be accurately restored by sequentially integrating the difference data by the integration circuit 314. Also, based on the reduction coefficient obtained in this way, the rearrangement processing unit 33
2, the DCT coefficient D of 8 rows and 8 columns is restored, and the inverse DC
By performing inverse DCT transform on this DCT coefficient D by the T transform unit 333, the image data of the corresponding block can be restored.

Here, the integration process by the integration circuit 314 is a simple arithmetic process, and includes inverse quantization processing and inverse DCT.
The amount of calculation is also very small compared to conversion processing. Therefore, when an image data compression device is configured with the predictive encoding means 115, it is possible to reduce the time required for the restoration process of the reduced image compared to the case where the above-mentioned transform encoding method is used. It is possible to speed up data restoration processing.

Furthermore, the transform encoding means 114 shown in FIG.
Instead, the image data compression device may be configured by including predictive encoding means 121 consisting of the above-described integration circuit 291 and encoding section 292. In this case, the set of reduced images obtained by the reduced image generation means 113 is input to the predictive encoding means 121 and encoded without losing information about the reduced images, so the image data restoration device side It becomes possible to accurately restore a set of reduced images.

As described above, the image data restoring device corresponding to the encoded data obtained by the image data compression device equipped with the predictive encoding means 121 uses the inverse quantization section of the reduced image restoring means 151 shown in FIG. 312 and the inverse DCT conversion section 313, an integration circuit 314 may be provided.

[0080]

[Effects of the Invention] As explained above, the present invention re-encodes as image data a set of reduction coefficients consisting only of components extracted from a coefficient matrix or a set of reduced images obtained from these reduction coefficients. , it becomes possible to obtain encoded data that includes only information regarding the selected spatial frequency component, and the time required to transmit the encoded data can be shortened, thereby increasing the speed of image data restoration processing.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the configuration of the invention according to claim 1.

FIG. 2 is a diagram showing the configuration of the invention according to claim 2.

FIG. 3 is a diagram showing the configuration of the invention according to claim 3.

FIG. 4 is a diagram showing the configuration of the invention according to claim 4.

FIG. 5 is a diagram showing the configuration of the invention according to claim 9.

FIG. 6 is a diagram showing the configuration of the invention according to claim 10.

FIG. 7 is a diagram showing the configuration of the invention according to claim 11.

FIG. 8 is a configuration diagram of an embodiment of an image data compression device of the present invention.

FIG. 9 is a configuration diagram of another embodiment of the image data compression device of the present invention.

FIG. 10 is a configuration diagram of an embodiment of an image data compression device according to an eighth aspect of the invention.

FIG. 11 is a configuration diagram of another embodiment of the image data restoration device of the invention according to claim 8.

FIG. 12 is a configuration diagram of an embodiment of an image data compression device according to a sixth aspect of the invention.

FIG. 13 is a configuration diagram of an embodiment of an image data restoring device according to a ninth aspect of the invention.

FIG. 14 is a configuration diagram of an embodiment of an image data restoration device according to a tenth aspect of the invention.

FIG. 15 is a configuration diagram of an embodiment of an image data compression device according to the third aspect of the invention.

FIG. 16 is a configuration diagram of an embodiment of an image data restoration device according to an eleventh aspect of the invention.

FIG. 17 is a configuration diagram of an embodiment of an image data compression device according to a seventh aspect of the invention.

FIG. 18 is a configuration diagram of another embodiment of the image data restoration device according to the eleventh aspect of the invention.

FIG. 19 is a configuration diagram of another embodiment of the image data restoration device according to the eleventh aspect of the invention.

FIG. 20 is a configuration diagram of an embodiment of an image data compression device of the invention according to claim 4;

FIG. 21 is a configuration diagram of another embodiment of the image data restoration device according to the eleventh aspect of the invention.

FIG. 22 is a configuration diagram of a conventional image data compression device.

FIG. 23 is a diagram showing an example of blocks.

FIG. 24 is a diagram showing an example of DCT coefficients D. FIG.

FIG. 25 is a diagram showing a quantization matrix.

FIG. 26 is a diagram showing a quantization coefficient DQU.

FIG. 27 is an explanatory diagram of zigzag scan.

FIG. 28 is a configuration diagram of a conventional image data restoration device.

FIG. 29 is an explanatory diagram of basis vectors.

[Explanation of symbols]

111 Orthogonal transformation means 112 Extraction means 113 Reduced image generation means 114, 115 Transformation encoding means 121, 122 Predictive encoding means 131 Feature analysis means 132 Coefficient position information holding means 141 Quantization means 142 Dequantization means 151 Reduced image restoration means 152, 162 Coefficient restoration means 153 Inverse orthogonal transformation means 154 Block restoration means 161 Reduction coefficient restoration means 211, 261, 331, 611 DCT transformation section 22
1,233,242,321 Buffer 231 Address calculation section 232 Read circuit 234,262,620 Linear quantization section 235 Multiplier 236,623 Quantization threshold holding section 241,313,333,731 Inverse DCT transformation section 2
63,292,631 Encoding unit 271 Selector 272 Demultiplexer 281 Distribution determining unit 282 Feature storage unit 283 Extraction instruction unit 291 Difference circuit 311, 711 Decoding unit 312, 720 Dequantization unit 314 Integration circuit 332 Reordering unit 334 Pixel data Input section 335 Arithmetic processing section 341 Image data holding section 351 Adder 621 DCT coefficient input section 622 Divider 632 Code table 712 Decoding table

Claims (11)

[Claims] 1. Orthogonal transformation means (111) for orthogonally transforming image data representing a multivalued image for each block of M×N pixels to obtain a coefficient matrix of M rows and N columns each consisting of transformation coefficients; Extraction means (112) for extracting predetermined components from the coefficient matrix corresponding to each block and generating K rows and L columns of reduction coefficients consisting only of the extracted components.
and a reduced image generating means (113) that performs inverse orthogonal transformation on the reduction coefficient to generate a reduced image consisting of K×L pixels corresponding to each block; An image data compression device characterized by comprising: conversion encoding means (114) for converting and encoding data and sending it out as encoded data. 2. The image data compression apparatus according to claim 1, wherein the transform encoding means (114) is replaced by predictive encoding means (114) for predictively encoding image data consisting of reduced images corresponding to each block. 121). 3. Orthogonal transformation means (111) for orthogonally transforming image data representing a multivalued image for each block to obtain the coefficient matrix, and extracting a plurality of components from the coefficient matrix corresponding to each block. extraction means (112
); and transform encoding means (115) for transform-coding a set of reduction coefficients corresponding to each block as image data and transmitting it as encoded data. 4. The image data compression apparatus according to claim 3, wherein the transform encoding means (115) is replaced by a predictive encoding means (115) for predictively encoding image data consisting of reduction coefficients corresponding to each block. 122). 5. The image data compression device according to claim 1, wherein the extraction means (11
2) is an image data compression device characterized in that the DC component of the coefficient matrix and the surrounding AC components are extracted as the predetermined component to generate a reduction coefficient. 6. The image data compression device according to claim 1, wherein the extraction means (11
2) is a feature analysis means (131) that analyzes the features of the multivalued image based on the coefficient matrix corresponding to each block.
), and is configured to determine a component to be extracted from the coefficient matrix according to an analysis result by the feature analysis means (131). 7. The image data compression device according to claim 1, wherein the extraction means (11
2) is a coefficient position information holding means (132) that outputs coefficient position information regarding the position of the component to be extracted on the coefficient matrix in response to input of hierarchy restoration information indicating the stage of hierarchy restoration.
An image data compression device comprising: a configuration that determines a component to be extracted from the coefficient matrix in accordance with input of the coefficient position information. 8. The image data compression device according to claim 1 or 3, wherein the extraction means (11
2) is a quantization means (2) which obtains quantized coefficients by quantizing the input coefficient matrix components using the corresponding quantization threshold.
141) and dequantization means (142) for dequantizing the quantized coefficients using a corresponding quantization threshold, and a reduction coefficient made of a transform coefficient obtained by dequantizing the quantized coefficients. An image data compression device characterized by being configured to generate image data. [Claim 9] Based on input encoded data,
reduced image restoring means (15) for restoring the set of reduced images;
1); coefficient restoring means (152) for restoring an M-by-N coefficient matrix corresponding to each block based on each of the reduced images; Inverse orthogonal transformation means (15) for restoring image data
3) An image data restoration device comprising: 10. Reduced image restoring means (1) for restoring the set of reduced images based on input encoded data.
51), and block restoring means (51) for estimating pixel data of other pixels from pixel data constituting each of the reduced images based on the encoded information and restoring the corresponding block.
154). 11. Reduction coefficient restoring means (161) for restoring the set of reduction coefficients corresponding to each block based on input encoded data; coefficient restoring means (162) for allocating positions on the matrix and restoring M-by-N coefficient matrices, and performing inverse orthogonal transformation on each of the coefficient matrices;
Inverse orthogonal transformation means (1
53) An image data restoration device comprising: