JP2002135596A - Image processing apparatus, image processing method and image data transfer method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To decompress compression image data in the unit of blocks. SOLUTION: Read addresses of color difference components U, V in a first block in a horizontal direction are stored in a start address memory M3. Huffman decoding, inverse quantization and inverse DCT are applied to a luminance component Y and the color difference components U, V for each block. Up-sampling is applied to decompressed image data of the color difference components U, V to extract a part corresponding to the decompressed image data of the luminance component Y. Color space conversion and color number reduction are applied to the luminance component Y and the color difference components U, V to obtain a decompressed image of a concerned block. When the decompression processing of the block by one line in the horizontal direction is finished, the coded data of the color difference components U, V are read again on the basis of the address information stored in the start address memory M3. Then the similar decompression processing to above is repeated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an image processing apparatus,
The present invention relates to an image processing method and an image data transfer method. More specifically, the present invention relates to an image processing apparatus and an image processing method for expanding progressively encoded compressed image data, and an image data transfer method applicable to the image processing apparatus and the image processing method.

[0002]

2. Description of the Related Art (Prior Art 1) There is an image processing apparatus for decompressing progressively encoded image data in units of blocks (see Japanese Patent Application No. 2000-273752). In this image processing apparatus, the following processing is performed for each component (luminance component, color difference component) of the compressed image data.

(1) The encoded data of each stage for a certain block is extracted from the compressed image data.

(2) Huffman decoding is performed on the extracted encoded data at each stage.

(3) The DCT coefficient data of the above block is obtained by combining the DCT coefficient data of each stage obtained by Huffman decoding.

(4) Inverse DCT is performed on the DCT coefficient data of the block.

(5) The color space of the expanded image data of the block obtained by the inverse DCT transform is converted from YUV (luminance component, color difference component) to RGB (three primary color components).

Thus, a color expanded image of the block is obtained. Then, the same processing is performed on all the blocks to obtain a color expanded image for one screen.

(Prior Art 2) In a compression coding method such as JPEG, Huffman coding is sometimes performed after downsampling the color difference component of the original image data to reduce the number of pixels in order to improve the compression ratio. In this case,
FIG. 36 shows a processing flow when the original image data is compressed by the progressive encoding method. Here, the luminance component Y and the color difference components U and V of the original image data are respectively divided into four blocks BY1-BY4.
It is assumed that it is divided into BU1-BU4, BV1-BV4.

Referring to FIG. 36, first, color difference components U, V
Is down-sampled by に in the vertical direction on the screen. That is, the number of pixels of the color difference components U and V is reduced to half in the vertical direction on the screen. Accordingly, the number of blocks of each of the color difference components U and V is two (BU13, BU24, B
V13, BV24). On the other hand, the number of blocks of the luminance component Y remains four (BY1-BY4).

Next, DCT transformation is performed on the luminance component Y and the color difference components U and V after downsampling. Then, the obtained DCT coefficient data is divided into a plurality of stages (here, n stages) for each block. Note that FIG.
In 6, the DCT coefficient data of the i-th stage of the block X is represented as “Xi”. Therefore, for example, the first-stage DCT coefficient data of the block BY1 is “BY1-
1 ".

Next, the DCT coefficient data is Huffman-coded for each stage. The coding order of the DCT coefficient data in each stage is as follows.
D of block BY2, block BY3, and block BY4
The order is the CT coefficient data. In the color difference components U and V, the DCT coefficient data of the blocks BU13 and BV13,
The order is the DCT coefficient data of the blocks BU24 and BV24.

As described above, compressed image data as shown in FIG. 36 is obtained.

[0014]

The flow of processing when the compressed image data progressively encoded as described in the above-mentioned prior art 2 is decompressed by the image processing apparatus shown in the prior art 1 is shown in FIG. 37.

Referring to FIG. 37, first, coded data BY1-1 of each stage for block BY1 of luminance component Y
To BY1-n and blocks BU13,
Encoded data BU13-1 to BU13-1 of each stage for BV13
BU13-n and BV13-1 to BV13-n are extracted from the compressed image data. Then, Huffman decoding is performed on the extracted encoded data at each stage.

Next, DCT coefficient data of each stage obtained by Huffman decoding is combined to form a block BY.
1, DC13 coefficient data of BU13 and BV13 are obtained.

Next, the blocks BY1, BU13, BV
The inverse DCT transform is performed on the thirteen DCT coefficient data.

Next, the color difference components U and V are up-sampled twice in the vertical direction on the screen. That is, the number of pixels of the color difference components U and V in the vertical direction on the screen is restored. As a result, the number of blocks of each of the color difference components U and V becomes two (BU1, BU3, BV1, and BV3).

As described above, for the color difference components U and V,
Blocks (BU1, BU3), (BV1, BV
3) can be extended together. On the contrary,
For the luminance component Y, only the block BY1 can be expanded first. And the block BY
The block of the luminance component Y expanded after 1 is the block BY2. Blocks BU3 and BV of color difference components U and V
The block BY3 corresponding to No. 3 is expanded because all blocks (BY1, B1) corresponding to one horizontal line on the screen are expanded.
After elongation of Y2).

As described above, when the compressed image data progressively encoded by down-sampling the color difference components U and V in the vertical direction on the screen and decompressing by the image processing apparatus shown in the prior art 1, Since the order of the blocks to be expanded is different between the luminance component Y and the color difference components U and V, the compressed image data cannot be expanded in block units.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and an object of the present invention is to provide an image processing apparatus capable of decompressing progressively encoded image data in units of blocks. It is.

[0022]

According to one aspect of the present invention, an image processing apparatus is configured to apply a second component to an original image including a first component and a second component in a vertical direction on a screen. The coefficient data obtained by performing orthogonal transformation on the first component and the second component after the down sampling for each block composed of a predetermined number of pixels is divided into a plurality of stages. Dividing and expanding compressed image data obtained by entropy-encoding coefficient data for each stage, comprising: extracting means, entropy decoding means, combining means, inverse orthogonal transform means, The image processing apparatus includes a restoration unit and a block extraction unit.

The extracting means converts the encoded data of each stage of the first component and the encoded data of each stage of the second component into a compressed image for each block of one line in the horizontal direction on the screen. Extract from data. The entropy decoding means performs entropy decoding on the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extracting means. The combining means combines the coefficient data of each stage of the first component obtained by the entropy decoding means to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding means. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component. The inverse orthogonal transform unit performs an inverse orthogonal transform on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combination unit. The restoring unit is a second unit that is obtained by the inverse orthogonal transform unit.
Is up-sampled in the vertical direction on the screen. The block extracting means extracts a portion corresponding to the decompressed image data of each block of the first component obtained by the inverse orthogonal transform means from the decompressed image data of the second component after being upsampled by the restoration means. I do.

After extracting the coded data at each stage for all the blocks for one line in the horizontal direction on the screen, the extracting means extracts the next one line from the one line for the first component. For each block, the coded data of each stage is extracted for each block, and for the second component, the coded data of each stage is again extracted for each block of the one-line block.

Preferably, the image processing apparatus further comprises:
It has storage means. The storage means, for the second component,
The position where the encoded data of each stage for the first block on the line is stored is stored. Further, the extracting means extracts the coded data of each stage for all the blocks on the line, and then, for each block on the line, for the second component based on the position stored in the storing means. The encoded data of each stage is extracted.

In the above-described image processing apparatus, first, the expanded image data of the first component and the second image of the portion corresponding to the expanded image data of the first component are obtained for each block of one horizontal line on the screen. The expanded image data of the component is obtained. Next, the expanded image data of the first component and the expanded image data of the second component corresponding to the expanded image data of the first component are obtained for each block of one line following the one line. As described above, according to the image processing apparatus, the progressively encoded compressed image data can be decompressed in block units.

According to another aspect of the present invention, an image processing apparatus down-samples a second component in a vertical direction on a screen with respect to an original image including a first component and a second component, The coefficient data obtained by performing the orthogonal transformation on the first component and the second component after down-sampling is divided into a plurality of stages for each block composed of a predetermined number of pixels. The compression image data obtained by entropy-encoding the coefficient data for each is expanded, and the extraction means, the entropy decoding means, the buffer memory, the combination means, the inverse orthogonal transform means, Means.

The extracting means is composed of 1 MCU (Minimum).
(Coded Unit) × For blocks included in one horizontal line on the screen, the coded data of each stage of the first component and the coded data of each stage of the second component are extracted from the compressed image data for each block. I do. The entropy decoding means performs entropy decoding on the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extracting means. The buffer memory is the first memory obtained by the entropy decoding means.
The coefficient data for the blocks other than the block on the last line (the bottom line in the vertical direction on the screen) among the coefficient data at each stage of the component is stored. The combination means combines the coefficient data for each block on the last line obtained by the entropy decoding means with respect to the first component to obtain coefficient data of the block, and constitutes one MCU together with the block. Are read from the buffer memory, and the read coefficient data are combined to obtain coefficient data of a block constituting one MCU.
On the other hand, the combination unit combines the coefficient data of each stage obtained by the entropy decoding unit with respect to the second component to obtain coefficient data of each block. The inverse orthogonal transform unit performs an inverse orthogonal transform on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combination unit. The restoration means up-samples the decompressed image data of each block of the second component obtained by the inverse orthogonal transformation means in the vertical direction on the screen.

In the above-described image processing apparatus, decompressed image data of the first component and decompressed image data of the second component of a block included in one MCU are obtained. That is, 1 MCU
In each case, decompressed image data is obtained. As described above, according to the image processing apparatus, the progressively encoded compressed image data can be expanded for each MCU.

According to still another aspect of the present invention, the image processing apparatus down-samples the second component in the vertical direction on the screen with respect to the original image including the first component and the second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. Expands the compressed image data obtained by entropy-encoding the coefficient data for each stage, and includes extraction means, entropy decoding means, combination means, inverse orthogonal transform means, buffer memory, Restoration means. The extracting means converts the encoded data of each stage of the first component and the encoded data of each stage of the second component into compressed image data for each block of one MCU × one horizontal line on the screen. Extract from The entropy decoding means performs entropy decoding on the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extracting means. The combining means combines the coefficient data of each stage of the first component obtained by the entropy decoding means to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding means. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component.
The inverse orthogonal transform unit performs an inverse orthogonal transform on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combination unit. The buffer memory is the first memory obtained by the inverse orthogonal transform means.
The expanded image data for the blocks other than the block on the last line among the expanded image data of the component is stored.
The restoration means up-samples the decompressed image data of each block of the second component obtained by the inverse orthogonal transformation means in the vertical direction on the screen.

When the decompressed image data for each block on the last line is obtained by the inverse orthogonal transform means for the first component, the image processing apparatus decompresses the block constituting one MCU together with the block. Reads image data from the buffer memory.

In the above image processing apparatus, the first component included in one MCU is expanded by the expanded image data read from the buffer memory and the expanded image data of the block on the last line obtained by the inverse orthogonal transform means. Image data is obtained. Further, decompressed image data of the second component included in the one MCU is obtained. As described above, according to the image processing apparatus, the progressively encoded compressed image data can be expanded for each MCU.

According to still another aspect of the present invention, the image processing apparatus downsamples the second component in the vertical direction on the screen with respect to the original image including the first component and the second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. It expands the compressed image data obtained by entropy-encoding the coefficient data for each stage, and includes extracting means, buffer memory, entropy decoding means, combination means, inverse orthogonal transform means, Restoration means.

The extraction means extracts, for each block, 1 MCU × the coded data of each stage of the first component and the coded data of each stage of the second component for each block included in one horizontal line on the screen. Extract from compressed image data. The buffer memory stores encoded data for blocks other than the block on the last line among the encoded data of the first component extracted by the extracting unit. The entropy decoding unit encodes, for the first component, the encoded data for each block on the last line extracted by the extraction unit and the encoded data stored in the buffer memory for the block that constitutes one MCU together with the block. While entropy decoding is performed on the data, the encoded data extracted by the extraction unit for the second component is entropy decoded. The combination means
The coefficient data of each block of the first component is obtained by combining the coefficient data of each stage of the first component obtained by the entropy decoding means, and each of the second components obtained by the entropy decoding means is obtained. The coefficient data of each block of the second component is obtained by combining the coefficient data of the stages. The inverse orthogonal transform unit performs an inverse orthogonal transform on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combination unit. The restoring unit is a second unit that is obtained by the inverse orthogonal transform unit.
Is up-sampled in the vertical direction on the screen.

In the above-described image processing apparatus, the first data contained in one MCU includes the encoded data of the first component read from the buffer memory and the encoded data of the block on the last line extracted by the extracting means. The encoded data of the component is obtained. Then, entropy decoding, combination of coefficient data, and inverse orthogonal transform are performed on the encoded data of the first component included in this 1 MCU. Thereby, decompressed image data of the first component included in the one MCU is obtained. In addition, the 1MC
The expanded image data of the second component included in U is obtained.
As described above, according to the image processing apparatus, the progressively encoded compressed image data can be expanded for each MCU.

According to still another aspect of the present invention, the image processing apparatus down-samples the second component in the vertical direction on the screen with respect to the original image including the first component and the second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. It expands the compressed image data obtained by entropy-encoding the coefficient data for each stage, and includes extraction means, entropy decoding means, buffer memory, combination means, inverse orthogonal transform means, The image processing apparatus includes a restoration unit and a block extraction unit.

The extracting means converts the encoded data of each stage of the first component and the encoded data of each stage of the second component for each block of one horizontal line on the screen into a compressed image. Extract from data. The entropy decoding means performs entropy decoding on the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extracting means. The buffer memory stores the encoded data of each stage of the second component obtained by the entropy decoding means. The combining means combines the coefficient data of each stage of the first component obtained by the entropy decoding means to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding means. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component. The inverse orthogonal transform unit performs an inverse orthogonal transform on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combination unit. The restoring unit is a second unit that is obtained by the inverse orthogonal transform unit.
Is up-sampled in the vertical direction on the screen. The block extracting means extracts a portion corresponding to the decompressed image data of each block of the first component obtained by the inverse orthogonal transform means from the decompressed image data of the second component after being upsampled by the restoration means. I do.

After extracting the encoded data at each stage for all the blocks for one line in the horizontal direction on the screen, the extracting means extracts the first component for the next one line after the one line. The encoded data at each stage is extracted for each block. Further, the combination means obtains coefficient data for all blocks of one line in the horizontal direction on the screen, and then converts the encoded data of each stage of each block stored in the buffer memory for the second component. The coefficient data of each block is obtained by combination.

In the above-described image processing apparatus, first, the expanded image data of the first component and the second image of the portion corresponding to the expanded image data of the first component for each block of one line in the horizontal direction on the screen. The expanded image data of the component is obtained. Next, decompressed image data of the first component is obtained for each block of one line following the one line. On the other hand, the encoded data of each block stored in the buffer memory is read, and combination processing, inverse orthogonal transformation processing, upsampling, and block extraction processing are performed. As a result, the first line for the next one line after the one line
The expanded image data of the second component corresponding to the expanded image data of the component is obtained. As described above, according to the image processing apparatus, the progressively encoded compressed image data can be decompressed in block units.

According to still another aspect of the present invention, the image processing apparatus down-samples the second component in the vertical direction on the screen with respect to the original image including the first component and the second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. Expands the compressed image data obtained by entropy-encoding the coefficient data for each stage, and includes extraction means, entropy decoding means, combination means, inverse orthogonal transform means, buffer memory, The image processing apparatus includes a restoration unit and a block extraction unit.

The extracting means converts the encoded data of each stage of the first component and the encoded data of each stage of the second component for each block of one horizontal line on the screen into a compressed image. Extract from data. The entropy decoding means performs entropy decoding on the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extracting means. The combining means combines the coefficient data of each stage of the first component obtained by the entropy decoding means to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding means. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component. The inverse orthogonal transform unit performs an inverse orthogonal transform on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combination unit. The buffer memory stores the decompressed image data of each block of the second component obtained by the inverse orthogonal transform means. The reconstructing means is a second means for reconstructing the second orthogonal transform
Is up-sampled in the vertical direction on the screen. The block extracting means extracts a portion corresponding to the decompressed image data of each block of the first component obtained by the inverse orthogonal transform means from the decompressed image data of the second component after being upsampled by the restoration means. I do.

After extracting the coded data at each stage for all the blocks for one line in the horizontal direction on the screen, the extracting means extracts the first component for the next one line after the one line. The encoded data at each stage is extracted for each block. In addition, the restoration means upsamples the decompressed image data for all the blocks of the second component for one line in the horizontal direction on the screen, and then decompresses each block of the second component stored in the buffer memory. The image data is up-sampled in the vertical direction on the screen.

In the above-described image processing apparatus, first, for each block of one line in the horizontal direction on the screen, the expanded image data of the first component and the second image of the portion corresponding to the expanded image data of the first component are obtained. The expanded image data of the component is obtained. Next, decompressed image data of the first component is obtained for each block of one line following the one line. On the other hand, the decompressed image data of each block of the second component stored in the buffer memory is read, and upsampling and block extraction processing are performed. As a result, decompressed image data of the second component corresponding to decompressed image data of the first component for one line following the one line is obtained. As described above, according to the image processing apparatus, the progressively encoded compressed image data can be decompressed in block units.

According to still another aspect of the present invention, the image processing apparatus down-samples the second component in the vertical direction on the screen with respect to the original image including the first component and the second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. It expands the compressed image data obtained by entropy-encoding the coefficient data for each stage, and includes extracting means, buffer memory, entropy decoding means, combination means, inverse orthogonal transform means, The image processing apparatus includes a restoration unit and a block extraction unit.

The extracting means converts the encoded data of each stage of the first component and the encoded data of each stage of the second component into a compressed image for each block of one horizontal line on the screen. Extract from data. The buffer memory stores the encoded data of each stage of the second component extracted by the extracting means. The entropy decoding means performs entropy decoding on the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extracting means. The combining means combines the coefficient data of each stage of the first component obtained by the entropy decoding means to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding means. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component. The inverse orthogonal transform unit performs an inverse orthogonal transform on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combination unit. The restoration means up-samples the decompressed image data of each block of the second component obtained by the inverse orthogonal transformation means in the vertical direction on the screen. The block extracting unit is configured to include, among the expanded image data of the second component after being up-sampled by the restoring unit, the first image obtained by the inverse orthogonal transform unit.
The part corresponding to the expanded image data of each block of the component is extracted.

After extracting the encoded data at each stage for all the blocks for one line in the horizontal direction on the screen, the extracting means extracts the first component for the next one line after the one line. The encoded data at each stage is extracted for each block. Further, the entropy decoding means performs entropy decoding of the encoded data of each stage for all the blocks for one line in the horizontal direction on the screen, and then, for each of the blocks stored in the buffer memory, for the second component. Is entropy-decoded for each stage of the encoded data.

In the image processing apparatus, first, the expanded image data of the first component and the second image of the portion corresponding to the expanded image data of the first component for each block of one horizontal line on the screen are first obtained. The expanded image data of the component is obtained. Next, decompressed image data of the first component is obtained for each block of one line following the one line. On the other hand, the encoded data of each block stored in the buffer memory is read, and entropy decoding, combination processing, inverse orthogonal transformation processing, upsampling, and block extraction processing are performed. As a result, decompressed image data of the second component corresponding to decompressed image data of the first component for one line following the one line is obtained. As described above, according to the image processing apparatus, the progressively encoded compressed image data can be decompressed in block units.

According to yet another aspect of the present invention, an image processing method downsamples a second component in a vertical direction on a screen from an original image including a first component and a second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. Decompressing the compressed image data obtained by entropy-encoding the coefficient data for each stage, the extracting step, the entropy decoding step, the combination step, the inverse orthogonal transformation step, the restoration step, And a block extracting step.

In the extraction step, the horizontal 1
For each block of lines, encoded data of each stage of the first component and encoded data of each stage of the second component are extracted from the compressed image data for each block. In the entropy decoding step, the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extraction step are entropy-decoded. In the combining step, the coefficient data of each stage of the first component obtained by the entropy decoding step are combined to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding step is obtained. By combining the coefficient data of each stage of the component
The coefficient data of each block of the component is obtained. In the inverse orthogonal transform step, the first orthogonal transform obtained by the combining step is performed.
The inverse orthogonal transform is performed on the coefficient data of each block of the component and the coefficient data of each block of the second component. In the restoration step, decompressed image data of each block of the second component obtained in the inverse orthogonal transformation step is up-sampled in the vertical direction on the screen. In the block extraction step, a portion corresponding to the decompressed image data of each block of the first component obtained in the inverse orthogonal transformation step is extracted from the decompressed image data of the second component after being upsampled in the restoration step. I do.

After extracting the encoded data of each stage for all the blocks of one line in the horizontal direction on the screen, the first component is added to the next one of the one line.
The encoded data of each stage is extracted for each block of the line block, and the encoded data of each stage is extracted for each block of the one-line block again for the second component.

In the above-described image processing method, first, the expanded image data of the first component and the second image of the portion corresponding to the expanded image data of the first component are obtained for each block of one line in the horizontal direction on the screen. The expanded image data of the component is obtained. Next, the expanded image data of the first component and the expanded image data of the second component corresponding to the expanded image data of the first component are obtained for each block of one line following the one line. As described above, according to the image processing method, the progressively encoded compressed image data can be expanded in units of blocks.

According to still another aspect of the present invention, an image processing method downsamples a second component in a vertical direction on a screen with respect to an original image including a first component and a second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. Decompressing compressed image data obtained by entropy-encoding coefficient data for each stage, comprising an extraction step, an entropy decoding step, a storage step, a combination step, an inverse orthogonal transformation step, Restoring step.

In the extracting step, for each block included in one MCU × one horizontal line on the screen, the encoded data of each stage of the first component and the second
The encoded data of each stage of the component is extracted from the compressed image data. In the entropy decoding step, the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extraction step are entropy-decoded. In the storing step, of the coefficient data of each stage of the first component obtained in the entropy decoding step, coefficient data for blocks other than the block on the last line is stored in the buffer memory. In the combining step, for the first component, coefficient data of each block on the last line obtained by the entropy decoding step is combined to obtain coefficient data of the block, and a block constituting one MCU together with the block Is read from the buffer memory, and the read coefficient data is combined to obtain coefficient data of the block constituting the one MCU, while the second component is obtained by the entropy decoding step. The coefficient data of each block is obtained by combining the coefficient data of the stages. In the inverse orthogonal transformation step, the coefficient data of each block of the first component obtained by the combination step and the second
Inverse orthogonal transform is performed on the coefficient data of each block of the component. In the restoration step, decompressed image data of each block of the second component obtained in the inverse orthogonal transformation step is up-sampled in the vertical direction on the screen.

In the above image processing method, decompressed image data of the first component and decompressed image data of the second component of a block included in one MCU are obtained. That is, 1 MCU
In each case, decompressed image data is obtained. As described above, according to the image processing method, the progressively encoded compressed image data can be expanded for each MCU.

According to still another aspect of the present invention, an image processing method downsamples a second component in a vertical direction on a screen with respect to an original image including a first component and a second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. Expands the compressed image data obtained by entropy-encoding the coefficient data for each stage, and includes an extraction step, an entropy decoding step, a combination step, an inverse orthogonal transformation step, a storage step, Restoring step.

In the extraction step, for each block included in one MCU × one horizontal line on the screen, the encoded data of each stage of the first component and the second
The encoded data of each stage of the component is extracted from the compressed image data. In the entropy decoding step, the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extraction step are entropy-decoded. In the combining step, the coefficient data of each stage of the first component obtained by the entropy decoding step are combined to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding step is obtained. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component.
In the inverse orthogonal transform step, inverse orthogonal transform is performed on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained in the combining step. In the storing step, expanded image data for blocks other than the block on the last line among the expanded image data of the first component obtained in the inverse orthogonal transformation step is stored in the buffer memory. In the restoration step, decompressed image data of each block of the second component obtained in the inverse orthogonal transformation step is up-sampled in the vertical direction on the screen.

When the decompressed image data for each block on the last line is obtained by the inverse orthogonal transformation step for the first component, 1
The decompressed image data of the blocks constituting one MCU is read from the buffer memory.

In the above image processing method, the decompression of the first component included in one MCU is performed by using the decompressed image data read from the buffer memory and the decompressed image data of the block on the last line obtained by the inverse orthogonal transformation step. Image data is obtained. In addition, the 1MC
The expanded image data of the second component included in U is obtained.
As described above, according to the image processing method, the progressively encoded compressed image data can be expanded for each MCU.

According to still another aspect of the present invention, an image processing method downsamples a second component in a vertical direction on a screen from an original image including a first component and a second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. Expands the compressed image data obtained by entropy-encoding the coefficient data for each stage, and includes an extraction step, a storage step, an entropy decoding step, a combination step, an inverse orthogonal transformation step, Restoring step.

In the extraction step, for each block included in one horizontal line on the MCU × 1 MCU, the coded data of each stage of the first component and the second
The encoded data of each stage of the component is extracted from the compressed image data. In the storing step, among the encoded data of the first component obtained in the extracting step, encoded data for blocks other than the block on the last line is stored in the buffer memory. In the entropy decoding step, for the first component, the encoded data for each block on the last line extracted by the extraction step and the encoded data for a block stored in the buffer memory and constituting the one MCU together with the block are stored. And entropy decoding the encoded data extracted by the extraction step for the second component. In the combining step, the coefficient data of each stage of the first component obtained by the entropy decoding step are combined to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding step is obtained. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component. In the inverse orthogonal transform step, inverse orthogonal transform is performed on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained in the combining step. In the restoration step, decompressed image data of each block of the second component obtained in the inverse orthogonal transformation step is up-sampled in the vertical direction on the screen.

In the above-described image processing method, the first MCU included in one MCU includes the encoded data of the first component read from the buffer memory and the encoded data of the block on the last line extracted in the extraction step. The encoded data of the component is obtained. Then, the first MCU included in this one MCU
Is subjected to entropy decoding, combination of coefficient data, and inverse orthogonal transform.
Thereby, decompressed image data of the first component included in the one MCU is obtained. In addition, decompressed image data of the second component included in the one MCU is obtained. As described above, according to the image processing method, the progressively encoded compressed image data can be expanded for each MCU.

According to yet another aspect of the present invention, an image processing method downsamples a second component in a vertical direction on a screen from an original image including a first component and a second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. Decompressing compressed image data obtained by entropy-encoding coefficient data for each stage, comprising an extraction step, an entropy decoding step, a storage step, a combination step, an inverse orthogonal transformation step, It includes a restoration step and a block extraction step.

In the extraction step, the horizontal 1
For each block of lines, encoded data of each stage of the first component and encoded data of each stage of the second component are extracted from the compressed image data for each block. In the entropy decoding step, the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extraction step are entropy-decoded. In the storage step, the encoded data of each stage of the second component obtained in the entropy decoding step is stored in the buffer memory. In the combining step, the coefficient data of each block of the first component obtained by the entropy decoding step is combined to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding means is obtained. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component. In the inverse orthogonal transform step, inverse orthogonal transform is performed on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained in the combining step. In the restoration step, decompressed image data of each block of the second component obtained in the inverse orthogonal transformation step is up-sampled in the vertical direction on the screen. In the block extraction step, a portion corresponding to the decompressed image data of each block of the first component obtained in the inverse orthogonal transformation step is extracted from the decompressed image data of the second component after being upsampled in the restoration step. I do.

After extracting the coded data of each stage for all the blocks of one line in the horizontal direction on the screen, for the first component, each block for the block of one line next to the one line is extracted. The encoded data of each stage is extracted every time. Also, after obtaining coefficient data for all blocks for one line in the horizontal direction on the screen, the second component is combined with the encoded data of each stage of each block stored in the buffer memory, and Is obtained.

In the above-described image processing method, first, the expanded image data of the first component and the second image of the portion corresponding to the expanded image data of the first component for each block of one horizontal line on the screen are first obtained. The expanded image data of the component is obtained. Next, decompressed image data of the first component is obtained for each block of one line following the one line. On the other hand, the encoded data of each block stored in the buffer memory is read, and combination processing, inverse orthogonal transformation processing, upsampling, and block extraction processing are performed. As a result, the first line for the next one line after the one line
The expanded image data of the second component corresponding to the expanded image data of the component is obtained. As described above, according to the image processing method, the progressively encoded compressed image data can be expanded in units of blocks.

According to yet another aspect of the present invention, an image processing method downsamples a second component in a vertical direction on a screen from an original image including a first component and a second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. Expands the compressed image data obtained by entropy-encoding the coefficient data for each stage, and includes an extraction step, an entropy decoding step, a combination step, an inverse orthogonal transformation step, a storage step, It includes a restoration step and a block extraction step.

In the extraction step, one of the horizontal
For each block of lines, encoded data of each stage of the first component and encoded data of each stage of the second component are extracted from the compressed image data for each block. In the entropy decoding step, the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extraction step are entropy-decoded. In the combining step, the coefficient data of each block of the first component obtained by the entropy decoding step is combined to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding means is obtained. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component. In the inverse orthogonal transform step, inverse orthogonal transform is performed on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained in the combining step. In the storage step, the decompressed image data of each block of the second component obtained in the inverse orthogonal transformation step is stored in the buffer memory. In the restoration step, decompressed image data of each block of the second component obtained in the inverse orthogonal transformation step is up-sampled in the vertical direction on the screen. In the block extraction step, a portion corresponding to the decompressed image data of each block of the first component obtained in the inverse orthogonal transformation step is extracted from the decompressed image data of the second component after being upsampled in the restoration step. I do.

After extracting the encoded data at each stage for all the blocks of one line in the horizontal direction on the screen, the first component is converted to the blocks of the next one line after the one line. The encoded data of each stage is extracted every time.

After up-sampling the decompressed image data for all the blocks of the second component for one horizontal line on the screen, the decompressed image data of each block of the second component stored in the buffer memory is obtained. Is up-sampled in the vertical direction on the screen.

In the above-mentioned image processing method, first, for each block of one line in the horizontal direction on the screen, the expanded image data of the first component and the second image of the portion corresponding to the expanded image data of the first component are obtained. The expanded image data of the component is obtained. Next, decompressed image data of the first component is obtained for each block of one line following the one line. On the other hand, the decompressed image data of each block of the second component stored in the buffer memory is read, and upsampling and block extraction processing are performed. As a result, decompressed image data of the second component corresponding to decompressed image data of the first component for one line following the one line is obtained. As described above, according to the image processing method, the progressively encoded compressed image data can be expanded in units of blocks.

According to yet another aspect of the present invention, an image processing method downsamples a second component in a vertical direction on a screen with respect to an original image including a first component and a second component. , The first component and the second component after down-sampling, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages for each block composed of a predetermined number of pixels. Expands the compressed image data obtained by entropy-encoding the coefficient data for each stage, and includes an extraction step, a storage step, an entropy decoding step, a combination step, an inverse orthogonal transformation step, It includes a restoration step and a block extraction step.

In the extraction step, the horizontal 1
For each block of lines, encoded data of each stage of the first component and encoded data of each stage of the second component are extracted from the compressed image data for each block. In the storing step, the encoded data of each stage of the second component extracted in the extracting step is stored in the buffer memory. In the entropy decoding step, the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extraction step are entropy-decoded. In the combining step, the coefficient data of each block of the first component obtained by the entropy decoding step is combined to obtain coefficient data of each block of the first component, and the second data obtained by the entropy decoding means is obtained. The coefficient data of each block of the second component is obtained by combining the coefficient data of each stage of the component. In the inverse orthogonal transform step, inverse orthogonal transform is performed on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained in the combining step.
In the restoration step, decompressed image data of each block of the second component obtained in the inverse orthogonal transformation step is up-sampled in the vertical direction on the screen. In the block extraction step, a portion corresponding to the decompressed image data of each block of the first component obtained in the inverse orthogonal transformation step is extracted from the decompressed image data of the second component after being upsampled in the restoration step. I do.

After extracting the coded data at each stage for all the blocks for one line in the horizontal direction on the screen, for the first component, each block for the block for one line following the one line The encoded data of each stage is extracted every time.

After the entropy decoding of the encoded data at each stage for all the blocks for one line in the horizontal direction on the screen, the second component is obtained at each stage of each block stored in the buffer memory. Entropy decode the encoded data.

In the above-described image processing method, first, the expanded image data of the first component and the second image of the portion corresponding to the expanded image data of the first component for each block of one horizontal line on the screen are first obtained. The expanded image data of the component is obtained. Next, decompressed image data of the first component is obtained for each block of one line following the one line. On the other hand, the encoded data of each block stored in the buffer memory is read, and entropy decoding, combination processing, inverse orthogonal transformation processing, upsampling, and block extraction processing are performed. As a result, decompressed image data of the second component corresponding to decompressed image data of the first component for one line following the one line is obtained. As described above, according to the image processing method, the progressively encoded compressed image data can be expanded in units of blocks.

According to still another aspect of the present invention, an image data transfer method comprises the steps of: converting image data including a first component and a second component vertically down-sampled on a screen into a predetermined number of image data; The transfer is performed in units of blocks composed of pixels. And the horizontal one on the screen
After transferring the image data for one line, for the first component, the data for one line following the one line in the horizontal direction is transferred, and for the second component, the data for one line in the horizontal direction is transferred.
Transfer the data for the line again.

In the image data transfer method, first, image data for one line in the horizontal direction on the screen is transferred.
The second component of the image data corresponds not only to the portion corresponding to the data of the first component for the one line in the horizontal direction on the screen but also to the data of the first component for the next one line. Part is also included. Thus, for the first component, the data for one line following the one line in the horizontal direction is transferred, while for the second component, the data for one line in the horizontal direction is transferred again. As a result, the data of the first component for the next one line of the one line in the horizontal direction and the data of the second component including the corresponding portion are transferred.

As described above, in the above-described image data transfer method, the first component is to transfer the data of one line following the one line in the horizontal direction, while the second component is to be transferred in the horizontal direction. The image data including the first component and the second component including the portion corresponding to the first component can be transferred in order to transfer the one line of data again.

Preferably, the second component to be transferred is up-sampled in the vertical direction on the screen, a portion corresponding to the first component is extracted from the up-sampled second component, and the extracted portion is extracted. To transfer.

According to the above-described image data transfer method, image data corresponding to the first component and the second component can be transferred.

According to still another aspect of the present invention, an image data transfer method comprises the steps of: converting image data including a first component and a second component vertically down-sampled on a screen into a predetermined number of image data; The transfer is performed in units of blocks composed of pixels. And for the first component,
Data of a predetermined number of lines in the horizontal direction on the screen is accumulated in the buffer memory. Next, a second component, a first component on the next horizontal line on the screen with respect to the first component stored in the buffer memory, and a first component stored in the buffer memory. The first component on the next one line and the first component on the same vertical line on the screen are transferred.

Preferably, the number (predetermined number) of horizontal lines on the screen of the first component stored in the buffer memory is determined in a state before the second component of one block is down-sampled. The number is one less than the number of vertical blocks on the screen included.

In the above image data transfer method, for the first component, data for a predetermined number of horizontal lines on the screen is stored in the buffer memory, and then the second component is
A first component on the next one horizontal line on the screen with respect to the first component stored in the buffer memory; and a first component on the next one of the first components stored in the buffer memory. , And the first component on the same vertical line on the screen, so that image data corresponding to the first component and the second component can be transferred.

According to yet another aspect of the present invention, an image data transfer method comprises the steps of: converting image data including a first component and a second component vertically down-sampled on a screen into a predetermined number of images; The transfer is performed in units of blocks composed of pixels. First, the image data for one line in the horizontal direction on the screen is transferred, and the second
Is stored in the buffer memory. Then
As for the first component, data for one line following the one line in the horizontal direction is transferred. On the other hand, for the second component, the data stored in the buffer memory is transferred.

Preferably, for the second component,
Among the data stored in the buffer memory, the data on the same vertical line on the screen as the first component of the next one line is transferred.

In the above-described image data transfer method, first, one line of image data in the horizontal direction on the screen is transferred, and the data of the second component is stored in the buffer memory. The second component stored in the buffer memory includes not only a portion corresponding to the data of the first component for one line in the horizontal direction on the screen but also data of the first component for the next one line. Corresponding parts are also included. Therefore, for the first component, data for one line following the one line in the horizontal direction is transferred, while for the second component,
Transfer the data stored in the buffer memory. As a result, the first line for the next line following the one line in the horizontal direction is obtained.
The data of the component and the data of the second component including the corresponding part are transferred.

As described above, in the image data transfer method, the data of the second component for one line in the horizontal direction on the screen is stored in the buffer memory, and the first component is
While the data of one line following the one line in the horizontal direction is transferred, the second component includes the first component and a corresponding part to transfer the data stored in the buffer memory. The image data including the second component can be transferred.

Preferably, the second component to be transferred is up-sampled in the vertical direction on the screen, and a portion corresponding to the first component to be transferred is extracted from the up-sampled second component. Transfer the part you did.

According to the above-described image data transfer method, image data corresponding to the first component and the second component can be transferred.

[0090]

Embodiments of the present invention will be described below in detail with reference to the drawings. In the drawings, the same or corresponding portions have the same reference characters allotted, and description thereof will not be repeated.

(First Embodiment) <Overall Configuration> FIG. 1 is a block diagram showing the overall configuration of an image processing apparatus according to a first embodiment of the present invention. FIG.
, The image processing apparatus includes an analyzer 101, a Huffman decoder 102, and selectors 103y and 103
u, 103v and combiners 104y, 104u, 1
04v, the inverse quantizer 105, and the inverse DCT transformer 106
, A color space converter 107, a color number reducer 108, decompressors 110u and 110v, a horizontal start controller 114, block extractors 111u and 111v, and input means iy1-
iy3, iu1-iu3, iv1-iv3, an encoding table TB1, a quantization table TB2, a compressed image memory M1, an expanded image memory M2, and a start position memory M3.

The compressed image memory M1 is a memory for storing compressed image data.

The analyzer 101 analyzes the header of the compressed image data stored in the compressed image memory M1, and obtains information necessary for the decompression processing, such as the encoding method, the quantization value, and the Huffman table. In addition, the analyzer 101 transfers the Huffman table to the encoding table TB1 and the quantization value to the quantization table TB2. When the compressed image data stored in the compressed image memory M1 is progressively coded, the analyzer 101 stores the coded data of each stage for each of the luminance component Y and the color difference components U and V. Detect the position that is being performed. Then, information on the position where the encoded data of each stage of the luminance component Y is stored is input to the input means iy1-iy3, and information on the position where the encoded data of each stage of the color difference components U and V is stored is input. Output to means iu1-iu3, iv1-iv3. On the other hand, when the compressed image data stored in the compressed image memory M1 is sequentially encoded,
The encoded data is transferred to the Huffman decoder 102.

The start position memory M3 stores information on the position where the coded data of each stage of the color difference components U and V of the first (left end) block in the horizontal direction on the screen is stored, and the luminance component Y of the block. Is a memory for storing information on the position where the encoded data of the first stage (DC component) is stored.

The input means iy1-iy3 extract the coded data of each stage of the luminance component Y of the compressed image data stored in the compressed image memory M1 based on the position information from the analyzer 101, and provide it to the selector 103y. Supply. The input means iu1-iu3 and iv1-iv3 extract encoded data of each stage of the color difference components U and V of the compressed image data stored in the compressed image memory M1 based on the position information from the analyzer 101. The signals are supplied to the selectors 103u and 103v.

The selectors 103y, 103u, 103v
Are input means iy1-iy3, iu1-iu, respectively.
3, iv1-iv3, and selectively supplies the coded data of each stage to the Huffman decoder 102.

The Huffman decoder 102 refers to the Huffman table stored in the encoding table TB1 and refers to the analyzer 101 or the selectors 103y, 103u, 103
Huffman decode the encoded data supplied by v.

The combiner 104y includes a selector 109
And a coefficient memory M4. The coefficient memory M4 includes storage areas d1-d3. The selector 109 selectively transfers the DCT coefficient data of the luminance component Y from the Huffman decoder 102 to the storage areas d1 to d3 of the coefficient memory M4. The DCT coefficient data of the luminance component Y stored in the storage areas d1 to d3 of the coefficient memory M4 is output to the inverse quantizer 105. Combiners 104u and 104y include selector 109 (not shown) and coefficient memory M4 (not shown), similarly to combiner 104y. The selectors included in the combiners 104u and 104v store the DCT coefficient data of the color difference components U and V from the Huffman decoder 102 in a storage area d1- of a coefficient memory M4 (not shown).
d3 (not shown).

The inverse quantizer 105 calculates the quantization table TB
2 with reference to the quantized value stored in
4y, 104u, and 104v, the DCT coefficient data of the luminance component Y and the DC of the color difference components U and V from the coefficient memory M4.
Dequantize the T coefficient data.

The inverse DCT transformer 106 has the inverse quantizer 10
5, the inverse DCT transform is performed on the DCT coefficient data of the luminance component Y and the DCT coefficient data of the color difference components U and V, which have been inversely quantized by 5.

The decompressors 110u and 110v upsample the expanded image data of the color difference components U and V obtained by the inverse DCT converter 106 twice in the vertical direction on the screen. That is, the number of pixels of the color difference components U and V in the vertical direction on the screen is restored.

When the horizontal start controller 114 completes the expansion processing of the last (right end) block in the horizontal direction on the screen,
Block extractors 111u and 111v and analyzer 101
Outputs a control signal CT. In response to the control signal CT, the analyzer 101 reads out the position information stored in the start position memory M3, and reads the corresponding input means i.
Output to y1, iu1-iu3, iv1-iv3.

The block extractors 111u and 111v respond to the control signal CT from the horizontal start controller 114, and the luminance component Y of the expanded image of the color difference components U and V upsampled by the decompressors 110u and 110v. The portion (block) corresponding to the decompressed image data is extracted and output.

The color space converter 107 is the inverse DCT converter 1
The YUV image data composed of the expanded image data of the luminance component Y obtained in step S06 and the expanded image data of the color difference components U and V obtained in the block extractors 111u and 111v is converted into three primary color components (R, G, and B). R)
Convert to GB image data.

The color number reducer 108 is a color space converter 107
, The number of colors of the decompressed image data (RGB image data) is reduced.

The expanded image memory M2 is provided with a color number reducer 10
A memory for storing decompressed image data whose number of colors has been reduced by eight.

<Compressed Image Data> Next, the compressed image data stored in the compressed image memory M1 shown in FIG. 1 will be described with reference to FIGS.

Referring to FIG. 2, original image data is represented by three primary color components (R, G, B). Each component R, G, B is
It is composed of 16 (4 × 4) pixels. The squares in the figure represent pixels, and the symbols represent pixel data. Then, each of the components R, G, and B is divided into four (2 × 2) blocks. One block is composed of four pixels (2 × 2). For example, the component R is composed of pixel data R11, R
12, R21, R22, a block composed of pixel data R13, R14, R23, R24, a block composed of pixel data R31, R32, R41, R42, and pixel data R33, R34, R
43 and R44. The following processing (1)-(4) is performed on the original image data represented by the three primary color components (R, G, B).

(1) Color Space Conversion The color space of the original image data is converted from the three primary color components (R, G, B) into a luminance component Y and color difference components U, V.

(2) Downsampling The chrominance components U and V are downsampled in the vertical direction on the screen by half. That is, a process of reducing the number of pixels of the color difference components U and V to １ ／ in the vertical direction on the screen is performed.

As a result, the color difference component U is
A block composed of 11, U12, U21, and U22;
A block composed of pixel data U31, U32, U41, U42, pixel data U13, U14, U23, U2
4 and pixel data U33, U
What was represented by blocks composed of U, U43, U43, U43, U44 is changed to pixel data U11, U12, U32, U3.
2 and pixel data U13, U1
4, U33, and U34.

Similarly, the color difference component V is the pixel data V1
1, V12, V32, V32 and a block composed of pixel data V13, V14, V33, V34.

Here, the pixel data Y1 of the luminance component Y
1, a block composed of Y12, Y21, Y22, a block composed of pixel data Y31, Y32, Y41, Y42, and pixel data U11, U12, U of a color difference component U.
31, U32, and a color difference component V
Of one pixel (Minum Code) in a block composed of the pixel data V11, V12, V31, and V32.
ed Unit). A block composed of pixel data Y13, Y14, Y23, and Y24 of the luminance component Y, pixel data Y33, Y34, Y43, and Y44.
, The pixel data U1 of the color difference component U
3, U14, U33, U34, and pixel data V13, V14, V33,
One MCU is composed of blocks composed of V34. That is, two blocks of the luminance component Y (one horizontal direction × two vertical directions on the screen) and one block of the color difference components U and V after down sampling (one horizontal direction × one vertical direction on the screen) ) Constitute one MCU.

(3) DCT transformation The luminance component Y and the down-sampled color difference components U,
DCT transforms the V image data. Thereby, the DCT coefficient data ya11-ya14, ya21 of the luminance component Y
-Ya24, yb11-yb14, yb21-yb2
4. DCT coefficient data u11-u1 of color difference components U and V
4, u21-u24, v11-v14, v21-v24
Is obtained. Then, the obtained DCT coefficient is quantized based on the quantization value stored in the quantization table.

Next, the quantized DCT coefficient data is divided into a plurality of stages (first to third stages) based on the frequency. Here, the DCT coefficient data of the first stage (DC) is the DCT coefficient data ya1 of the DC component of each block.
1, ya21, yb11, yb21, u11, u21,
v11 and v21. The DCT coefficient data of the second stage (AC1) is the DCT coefficient data (ya1) of each block.
2, ya13), (ya22, ya23), (yb1
2, yb13), (yb22, yb23), (u12,
u13), (u22, u23), (v12, v13),
(V22, v23). DC of the third stage (AC2)
The T coefficient data is DCT coefficient data ya1 of each block.
4, ya24, yb14, yb24, u14, u24,
v14 and v24.

(4) Huffman coding The quantized DCT coefficient data is Huffman-coded for each stage. Specifically, as shown in FIG. 3, the DCT coefficient data of the first stage (DC) of the luminance component Y is for each MCU, that is, the DCT coefficient data (ya11, yb1
1) Huffman coding is performed in the order of (ya21, yb21). The DCT coefficient data of the second stage (AC1) of the luminance component is in the block order in the horizontal direction on the screen, ie, D
CT coefficient data (ya12, ya13), (ya22,
ya23), (yb12, yb13), (yb22, y
b23) is performed in the order of Huffman coding. The DCT coefficient data of the third stage (AC2) of the luminance component Y is arranged in the block order in the horizontal direction on the screen, that is, the DCT coefficient data ya
Huffman coding is performed in the order of 14, ya24, yb14, yb24.

The DCT coefficient data of the first stage (DC) of the color difference component U is Huffman-coded for each MCU, that is, in the order of the DCT coefficient data u11 and u21. The DCT coefficient data of the second stage (AC1) of the color difference component U is Huffman-coded in block order in the horizontal direction on the screen, that is, in the order of DCT coefficient data (u12, u13), (u22, u23). Third stage of color difference component U (AC2)
Are subjected to Huffman coding in the block order in the horizontal direction on the screen, that is, in the order of DCT coefficient data u14 and u24.

The DCT coefficient data of the first stage (DC) of the color difference component V is Huffman-coded for each MCU, that is, in the order of DCT coefficient data v11 and v21. The DCT coefficient data of the second stage (AC1) of the color difference component V is Huffman-coded in the block order in the horizontal direction on the screen, that is, in the order of DCT coefficient data (v12, v13), (v22, v23). Third stage of color difference component V (AC2)
Are subjected to Huffman coding in the block order in the horizontal direction on the screen, that is, in the order of DCT coefficient data v14 and v24.

In this way, the progressively encoded compressed image data as shown in FIG. 3 is obtained. Although not shown in FIG. 3, the compressed image data includes
Table data (Huffman table, quantization value), frame header, and other data necessary for decompression processing are included. The frame header includes information indicating an encoding method of the compressed image data (whether the data is sequentially encoded data or progressively encoded data).

<Decompression Processing> Next, the decompression processing by the image processing apparatus shown in FIG. 1 will be described. FIG. 4 is a flowchart illustrating a procedure of a decompression process performed by the image processing apparatus illustrated in FIG. Hereinafter, description will be made with reference to FIG. 4 and FIG. In the following description, for an expanded image composed of 2 × 2 blocks, (1) an upper left block expansion process on the screen, (2) an upper right block expansion process on the screen, (3) Decompression processing of the lower left block on the screen,
(4) Decompression processing of the lower right block on the screen will be described separately.

(1) Decompression processing of upper left block on screen First, in step ST701, the analyzer 101
The frame header of the compressed image data stored in the compressed image memory M1 is analyzed to obtain information such as an encoding method (progressive encoding or sequential encoding) and an image size. As a result, it is recognized that the compressed image data is progressively encoded data. Further, the analyzer 101 extracts a quantization value from the table data of the compressed image data, transfers the quantization value to the quantization table TB2, extracts a Huffman table, and encodes the encoding table TB1.
Transfer to

Next, in step ST702, the analyzer 101 identifies the scan header of each stage of the compressed image data, and stores the leading encoded data of each stage of the luminance component Y and the color difference components U and V. Detect the position.

Specifically, the analyzer 101 calculates the luminance component Y
Of the first stage (DC) of the encoded data hya11, hyb
11 (see FIG. 3) is a luminance component Y of the compressed image data.
The first byte of the coded data hya12, hya13 (see FIG. 3) of the second stage (AC1) is the first byte of the luminance component Y of the compressed image data; Step (AC2) encoded data hya1
4 (see FIG. 3) is the number of bytes from the beginning of the luminance component Y of the compressed image data, and the first stage (DC) encoded data hu11 (see FIG. 3) of the color difference component U is compressed. The number of the byte from the head of the color difference component U of the image data, the head of the encoded data hu12, hu13 (see FIG. 3) of the second stage (AC1) of the color difference component U is the head of the color difference component U of the compressed image data. The first byte of the coded data hu14 (see FIG. 3) of the third stage (AC2) of the color difference component U is the number byte from the top of the color difference component U of the compressed image data, and the color difference component The first stage of DC (DC)
The head of the encoded data hv11 (see FIG. 3) is the number of bytes from the beginning of the color difference component V of the compressed image data, and the encoded data hv1 of the second stage (AC1) of the color difference component V
2, the head of hv13 (see FIG. 3) is the number of bytes from the head of the chrominance component V of the compressed image data, and the head of the coded data hv14 (see FIG. 3) of the third stage (AC2) of the chrominance component v Detects position information indicating the number of the byte from the head of the color difference component V of the compressed image data.

The analyzer 101 inputs the position information of the coded data of the first to third stages of the luminance component Y to the input means iy1 to iy3, respectively, and inputs the position information of the first to third stages of the color difference component U. Means for inputting the position information of the
u1-iu3 is provided with the position information of the coded data of the first to third stages of the color difference component V, respectively, as input means iv1-iv.
Output to 3.

Next, in step ST401, as shown in FIG. 5, the current reading position of the color difference components U and V, that is, the first to third code of the color difference components U and V detected in step ST702. The position information of the encoded data and the current read position of the first stage (DC) of the luminance component Y, that is, the position information of the encoded data of the first stage of the luminance component Y detected in step ST702 are used for the start position. It is stored in the memory M3.

Next, in step ST703, the input means iy1-iy3, based on the position information from the analyzer 101, encode the encoded data hya11, (hya12, hya13) of the first to third stages of the luminance component Y. hy
a14 (see FIG. 5) is extracted from the compressed image memory M1 and transferred to the selector 103y. At this time, the input means iy1 skips the encoded data hyb11 of the first stage and skips only the encoded data hya11 to the selector 1
03y. Also, the input means iu1-iu3 is
Based on the position information from the analyzer 101, the encoded data hu11, (hu1) of the first to third stages of the color difference component U
2, hu13) and hu14 (see FIG. 5) are extracted from the compressed image memory M1 and transferred to the selector 103u. The input means iv1-iv3, based on the position information from the analyzer 101, encode the first-third stage encoded data hv11, (hv12, hv13), hv14 of the color difference component V.
(See FIG. 5) is extracted from the compressed image memory M1 and transferred to the selector 103v.

The analyzer 101 is provided with a selector 103
The position information is incremented by the number of bytes of the encoded data transferred to y, 103u, and 103v. Thereby, the input means iy1-iy3, iu1-iu3, iv1
The position information given to -iv3 indicates the number of bytes from the beginning of the compressed image data at the beginning of the encoded data at each stage of the next block.

Next, in step ST704, the selectors 103y, 103u, and 103v enter the input means iy
, Iu1, iv1, the first-stage encoded data hya11, hya of the luminance component Y and the chrominance components U, V
u11, hv11 (see FIG. 5)
Feed to 2. The Huffman decoder 102 refers to the Huffman table stored in the encoding table TB1, and performs Huffman decoding of the first-stage encoded data hya11, hu11, hv11 of the luminance component Y and the color difference components U, V. I do. Thus, the first-stage DCT coefficient data ya11, u1 of the luminance component Y and the color difference components U, V
1, v11 are obtained.

Next, in step ST705, the selector 1 of the combiners 104y, 104u, 104v
09 is a first-stage DC component of the luminance component Y and the color difference components U and V obtained by the Huffman decoder 102, respectively.
T coefficient data ya11, u11, v11 are stored in a coefficient memory M
4 to the storage area d1. The coefficient memory M4 is 2 ×
It has a capacity to store two DCT coefficient data. The storage area d1 stores the first-stage DCT coefficient data, the storage area d2 stores the second-stage DCT coefficient data, and the storage area d3 stores the third-stage DCT coefficient data.

Then, in step ST706, it is determined whether or not the DCT coefficient data of all stages has been stored in coefficient memory M4. If DCT coefficient data of all stages has been stored, the process proceeds to step ST707. D for all stages
If the CT coefficient data is not stored, step S
It returns to T704. Here, since the DCT coefficient data of all the stages has not been stored yet, the process returns to step ST704.

Then, in step ST704, the selectors 103y, 103u, 103v provide the input means iy
, Iu2, iv2, the second stage of encoded data (hya12, hya12,
hya13), (hu12, hu13), (hv12,
hv13) (see FIG. 5) to the Huffman decoder 102. Then, the Huffman decoder 102 refers to the Huffman table stored in the encoding table TB1, and encodes the second-stage encoded data (hya12, hya13), (hu12, hya13, lu) of the luminance component Y and the color difference components U, V. h
u13) and (hv12, hv13) are Huffman-decoded. As a result, the second component of the luminance component Y and the color
Stage DCT coefficient data (ya12, ya13), (u
12, u13) and (v12, v13) are obtained.

Next, in step ST705, the selector 1 of the combiners 104y, 104u, 104v
09 is the DC component of the second stage of the luminance component Y and the color difference components U and V obtained by the Huffman decoder 102, respectively.
T coefficient data (ya12, ya13), (u12, u1)
3), (v12, v13) are transferred to the storage area d2 of the coefficient memory M4. DCT coefficient data (ya12, ya13) of the second stage of the luminance component Y and the color difference components U, V,
(U12, u13) and (v12, v13) are stored in the storage area d2 in zigzag scan order.

Then, the process returns to step ST704 again according to the determination in step ST706.

In step ST704, the selector 1
03y, 103u, 103v are input means iy3, iu
3 and iv3, the encoded data hya14 and hu1 of the third stage of the luminance component Y and the color difference components U and V
4 and hv14 (see FIG. 5) to the Huffman decoder 102. Then, the Huffman decoder 102 performs the Huffman decoding on the third-stage encoded data hya14, hu14, hv14 of the luminance component Y and the color difference components U, V with reference to the Huffman table stored in the encoding table TB1. I do. Thereby, the luminance component Y and the color difference components U, V
DCT coefficient data ya14, u14, v1 of the third stage
4 is obtained.

Next, in step ST705, the selector 1 of the combiners 104y, 104u, 104v
09 is the third-stage DC of the luminance component Y and the color difference components U and V obtained by the Huffman decoder 102, respectively.
The T coefficient data ya14, u14, and v14 are transferred to the storage area d3 of the coefficient memory M4. Thereby, as shown in FIG. 5, for the luminance component Y and the color difference components U and V,
2 × 2 DCT coefficient data ya11-ya14, u11-u14, v11- arranged in zigzag scan order
v14 is obtained.

As described above, the DCT coefficient data ya11-ya14, u of the luminance component Y and the color difference components U, V
11-u14 and v11-v14 are obtained. Then, the process proceeds to step ST707 according to the determination in step ST706.

Next, in step ST707, the inverse quantizer 105 sets the combination units 104y, 104u,
DCT coefficient data ya11-ya14, u11 of the luminance component Y and the color difference components U, V obtained by 104v
-U14, v11-v14 are inversely quantized based on the quantization values stored in the quantization table TB2.

Next, in step ST708, the inverse DCT transformer 106 performs DCT coefficient data ya11-ya14, u11-u14, v11 of the luminance component Y and the chrominance components U, V dequantized by the dequantizer 105.
-V14 is subjected to inverse DCT transform. Thereby, as shown in FIG. 6, expanded image data (Y11, Y12, Y21, Y22) of the luminance component Y and the color difference components U, V,
(U11, U12, U31, U32), (V11, V1
2, V31, V32) are obtained.

Next, in step ST402, the decompressors 110u and 110v, as shown in FIG. 6, expand the image data (U11, U12, U31, U31, U31) of the color difference components U and V.
U32) and (V11, V12, V31, V32) are up-sampled twice in the vertical direction on the screen. That is, the number of pixels of the expanded image data of the color difference components U and V in the vertical direction on the screen is restored.

As a result, the expanded image data of the color difference component U is represented by a block composed of the pixel data U11, U12, U31, and U32.
1, U12, U11, and U12, and a block composed of pixel data U31, U32, U31, and U32.

Similarly, the expanded image data of the color difference component V is
What was represented by a block composed of the pixel data V11, V12, V31, and V32 has been changed to pixel data V11,
It is represented by a block composed of V12, V11, V12 and a block composed of pixel data V31, V32, V31, V32.

Next, in step ST403, the block extractor 111u, as shown in FIG.
Among the expanded image data of the color difference component U upsampled by 10u, the pixel data U11, U12, U
A block composed of U11 and U12 is extracted and output. As shown in FIG. 6, the block extractor 111v outputs the pixel data V1 out of the expanded image data of the color difference component V up-sampled by the decompressor 110v.
1, V12, V11, and V12 are extracted and output.

Next, in step ST709, the color space converter 107, as shown in FIG. 6, expands the image data (Y11, Y12, Y21, Y22) of the luminance component Y obtained by the inverse DCT converter 106, and Decompressed image data (U11, U12, U11, U12) of color difference components U, V from block extractors 111u, 111v, (V
11, V12, V11, V12) to the three primary color components (R,
G, B) decompressed image data (R11, R12, R21,
R22), (G11, G12, G21, G22), (B
11, B12, B21, B22).

Next, in step ST710, the color number reducer 108 expands the expanded image data (R1
1, R12, R21, R22), (G11, G12, G
21, G22), (B11, B12, B21, B22)
Reduce the number of colors. At this time, processing is performed using a technique such as an error diffusion method so that the appearance of the image does not differ as much as possible before and after the color number reduction processing. A display device of a small terminal such as a mobile phone can often display only a small amount of color information with low resolution. In an image processing apparatus mounted on such a small terminal, it is necessary to provide the color number reduction unit 108.

Next, in step ST711, the expanded image data of the three primary color components (R11,
R12, R21, R22), (G11, G12, G2
1, G22) and (B11, B12, B21, B22) are transferred from the color number reducer 108 to the expanded image memory M2. As a result, the expanded image data of the upper left block on the screen among the expanded images composed of 2 × 2 blocks is obtained.

[0146] Next, in step ST404, it is determined whether or not the decompression processing of one horizontal line block on the screen has been completed. If it is determined that the processing has been completed, the process proceeds to step ST405. If it is determined that the processing has not been completed, the process returns to step ST703. Here, 2
Since the decompression process of the upper right block on the screen in the decompressed image composed of × 2 blocks has not been completed yet,
It returns to step ST703.

(2) Decompression processing of upper right block on screen Then, in step ST703, input means iy1
−iy3 is based on the position information from the analyzer 101,
First-third-stage encoded data hya of luminance component Y
21, (hya22, hya23), hya24 (FIG. 7)
Selector) extracted from the compressed image memory M1.
03y. At this time, the input means iy1
Transfers the first-stage encoded data hyb21 to the selector 103y, skipping the first-stage encoded data hyb21. Also, input means iu1-iu3, iv1-i
v3 is the encoded data [hu of the first to third stages of the color difference components U and V based on the position information from the analyzer 101.
21, (hu22, hu23), hu24], [hv2
1, (hv22, hv23), hv24] (see FIG. 7).
Is extracted from the compressed image memory M1 and the selector 103
u, 103v.

Then, the analyzer 101 is provided with a selector 103
The position information is incremented by the number of bytes of the encoded data transferred to y, 103u, and 103v.

Next, steps ST704 to ST706
In the above, Huffman decoding processing and DCT coefficient data combination processing are performed in the same manner as described above.
Thereby, as shown in FIG. 7, for the luminance component Y and the color difference components U and V, 2 × 2 pieces of DCT coefficient data ya21-ya24, u2 arranged in zigzag scan order.
1-u24, v21-v24 are obtained. Then, the process proceeds to step ST707 according to the determination in step ST706.

Next, steps ST707 and ST708
In the same manner as described above, inverse quantization and inverse DC
T conversion is performed. Thereby, as shown in FIG. 8, the expanded image data (Y1) of the luminance component Y and the color difference components U and V
3, Y14, Y23, Y24), (U13, U14, U
33, U34), (V13, V14, V33, V34)
Is obtained.

Next, in step ST402, the decompressors 110u and 110v, as shown in FIG. 8, decompressed image data (U13, U14, U33,
U34) and (V13, V14, V33, V34) are up-sampled twice in the vertical direction on the screen. As a result, the expanded image data of the color difference component U becomes the pixel data U1
3, U14, U13, and U14, and a block composed of pixel data U33, U34, U33, and U34. Similarly, the expanded image data of the color difference component V includes pixel data V13, V14, V13, and V1.
4 and pixel data V33, V3
4, V33, and V34.

Next, in step ST403, the block extractor 111u, as shown in FIG.
Among the expanded image data of the color difference component U upsampled by 10u, the pixel data U13, U14, U
13 and U14 are extracted and output. As shown in FIG. 8, the block extractor 111v outputs the pixel data V1 out of the expanded image data of the color difference component V upsampled by the decompressor 110v.
3, V14, V13, and V14 are extracted and output.

Next, in step ST709, the color space converter 107 expands the image data (Y13, Y14, Y13) of the luminance component Y obtained by the inverse DCT converter 106.
Y23, Y24), and the block extractor 111u, 1
11v, the expanded image data of the color difference components U and V (U1
3, U14, U13, U14), (V13, V14, V
13, V14) with the expanded image data (R1
3, R14, R23, R24), (G13, G14, G
23, G24), (B13, B14, B23, B24)
Convert to

Next, in step ST710, the color number reducer 108 expands the expanded image data (R1
3, R14, R23, R24), (G13, G14, G
23, G24), (B13, B14, B23, B24)
Reduce the number of colors.

Next, in step ST711, the decompressed image data (R13,
R14, R23, R24), (G13, G14, G2
3, G24) and (B13, B14, B23, B24) are transferred from the color number reducer 108 to the expanded image memory M2. As a result, of the expanded image composed of 2 × 2 blocks, expanded image data of the upper right block on the screen is obtained.

Next, in step ST404, it is determined whether or not the expansion processing of one horizontal line block on the screen is completed. Here, since the decompression processing of the upper left and upper right blocks on the screen of the decompressed image composed of 2 × 2 blocks has been completed, step S
Proceed to T405.

Next, in step ST405,
It is determined whether or not decompression processing has been completed for all blocks included in (1 MCU) × (one line in the horizontal direction on the screen). If it is determined that the process has been completed, the process proceeds to step ST712. If it is determined that the processing has not been completed, the process proceeds to step ST406. Here, since the decompression processing of the lower left and lower right blocks on the screen has not been completed (see FIG. 2), the process proceeds to step ST406.

[0158] Next, in step ST406, the horizontal start controller 114 controls the control signal C indicating that the decompression processing of one horizontal line block on the screen is completed.
Let T be an analyzer 101 and block extractors 111u, 11
Output to 1v. In response to this, the analyzer 101 determines the position information stored in the start position memory M3 [the encoded data hu11, (hu12, hu12, hu12,
hu13), hu14, hv11, (hv12, hv1)
3), the position information of hv14, the position information of the encoded data (hya11, hyb11) of the first stage of the luminance component Y], and the corresponding input means iy1, iu1-iu
3, iv1-iv3. Further, in response to the control signal CT, the block extractors 111u and 111v switch blocks to be extracted. Then, the process returns to step ST703.

(3) Decompression processing of lower left block on screen Next, in step ST703, input means iy1
−iy3 is based on the position information from the analyzer 101,
First stage to third stage encoded data hyb of luminance component Y
11, (hyb12, hyb13), hyb14 (FIG. 9)
Selector) extracted from the compressed image memory M1.
03y. At this time, the input means iy1
Transfers the first-stage encoded data hya11 to the selector 103y, skipping the first-stage encoded data hya11. Also, input means iu1-iu3, iv1-i
v3 is the encoded data [hu of the first to third stages of the color difference components U and V based on the position information from the analyzer 101.
11, (hu12, hu13), hu14], [hv1
1, (hv12, hv13), hv14] (see FIG. 9).
Is extracted from the compressed image memory M1 and the selector 103
u, 103v. Then, the analyzer 101 increments the position information by the number of bytes of the encoded data transferred to the selectors 103y, 103u, and 103v.

Next, steps ST704 to ST706
In the same manner as described above, Huffman decoding processing and DCT coefficient data combination processing are performed. Thereby, as shown in FIG. 9, the luminance component Y and the color difference components U and V are arranged in a zigzag scan order.
× 2 DCT coefficient data yb11−yb14, u11
-U14, v11-v14 are obtained. Then, the process proceeds to step ST707 according to the determination in step ST706.

Next, steps ST707 and ST708
In the same manner as described above, inverse quantization and inverse DC
T conversion is performed. Thereby, as shown in FIG.
Decompressed image data of the luminance component Y and the color difference components U and V (Y
31, Y32, Y41, Y42), (U11, U12,
U31, U32), (V11, V12, V31, V3
2) is obtained.

Next, in step ST402, the decompressors 110u and 110v, as shown in FIG. 10, expand the image data (U11, U12, U3) of the color difference components U and V.
1, U32) and (V11, V12, V31, V32) are up-sampled twice in the vertical direction on the screen. Thus, the expanded image data of the color difference component U is represented by a block composed of the pixel data U11, U12, U11, U12 and a block composed of the pixel data U31, U32, U31, U32. Similarly, the expanded image data of the color difference component V includes pixel data V11, V12, V11,
V12 and the pixel data V31,
It is represented by a block composed of V32, V31, and V32.

Next, in step ST403, the block extractor 111u outputs the chrominance component U upsampled by the reconstructor 110u as shown in FIG.
Pixel data U31, U32,
A block composed of U31 and U32 is extracted and output. Further, as shown in FIG. 10, the block extractor 111v outputs the pixel data V3 out of the expanded image data of the color difference component V upsampled by the decompressor 110v.
1, V32, V31, and V32 are extracted and output.

Next, in step ST 709, the color space converter 107 expands the expanded image data (Y 31, Y 32, Y 32) of the luminance component Y obtained by the inverse DCT converter 106.
Y41, Y42), and the block extractor 111u, 1
11v, the expanded image data of the color difference components U and V (U3
1, U32, U31, U32), (V31, V32, V
31, V32) with the expanded image data (R3
1, R32, R41, R42), (G31, G32, G
41, G42), (B31, B32, B41, B42)
Convert to

Next, in step ST710, the number-of-colors reducer 108 expands the expanded image data (R3
1, R32, R41, R42), (G31, G32, G
41, G42), (B31, B32, B41, B42)
Reduce the number of colors.

Next, in step ST711, the expanded image data (R31,
R32, R41, R42), (G31, G32, G4
1, G42) and (B31, B32, B41, B42) are transferred from the color number reducer 108 to the expanded image memory M2. As a result, of the decompressed image composed of 2 × 2 blocks, the decompressed image data of the lower left block on the screen is obtained.

Next, in step ST404, it is determined whether or not the extension processing of one horizontal line block on the screen is completed. Here, since the decompression processing of the lower right block on the screen among the decompressed images composed of 2 × 2 blocks has not been completed, step S
It returns to T703.

(4) Decompression processing of lower right block on screen Then, in step ST703, input means iy1
−iy3 is based on the position information from the analyzer 101,
First stage to third stage encoded data hyb of luminance component Y
21, (hyb22, hyb23), hyb24 (FIG. 1)
1) is extracted from the compressed image memory M1 and transferred to the selector 103y. At this time, the input means iy1
Transfers the first-stage encoded data hya21 to the selector 103y, skipping the first-stage encoded data hya21. Also, input means iu1-iu3, iv1-i
v3 is the encoded data [hu of the first to third stages of the color difference components U and V based on the position information from the analyzer 101.
21, (hu22, hu23), hu24], [hv2
1, (hv22, hv23), hv24] (see FIG. 11) from the compressed image memory M1 and extract the selector 10
3u, 103v. Then, the analyzer 101
The position information is incremented by the number of bytes of the encoded data transferred to the selectors 103y, 103u, and 103v.

Next, steps ST704 to ST706
In the above, Huffman decoding processing and DCT coefficient data combination processing are performed in the same manner as described above.
Thereby, as shown in FIG. 11, 2 × 2 DCT coefficient data yb21-yb24, u arranged in zigzag scan order for the luminance component Y and the color difference components U, V
21-u24 and v21-v24 are obtained. Then, the process proceeds to step ST707 according to the determination in step ST706.

Next, steps ST707 and ST708
In the same manner as described above, inverse quantization and inverse DC
T conversion is performed. Thereby, as shown in FIG.
Decompressed image data of the luminance component Y and the color difference components U and V (Y
33, Y34, Y43, Y44), (U13, U14,
U33, U34), (V13, V14, V33, V3
4) is obtained.

Next, in step ST402, the decompressors 110u and 110v, as shown in FIG. 12, expand the image data (U13, U14, U3) of the color difference components U and V.
3, U34) and (V13, V14, V33, V34) are upsampled twice in the vertical direction on the screen. Thus, the expanded image data of the color difference component U is represented by a block including the pixel data U13, U14, U13, and U14 and a block including the pixel data U33, U34, U33, and U34. Similarly, the expanded image data of the color difference component V includes pixel data V13, V14, V13,
V14 and the pixel data V33,
It is represented by a block composed of V34, V33, and V34.

Next, in step ST403, as shown in FIG. 12, the block extractor 111u performs the chrominance component U upsampled by the reconstructor 110u.
Pixel data U33, U34,
The block composed of U33 and U34 is extracted and output. As shown in FIG. 12, the block extractor 111v outputs the pixel data V3 of the expanded image data of the color difference component V upsampled by the decompressor 110v.
3, V34, V33, and V34 are extracted and output.

Then, in step ST709, the color space converter 107 expands the expanded image data (Y33, Y34, Y34) of the luminance component Y obtained by the inverse DCT converter 106.
Y43, Y44), and the block extractor 111u, 1
11v, the expanded image data of the color difference components U and V (U3
3, U34, U33, U34), (V33, V34, V
33, V34) with the expanded image data (R3
3, R34, R43, R44), (G33, G34, G
43, G44), (B33, B34, B43, B44)
Convert to

Next, in step ST710, the color number reducer 108 expands the expanded image data (R3
3, R34, R43, R44), (G33, G34, G
43, G44), (B33, B34, B43, B44)
Reduce the number of colors.

Next, in step ST711, the expanded image data of the three primary color components (R33,
R34, R43, R44), (G33, G34, G4
3, G44) and (B33, B34, B43, B44) are transferred from the color number reducer 108 to the expanded image memory M2. As a result, of the decompressed image composed of 2 × 2 blocks, decompressed image data of the lower right block on the screen is obtained.

Next, in step ST404, it is determined whether or not the extension processing of one horizontal line block on the screen is completed. Here, since the decompression processing of the lower left and lower right blocks on the screen in the decompressed image composed of 2 × 2 blocks has been completed, step S
Proceed to T405.

Next, in step ST405,
It is determined whether or not decompression processing has been completed for all blocks included in (1 MCU) × (one line in the horizontal direction on the screen). Here, since the decompression processing of all blocks included in (1 MCU) × (one line in the horizontal direction on the screen) has been completed (see FIG. 2), step S
Proceed to T712.

Next, in step ST712, all the blocks (here, 2 ×
It is determined whether or not the decompression process has been completed for two (2). When the decompression processing is completed for all the blocks, the processing ends. Otherwise, the process returns to step ST401. Here, since the decompression processing of all blocks has been completed, the processing is terminated.

As described above, the decompressed image data of the progressively encoded compressed image data is obtained. When displaying the decompressed image data on the display device, the decompressed image data composed of 2 × 2 blocks in the order of the upper left block, the upper right block, the lower left block, and the lower right block is displayed on the screen. Is displayed.

Here, the expansion processing of the progressively encoded compressed image data has been described. When decompressing the sequentially encoded compressed image data, input means iy1-iy3, iu1-iu3, iv
The encoded data is sequentially supplied to the Huffman decoder 102 directly from the compressed image memory M1 without going through 1-iv3. Then, Huffman decoding, inverse quantization, inverse DCT transform, upsampling, color space conversion, and color number reduction processing are performed. As a result, decompressed image data of the sequentially encoded compressed image data is obtained.

<Effect> As described above, the image processing apparatus according to the first embodiment of the present invention provides the start position memory M3
, Horizontal start controller 114, and decompressors 110u and 110
v and the block extractors 111u and 111v, even if the color difference components U and V of the progressively encoded compressed image data are down-sampled in the vertical direction on the screen, the compressed image data Can be expanded block by block.

Since the above effects can be obtained without providing a special buffer memory, the required memory capacity can be reduced.

(Second Embodiment) <Overall Configuration> FIG. 13 is a block diagram showing the overall configuration of an image processing apparatus according to a second embodiment of the present invention. Referring to FIG. 13, this image processing apparatus includes a start position memory M3, a horizontal start controller 114, and block extractors 111u and 111v in the image processing apparatus shown in FIG.
And a selector 112 and a buffer memory M5. Other configurations are the same as those of the image processing apparatus shown in FIG.

The selector 112 selects the Huffman decoder 10
2, and selectively transfers the received DCT coefficient data to the buffer memory M5 or the combiner 104y. Specifically, the selector 112 selects the last line (the bottom line in the vertical direction on the screen) of the DCT coefficient data of the luminance component Y of the block included in (1 MCU) × (one line in the horizontal direction on the screen). The DCT coefficient data of the second and third stages for blocks other than the block on the line
Transfer to 5. On the other hand, the selector 112 is (1MCU)
× (one horizontal line on the screen) of the DCT coefficient data of the luminance component Y of the block included in the DCT coefficient data of each stage with respect to the block on the last line and the first except for the block on the final line. The DCT coefficient data of the stage is transferred to the combiner 104y.

The buffer memory M5 stores the DCT coefficient data of the luminance component Y transferred from the selector 112.

<Decompression Processing> Next, the decompression processing by the image processing apparatus shown in FIG. 13 will be described. It is assumed that the compressed image data M1 (progressively encoded compressed image data) as shown in FIG. 3 is stored in the compressed image memory M1.

FIG. 14 is a flowchart showing the procedure of the decompression process by the image processing apparatus shown in FIG. Hereinafter, description will be made with reference to FIG. 14 and FIG.

First, in steps ST701 and ST702, in the same manner as described in the first embodiment (FIG. 4), processing of header analysis, acquisition of a table, and detection of the position of encoded data at each stage are performed. Done.

Next, in step ST1401,
The input means iy2 and iy3, based on the position information from the analyzer 101, encode the encoded data (hya12, hya13) and hya14 of the second and third stages of the luminance component Y.
(See FIG. 15) is extracted from the compressed image memory M1 and transferred to the selector 103y. Then, the analyzer 101
The position information is incremented by the number of bytes of the encoded data transferred to the selector 103y. As a result, the position information given to the input means iy2 and iy3 indicates the number of bytes from the head of the compressed image data at which the head of the encoded data of each stage of the next block is.

Next, in step ST1402,
The selector 103y outputs the encoded data (hya12, hya13), hya14 (FIG. 15) of the second and third stages of the luminance component Y supplied from the input means iy2, iy3.
) To the Huffman decoder 102. The Huffman decoder 102 performs Huffman decoding on the encoded data. Thereby, the DCT coefficient data ya12-ya14 of the second and third stages of the luminance component Y is obtained.

Next, in step ST1403,
As shown in FIG. 15, the DCT of the second and third stages of the luminance component Y obtained by the Huffman decoder 102
The coefficient data ya12-ya14 is stored in the buffer memory M5.

Next, in step ST1404,
Of the DCT coefficient data of the luminance component Y of the block included in (1 MCU) × (one line in the horizontal direction on the screen)
It is determined whether the DCT coefficient data of the second and third stages for all blocks except for the block on the last line has been stored in the buffer memory M5. If it is determined that it has been stored, the process proceeds to step ST1405. On the other hand, when it is determined that it is not stored, the process returns to step ST1401. Here, since the DCT coefficient data ya22-ya24 of the second and third stages for the upper right block on the screen has not been stored in the buffer memory M5 yet (see FIG. 2), step ST14 is performed.
Return to 01.

Steps ST1401-ST14
03, the same processing as described above is performed, and FIG.
As shown in FIG. 5, the DCT coefficient data ya22 to ya24 of the second and third stages of the luminance component Y obtained by the Huffman decoder 102 are stored in the buffer memory M5.

Next, in step ST1404,
The determination as described above is performed. Here, since the DCT coefficient data of the second and third stages for the upper left and upper right blocks on the screen composed of 2 × 2 blocks has been stored in the buffer memory M5, step ST
Proceed to 1405.

Next, in step ST1405, the input means iy1 to iy3, based on the position information from the analyzer 101, encode the encoded data (hya11, hyb11), (hyb12, h
yb13) and hyb14 (see FIG. 16) are extracted from the compressed image memory M1 and transferred to the selector 103y. The input means iu1-iu3, iv1-iv3, based on the position information from the analyzer 101, encodes the chrominance components U, V in the first to third stages of the encoded data [hu11, (h
u12, hu13), hu14], [hv11, (hv
12, hv13), hv14] (see FIG. 16) from the compressed image memory M1 and extract the selectors 103u, 103
v. Then, the analyzer 101 selects the selector 10
The position information is incremented by the number of bytes of the encoded data transferred to 3y, 103u, and 103v.

Next, in step ST1406,
The selector 103y outputs the encoded data (hya11, hyb11), (hyb12, hyb1) of the first to third stages of the luminance component Y supplied from the input means iy1 to iy3.
3), supply the hyb 14 to the Huffman decoder 102. The Huffman decoder 102 performs Huffman decoding on the encoded data. Thereby, the DC of the luminance component Y
The T coefficient data ya11, yb11-yb14 are obtained. Then, the selector 112 selects the Huffman decoder 10
2, the DCT coefficient data ya of the luminance component Y obtained by
11, yb11-yb14 are transferred to the combiner 104y.

The selectors 103u and 103v provide first-third-stage encoded data [hu11, (hu12, hu13), of the first-third-stage color difference components U and V supplied from the input means iu1-iu3 and iv1-iv3. hu14],
[Hv11, (hv12, hv13), hv14] (see FIG. 16) are supplied to the Huffman decoder 102. The Huffman decoder 102 performs Huffman decoding on the encoded data. As a result, DCT coefficient data u11-u14 and v11-v14 of the color difference components U and V are obtained.

Next, in step ST1407,
The DCT coefficient data ya12-ya14 of the luminance component Y is read from the buffer memory M5,
y (see FIG. 16).

Next, in step ST1408,
Selector 109 of combiner 104y (see FIG. 1)
Is the DCT coefficient data (ya) from the buffer memory M5.
12, ya13) and ya14 are transferred to storage areas d2 and d3 (see FIG. 1) of the coefficient memory M4, and the selector 112
Is transferred to the storage area d1 (see FIG. 1) of the coefficient memory M4. As a result, as shown in FIG.
× 2 pieces of DCT coefficient data ya11-ya14 are obtained. The selector 109 of the combiner 104y (see FIG. 1) outputs the DCT coefficient data y from the selector 112.
b11, (yb12, yb13), and yb14 are respectively transferred to the storage areas d1-d3 (see FIG. 1) of the coefficient memory M4. As a result, as shown in FIG. 16, 2 × 2 pieces of DCT coefficient data y arranged in the zigzag scan order
b11-yb14 is obtained.

The selector 109 of the combiners 104u and 104v (see FIG. 1) outputs the DCT coefficient data [u11, (u12, u13), u] from the Huffman decoder 102.
14], [v11, (v12, v13), v14]
Each is transferred to the storage area d1-d3 (see FIG. 1) of the coefficient memory M4. As a result, as shown in FIG. 16, 2 × 2 pieces of DCT coefficient data u11-u14 and v11-v14 arranged in zigzag scan order are obtained.

Next, in step ST707, the inverse quantizer 105 sets the combination units 104y, 104u,
DCT coefficient data ya11-ya14, yb1 of the luminance component Y and the color difference components U, V obtained by 104v
1-yb14, u11-u14, v11-v14 are inversely quantized based on the quantization values stored in the quantization table TB2.

Next, in step ST708, the inverse DCT transformer 106 performs DCT coefficient data ya11-ya14, yb11-yb14, u of the luminance component Y and the chrominance components U, V dequantized by the dequantizer 105.
11-u14 and v11-v14 are subjected to inverse DCT. Thereby, as shown in FIG. 17, the expanded image data [(Y11, Y1) of the luminance component Y and the color difference components U, V
2, Y21, Y22), (Y31, Y32, Y41, Y
42)], (U11, U12, U31, U32), (V
11, V12, V31, V32) are obtained. Note that the inverse DCT converter 106 calculates the DCT coefficient data ya11-ya14 and yb11-yb14 of the luminance component Y
Inverse DCT transform is performed for each 2 × 2 DCT coefficient data. That is, after performing the inverse DCT transform on the DCT coefficient data ya11-ya14, the DCT coefficient data yb
Inverse DCT is performed on 11-yb14.

Next, in step ST402, the decompressors 110u and 110v, as shown in FIG. 17, expand the image data (U11, U12, U3) of the color difference components U and V.
1, U32) and (V11, V12, V31, V32) are up-sampled twice in the vertical direction on the screen.

Then, in step ST709, as shown in FIG. 17, the color space converter 107 expands the luminance component Y obtained by the inverse DCT converter 106 [(Y11, Y12, Y21, Y22), (Y31,
Y32, Y41, Y42)], and the decompressor 110u,
Decompressed image data of color difference components U and V from 110v [(U
11, U12, U11, U12), (U31, U32,
U31, U32)], [(V11, V12, V11, V
12), (V31, V32, V31, V32)]
Expanded image data of primary color components [(R11, R12, R2
1, R22), (R31, R32, R41, R4
2)], [(G11, G12, G21, G22), (G
31, G32, G41, G42)], [(B11, B1
2, B21, B22), (B31, B32, B41, B
42)].

Next, in step ST710, the color number reduction unit 108 reduces the number of colors of the expanded image data of the three primary color components obtained by the color space converter 107.

Next, in step ST711, the expanded image data [(R1
1, R12, R21, R22), (R31, R32, R
41, R42)], [(G11, G12, G21, G2
2), (G31, G32, G41, G42)], [(B
11, B12, B21, B22), (B31, B32,
B41, B42)] are transferred to the expanded image memory M2. As a result, decompressed image data (decompressed image data for one MCU: see FIG. 2) of the upper left and lower left blocks on the screen out of the decompressed image composed of 2 × 2 blocks is obtained.

Next, in step ST1409,
It is determined whether the decompression processing has been completed for all blocks included in (1 MCU) × (one line in the horizontal direction on the screen). If it is determined that the processing has been completed, the process proceeds to step ST712. On the other hand, if it is determined that the processing has not been completed, the process returns to step ST1405. Here, 2
Since the decompression process for the upper right and lower right blocks on the screen of the decompressed image composed of × 2 blocks has not been completed yet (see FIG. 2), the process returns to step ST1405.

Then, in step ST1405,
The input means iy1-iy3, based on the position information from the analyzer 101, encode the encoded data (hya21, hyb21), (hyb22, hyb22,
hyb23) and hyb24 (see FIG. 18) are extracted from the compressed image memory M1 and transferred to the selector 103y.
The input means iu1-iu3, iv1-iv3 provide the color difference components U, V based on the position information from the analyzer 101.
Of the first to third stages of the encoded data [hu21,
(Hu22, hu23), hu24], [hv21,
(Hv22, hv23), hv24] (see FIG. 18) are extracted from the compressed image memory M1 and the selectors 103u,
Transfer to 103v.

Next, in step ST1406,
The selector 103y outputs the encoded data (hya21, hyb21), (hyb22, hyb2) of the first to third stages of the luminance component Y supplied from the input means iy1 to iy3.
3), supply hyb 24 to the Huffman decoder 102; The Huffman decoder 102 performs Huffman decoding on the encoded data. Thereby, the DC of the luminance component Y
T coefficient data ya21, yb21-yb24 are obtained. Then, the selector 112 selects the Huffman decoder 10
2, the DCT coefficient data ya of the luminance component Y obtained by
21, yb21-yb24 are transferred to the combiner 104y.

The selectors 103u and 103v provide first-third-stage encoded data [hu21, (hu22, hu23), of the first-third-stage color difference components U and V supplied from the input means iu1-iu3 and iv1-iv3. hu24],
[Hv21, (hv22, hv23), hv24] are supplied to the Huffman decoder 102. Huffman decoder 1
02 performs Huffman decoding on these encoded data.
Thereby, the DCT coefficient data u21 of the color difference components U and V
-U24, v21-v24 are obtained.

Next, in step ST1407,
The DCT coefficient data ya22-ya24 of the luminance component Y is read from the buffer memory M5 (see FIG. 18) and supplied to the combiner 104y.

Then, in step ST1408,
In the same manner as described above, the combination processing of the DCT coefficient data is performed. Thereby, as shown in FIG.
2 × 2 DCT coefficient data ya21-ya24, yb21-yb24, u2 arranged in zigzag scan order
1-u24, v21-v24 are obtained.

Next, in step ST707, the inverse quantizer 105 sets the combination units 104y, 104u,
DCT coefficient data (ya21-ya24, yb) of the luminance component Y and the color difference components U and V obtained by
21-yb24), u21-u24, v21-v24
Is inversely quantized based on the quantization value stored in the quantization table TB2.

Next, in step ST708, the inverse DCT transformer 106 performs the DCT coefficient data (ya21-ya24, yb21-yb2) of the luminance component Y and the color difference components U and V dequantized by the dequantizer 105.
4) Inverse DC for u21-u24, v21-v24
Perform T conversion. Thereby, as shown in FIG. 19, the expanded image data [(Y1) of the luminance component Y and the color difference components U and V are obtained.
3, Y14, Y23, Y24), (Y33, Y34, Y
43, Y44)], (U13, U14, U33, U3
4), (V13, V14, V33, V34) are obtained.

Next, in step ST402, the decompressors 110u and 110v, as shown in FIG. 19, decompress image data (U13, U14, U3) of the color difference components U and V.
3, U34) and (V13, V14, V33, V34) are upsampled twice in the vertical direction on the screen.

Next, in step ST709, the color space converter 107 expands the image data [(Y13, Y14, Y23, Y24), of the luminance component Y obtained by the inverse DCT converter 106, as shown in FIG. (Y33,
Y34, Y43, Y44)], and a decompressor 110u,
Decompressed image data of color difference components U and V from 110v [(U
13, U14, U13, U14), (U33, U34,
U33, U34)], [(V13, V14, V13, V
14), (V33, V34, V33, V34)]
Expanded image data of primary color components [(R13, R14, R2
3, R24), (R33, R34, R43, R4
4)], [(G13, G14, G23, G24), (G
33, G34, G43, G44)], [(B13, B1
4, B23, B24), (B33, B34, B43, B
44)].

Next, in step ST710, the color number reducer 108 reduces the number of colors of the expanded image data of the three primary color components obtained by the color space converter 107.

Next, in step ST711, the expanded image data of the three primary color components [(R1
3, R14, R23, R24), (R33, R34, R
43, R44)], [(G13, G14, G23, G2
4), (G33, G34, G43, G44)], [(B
13, B14, B23, B24), (B33, B34,
B43, B44)] are transferred to the expanded image memory M2. As a result, decompressed image data (decompressed image data for one MCU: see FIG. 2) of the upper right and lower right blocks on the screen out of the decompressed image composed of 2 × 2 blocks is obtained.

Next, in step ST1409,
It is determined whether the decompression processing has been completed for all blocks included in (1 MCU) × (one line in the horizontal direction on the screen). Here, since the decompression processing has been completed for all blocks included in (1 MCU) × (one line in the horizontal direction on the screen) (see FIG. 2), the process proceeds to step ST712.

Next, in step ST712, all the blocks (here, 2 ×
It is determined whether or not the decompression process has been completed for two (2). When the decompression processing is completed for all the blocks, the processing ends. If not, the process returns to step ST1401. Here, since the decompression processing has been completed for all the blocks, the processing is terminated.

As described above, the decompressed image data of the progressively encoded compressed image data is obtained. When displaying the decompressed image data on the display device, first, decompressed image data of the upper left and lower left blocks on the screen,
Next, the decompressed image data composed of 2 × 2 blocks is displayed in the order of the decompressed image data of the upper right and lower right blocks on the screen. That is, decompressed image data is displayed for each MCU.

Here, the expansion processing of the progressively encoded compressed image data has been described. When decompressing the sequentially encoded compressed image data, input means iy1-iy3, iu1-iu3, iv
The encoded data is sequentially supplied to the Huffman decoder 102 directly from the compressed image memory M1 without going through 1-iv3. Then, Huffman decoding, inverse quantization, inverse DCT transform, upsampling, color space conversion, and color number reduction processing are performed. As a result, decompressed image data of the sequentially encoded compressed image data is obtained.

<Effects> As described above, the image processing apparatus according to the second embodiment of the present invention includes the selector 112, the buffer memory M5, and the decompressors 110u and 110v. Even when the color difference components U and V of the image data are down-sampled in the vertical direction on the screen, the compressed image data can be expanded for each MCU.

Further, there is no need to repeatedly perform encoded data extraction processing, Huffman decoding processing, and inverse DCT transform processing on the same block.

Here, the selector 112 and the buffer memory M5 are provided between the Huffman decoder 102 and the combiner 104y.
0, the inverse DCT converter 106 and the color space converter 1
07 or as shown in FIG.
It may be provided between y and the Huffman decoder 102 or between the inverse quantizer 105 and the inverse DCT transformer 106 as shown in FIG. In this case, the same effect as described above can be obtained.

(Third Embodiment) <Overall Configuration> FIG. 23 is a block diagram showing the overall configuration of an image processing apparatus according to a third embodiment of the present invention. Referring to FIG. 23, this image processing apparatus includes a buffer memory M6 and selectors 113u and 113v instead of start position memory M3 in the image processing apparatus shown in FIG. Other configurations are the same as those of the image processing apparatus shown in FIG.

The buffer memory M 6 stores the DC component of each stage of the color difference components U and V obtained by the Huffman decoder 102.
T coefficient data (u11-u14, u21-u24),
(V11-v14, v21-v24) and the luminance component Y
Of the first stage DCT coefficient data (y11, yb21).

The selectors 113u and 113v output the DCT coefficient data of the color difference components U and V from the Huffman decoder 102 or the color difference components U and V stored in the buffer memory M6.
V DCT coefficient data is selectively combined by a combiner 104.
u, 104v.

<Decompression Processing> Next, the decompression processing by the image processing apparatus shown in FIG. 23 will be described. It is assumed that the compressed image data M1 (progressively encoded compressed image data) as shown in FIG. 3 is stored in the compressed image memory M1.

FIG. 24 is a flowchart showing the procedure of the decompression process by the image processing apparatus shown in FIG. Hereinafter, description will be made with reference to FIGS. 24 and 23. In addition,
In the following description, for an expanded image composed of 2 × 2 blocks, (1) an upper left block expansion process on the screen, (2) an upper right block expansion process on the screen, (3)
The process of expanding the lower left block on the screen and the process of expanding the lower right block on the screen (4) will be described separately.

(1) Decompression processing of upper left block on screen First, in steps ST701 and ST702, the first
In the same manner as described in the embodiment (FIG. 4),
Processing of header analysis, acquisition of a table, and detection of the position of encoded data at each stage are performed.

Next, in step ST703, the input means iy1-iy3, based on the position information from the analyzer 101, encode the encoded data (hya11, hyb11), (hya12) of the first to third stages of the luminance component Y. , H
ya13) and hya14 are extracted from the compressed image memory M1 and transferred to the selector 103y. Also, input means i
u1-iu3, iv1-iv3 are based on the position information from the analyzer 101, and the first to third stages of the color difference components U, V
Stage encoded data [hu11, (hu12, hu1
3), hu14], [hv11, (hv12, hv1
3), hv14] is extracted from the compressed image memory M1 and transferred to the selectors 103u and 103v. Then, the analyzer 101 increments the position information by the number of bytes of the encoded data transferred to the selectors 103y, 103u, and 103v.

Next, in step ST704a,
The selector 103y outputs the encoded data (hya11, hyb11), (hya12, hya1) of the first to third stages of the luminance component Y supplied from the input means iy1 to iy3.
3), hya 14 (see FIG. 25) is converted to Huffman decoder 1
02. The Huffman decoder 102 performs Huffman decoding on the encoded data. Thereby, the DCT coefficient data ya11-ya14, yb1 of the luminance component Y
1 is obtained.

The selectors 103u and 103v provide first-third-stage encoded data [hu11, (hu12, hu13), of the first-third-stage color difference components U and V supplied from the input means iu1-iu3, iv1-iv3. hu14],
[Hv11, (hv12, hv13), hv14] are supplied to the Huffman decoder 102. Huffman decoder 1
02 performs Huffman decoding on these encoded data.
Thereby, the DCT coefficient data u11 of the color difference components U and V is obtained.
-U14, v11-v14 are obtained.

Next, in step ST2301,
As shown in FIG. 25, the DCT coefficient data yb11 of the luminance component Y and the DCT coefficient data u11-u14, v of the color difference components U and V obtained by the Huffman decoder 102
11-v14 is stored in the buffer memory M6.

DCT coefficient data u11 of color difference components U and V
−u14, v11−v14 are stored in the buffer memory M6 and also transferred to the selectors 113u, 113v. DCT coefficient data ya11-y of luminance component Y
a14 is transferred to the combiner 104y.

Next, in step ST705a,
Selector 109 of combiner 104y (see FIG. 1)
Is the DCT coefficient data y from the Huffman decoder 102
a11, (ya12, ya13), ya14 are respectively transferred to the storage areas d1-d3 (see FIG. 1) of the coefficient memory M4. Thereby, as shown in FIG. 25, 2 × 2 DCT coefficient data y arranged in zigzag scan order
a11-ya14 is obtained.

The selector 109 (see FIG. 1) of the combiners 104u and 104v is connected to the Huffman decoder 10
DCT coefficient data [u11, (u12, u1)
3), u14], [v11, (v12, v13), v1
4] in the storage areas d1-d3 of the coefficient memory M4, respectively.
(See FIG. 1). Thereby, as shown in FIG. 25, 2 × 2 DCs arranged in zigzag scan order
T coefficient data u11-u14 and v11-v14 are obtained.

Next, in step ST707a,
The inverse quantizer 105 includes combination units 104y and 104
u, 104V, the DCT coefficient data ya11-ya14, u1 of the luminance component Y and the color difference components U, V obtained by
1-u14, v11-v14 are converted to the quantization table TB2.
Is inversely quantized based on the quantized value stored in.

Next, in step ST708a,
The inverse DCT transformer 106 performs a DCT on the luminance component Y and the color difference components U and V, which have been inversely quantized by the inverse quantizer 105.
Coefficient data ya11-ya14, u11-u14, v1
1-v14 is subjected to inverse DCT. This allows
As shown in FIG. 26, the luminance component Y and the color difference components U and V
Decompressed image data (Y11, Y12, Y21, Y2
2), (U11, U12, U31, U32), (V1
1, V12, V31, V32) are obtained.

Next, in step ST402a,
As shown in FIG. 26, the decompressors 110u and 110v decompress the expanded image data (U11, U12, U3) of the color difference components U and V.
1, U32) and (V11, V12, V31, V32) are up-sampled twice in the vertical direction on the screen.

Next, in step ST403a,
As shown in FIG. 26, the block extractor 111u outputs the pixel data U11 and U1 of the expanded image data of the color difference component U up-sampled by the decompressor 110u.
A block composed of U2, U11 and U12 is extracted and output. As shown in FIG. 26, the block extractor 111v extracts a block composed of pixel data V11, V12, V11, and V12 from the expanded image data of the color difference component V up-sampled by the reconstructor 110v. Output.

Next, in step ST709a,
The color space converter 107 has an inverse DCT as shown in FIG.
Decompressed image data (Y11, Y12, Y21, Y22) of luminance component Y obtained by converter 106 and decompressed image data (U11, U12, U11, U12) of color difference components U, V from decompressors 110u, 110v. , (V1
, V12, V11, V12), the expanded image data (R11, R12, R21, R22) of three primary color components, (G1
1, G12, G21, G22), (B11, B12, B
21, B22).

Next, in step ST710a,
The color number reducer 108 reduces the number of colors of the expanded image data of the three primary color components obtained by the color space converter 107.

Next, in step ST711a,
The decompressed image data of the three primary color components (R1
1, R12, R21, R22), (G11, G12, G
21, G22), (B11, B12, B21, B22)
Is transferred to the expanded image memory M2. This gives 2
The decompressed image data of the upper left block on the screen among the decompressed images composed of × 2 blocks is obtained.

Next, in step ST2302,
It is determined whether the decompression processing of one horizontal line block on the screen is completed. If it is determined that the processing has been completed, the process proceeds to step ST2303. If it is determined that the processing has not been completed, the process returns to step ST703. Here, since the decompression processing of the upper right block on the screen among the decompressed images composed of 2 × 2 blocks has not been completed, the process returns to step ST703.

(2) Decompression processing of upper right block on screen Then, in step ST703, input means iy1
−iy3 is based on the position information from the analyzer 101,
First-third encoded data (hy
a21, hyb21), (hya22, hya23),
hya24 (see FIG. 27) is extracted from the compressed image memory M1 and transferred to the selector 103y. Further, the input means iu1-iu3, iv1-iv3, based on the position information from the analyzer 101, encodes the chrominance components U, V in the first to third stages of the encoded data [hu21, (hu22, hu2).
3), hu24], [hv21, (hv22, hv2
3), hv24] is extracted from the compressed image memory M1 and transferred to the selectors 103u and 103v. Then, the analyzer 101 increments the position information by the number of bytes of the encoded data transferred to the selectors 103y, 103u, and 103v.

Next, in step ST704a,
The selector 103y outputs the encoded data (hya21, hyb21), (hya22, hya2) of the first to third stages of the luminance component Y supplied from the input means iy1 to iy3.
3), and supply hya 24 to the Huffman decoder 102. The Huffman decoder 102 performs Huffman decoding on the encoded data. Thereby, the DC of the luminance component Y
T coefficient data ya21-ya24, yb21 are obtained.

Further, the selectors 103u and 103v provide coded data [hu21, (hu22, hu23), hu21, (hu23, hu23), hu24],
[Hv21, (hv22, hv23), hv24] are supplied to the Huffman decoder 102. Huffman decoder 1
02 performs Huffman decoding on these encoded data.
Thereby, the DCT coefficient data u21 of the color difference components U and V
-U24, v21-v24 are obtained.

Next, in step ST2301,
As shown in FIG. 27, the DCT coefficient data yb21 of the luminance component Y and the DCT coefficient data u21-u24, v of the color difference components U and V obtained by the Huffman decoder 102
21-v24 are stored in the buffer memory M6.

The DCT coefficient data u21 of the color difference components U and V
−u24, v21−v24 are stored in the buffer memory M6 and transferred to the selectors 113u, 113v. DCT coefficient data ya21-y of luminance component Y
a24 is transferred to the combiner 104y.

Next, in step ST705a,
The same processing as described above is performed, and as shown in FIG. 27, 2 × 2 DCTs arranged in zigzag scan order
Coefficient data ya21-ya24, u21-u24, v2
1-v24 is obtained.

Next, in step ST707a,
The inverse quantizer 105 includes combination units 104y and 104
u, 104V, the DCT coefficient data ya21-ya24, u2 of the luminance component Y and the color difference components U, V obtained by
1-u24, v21-v24 are converted to the quantization table TB2.
Is inversely quantized based on the quantized value stored in.

Next, in step ST708a,
The inverse DCT transformer 106 performs a DCT on the luminance component Y and the color difference components U and V, which have been inversely quantized by the inverse quantizer 105.
Coefficient data ya21-ya24, u21-u24, v2
1-v24 is subjected to inverse DCT transformation. This allows
As shown in FIG. 28, the luminance component Y and the color difference components U and V
Decompressed image data (Y13, Y14, Y23, Y2
4), (U13, U14, U33, U34), (V1
3, V14, V33, V34) are obtained.

Next, in step ST402a,
As shown in FIG. 28, the decompressors 110u and 110v decompress the expanded image data (U13, U14, U3) of the color difference components U and V.
3, U34) and (V13, V14, V33, V34) are upsampled twice in the vertical direction on the screen.

Next, in step ST403a,
As shown in FIG. 28, the block extractor 111u outputs the pixel data U13 and U1 of the expanded image data of the color difference component U up-sampled by the decompressor 110u.
The block composed of U4, U13, and U14 is extracted and output. Further, as shown in FIG. 28, the block extractor 111v extracts a block composed of pixel data V13, V14, V13, and V14 from the expanded image data of the color difference component V up-sampled by the restorer 110v. Output.

Next, in step ST709a,
As shown in FIG. 28, the color space converter 107
Decompressed image data (Y13, Y14, Y23, Y24) of luminance component Y obtained by converter 106, and decompressed image data (U13, U14, U13, U14) of color difference components U, V from decompressors 110u, 110v. , (V1
3, V14, V13, and V14), and decompressed image data (R13, R14, R23, R24) of three primary color components, (G1
3, G14, G23, G24), (B13, B14, B
23, B24).

Next, in step ST710a,
The color number reducer 108 reduces the number of colors of the expanded image data of the three primary color components obtained by the color space converter 107.

Next, in step ST711a,
The decompressed image data of the three primary color components (R1
3, R14, R23, R24), (G13, G14, G
23, G24), (B13, B14, B23, B24)
Is transferred to the expanded image memory M2. This gives 2
The expanded image data of the upper right block on the screen among the expanded images composed of × 2 blocks is obtained.

Next, in step ST2302,
It is determined whether the decompression processing of one horizontal line block on the screen is completed. Here, since the expansion processing of the upper left and upper right blocks on the screen in the expanded image composed of 2 × 2 blocks has been completed, the process proceeds to step ST2303. At this time, the horizontal start controller 114
Outputs to the block extractors 111u and 111v a control signal CT indicating that the decompression processing of one horizontal line block on the screen is completed. In response to the control signal CT, the block extractors 111u and 111v switch blocks to be extracted.

(3) Decompression processing of lower left block on screen Next, in step ST2303, input means iy
2, iy3, based on the position information from the analyzer 101, encodes the encoded data (hyb12, hyb13) and hyb14 (see FIG. 29) of the second and third stages of the luminance component Y from the compressed image memory M1. Extract and Selector 10
Transfer to 3y. Then, the analyzer 101 selects the selector 1
The position information is incremented by the number of bytes of the encoded data transferred to 03y.

Next, in step ST704b,
The selector 103y supplies the Huffman decoder 102 with the encoded data (hyb12, hyb13) and hyb14 of the second and third stages of the luminance component Y supplied from the input means iy2 and iy3. Huffman decoder 102
Performs Huffman decoding on these encoded data. Thereby, the DCT coefficient data yb12-yb of the luminance component Y
14 is obtained.

Next, in step ST2304,
DC of color difference components U and V stored in buffer memory M6
The T coefficient data u11-u14, v11-v14, and the first-stage DCT coefficient data yb11 of the luminance component Y are read (see FIG. 29).

Next, in step ST705b,
Selector 109 of combiner 104y (see FIG. 1)
Stores the DCT coefficient data yb11 read from the buffer memory M6 and the DCT coefficient data (yb12, yb13) and yb14 from the Huffman decoder 102 in the storage areas d1-d3 (see FIG. 1) of the coefficient memory M4, respectively. Forward. Thereby, as shown in FIG. 29, 2 × 2 DCT coefficient data y arranged in zigzag scan order
b11-yb14 is obtained.

The selector 109 (see FIG. 1) of the combiners 104u and 104v outputs the DCT coefficient data [u11, (u12,
u13), u14], [v11, (v12, v13),
v14] is stored in the storage area d1-
d3 (see FIG. 1). As a result, as shown in FIG. 29, 2 × 2 pieces of DCT coefficient data u11-u14 and v11-v14 arranged in the zigzag scan order are obtained.

Next, in step ST707b,
The inverse quantizer 105 includes combination units 104y and 104
u, 104V, the DCT coefficient data ya11-ya14, u1 of the luminance component Y and the color difference components U, V obtained by
1-u14, v11-v14 are converted to the quantization table TB2.
Is inversely quantized based on the quantized value stored in.

Next, in step ST708b,
The inverse DCT transformer 106 performs a DCT on the luminance component Y and the color difference components U and V, which have been inversely quantized by the inverse quantizer 105.
Coefficient data ya11-ya14, u11-u14, v1
1-v14 is subjected to inverse DCT. This allows
As shown in FIG. 30, the luminance component Y and the color difference components U and V
Decompressed image data (Y31, Y32, Y41, Y4
2), (U11, U12, U31, U32), (V1
1, V12, V31, V32) are obtained.

Next, in step ST402b,
As shown in FIG. 30, the decompressors 110u and 110v decompress the expanded image data (U11, U12, U3) of the color difference components U and V.
1, U32) and (V11, V12, V31, V32) are up-sampled twice in the vertical direction on the screen.

Next, in step ST403b,
As shown in FIG. 30, the block extractor 111u outputs the pixel data U31 and U3 of the expanded image data of the color difference component U up-sampled by the decompressor 110u.
2, U31 and U32 are extracted and output. Further, as shown in FIG. 30, the block extractor 111v extracts a block composed of pixel data V31, V32, V31, V32 from the expanded image data of the color difference component V up-sampled by the restorer 110v. Output.

Next, in step ST709b,
The color space converter 107 has an inverse DCT as shown in FIG.
Decompressed image data of luminance component Y obtained by converter 106 (Y31, Y32, Y41, Y42) and decompressed image data of color difference components U, V from decompressors 110u, 110v (U31, U32, U31, U32) , (V3
, V32, V31, V32), the expanded image data (R31, R32, R41, R42) of three primary color components, (G3
1, G32, G41, G42), (B31, B32, B
41, B42).

Next, in step ST710b,
The color number reducer 108 reduces the number of colors of the expanded image data of the three primary color components obtained by the color space converter 107.

Next, in step ST711b,
Decompressed image data of three primary color components (R3
1, R32, R41, R42), (G31, G32, G
41, G42), (B31, B32, B41, B42)
Is transferred to the expanded image memory M2. This gives 2
The expanded image data of the lower left block on the screen among the expanded images composed of × 2 blocks is obtained.

Next, in step ST2305,
It is determined whether the decompression processing of one horizontal line block on the screen is completed. If it is determined that the process has been completed, the process proceeds to step ST2306. On the other hand, if it is determined that the processing has not been completed, the process returns to step ST2303. Here, since the decompression processing of the lower right block on the screen among the original image data composed of 2 × 2 blocks has not been completed, the process returns to step ST2303.

(4) Decompression process of lower right block on screen Then, in step ST2303, input means iy
2, iy3, based on the position information from the analyzer 101, encodes the encoded data (hyb22, hyb23) and hyb24 (see FIG. 31) of the second and third stages of the luminance component Y from the compressed image memory M1. Extract and Selector 10
Transfer to 3y.

Next, in step ST704b,
In the same manner as described above, DCT coefficient data yb22-yb24 of the luminance component Y is obtained.

Next, in step ST2304,
DC of color difference components U and V stored in buffer memory M6
The T coefficient data u21-u24, v21-v24, and the first-stage DCT coefficient data yb21 of the luminance component Y are read (see FIG. 31).

Next, in step ST705b,
The same processing as described above is performed, and as shown in FIG. 31, 2 × 2 DCTs arranged in zigzag scan order
Coefficient data yb21-yb24, u21-u24, v2
1-v24 is obtained.

Next, in step ST707b,
The inverse quantizer 105 includes combination units 104y and 104
u, DCT coefficient data yb21-yb24, u2 of the luminance component Y and the color difference components U, V obtained by 104v
1-u24, v21-v24 are converted to the quantization table TB2.
Is inversely quantized based on the quantized value stored in.

Then, in step ST708b,
The inverse DCT transformer 106 performs a DCT on the luminance component Y and the color difference components U and V, which have been inversely quantized by the inverse quantizer 105.
Coefficient data yb21-yb24, u21-u24, v2
1-v24 is subjected to inverse DCT transformation. This allows
As shown in FIG. 32, the luminance component Y and the color difference components U and V
Decompressed image data (Y33, Y34, Y43, Y4
4), (U13, U14, U33, U34), (V1
3, V14, V33, V34) are obtained.

Next, in step ST402b,
As shown in FIG. 32, the decompressors 110u and 110v decompress the expanded image data (U13, U14, U3) of the color difference components U and V.
3, U34) and (V13, V14, V33, V34) are upsampled twice in the vertical direction on the screen.

Next, in step ST403b,
As shown in FIG. 32, the block extractor 111u outputs the pixel data U33 and U3 of the expanded image data of the color difference component U up-sampled by the decompressor 110u.
4, U33 and U34 are extracted and output. Further, as shown in FIG. 32, the block extractor 111v extracts a block composed of pixel data V33, V34, V33, V34 from the expanded image data of the color difference component V up-sampled by the restorer 110v. Output.

Next, in step ST709b,
The color space converter 107 has an inverse DCT as shown in FIG.
Decompressed image data of luminance component Y obtained by converter 106 (Y33, Y34, Y43, Y44) and decompressed image data of color difference components U, V from decompressors 110u, 110v (U33, U34, U33, U34) , (V3
, V34, V33, V34), and decompressed image data (R33, R34, R43, R44) of three primary color components, (G3
3, G34, G43, G44), (B33, B34, B
43, B44).

Next, in step ST710b,
The color number reducer 108 reduces the number of colors of the expanded image data of the three primary color components obtained by the color space converter 107.

Next, in step ST711b,
Decompressed image data of three primary color components (R3
3, R34, R43, R44), (G33, G34, G
43, G44), (B33, B34, B43, B44)
Is transferred to the expanded image memory M2. This gives 2
The expanded image data of the lower right block on the screen among the expanded images composed of × 2 blocks is obtained.

Next, in step ST2305,
It is determined whether the decompression processing of one horizontal line block on the screen is completed. Here, since the expansion processing of the lower left and lower right blocks on the screen among the original image data composed of 2 × 2 blocks has been completed, the process proceeds to step ST2306.

Next, in step ST2306,
It is determined whether the decompression processing has been completed for all blocks included in (1 MCU) × (one line in the horizontal direction on the screen). If it is determined that the processing has been completed, the process proceeds to step ST712. On the other hand, if it is determined that the processing has not been completed, the process returns to step ST2303. here,
Since decompression processing has been completed for all blocks included in (1 MCU) × (one line in the horizontal direction on the screen) (see FIG. 2), the process proceeds to step ST712.

Next, in step ST712, all the blocks (here, 2 ×
It is determined whether or not the decompression process has been completed for two (2). When the decompression processing is completed for all the blocks, the processing ends. If not, the process returns to step ST703. Here, since the decompression processing has been completed for all the blocks, the processing is terminated.

As described above, the decompressed image data of the progressively encoded compressed image data is obtained. When displaying the decompressed image data on the display device, the decompressed image data composed of 2 × 2 blocks in the order of the upper left block, the upper right block, the lower left block, and the lower right block is displayed on the screen. Is displayed.

Here, the expansion processing of the progressively encoded compressed image data has been described. When decompressing the sequentially encoded compressed image data, input means iy1-iy3, iu1-iu3, iv
The encoded data is sequentially supplied to the Huffman decoder 102 directly from the compressed image memory M1 without going through 1-iv3. Then, Huffman decoding, inverse quantization, inverse DCT transform, upsampling, color space conversion, and color number reduction processing are performed. As a result, decompressed image data of the sequentially encoded compressed image data is obtained.

<Effects> As described above, the image processing apparatus according to the third embodiment of the present invention includes the buffer memory M6
, Selectors 113u and 113v, and horizontal start controller 1
14, the decompressors 110u and 110v, and the block extractors 111u and 111v are provided, so that the color difference components U and V of the progressively encoded compressed image data are down-sampled in the vertical direction on the screen. However, the compressed image data can be expanded for each block.

Further, there is no need to repeatedly perform encoded data extraction processing and Huffman decoding processing on the same block.

Here, the buffer memory M6 and the selectors 113u and 113v are connected to the Huffman decoder 1
02 and the combiners 104y, 104u, 104v, but instead of this, an inverted D
It may be provided between the CT converter 106 and the restoration units 110u and 110v. In this case, the data stored in the buffer memory M6 is not the DCT coefficient data (frequency component) but the expanded image data (brightness component). Therefore,
Decompressed image data of color difference components U and V (U11, U12, U
31, U32), (U13, U14, U33, U3
4), (V11, V12, V31, V32), (V1
3, V14, V33, V34) need not be stored in the buffer memory M6, and the decompressed image data (U31, U32), (U33, U34),
Only (V31, V32) and (V33, V34) need to be stored in the buffer memory M6. Therefore, the capacity of the buffer memory M6 can be reduced.

As shown in FIG. 34, a buffer memory M6 and selectors 113u and 113v may be provided between the inverse DCT converter 106 and the decompressors 110u and 110v.

As shown in FIG. 35, the inverse quantizer 1
05 between the DCT 05 and the inverse DCT converter 106
6 and selectors 113u and 113v may be provided.

In the description of the first to third embodiments, the color difference components U and V of the progressively encoded compressed image data are down-sampled by a factor of に in the vertical direction on the screen. Although the case has been described, the image processing apparatus according to the present invention can be similarly applied to a case where downsampling is performed at another magnification (for example, 1/4).

Also, DCT is used as orthogonal transform / inverse orthogonal transform.
Although the case where Huffman coding / decoding is used as entropy coding / decoding using transform / inverse DCT transform has been described, other orthogonal transform / inverse orthogonal transform and entropy coding / decoding may be used. Good.

Also, as a method of implementing progressive coding, a case has been described where quantized DCT coefficient data is divided into a plurality of stages based on frequency. Instead, other implementation methods, such as quantization, are used. DC
A method of dividing the T coefficient data into a plurality of stages for each bit may be used.

Further, the image processing apparatus according to the first to third embodiments does not always require a hardware configuration, and can be realized by software.

[0299]

According to the image processing apparatus and the image processing method of the present invention, it is possible to decompress progressively encoded compressed image data in block units or MCU units.

According to the image data transfer method of the present invention, image data corresponding to the first component and the second component can be transferred.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an overall configuration of an image processing apparatus according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a flow of a process for creating progressively encoded compressed image data from original image data.

FIG. 3 is a flowchart illustrating a flow of a process of creating progressively encoded compressed image data from original image data.

FIG. 4 is a flowchart illustrating a procedure of a decompression process performed by the image processing apparatus illustrated in FIG. 1;

FIG. 5 is a diagram illustrating a state of data at each stage of a decompression process performed by the image processing apparatus illustrated in FIG. 1;

FIG. 6 is a diagram illustrating a state of data at each stage of a decompression process performed by the image processing apparatus illustrated in FIG. 1;

FIG. 7 is a diagram illustrating a state of data at each stage of a decompression process performed by the image processing apparatus illustrated in FIG. 1;

FIG. 8 is a diagram illustrating a state of data at each stage of a decompression process performed by the image processing apparatus illustrated in FIG. 1;

FIG. 9 is a diagram showing a state of data at each stage of the decompression process by the image processing apparatus shown in FIG. 1;

10 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 1;

11 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 1;

12 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 1;

FIG. 13 is a block diagram illustrating an overall configuration of an image processing apparatus according to a second embodiment of the present invention.

FIG. 14 is a flowchart illustrating a procedure of a decompression process performed by the image processing apparatus illustrated in FIG. 13;

15 is a diagram illustrating a state of data at each stage of the decompression process by the image processing apparatus illustrated in FIG.

16 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 13;

17 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 13;

18 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG.

19 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 13;

FIG. 20 is a block diagram illustrating an overall configuration of an image processing apparatus according to a modification of the second embodiment of the present invention.

FIG. 21 is a block diagram illustrating an overall configuration of an image processing apparatus according to a modification of the second embodiment of the present invention.

FIG. 22 is a block diagram illustrating an overall configuration of an image processing apparatus according to a modification of the second embodiment of the present invention.

FIG. 23 is a block diagram illustrating an overall configuration of an image processing apparatus according to a third embodiment of the present invention.

FIG. 24 is a flowchart illustrating a procedure of a decompression process performed by the image processing apparatus illustrated in FIG. 23;

25 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 23.

26 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG.

FIG. 27 is a diagram illustrating a state of data at each stage of the decompression process by the image processing apparatus illustrated in FIG. 23;

FIG. 28 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 23;

FIG. 29 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 23;

30 is a diagram illustrating a state of data at each stage of a decompression process by the image processing device illustrated in FIG. 23;

31 is a diagram illustrating a state of data at each stage of the decompression process by the image processing device illustrated in FIG. 23.

32 is a diagram illustrating a state of data in each stage of the decompression process by the image processing device illustrated in FIG.

FIG. 33 is a block diagram showing an overall configuration of an image processing device according to a modification of the third embodiment of the present invention.

FIG. 34 is a block diagram showing an overall configuration of an image processing device according to a modification of the third embodiment of the present invention.

FIG. 35 is a block diagram showing the overall configuration of an image processing device according to a modification of the third embodiment of the present invention.

FIG. 36 is a diagram illustrating a flow of processing when original image data is compressed by a progressive encoding method when a chrominance component is down-sampled.

FIG. 37 is a diagram illustrating a flow of processing when the compressed image data progressively encoded as described in the related art 2 is decompressed by the image processing apparatus illustrated in the related art 1.

[Explanation of symbols]

Reference Signs List 101 analyzer 102 Huffman decoder 103y, 103u, 103v selector 104y, 104u, 104v combiner 106 inverse DCT converter 110u, 110v decompressor 111u, 111v block extractor 114 horizontal start controller iy1-iy3, iu1-iu3 iv1-iv3 input means 112, 113u, 113v selector M3 start position memory M5, M6 buffer memory

Claims (22)

[Claims] 1. An original image including a first component and a second component is down-sampled in a vertical direction on a screen with respect to an original image, and the first component and a second component after the down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing apparatus for decompressing compressed image data, comprising: for one horizontal line block on a screen, each block of encoded data of a first component and each stage of a second component; Extracting the encoded data of the first component from the compressed image data, and entropy encoding the encoded data of each stage of the first component and the encoded data of the second component extracted by the extracting unit. Coding means, and coefficient data of each stage of the first component obtained by the entropy decoding means are combined to obtain coefficient data of each block of the first component, and obtained by the entropy decoding means. Combining means for combining the obtained coefficient data of each stage of the second component to obtain coefficient data of each block of the second component; and coefficient data of each block of the first component obtained by the combining means, and Means for performing an inverse orthogonal transform on the coefficient data of each block of the second component; and up-sampling the expanded image data of each block of the second component obtained by the inverse orthogonal transform means in the vertical direction on the screen. A restoring unit, among the expanded image data of the second component up-sampled by the restoring unit, Block extracting means for extracting a portion corresponding to the decompressed image data of each block of the first component obtained as described above, wherein the extracting means performs each step for all the blocks of one horizontal line on the screen. After extracting the encoded data of the first component, the encoded data of each stage is extracted for each block for the block of one line following the one line for the first component. Said 1
An image processing apparatus for extracting encoded data of each stage for each block of blocks for a line. 2. The image processing apparatus according to claim 1, further comprising: a unit that stores, for the second component, a position where encoded data of each stage with respect to a first block on the line is stored. After extracting the encoded data at each stage for all the blocks on the line, the extracting unit extracts, for each block on the line, a second component based on the position stored in the storage unit. An image processing apparatus for extracting encoded data of stages. 3. An original image including a first component and a second component is down-sampled in a vertical direction on a screen with respect to an original image, and the first component and a second component after down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing apparatus for decompressing compressed image data, comprising: 1MCU × encoded data of each stage of a first component and a second component of a block included in one horizontal line on a screen. Means for extracting the encoded data of each stage from the compressed image data; encoded data of each stage of the first component and encoded data of each stage of the second component extracted by the extracting means Means for performing entropy decoding; a buffer memory for storing coefficient data for blocks other than the block on the last line among the coefficient data of each stage of the first component obtained by the entropy decoding means; For the component, the coefficient data of each block on the last line obtained by the entropy decoding means is combined to obtain the coefficient data of the block, and the coefficient data of each stage for the block constituting one MCU together with the block Is read from the buffer memory, and the read coefficient data is combined to obtain coefficient data of a block constituting the one MCU, while the coefficient data of each stage obtained by the entropy decoding unit is obtained for the second component. Combining the coefficient data of each block Means for performing inverse orthogonal transformation on coefficient data of each block of the first component and coefficient data of each block of the second component obtained by the combination means; An image processing apparatus comprising: a restoration unit for upsampling the obtained expanded image data of each block of the second component in a vertical direction on a screen. 4. An original image including a first component and a second component is down-sampled in a vertical direction on a screen by a second component, and the first component and a second component after down-sampling are sampled. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing apparatus for decompressing compressed image data, comprising: 1MCU × encoded data of each stage of a first component and a second component of a block included in one horizontal line on a screen. Means for extracting the encoded data of each stage from the compressed image data; encoded data of each stage of the first component and encoded data of each stage of the second component extracted by the extracting means Means for performing entropy decoding; and coefficient data of each stage of the first component obtained by the entropy decoding means are combined to obtain coefficient data of each block of the first component, and the entropy decoding means Combining means for combining the obtained coefficient data of each stage of the second component to obtain coefficient data of each block of the second component; and coefficient data of each block of the first component obtained by the combining means; Means for performing an inverse orthogonal transform on the coefficient data of each block of the second component, and for the blocks excluding the block on the last line in the expanded image data of the first component obtained by the inverse orthogonal transform means A buffer memory for storing decompressed image data, and decompression of each block of the second component obtained by the inverse orthogonal transformation means. Restoration means for up-sampling the image data in the vertical direction on the screen, wherein the image processing apparatus obtains expanded image data for each block on the last line by the inverse orthogonal transformation means for the first component. An image processing device that reads out, from the buffer memory, decompressed image data of a block that constitutes one MCU together with the block. 5. An original image including a first component and a second component, wherein the second component is down-sampled in a vertical direction on a screen, and the first component and the second component after the down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing apparatus for decompressing compressed image data, comprising: 1MCU × encoded data of each stage of a first component and a second component of a block included in one horizontal line on a screen. Means for extracting the encoded data of each stage from the compressed image data; and, of the encoded data of the first component extracted by the extracting means, A buffer memory for storing encoded data to be encoded, for the first component, encoded data for each block on the last line extracted by the extraction means, and one MCU together with the block stored in the buffer memory. Means for performing entropy decoding on the encoded data for the blocks constituting the second component, and for the second component, means for entropy decoding the encoded data extracted by the extracting means, and a second component obtained by the entropy decoding means. The coefficient data of each block of the first component is obtained by combining the coefficient data of each stage of the first component, and the coefficient data of each stage of the second component obtained by the entropy decoding means is combined. Combination means for obtaining coefficient data of each block of the two components; and said combination means Means for performing an inverse orthogonal transform on the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by An image processing apparatus comprising: a restoration unit that upsamples expanded image data of each block in a vertical direction on a screen. 6. An original image including a first component and a second component is down-sampled in a vertical direction on a screen with respect to a second component, and the first component and a second component after down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing apparatus for decompressing compressed image data, comprising: for one horizontal line block on a screen, each block of encoded data of a first component and each stage of a second component; Extracting the encoded data of the first component from the compressed image data, and entropy encoding the encoded data of each stage of the first component and the encoded data of the second component extracted by the extracting unit. Encoding means; a buffer memory for storing encoded data of each step of the second component obtained by the entropy decoding means; and a buffer memory for storing each step of the first component obtained by the entropy decoding means. The coefficient data of each block of the first component is obtained by combining the coefficient data, and the coefficient data of each block of the second component obtained by the entropy decoding means is combined to obtain the coefficient data of each block of the second component. Combining means for obtaining coefficient data; means for performing inverse orthogonal transform on coefficient data of each block of the first component and coefficient data of each block of the second component obtained by the combining means; Restoration means for up-sampling the expanded image data of each block of the second component obtained by the conversion means in the vertical direction on the screen; Block extraction for extracting a portion corresponding to the expanded image data of each block of the first component obtained by the inverse orthogonal transform unit, from the expanded image data of the second component after being upsampled by the restoration unit. Means for extracting the coded data at each stage for all the blocks of one line in the horizontal direction on the screen, and then extracting the first component for one line following the one line. After extracting the coded data of each stage for each block of the block, the combination means obtains coefficient data for all blocks of one horizontal line on the screen,
An image processing apparatus for obtaining coefficient data of each block by combining the encoded data of each stage of each block stored in the buffer memory with respect to the component (1). 7. An original image including a first component and a second component, the second component is down-sampled in a vertical direction on a screen, and the first component and the second component after the down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing apparatus for decompressing compressed image data, comprising: for one horizontal line block on a screen, each block of encoded data of a first component and each stage of a second component; Extracting the encoded data of the first component from the compressed image data, and entropy encoding the encoded data of each stage of the first component and the encoded data of the second component extracted by the extracting unit. Coding means, and coefficient data of each stage of the first component obtained by the entropy decoding means are combined to obtain coefficient data of each block of the first component, and obtained by the entropy decoding means. Combining means for obtaining the coefficient data of each block of the second component by combining the coefficient data of each stage of the second component obtained, and the coefficient data of each block of the first component obtained by the combining means, Means for performing inverse orthogonal transform on coefficient data of each block of the second component; buffer memory for storing expanded image data of each block of the second component obtained by the inverse orthogonal transform means; Restoration means for up-sampling the decompressed image data of each block of the second component obtained by the conversion means in the vertical direction on the screen; Block extracting means for extracting a portion corresponding to the decompressed image data of each block of the first component obtained by the inverse orthogonal transform means, out of the decompressed image data of the second component upsampled by the restoration means After extracting the encoded data of each stage for all the blocks for one line in the horizontal direction on the screen, the extracting means extracts the first component for the next one line after the one line. For each block, the encoded data of each stage is extracted for each block, and the restoration means up-samples the decompressed image data for all blocks of the second component for one line in the horizontal direction on the screen. The expanded image data of each block of the second component stored in the buffer memory is vertically up-sampled on the screen. Image processing device. 8. An original image including a first component and a second component is down-sampled in a vertical direction on a screen, and the first component and the second component after the down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing apparatus for decompressing compressed image data, comprising: for one horizontal line block on a screen, each block of encoded data of a first component and each stage of a second component; Means for extracting the encoded data of the compressed image data from the compressed image data; buffer memory for storing the encoded data of each stage of the second component extracted by the extracting means; Means for entropy decoding the encoded data of each stage of the first component and the encoded data of each stage of the second component thus obtained, and The coefficient data of each block of the first component is obtained by combining the coefficient data, and the coefficient data of each block of the second component obtained by the entropy decoding means is combined to obtain the coefficient data of each block of the second component. Combining means for obtaining coefficient data; means for performing inverse orthogonal transform on coefficient data of each block of the first component and coefficient data of each block of the second component obtained by the combining means; Restoration means for up-sampling the expanded image data of each block of the second component obtained by the conversion means in the vertical direction on the screen; Therefore, a block extraction unit that extracts a portion corresponding to the expanded image data of each block of the first component obtained by the inverse orthogonal transform unit, out of the expanded image data of the second component after being upsampled, After extracting the encoded data at each stage for all the blocks for one line in the horizontal direction on the screen, the extracting means performs the following for the first component, and for the block for one line next to the one line. The encoded data of each stage is extracted for each block.
After entropy decoding the encoded data of each stage for all the blocks of the line, entropy decoding of the encoded data of each stage of each block stored in the buffer memory for the second component. Image processing apparatus. 9. An original image including a first component and a second component is downsampled in a vertical direction on a screen with respect to an original image, and the first component and a second component after the downsampling are sampled. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing method for decompressing compressed image data, comprising: for each block of one horizontal line on a screen, encoded data of each stage of a first component and each stage of a second component for each block; Extracting the encoded data of the first component and the encoded data of each stage of the second component extracted by the extracting step. A step of Ropi decoded, first obtained by the entropy decoding step 1
The coefficient data of each block of the first component is obtained by combining the coefficient data of each stage of the component of the second component, and the coefficient data of each stage of the second component obtained by the entropy decoding step is combined. Combining the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combining step by performing an inverse orthogonal transformation. A restoring step of up-sampling the expanded image data of each block of the second component obtained in the inverse orthogonal transformation step in a vertical direction on a screen; In the expanded image data of the component, each block of the first component obtained in the inverse orthogonal transformation step is used. Extracting a portion corresponding to the decompressed image data of step (a). After extracting the encoded data of each stage for all the blocks for one line in the horizontal direction on the screen, The coded data of each stage is extracted for each block for the block of one line following the one line, and the second component is coded for each block of the block for one line again for each block. An image processing method characterized by extracting data. 10. An original image including a first component and a second component, the second component is down-sampled in a vertical direction on a screen, and the first component and the second component after the down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing method for decompressing compressed image data, comprising: 1 MCU × a block included in one horizontal line on a screen, encoded data of each stage of a first component and a second component Extracting the encoded data of each stage from the compressed image data; and the encoded data of each stage of the first component and the code of each stage of the second component extracted by the extracting step. A step of entropy decoding the data, the first of the obtained by the entropy decoding step
Storing, in the buffer memory, coefficient data for blocks other than the block on the last line of the coefficient data of each stage of the component, and for the first component on the last line obtained by the entropy decoding step. The coefficient data of each block is obtained by combining the coefficient data of each block, and the coefficient data of each stage for the block that constitutes one MCU together with the block is read from the buffer memory, and the read coefficient data is combined, Combining the coefficient data of each block obtained by the entropy decoding step for the second component while obtaining the coefficient data of the blocks making up one MCU, and obtaining the coefficient data of each block; Gained by Applying an inverse orthogonal transform to the obtained coefficient data of each block of the first component and the coefficient data of each block of the second component; and each block of the second component obtained by the inverse orthogonal transform step. A decompression step of upsampling the decompressed image data in the vertical direction on the screen. 11. An original image including a first component and a second component is down-sampled in a vertical direction on a screen with respect to a second component, and the first component and the second component after the down-sampling are sampled. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing method for decompressing compressed image data, comprising: 1 MCU × a block included in one horizontal line on a screen, encoded data of each stage of a first component and a second component Extracting the encoded data of each stage from the compressed image data; and the encoded data of each stage of the first component and the code of each stage of the second component extracted by the extracting step. A step of entropy decoding the data, the first of the obtained by the entropy decoding step
The coefficient data of each block of the first component is obtained by combining the coefficient data of each stage of the component of the second component, and the coefficient data of each stage of the second component obtained by the entropy decoding step is combined. Combining the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combining step by performing an inverse orthogonal transformation. Storing the decompressed image data for blocks other than the block on the last line in the decompressed image data of the first component obtained in the inverse orthogonal transform step in a buffer memory; The obtained expanded image data of each block of the second component is vertically up-sampled on the screen. When the decompressed image data for each block on the last line is obtained by the inverse orthogonal transformation step for the first component, the decompressed image data of a block constituting one MCU together with the block. Reading from the buffer memory. 12. An original image including a first component and a second component, the second component is down-sampled in a vertical direction on a screen, and the first component and the second component after the down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing method for decompressing compressed image data, comprising: 1 MCU × a block included in one horizontal line on a screen, encoded data of each stage of a first component and a second component Extracting the encoded data at each stage from the compressed image data; and removing the block on the last line from the encoded data of the first component obtained by the extracting step. Storing encoded data for a block in the buffer memory; and for the first component, encoded data for each block on the last line extracted by the extracting step, and the block stored in the buffer memory. And entropy-decoding the encoded data for the blocks making up one MCU, while entropy-decoding the encoded data extracted by the extraction step for the second component; The first obtained
The coefficient data of each block of the first component is obtained by combining the coefficient data of each stage of the component of the second component, and the coefficient data of each stage of the second component obtained by the entropy decoding step is combined. Combining the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combining step by performing an inverse orthogonal transformation. And a restoration step of up-sampling the expanded image data of each block of the second component obtained in the inverse orthogonal transformation step in a vertical direction on a screen. 13. An original image including a first component and a second component, wherein the second component is down-sampled in a vertical direction on a screen, and the first component and the second component after the down-sampling are obtained. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing method for decompressing compressed image data, comprising: for each block of one horizontal line on a screen, encoded data of each stage of a first component and each stage of a second component for each block; Extracting the encoded data of the first component and the encoded data of each stage of the second component extracted by the extracting step. A step of Toropi decoding, second obtained by said entropy decoding step
Storing the coded data of each stage of the component in the buffer memory; and the first obtained by the entropy decoding step.
The coefficient data of each block of the first component is obtained by combining the coefficient data of each stage of the component of the first component, and the coefficient data of each stage of the second component obtained by the entropy decoding means is combined with the second component. Combining the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combining step by performing an inverse orthogonal transformation. A restoring step of up-sampling the expanded image data of each block of the second component obtained in the inverse orthogonal transformation step in a vertical direction on a screen; Of the decompressed image data of the component, each block of the first component obtained in the inverse orthogonal transformation step is Extracting a portion corresponding to long image data. After extracting coded data at each stage for all blocks of one line in the horizontal direction on the screen, For each block of the next one line of the line, the coded data of each stage is extracted, and coefficient data is obtained for all the blocks of one horizontal line on the screen. An image processing apparatus for obtaining coefficient data of each block by combining encoded data of each block of each stage stored in the buffer memory. 14. An original image including a first component and a second component, the second component is down-sampled in a vertical direction on a screen, and the first component and the second component after the down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing method for decompressing compressed image data, comprising: for each block of one horizontal line on a screen, encoded data of each stage of a first component and each stage of a second component for each block; Extracting the encoded data of the first component and the encoded data of each stage of the second component extracted by the extracting step. A step of Toropi decoded, first obtained by the entropy decoding step 1
The coefficient data of each stage of the first component is obtained by combining the coefficient data of each stage of the component (i), and the coefficient data of each stage of the second component obtained by the entropy decoding means is combined with the second component. Combining the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combining step by performing an inverse orthogonal transformation. Storing the decompressed image data of each block of the second component obtained in the inverse orthogonal transform step in a buffer memory; decompressing each block of the second component obtained in the inverse orthogonal transform step A restoration step of up-sampling the image data in the vertical direction on the screen; A block extraction step of extracting a portion corresponding to the expanded image data of each block of the first component obtained by the inverse orthogonal transformation step, from the expanded image data of the second component after the pulling, After extracting the encoded data of each stage for all the blocks for one line in the horizontal direction on the screen, for the first component, for each block of the block for one line next to the one line, After extracting the encoded data and up-sampling the decompressed image data for all the blocks of the second component for one line in the horizontal direction on the screen, each block of the second component stored in the buffer memory is extracted. An image processing method characterized by up-sampling expanded image data in a vertical direction on a screen. 15. An original image including a first component and a second component, the second component is down-sampled in a vertical direction on a screen, and the first component and the second component after the down-sampling. In contrast, for each block composed of a predetermined number of pixels, the coefficient data obtained by performing the orthogonal transformation is divided into a plurality of stages, and the coefficient data is obtained by entropy coding the coefficient data for each stage. An image processing method for decompressing compressed image data, comprising: for each block of one horizontal line on a screen, encoded data of each stage of a first component and each stage of a second component for each block; Extracting the encoded data of the compressed image data from the compressed image data; and storing the encoded data of each stage of the second component extracted in the extracting step in a buffer memory; Entropy decoding the encoded data of each stage of the first component and the encoded data of each stage of the second component extracted by the extracting step; and the first obtained by the entropy decoding step.
The coefficient data of each block of the first component is obtained by combining the coefficient data of each stage of the component of the first component, and the coefficient data of each stage of the second component obtained by the entropy decoding means is combined with the second component. Combining the coefficient data of each block of the first component and the coefficient data of each block of the second component obtained by the combining step by performing an inverse orthogonal transformation. A restoring step of up-sampling the expanded image data of each block of the second component obtained in the inverse orthogonal transformation step in a vertical direction on a screen; Of the decompressed image data of the component, each block of the first component obtained in the inverse orthogonal transformation step is Extracting a portion corresponding to long image data. After extracting coded data at each stage for all blocks of one line in the horizontal direction on the screen, After extracting the coded data of each stage for each block of the next one line block of the line, and entropy decoding the coded data of each stage for all the blocks of one horizontal line on the screen, And an entropy decoding of encoded data of each stage of each block stored in the buffer memory for the second component. 16. A method of transferring image data including a first component and a second component down-sampled in the vertical direction on a screen in a block unit composed of a predetermined number of pixels, After transferring one line of image data in the horizontal direction on the screen, for the first component, the data for the next one line after the one line in the horizontal direction is transferred, and for the second component, An image data transfer method, wherein the data for one line in the direction is transferred again. 17. The image data transfer method according to claim 16, wherein a second component to be transferred is up-sampled in a vertical direction on a screen, and said first component of said up-sampled second component is up-sampled.
A method for extracting a portion corresponding to the component (1), and transferring the extracted portion. 18. A method for transferring image data including a first component and a second component vertically down-sampled on a screen in a block unit composed of a predetermined number of pixels, For the first component, data for a predetermined number of lines in the horizontal direction on the screen is stored in a buffer memory, and then the second component and the first component stored in the buffer memory are stored on the screen. A first component on the next one line in the horizontal direction, and a first component on the next one of the first components stored in the buffer memory, which is the same as the first component on the screen in the vertical direction. An image data transfer method comprising transferring a first component on a line. 19. The image data transfer method according to claim 18, wherein the number (predetermined number) of horizontal lines on the screen of the first component stored in the buffer memory is the second one for one block. Wherein the number of components is one less than the number of vertical blocks on the screen included in the state before down-sampling. 20. A method of transferring image data including a first component and a second component vertically down-sampled on a screen in a block unit composed of a predetermined number of pixels, After transferring the image data for one line in the horizontal direction on the screen and storing the data of the second component in the buffer memory, the first component is transferred for the next one line in the horizontal direction. And transferring the data stored in the buffer memory for the second component. 21. The image data transfer method according to claim 20, wherein the second component and the first component for the next one line of the data stored in the buffer memory are displayed on a screen. An image data transfer method comprising transferring data on the same line in the vertical direction. 22. The data transfer method according to claim 20, wherein a second component to be transferred is up-sampled in a vertical direction on a screen, and the first component to be transferred among the up-sampled second components is transferred. A method for extracting a portion corresponding to the component (1), and transferring the extracted portion.