JP4713970B2 - Image processing device - Google Patents

Description

  The present invention relates to an image processing apparatus, and more particularly to an image processing apparatus capable of performing compression processing of image data.

  In the image processing apparatus, as a technique for reducing the data transfer amount of the bus and the memory capacity, a technique that enables pipeline processing of a plurality of image processes including an enlargement / reduction process is proposed in Patent Document 1. Yes. In the technique disclosed in Patent Document 1, an image processing apparatus is configured so that a plurality of image processing units and subsequent compression processing units are directly connected via a buffer. Interpolation pixel data necessary for image processing and data having column direction data for MCU (Minimum Coded Unit) which is the minimum unit necessary for compression processing are input together. Here, the MCU is data composed of, for example, data of 8 pixels × 8 pixels or 16 pixels × 16 pixels.

In Patent Document 1, it is not necessary to transfer processing data to a storage device such as an SDRAM in the processing from the image processing unit to the compression processing unit, so that the data transfer amount of the bus can be reduced.
JP 2000-311241 A

  In the method of Patent Document 1, the width in the column direction of data input to the compression processing unit after image processing is necessarily limited to one MCU. Here, the number of interpolation pixels tends to increase due to the recent advancement of image processing technology. On the other hand, since the number of interpolated pixels does not depend on the data width after processing, increasing the width in the column direction of input data when performing compression processing leads to higher efficiency of image processing.

  The present invention has been made in view of the above circumstances, and provides an image processing apparatus capable of performing more efficient image processing by increasing the column-direction width of image data that can be input to a compression processing unit. Objective.

In order to achieve the above object, an image processing apparatus according to a first aspect of the present invention includes a plurality of serially connected image processing units for performing spatial image processing on supplied image data. The image data processed by the image processing unit is scanned with the block unit image data in units of blocks such that the number of input pixels in the column direction is N (N ≧ 2) times the minimum unit of compression processing. A compression processing unit for sequentially inputting one block line arranged for the length of each block and sequentially compressing the input block-unit image data to obtain block-unit compressed image data for one block line, and the compression The first rearrangement unit that sequentially reads out the compressed image data for each block line obtained by the processing unit and performs the first rearrangement process, and the first rearrangement unit rearranges the compressed image data. A storage unit having a first storage area in which compressed image data for each block line is sequentially written, and the compressed image data written in the first storage area of the storage unit is read, A second rearrangement unit that sequentially writes the compressed image data in units of blocks corresponding to one block line in which the rearrangement order is changed to a second storage area different from the first storage area. It is characterized by comprising.

  According to the first aspect, the column direction width of the image data that can be input to the compression processing unit can be increased, and accordingly, the column direction width of the image data that can be input to the image processing unit can also be increased. . In addition, by increasing the column direction width of image data that can be input to the compression processing unit, compression processing is performed in an order that is not suitable for later reproduction, but this may be rearranged in an order that is suitable for later reproduction. it can.

  ADVANTAGE OF THE INVENTION According to this invention, the image processing apparatus which can perform more efficient image processing can be provided by enlarging the column direction width | variety of the image data which can be input into a compression process part.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[First Embodiment]
FIG. 1 is a diagram showing a main configuration of an image processing apparatus according to the first embodiment of the present invention. The image processing apparatus illustrated in FIG. 1 includes an imaging unit 101, a preprocessing unit 102, a DMA 103, a bus 104, a storage device 105, a DMA 106, an image processing unit 107, a compression processing unit 108, and a DMA 109. It is configured. Further, the image processing unit 107 can perform a plurality of different image processes including spatial image processing such as enlargement / reduction processing (in FIG. 1, the image processing unit A 107a and the image processing unit 107). Part B107b) is connected directly.

  In FIG. 1, an image signal obtained by being imaged by the imaging unit 101 is subjected to predetermined analog processing and digital processing in a preprocessing unit 102. The image data generated by these processes is transferred to the bus 104 via the DMA 103 and written in the storage device 105 configured by SDRAM or the like.

  The image data written in the storage device 105 is input to the image processing unit 107 via the DMA 106. Here, in the first embodiment, the number of input pixels in the column direction is an integer multiple of the minimum unit data (MCU) necessary for the subsequent compression processing (for example, four times here) from the storage device 105 to the image processing unit 107. ) And image data composed of interpolated pixel data for interpolating the pixel data to be deleted in the image processing unit 107 (for such image data, this is referred to as image data for one block line). Entered. In order to input such image data for one block line, the image data stored in the storage device 105 is obtained by adding the interpolation pixel data to the data width four times the MCU in the column direction pixel by pixel. Image data is read by reading only the width and repeating this reading for the length of the scanning line in the row direction. By inputting the image data for each unit read in this way to the image processing unit 107, the interpolated pixel data portion is deleted by image processing in the image processing unit A 107a and the image processing unit B 107b, and finally. Only image data for each unit, in which the number of input pixels in the column direction is four times that of the MCU, is input to the compression processing unit 108. The compression processing unit 108 performs compression processing on the input image data, and then writes the compressed image data obtained thereby to the storage device 105 via the DMA 109.

  Note that the image processing in the image processing unit 107 does not necessarily need to be performed entirely, and can be bypassed so as not to perform unnecessary image processing. In the example of FIG. 1, the processing in the image processing unit B 107b can be bypassed.

  FIG. 2 is a diagram illustrating a configuration of the compression processing unit 108 which is a main part of the image processing apparatus according to the first embodiment of the present invention. The compression processing unit 108 illustrated in FIG. 2 includes a JPEG core I / F unit 201, a JPEG core 202, a first rearrangement unit 203, and a second rearrangement unit 204.

  The JPEG core I / F unit 201 is provided with four buffers (BUF-A to BUF-D), and the image data input to the JPEG core I / F unit 201 is the buffers BUF-A to BUF. Stored sequentially in -D. Here, each of the buffers BUF-A to BUF-D has a capacity sufficient to store the image data input to the JPEG core I / F unit 201 by one compression unit. This one compression unit refers to non-compressed image data in which the number of pixels in the column direction is 1 MCU and the number of pixels in the row direction is an integer multiple of 1 MCU. A restart marker is added to the end portion of the image data in one compression unit during JPEG compression. Here, the restart marker is data for indicating a delimiter of compression processing. That is, if the restart marker interval is 2, that is, if a restart marker is added every 2 MCUs during compression, one compression unit is image data for 2 MCUs as shown in FIG. If a restart marker is added every 4 MCUs during compression, one compression unit is image data for 4 MCUs as shown in FIG.

  Hereinafter, description will be continued by numbering the compression units in the image data input to the JPEG core I / F unit 201 as shown in FIG. 4, the compression units 1-1, 1-2, 1-3, and 1-4 at the upper left end are input to the JPEG core I / F unit 201 in parallel as shown in FIG. . Here, if the image data consisting of the compression unit image data 1-1, 1-2, 1-3, and 1-4 is referred to as block unit image data, the block unit image data is arranged in the row direction. Data is input for each pixel in the direction indicated by the arrow in FIG. 5, the first 1 MCU image data in the column direction is stored in the buffer BUF-A, and the next 1 MCU image data is stored in the buffer BUF-B. The data is stored in the buffer BUF-C and the buffer BUF-D, and this input operation is repeated until the number of pixels in the row direction reaches the number of pixels in one compression unit. These four compression units are separately stored in the buffers BUF-A to BUF-D as shown in FIG. When image data for one compression unit is stored in each of the buffers BUF-A to BUF-D, first, image data for one compression unit is input from the buffer BUF-A to the JPEG core 202, and thereafter the buffer BUF-B. , Buffer BUF-C, buffer BUF-D and image data for one compression unit are sequentially input to the JPEG core 202. When image data for one compression unit is output from each of the buffers BUF-A to BUF-D, the next four compression units 2-1, 2-2, 2-3, 2-4 are parallelized to the JPEG core I. / F unit 201 is input. Thereafter, similarly, four compression units are input in parallel to the JPEG core I / F unit 201, and four compression units are stored in four buffers and input to the JPEG core 202 as soon as possible. As a result, image data for each compression unit is input to the JPEG core 202 in the order shown in FIG.

  The JPEG core 202 performs JPEG compression processing every time image data for one compression unit is input. Each time image data for one compression unit is compressed, a restart marker is added to generate compressed image data. Here, the restart markers are numbered with 8 modulo. That is, each time compressed image data for one compression unit is generated, restart markers numbered in the order of FFD0, FFD1, FFD2, FFD3, FFD4, FFD5, FFD6, FFD7, FFD0, FFD1,. Added.

  Here, when the compressed image data input and compressed to the JPEG core 202 in the order shown in FIG. 7 is written as it is into the storage device 105 in the order shown in FIG. 4 is read out in the row direction, the compressed image data stored in the order of FIG. 7 cannot be decompressed in the original image order shown in FIG. Therefore, after the rearrangement of the compressed image data is performed using the first rearrangement unit 203 and the second rearrangement unit 204, the storage device 105 stores them.

  FIG. 8A is a diagram illustrating an internal configuration of the first rearrangement unit 203 and the second rearrangement unit 204.

  First, the first rearrangement unit 203 that performs the first rearrangement process includes a compressed image data reading unit 203a, a restart marker detection unit 203b, a restart marker rewriting unit 203c, a selection unit 203d, and a selection unit. 203e and a compressed image data writing unit 203f.

  The compressed image data reading unit 203 a sequentially reads compressed image data in a compression unit for one block line from the JPEG core 202. The read compressed image data is input to the restart marker detection unit 203b. The restart marker detection unit 203b detects the restart marker in the compressed image data input from the compressed image data reading unit 203a and counts the number of restart markers.

The restart marker rewriting unit 203c rewrites the marker number of the restart marker detected by the restart marker detection unit 203b. The marker number at the time of rewriting is
Marker number = ((number of compression units in row direction × current compression unit position in row direction) + current compression unit position in column direction)% 8 (Formula 1)
It is determined based on the following formula. Here, the number of compression units in the row direction in (Equation 1) indicates the number of compression units arranged in the row direction. In the example of FIG. The current row direction compression unit position indicates the row direction position (that is, what column) of the compressed image data for each compression unit to be rewritten by the restart marker. For example, the position of the compressed image data obtained by compressing the compression unit 1-1 (hereinafter referred to as compressed image data 1-1. The same applies to other compressed image data) (row, column) = (0, 0), the position of the compressed image data 1-2 in the row direction is 1. Also, the row direction position of the compressed image data 2-1 is zero. Furthermore, the current column direction compression unit position indicates the column direction position (that is, what number line) of the compressed image data for each unit to be rewritten by the restart marker. The position in the column direction of the compressed image data 1-2 is 0. The position in the column direction of the compressed image data 2-1 is 1. In addition, “% 8” in (Expression 1) means that 8 modulo operations are performed.

  For example, when the number of compression units in the row direction is 10 and the restart marker to be rewritten is the compressed image data 1-2, the marker number of the restart marker after rewriting is marker number = ((10 × 1) +0. )% 8 = 2. That is, by performing the rewriting, the marker number of the compressed image data 1-2 in which the original marker number is FFD1 becomes FFD2.

  In the selection unit 203d, either the compressed image data read by the compressed image data reading unit 203a or the restart marker rewritten by the restart marker rewriting unit 203c is selectively output to the selection unit 203e. That is, the selection unit 203d outputs the compressed image data read by the compressed image data reading unit 203a until the restart marker is detected by the restart marker detection unit 203b, and the restart marker detection unit 203b restarts. When the marker is detected, the restart marker rewritten by the restart marker rewriting unit 203c is output. As a result, the restart marker whose marker number has been rewritten by the restart marker rewriting unit 203c is added to the end of the compressed image data for each compression unit.

  In the selection unit 203e, the compressed image data output from the selection unit 203d is selectively output to the code buffer BUF-a to the code buffer BUF-d. First, the selection unit 203e outputs compressed image data to the code buffer BUF-a until a restart marker is detected by the restart marker detection unit 203b. When the restart marker is detected by the restart marker detection unit 203b, the restart marker rewritten by the restart marker rewriting unit 203c is output to the code buffer BUF-a, and then the output destination of the compressed image data is switched. Is done. The compressed image data is output to the code buffer BUF-b until a restart marker is detected next time. When the restart marker is detected by the restart marker detection unit 203b, the restart marker rewritten by the restart marker rewriting unit 203c is output to the code buffer BUF-b, and then the output destination of the compressed image data is switched. Is done. The compressed image data is output to the code buffer BUF-c until a restart marker is detected next time. When the restart marker detecting unit 203b detects a restart marker, the restart marker rewritten by the restart marker rewriting unit 203c is output to the code buffer BUF-c, and then the output destination of the compressed image data is switched. Is done. The compressed image data is output to the code buffer BUF-d until a restart marker is detected next time.

  Similarly, the output destination is switched every time the restart marker is detected by the restart marker detector 203b. Thus, compressed image data for each compression unit compressed through the buffer BUF-A system is stored in the code buffer BUF-a, and compressed in the code buffer BUF-b through the buffer BUF-B system. Compressed image data for each compression unit is stored, compressed image data for each compression unit compressed through the buffer BUF-C system is stored in the code buffer BUF-c, and buffer BUF-d is stored in the code buffer BUF-d. Compressed image data for each compression unit compressed through the system D is stored. Here, the code buffer BUF-a to the code buffer BUF-d each have a capacity that is an integral multiple of the capacity that the compressed data writing unit 203f can write in one write operation. When the compressed image data is fully stored in the buffer, the compressed image data stored in the full code buffer is output to the compressed image data writing unit 203f.

  The compressed image data writing unit 203f performs a first rearrangement process of writing the compressed image data output from the code buffer BUF-a to the code buffer BUF-d into the first storage area 105a of the storage device 105 for one block line. Done. At this time, the compressed image data is sequentially written from the full code buffer regardless of the buffer number. First, the compressed image data writing unit 203f, at the start of the first rearrangement process, writes the write areas corresponding to the code buffers in the storage device 105 (write areas A-1 to D-1 shown in FIG. 8A). After that, the compressed image data stored in the full code buffer is sequentially written to the corresponding write area, and the next write area corresponding to the first full write area is currently secured. It is ensured that the last writing area is continuously provided. At this time, the value of the start address of the newly secured write area is written to the end of the corresponding previous write area.

  For example, the example of FIG. 8A shows an example in which the write area C-1 is fullest earliest. In this case, the next write area C-2 is secured after the write area D-1. At the same time, the value of the start address of the write area C-2 is written at the end of the write area C-1. Thereafter, similarly, the next write area corresponding to the full write area is secured, and the value of the start address of the newly secured write area is written at the end of the previous write area.

  When the first rearrangement process is completed, the compressed image data for one block line is written in the first storage area 105a of the storage device 105.

  Next, the second rearrangement unit 204 that performs the second rearrangement process includes a compressed image data reading unit 204a and a compressed image data writing unit 204b.

  The compressed image data reading unit 204 a reads the compressed image data written in the first storage area 105 a of the storage device 105. At this time, first, the compressed image data written in the writing area A-1 is read. Next, the compressed image data written in the writing area A-2 is read by referring to the value of the head address written in the end portion of the writing area A-1. Thereafter, similarly, the value of the head address written at the end of the writing area is referred to, and the compressed image data of the system of the code buffer BUF-a is sequentially read out. Next, the compressed image data written in the writing area B-1 is read out. After that, the compressed image data written in the writing area B-2 is read by referring to the value of the head address written in the end portion of the writing area B-1. Thereafter, similarly, the value of the head address written in the end portion of the writing area is referred to, and the compressed image data of the system of the code buffer BUF-b is sequentially read out.

  Thereafter, similarly, the compressed image data of the code buffer BUF-c and the compressed image data of the code buffer BUF-d are read out.

  In the compressed image data writing unit 204b, the compressed image data read by the compressed image data reading unit 204a is written in the second storage area 105b different from the first storage area 105a of the storage device 105 in the same order.

  By such processing, the compressed image data is written in the storage device 105 in an order corresponding to the arrangement order shown in FIG.

  In the example of FIG. 8A, the compressed image data that has been sorted by the first sorting unit 203 and the second sorting unit 204 are both stored in the storage device 105. 8B, the compressed image data that has been rearranged by the first rearrangement unit 203 is written in the storage device 105, and the second rearrangement unit 204 performs the rearrangement. You may make it write image data in the external medium 110 different from the memory | storage device 105. FIG. For example, a memory card or a hard disk is used as the external medium 110.

  Such a process is repeated until the restart marker for one block line is counted. When the restart marker for one block line is counted, reading by the compressed image data reading unit 203a is once finished.

  Thereafter, the processing moves to the next block line, and such processing is repeated until the end of all the block lines, that is, until the processing for one frame is completed.

  Next, the effect of 1st Embodiment is demonstrated with reference to Fig.9 (a) and FIG.9 (b). As described above, in the image processing unit A 107a and the image processing unit B 107b, the interpolation pixel data portion is deleted by image processing. In general image processing, when the image data for one block line is read out from the storage device 105, interpolation is performed as shown in FIGS. Pixel data is read so as to overlap. If such reading is performed, only the upper and lower ends of the image data stored in the storage device 105 are finally cut as interpolation pixel data. The interpolated pixel data deleted at this time does not depend on the width of the processed image data. Actually, the left end and the right end in the image data are also removed, but these influences are negligible.

  Here, if the number of input pixels in the column direction that can be input to the image processing unit 107 with a single input is small, as shown in FIG. 9A, 1 is input to the image processing unit to process image data for one frame. The number of times the image data for the block line is input also increases. In this case, the amount of interpolated pixel data that must be read redundantly increases, the total amount of data input to the image processing unit 107 increases, and image processing efficiency decreases.

  On the other hand, in the first embodiment, by allowing image data of a plurality of compression units to be input to the compression processing unit 108, one block line input to the image processing unit 107 as shown in FIG. 9B. The width in the column direction of the image data can be increased. As a result, it is possible to reduce the amount of interpolation pixel data that must be read redundantly, reduce the total amount of input data, and improve the efficiency of image processing.

  In addition, by inputting image data of a plurality of compression units to the compression processing unit 108, image data in the correct arrangement order cannot be restored at the time of decompression at the time of image reproduction. The first embodiment Then, by rearranging the compressed image data obtained for each compression unit during the compression process, the image data in the correct arrangement order is restored at the time of decompression.

[Second Embodiment]
FIG. 10 is a diagram illustrating a configuration of the first rearrangement unit and the second rearrangement unit in the image processing apparatus according to the second embodiment of the present invention. The second embodiment is an example in which storage areas corresponding to code buffer BUF-a to code buffer BUF-d are secured in advance by dividing the first storage area in accordance with the number of code buffers. is there. Here, FIG. 10A stores the first storage area 105c in which the compressed image data rearranged by the first rearrangement unit is stored and the compressed image data rearranged by the second rearrangement unit. It is a figure which shows the example which provided the 2nd storage area 105d to be provided in the same memory | storage device 105. FIG. On the other hand, in FIG. 10B, the storage device 105 is provided with a first storage area 105 c in which the compressed image data rearranged by the first rearrangement unit 203 is stored, and the second rearrangement unit 204 performs rearrangement. It is a figure which shows the example which provided the 2nd storage area 105d in which the compressed image data by which it was stored in the external medium 110. FIG. Other configurations are the same as those described in the first embodiment.

  As shown in FIGS. 10A and 10B, in the second embodiment, the write area A is stored in the storage area 105c in which the compressed image data rearranged by the first rearrangement unit 203 is stored. Four write areas of .about.write area D are secured in advance. In such a second embodiment, the compressed image data output from the code buffer BUF-a to code buffer BUF-d are sequentially stored in the corresponding write areas. That is, the compressed image data writing unit 203f sequentially writes the compressed image data input from the code buffer BUF-a from the top address of the write area A, and the compressed image data input from the code buffer BUF-b is written to the write area B. Are sequentially written from the head address of the writing area C, the compressed image data input from the code buffer BUF-c is sequentially written from the head address of the writing area C, and the compressed image data input from the code buffer BUF-d is Sequentially written from the address.

  The compressed image data reading unit 204a sequentially reads the compressed image data written in the first storage area 105c of the storage device 105 in the order of the writing area A, the writing area B, the writing area C, and the writing area D. The compressed image data writing unit 204b writes the compressed image data read by the compressed image data reading unit 204a to the second storage area 105d in the same order.

  When the first rearrangement process is completed, the compressed image data (1-1, 2-1) in the compression unit stored in the buffer BUF-A of the JPEG core I / F unit 201 and compressed in the writing area A is completed. ,..., H- (v-3)) are stored, and in the writing area B, compressed image data (1-2, 1-2) stored in the buffer BUF-B of the JPEG core I / F unit 201 and compressed. 2-2,..., H- (v-2)) are stored, and similarly, in the write area C and the write area D, compressed image data (in a compression unit stored in the buffer BUF-C and compressed) 1-3, 2-3,..., H- (v-1)), compressed image data (1-4, 2-4,..., Hv) stored in the buffer BUF-D and compressed. ) Are stored respectively.

  In the second embodiment as described above, it is not necessary to store the start address of the next area in the end portion when performing the first rearrangement process.

  Although the present invention has been described based on the above embodiments, the present invention is not limited to the above-described embodiments, and various modifications and applications are naturally possible within the scope of the gist of the present invention. For example, in each of the embodiments described above, the number of input pixels in the column direction of the image data input from the storage device 105 to the image processing unit 107 is described as four times that of the MCU, but this number may be changed. Further, the processing speed may be increased by a so-called butterfly configuration in which each of BUF-A to BUF-D is configured by two buffers, that is, an input side buffer and an output side buffer.

  Further, the above-described embodiments include various stages of the invention, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, even if some constituent requirements are deleted from all the constituent requirements shown in the embodiment, the problem described in the column of the problem to be solved by the invention can be solved, and the effect described in the column of the effect of the invention Can be extracted as an invention.

1 is a diagram illustrating a main configuration of an image processing apparatus according to a first embodiment of the present invention. It is the figure shown about the structure of the compression process part which is the principal part of the image processing apparatus which concerns on the 1st Embodiment of this invention. It is a figure for demonstrating a compression unit. It is the figure shown about the image data divided | segmented for every compression unit. It is a figure shown about the input order when the image data of a compression unit is input into an image process part. It is a figure which shows a mode when the image data of a compression unit are stored in buffer BUF-A-BUF-D. It is a figure shown about the input order when the image data for every compression unit is input into a JPEG core. It is a figure which shows the structure inside the 1st rearrangement part and 2nd rearrangement part in 1st Embodiment. It is a figure for demonstrating the effect of 1st Embodiment. It is a figure which shows the structure inside the 1st rearrangement part in 2nd Embodiment, and a 2nd rearrangement part.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 101 ... Imaging part (CCD), 102 ... Pre-processing part, 103, 106, 109 ... DMA, 104 ... Bus, 105 ... Storage device, 107 ... Image processing part, 108 ... Compression processing part, 110 ... External media, 201 ... JPEG core I / F unit, 202 ... JPEG core, 203 ... first rearrangement unit, 204 ... second rearrangement unit

Claims (8)

A plurality of serially connected image processing units for performing spatial image processing on the supplied image data;
The image data processed by the image processing unit is converted into image data of the block unit in block units such that the number of input pixels in the column direction is N (N ≧ 2) times the minimum unit of compression processing. A compression processing unit which sequentially inputs one block line arranged in length and obtains compressed image data in block units for one block line by sequentially compressing the input block unit image data;
A first rearrangement unit that sequentially reads out the compressed image data in block units for one block line obtained by the compression processing unit and performs a first rearrangement process;
A storage unit having a first storage area in which compressed image data in units of blocks for the one block line rearranged by the first rearrangement unit is sequentially written;
The compressed image data written in the first storage area of the storage unit is read out, subjected to a second rearrangement process, and the compressed image data in block units for one block line whose arrangement order has been changed is changed to the first A second rearrangement unit for sequentially writing to a second storage area different from the storage area;
An image processing apparatus comprising: The block unit image data includes N units of compression unit image data consisting of image data corresponding to the number of pixels that is M times the minimum unit of the compression process in the column direction and M units of the minimum unit of the compression process in the column direction. It is an array of pieces,
The compression processing unit includes a single or a plurality of buffers for storing the block-unit image data,
When obtaining the compressed image data in units of blocks for the one block line, the compression processing unit obtains image data composed of one pixel in the row direction and N pixels as the minimum unit of the compression processing in the column direction. Sequential storage in the buffer is repeated in the row direction, and N pieces of compression unit image data for each block unit are stored in the buffer, and N pieces of compression unit image data for each stored block unit are stored. Each time compression processing is performed sequentially from the image data of the first compression unit in the column direction, and the compression processing of the image data of one compression unit is completed, the image data of the compression unit is added to the compressed image data. The compressed marker data for each block unit is obtained by adding a restart marker indicating that the compression processing has been completed, and the compressed image data of the obtained compression unit for each block unit is obtained. The image processing apparatus according to claim 1, characterized in that sequentially outputs the one block line of the first sorting section the data. The first rearrangement unit includes K (1 ≦ 1) from the top of the compression unit image data of each block among the compressed image data for each compression unit of one block line sequentially output from the compression processing unit. K ≦ N) including N code buffers for inputting and storing only compressed image data for each compression unit corresponding to image data of the compression unit, and for each compression unit stored in the N code buffers The compressed image data is written in correspondence with each of a plurality of storage areas obtained by dividing the first storage area of the storage unit so as to correspond to the code buffer.
The second rearrangement unit changes the compressed image data for each compression unit whose arrangement order is changed by the first rearrangement process and written in the plurality of storage areas, for each compression unit corresponding to the same code buffer. The image processing apparatus according to claim 2, wherein only the compressed image data is sequentially read out and written in the second storage area in the same order as it is, so that the second rearrangement process is performed. The first storage area is composed of a single storage area, and the single storage area is divided into a plurality of storage areas corresponding to the N code buffers,
The first rearrangement unit sequentially stores compressed image data of compression units output from the N code buffers in the plurality of corresponding storage areas,
The second rearrangement unit sequentially reads out compressed image data of compression units stored in the plurality of storage areas corresponding to the same code buffer, and writes the compressed image data in the second storage area. The image processing apparatus according to claim 3. The first storage area is divided into N storage areas corresponding to the N code buffers on a one-to-one basis,
The first rearrangement unit sequentially stores compressed image data of a compression unit output from the N code buffers in a corresponding storage area among the N storage areas;
The second rearrangement unit sequentially reads out the compressed image data of the compression unit stored in each of the N storage areas for each compressed image data of the compression unit of the same storage area, and performs the second storage. The image processing apparatus according to claim 3, wherein the image processing apparatus writes in the area.   The first rearrangement unit rewrites the restart marker when performing the first rearrangement process on the compressed image data of the compression unit for one block line. Item 3. The image processing apparatus according to Item 2.   The image processing according to claim 1, wherein the second storage area is a storage area provided in the storage unit separately from the first storage area. apparatus.   The image processing apparatus according to claim 1, wherein the second storage area is a storage area in a storage device provided separately from the storage unit.

Images