JPH1127513A - Image-processing unit and image-processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a processing unit to cope with requests for high image- processing speed with simple management by merging converted image data obtained through image processing of a divided original image in the unit of a prescribed block and outputting the result as a single conversion image. SOLUTION: An original image read from an image storage section 11 in the case of compression is given to an image sorter 12, where the image is divided into, e.g. two, an image of a left-half is given to a codec A:13 and an image of a right half is given to codec B:14. The codecs A:13, B:14 respectively compresses the left half and the right half of the original image. In this case, the codecs A:13, B:14 respectively add an EOL code to each line and outputs the result. A code sorter discriminates the EOL to merge both codes and rearranges the data in the order of a 1st line code of the left image →2nd line code of the left image → 2nd line code of the right image ... and sends the result to a code storage section 16. In the case of expansion, the codec is set to the expansion operation so as to obtain a reverse data flow.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus and an image processing method, and more particularly to an image processing apparatus and an image processing method capable of processing image information at a high speed, such as compression and decompression.

[0002]

2. Description of the Related Art In recent years, in the field of image information processing, high-speed processing has been cited as an important function of a request for a compression / decompression processing apparatus that executes compression / decompression processing of image information.
For this reason, in the conventional compression / decompression processing device, in order to meet the demand exceeding the single processing performance of the compression / decompression unit (hereinafter, codec), a plurality of codecs are connected, the original image information is divided, and compression is performed by each codec. By performing the decompression processing, the speed was increased.

FIG. 2 shows an example of a compression / expansion apparatus using only one codec as a basic configuration of a conventional compression / expansion processing apparatus. That is, the compression / expansion apparatus includes an image storage unit (for example, a memory) 1 for storing original image information, a codec 2 for performing compression / expansion processing on the original image information from the image storage unit 1, and a compression / expansion using the codec 2. A code storage unit (for example, HDD) 3 for storing processed codes is connected, and transfer control is performed by a CPU or a DMA (not shown).

In such a compression / decompression device, the data flow during the compression operation is as follows: the image storage unit 1 → the codec (compression operation) 2 → the code storage unit 3 by using the codec 2 as the compression operation.

[0005] By this processing, the original image information on the image storage unit (for example, memory) 1 is compressed and stored in a code file on the code storage unit (for example, HDD) 3, or conversely, the code storage unit (for example, HDD) 3 The above code file can be expanded to be decoded original image information.

Here, the specific configuration of the codec 2 differs depending on the compression / decompression method. For example, the JBIG coding method (ISO / IEC11544, ITU-TT.8) is used.
In the case of the sequential operation 2), as shown in FIG. 4, the image I / F 31, the line memory 32, the typical prediction unit 33, the context unit 34, the encoding / decoding unit 35, and the code I / F 36 are realized. can do.

The details of this will be described later in the description of the present invention. The data flow at the time of the decompression operation is the image storage unit 1 ← the codec (decompression operation) 2 ← the code storage unit 3 as the data flow reverse to the data flow at the time of the compression operation by performing the codec 2 as the decompression operation. .

Assuming that the processing performance of the codec 2 is, for example, 10 M pixels / s, the configuration shown in FIG. Cannot be realized.

For this reason, conventionally, as shown in FIG. 3, the number of codecs has been increased and the divided image is processed by each codec. In FIG.
The configuration when two codecs are used is shown.

In the data flow during the compression operation, two flows flow in parallel as follows. (1) Image storage unit 1 → codec (compression operation) 2 → code storage unit 3 (2) Image storage unit 1 → codec (compression operation) 2 → code storage unit 3 As a result of this processing, as shown in FIG. The image is divided into two and converted into two code files corresponding to each part.

Here, an example of vertically dividing into two parts is shown. However, the speeding up of the compression / expansion processing by such a method has the following problems.

Since the number of transfer channels in the image storage unit 3 → codec 2 and codec 2 → code storage unit increases,
This means that the processing of the CPU or DMA for controlling this becomes complicated.

That is, in order to increase the speed of the conventional compression / expansion processing, since there are a plurality of code files, management becomes complicated, and it is necessary to synchronize when reading out a plurality of code files at the time of decompression. There was a disadvantage.

[0014]

As described above, in order to increase the speed of the conventional compression / decompression processing, since there are a plurality of code files, the management becomes complicated, and the processing of reading a plurality of code files at the time of decompression is performed. In this case, there is a disadvantage that synchronization is required.

Therefore, the present invention has been made in view of the above points, and is intended to improve the speed of image processing by improving the disadvantages of speeding up the conventional compression / expansion processing. It is an object of the present invention to provide an image processing apparatus and an image processing method which can manage the image easily.

[0016]

According to the present invention, in order to solve the above-mentioned problems, an image storage unit for storing an original image,
A plurality of image processing units for performing a predetermined conversion process on the original image from the image storage unit; an image sorter unit connecting the image storage unit and the plurality of image processing units; and the plurality of image processing units A conversion image storage unit for storing the conversion image data from the image processing unit; and a conversion image sorter unit for connecting the plurality of image processing units and the conversion image storage unit. The image sorter unit divides an image into a plurality of images and sends the divided images to the plurality of image processing units. The plurality of image processing units perform image processing on each image and output converted image data, and the converted image sorter An image processing apparatus is provided, wherein the unit merges the converted images from the plurality of image processing units in units of predetermined blocks and outputs the merged images to the converted image storage unit as one converted image.

According to the present invention, in order to solve the above problems, an image storage unit for storing an original image, and a plurality of image processing units for performing a predetermined conversion process on the original image from the image storage unit An image sorter unit that connects the image storage unit and the plurality of image processing units, a converted image storage unit that stores converted image data from the plurality of image processing units, the plurality of image processing units, A conversion image sorter unit for connecting to a conversion image storage unit, and the image sorter unit vertically reads the original image read out by raster scan from the image storage unit.
Divided and sent to the plurality of image processing units, the plurality of image processing units perform image processing on each image and output converted image data, and the converted image sorter unit outputs the converted image data from the plurality of image processing units. There is provided an image processing apparatus characterized in that a converted image is merged in units of predetermined blocks and output as one converted image to a converted image storage unit.

According to the present invention, in order to solve the above-mentioned problems, an image storage section for storing an original image or a decoded image, and a compression / expansion of the original image or the decoded image stored in the image storage section are provided. A plurality of codecs for performing predetermined encoding processing, an image sorter unit connecting the image storage unit and the plurality of codecs, a code storage unit storing codes from the plurality of codecs, and the plurality of codecs And a code sorter unit connecting the code storage unit and
During the compression operation, the image sorter unit divides the image read by the raster scan from the image storage unit into a plurality of parts and sends the plurality of codecs to the plurality of codecs, and the plurality of codecs indicates a code break for each block of a certain size. The code sorter unit merges the codes from the plurality of codecs on a block-by-block basis and outputs them as one code to the code storage unit. The code sorter unit divides the code from the storage unit into the block units and sequentially sends the divided code units to the plurality of codecs, the plurality of codecs decompresses the respective codes, and the image sorter unit decompresses the code by the plurality of codecs. Rearranged images in block units, and arranges them in the same order as the raster scan of the original image, and outputs them to the image storage unit. The image processing apparatus is provided, wherein the door.

According to the present invention, in order to solve the above-mentioned problems, an image storage unit for storing an original image or a decoded image, and a compression / expansion of the original image or the decoded image stored in the image storage unit are provided. A plurality of codecs for performing predetermined encoding processing, an image sorter unit connecting the image storage unit and the plurality of codecs, a code storage unit storing codes from the plurality of codecs, and the plurality of codecs And a code sorter unit connecting the code storage unit and
During the compression operation, the image sorter unit vertically divides an image read out by raster scanning from the image storage unit and sends it to the plurality of codecs, and the codec indicates a code break for each block of a certain size. Performing an encoding process by adding a code, the code sorter unit outputs the code from the plurality of codecs to the code storage unit as one code by merging the blocks in units, and at the time of decompression operation, The code from the code storage unit is divided by the code sorter unit in the block unit and sequentially sent to the plurality of codecs, the plurality of codecs decompresses the respective codes, and the image sorter unit is controlled by the plurality of codecs. The decompressed image is rearranged in units of the block, and output to the image storage unit in the same arrangement as the raster scan of the original image. The image processing apparatus is provided, characterized in that.

According to the present invention, in order to solve the above-mentioned problems, an image storage unit for storing an original image and a predetermined conversion process for enlargement / reduction are performed on the original image from the image storage unit. A plurality of image processing units; an image sorter unit that connects the image storage unit and the plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; An image processing unit and a conversion image sorter unit for connecting the conversion image storage unit, wherein the image sorter unit divides an original image read by raster scan from the image storage unit into a plurality of image processing units, and The plurality of image processing units perform image processing on each image and output converted image data, and the converted image sorter unit converts the converted images from the plurality of image processing units into predetermined blocks. As a unit The image processing apparatus is provided for and outputting the converted image storage unit merges as one converted image.

According to the present invention, in order to solve the above-mentioned problem, an image storage unit for storing an original image, and a predetermined conversion process for enlargement / reduction of the original image from the image storage unit are performed. A plurality of image processing units; an image sorter unit that connects the image storage unit and the plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; An image processing unit and a converted image storage unit for connecting the converted image storage unit, wherein the image sorter unit vertically divides the original image read by raster scan from the image storage unit into n images, Sent to a processing unit, the plurality of image processing units perform image processing on each image to output converted image data, and the converted image sorter unit converts the converted images from the plurality of image processing units into a predetermined block. In units of The image processing apparatus is provided for and outputting the converted image storage unit merges as one converted image.

According to the present invention, in order to solve the above-mentioned problem, an image storage unit for storing an original image and a predetermined conversion process for enlargement / reduction of the original image from the image storage unit are performed. A plurality of image processing units; an image sorter unit that connects the image storage unit and the plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; An image processing unit and a converted image sorter unit that connects the converted image storage unit, and divides the original image read by the raster scan from the image storage unit into a plurality of images that are overlapped by the image sorter unit. To the plurality of image processing units as a scaling processing unit, the plurality of image processing units as the scaling processing unit performs a scaling process on each image and outputs a converted image, The conversion image sorter unit merges the conversion images from the plurality of image processing units as the scaling processing unit while deleting the conversion images for the overlap, and outputs the merged images as one conversion image to the conversion image storage unit. An image processing apparatus is provided.

According to the present invention, in order to solve the above-mentioned problems, an image storage unit for storing an original image and a predetermined conversion process for enlargement / reduction are performed on the original image from the image storage unit. A plurality of image processing units; an image sorter unit that connects the image storage unit and the plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; An image processing unit and a conversion image sorter unit for connecting the conversion image storage unit, wherein the image sorter unit vertically divides the original image read out by raster scan from the image storage unit into n overlaps The number of pixels of the plurality of image processing units as the enlargement / reduction processing unit is divided into a plurality of overlapping images as the number of reference pixels of the conversion method of the plurality of image processing units as the enlargement / reduction processing unit. The plurality of image processing units as the scaling processing unit perform a scaling process on each image to output a converted image, and the converted image sorter unit includes the plurality of image processing units as the scaling processing unit. The conversion image from the image processing unit is merged while deleting the conversion image for the overlap by the number of overlap pixels × the conversion rate, and the merged image is output to the conversion image storage unit as one conversion image. An image processing device is provided.

According to the present invention, in order to solve the above-mentioned problems, an image storage unit for storing an original image and a predetermined conversion process for OCR processing are performed on the original image from the image storage unit. A plurality of image processing units; an image sorter unit that connects the image storage unit and the plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; An image processing unit and a converted image sorter unit for connecting the converted image storage unit, wherein the image sorter unit divides the original image read by the raster scan from the image storage unit into a plurality of pieces and performs OCR processing. Send to a plurality of image processing units, the plurality of image processing units perform image processing on each image and output converted image data, the converted image sorter unit, the converted image from the plurality of image processing units, Predetermined The image processing apparatus is provided for and outputting the converted image storage unit by merging the lock as a unit as one of the converted image.

According to the present invention, in order to solve the above-mentioned problems, an image storage unit for storing an original image and a predetermined conversion process for OCR processing are performed on the original image from the image storage unit. A plurality of image processing units; an image sorter unit that connects the image storage unit and the plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; An image processing unit and a converted image sorter unit for connecting the converted image storage unit, wherein the original image read by raster scanning from the image storage unit is subjected to OCR processing by the plurality of image processing units. The image is overlapped by the number of pixels of the maximum character size to be recognized and divided vertically by n and sent to the plurality of image processing units that perform the OCR process, and the plurality of image processing units perform image processing on each image. The converted image sorter unit outputs the converted images from the plurality of image processing units while deleting one of the characters double recognized at the boundary of the overlapped part as a conversion result sorter unit. And an image processing apparatus for merging a predetermined block as a unit and outputting the merged image as one converted image to a converted image storage unit.

According to the present invention, in order to solve the above-mentioned problems, a step of storing an original image in an image storage unit;
When a predetermined conversion process is performed on the original image by a plurality of image processing units, a step of connecting the image storage unit and the plurality of image processing units by an image sorter unit; Storing the converted image data in a converted image storage unit, and connecting the plurality of image processing units and the converted image storage unit by a converted image sorter unit, and reading the converted image data by raster scan from the image storage unit The image sorter unit divides the original image into a plurality of images and sends the divided images to the plurality of image processing units.The plurality of image processing units perform image processing on each of the images and output converted image data. The image sorter unit merges the converted images from the plurality of image processing units in units of a predetermined block and outputs the merged images as one converted image to the converted image storage unit. Image processing method is provided.

According to the present invention, in order to solve the above-mentioned problems, a step of storing an original image or a decoded image in an image storage unit, and a step of compressing the original image or the decoded image stored in the image storage unit. Performing a predetermined encoding process for decompression by a plurality of codecs; connecting the image storage unit and the plurality of codecs by an image sorter unit; and storing codes from the plurality of codecs in a code storage unit And a step of connecting the plurality of codecs and a code storage unit by a code sorter unit. During a compression operation, the image sorter unit divides an image read by raster scan from the image storage unit into a plurality of images. To the plurality of codecs, and the plurality of codecs indicate a code break for each block of a certain size. The code sorter unit adds the codes from the plurality of codecs and outputs them as one code to the code storage unit by merging the blocks in units of the blocks. The code sorter unit divides the code from the unit into blocks, sequentially sends the codec to the plurality of codecs, the plurality of codecs expands respective codes, and the image sorter unit is expanded by the plurality of codecs. The image processing method is characterized in that the images are rearranged in block units and output to the image storage unit in the same arrangement as the raster scan of the original image.

According to the present invention, in order to solve the above-mentioned problem, a step of storing an original image in an image storage unit;
Performing a predetermined conversion process for scaling up and down on the original image from the image storage unit by a plurality of image processing units; and connecting the image storage unit and the plurality of image processing units by an image sorter unit A step of storing converted image data from the plurality of image processing units in a converted image storage unit, and a step of connecting the plurality of image processing units and the converted image storage unit by a converted image sorter unit. Then, the image sorter unit divides the original image read by the raster scan from the image storage unit into a plurality of pieces and sends the divided images to the plurality of image processing units, and the plurality of image processing units perform image processing on each image. Output converted image data, and the converted image sorter unit merges the converted images from the plurality of image processing units in units of predetermined blocks as one converted image. Image processing method and outputting the conversion image storage unit is provided.

According to the present invention, in order to solve the above-mentioned problem, a step of storing an original image in an image storage unit;
Performing a predetermined conversion process for the OCR process on the original image from the image storage unit by a plurality of image processing units; and connecting the image storage unit and the plurality of image processing units by an image sorter unit A step of storing converted image data from the plurality of image processing units in a converted image storage unit, and a step of connecting the plurality of image processing units and the converted image storage unit by a converted image sorter unit. The image sorter unit divides the original image read by the raster scan from the image storage unit into a plurality of
The plurality of image processing units perform processing, the plurality of image processing units perform image processing on each image and output converted image data, and the converted image sorter unit outputs the converted image data from the plurality of image processing units. An image processing method is provided in which a converted image is merged in units of predetermined blocks and output as one converted image to a converted image storage unit.

[0030]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1A is a block diagram showing a configuration of a compression / decompression apparatus to which an image processing apparatus and an image processing method according to a first embodiment of the present invention are applied.

In FIG. 1A, reference numeral 11 denotes an image storage unit (for example, a memory) for storing original image information;
12, an image sorter for rearranging image data, 13,
Reference numeral 14 denotes codecs A and B for performing compression / decompression processing, reference numeral 15 denotes a code sorter for rearranging code data, and reference numeral 16 denotes a code storage unit (for example, HDD) for storing code data. Transfer control is being performed.

The image storage section 11 can be constituted by a memory, for example. During compression, CPU or D
Under the control of the MA, an original image is read by raster scanning.

At the time of decompression, a restored image is written by raster scanning. The image sorter 12 has an interface (I / F) with the image storage unit 11 and a codec A13.
In compression, the original image information is input by raster scan from the image storage unit, and the original image is divided into n parts, for example, vertically in the image sorter 12, and one of the pieces of image information is encoded by the codec A13. Output from the I / F with
The other image information is output from the I / F with the codec B14.

At the time of decompression, the reverse operation is performed. The inside of the image sorter 12 will be described later in detail. The codecs A13 and B14 are, for example, codecs for JBIG sequential operation, and include an image I / F and a code I / F.
During compression, an image is input from the image I / F, a compression operation is internally performed, and a code is output from the code I / F.

At the time of decompression, the reverse operation is performed. The code sorter 15 has an I / F with the codecs A13 and B14 and an I / F with the code storage unit 16. At the time of compression, the respective codes are input from the I / F with the codecs A13 and B14. The codes are merged internally and output from the I / F with the code storage unit 16.

At the time of decompression, the reverse operation is performed. The inside of the code sorter 15 will be described later in detail. The code storage unit 16 can be constituted by, for example, an HDD or a memory.

At the time of compression, a code file is formed under the control of the CPU or DMA. At the time of decompression, the code file is read and sent to the code sorter 15.

Next, as an example of the codecs A13 and B14, the case of the JBIG system is shown in FIG. 4, reference numeral 31 denotes an image I / F, 32 denotes a line memory,
33 is a typical prediction section, 34 is a context section, 35 is an encoding / decoding section, and 36 is a code I / F.

An image I / F 31 is an I / F of binary image data.
Is used to exchange image data with the image sorter 12. The line memory 32 is a memory for storing reference images required for templates of two or three lines.

In the typical prediction mode, the typical prediction unit 33 performs comparison between an encoding / decoding line and a line immediately above and generation of a pseudo pixel. The context section 34 generates a JBIG lowest-resolution standard context using a 2-pixel or 3-line 10-pixel template, and loads / stores the contents in a context table RAM.

The encoding / decoding section 35 performs arithmetic / decoding operations of arithmetic codes, and further has a probability estimation table ROM, a stuff byte addition / deletion function, and a marker code addition / detection function.

The code I / F 36 is an I / F for inputting / outputting code data, and is used for exchanging code data with the code sorter 15. (Operation During Image Sorter / Compression) Next, the configuration and operation inside the image sorter 12 will be described.

First, the state at the time of compression is shown in FIG. The image sorter 12 includes a line buffer 21a, an I / F 22a with the image storage unit 11, and an I / F with the codec A13.
F23a, I / F 24a with codec B14, input side (from I / F with image storage unit 11 to line buffer 21a
Sequencer P25a that controls the output (from the output of the line buffer 21a to the codec A13).
And I / F with B14)
26a.

The line buffer 21a has a size capable of storing image data for four lines, with one half of the main scanning of the original image as one line. Here, the four lines of the line buffer 21a are referred to as lines 1, 2, 3, and 4, respectively.

It is assumed that the lines 1, 2, 3, and 4 can perform input / output parallel operation by FIFO control. The data flow in this case is as follows.

The I / F 22a with the image storage unit 11 → the line buffer 21a → the I / F 23a with the codec A13 →
The I / F 24a to the codec B14 and the sequencer P25a have １ ／ of the original image set as a line length from a CPU or the like in advance, and perform the following control.

First, image data from the I / F 22a with the image storage unit 11 is input to the line 1 of the line buffer 21a. Thus, when the input image data length becomes the line length, the image data is newly input to line 2.

In this way, the input is performed in the order of line 1 → 2 → 3 → 4 → 1... Using the line length as a unit. When inputting to each line for the second time or later, it is to be performed after confirming that the input area is empty.

The sequencer Q26a has a codec A13
The following control is performed on the I / Fs 23a and 24a with the B14 and B14. That is, I / O with codec A13
Regarding F23a, first, reading is performed from line 1 of the line buffer 21a, and after reading the line length of line 1, next, line 2 is skipped to line 3
From the line length.

In this way, reading is performed by repeating the odd lines 1 → 3 → 1 → 3... In units of line length, and output to the codec A13. On the other hand, the I / F 24a with the codec B14 is similarly read out by repeating the even-numbered line 2 → 4 → 2 → 4.

With the above configuration and control, the image sorter 12 inputs the data from the image storage unit 11 in the order of 1 → 2 → 3 → 4 → 1... , Line 1 → 3 → 1 for codec A13
.. Are read out in the order of → 3..., And are read out and output to the codec B14 in the order of lines 2 → 4 → 2 → 4.

That is, the image is input from the image storage unit 11 by a normal raster scan, and the image is divided into n parts, for example, two vertically.
It is possible to realize a control in which the data is divided and output to the codecs A13 and B14.

(Image Sorter / Compression Operation) Subsequently, FIG. 5B shows the operation when the image sorter 12 is compressed. The image sorter 12 includes a line buffer 21b, an image storage unit 11,
22b with I / F23 and I / F23 with codec A13
b, I / F 24b with codec B, output side (from output of line buffer 21b to I / F 22 with image storage unit 11)
b), and the output side (I / Fs 23b and 24 with the codecs A13 and B14).
b to the input of the line buffer 21b).

The line buffer 21b, the I / F 22b with the image storage unit 11, and the I / F with the codecs A13 and B14
F23b and F23b are the same as in the compression except that the direction of the data flow is reversed.

The data flow in this case is as follows. The I / F 22b with the image storage unit 11 ← the line buffer 21b ← the I / F 23b with the codec A13 ← the I / F 24b with the codec B14 The sequencer S26b performs the following control on the input of the line buffer 21b.

That is, I / F2 with codec A13
3b, first, the image data is input to the line 1 of the line buffer 21b, and when the input image data length becomes the line length, the image data is skipped over the line 2 and newly added to the line 3.
To enter.

In this manner, the odd-numbered lines 1 → 3 → 1 → 3... Are input in the order of the line length. Regarding the I / F 24b with the codec B14, first, the signal is input to the line 2, and similarly, the even line 2 → 4 → 2 → 4.
・ Enter in order.

The sequencer R26b is connected to the line buffer 2
The following control is performed on the output of 1b and the I / F 22b with the image storage unit 11. The data is read from the line 1 of the line buffer 21b and output to the image storage unit 11. When the image data length becomes the main scanning length, the data is newly read from the line 2.

Hereinafter, lines 1 → 2 →
Are output to the image storage unit 11 in the order of 3 → 4 → 1. With the above configuration and control, the image sorter 11
I / Fs 23b and 24b with codecs A13 and B14
For example, the decoded images divided into n, for example, vertically divided into two, can be merged and rearranged in the order of the original images.

(Operation at the time of code sorter / compression) FIG. 6A shows the operation of the code sorter 15 at the time of compression. (Operation at the time of code sorter / expansion) FIG. 6B shows the operation of the code sorter 15 at the time of expansion.

The compression and decompression of the code sorter 15 shown in FIGS. 6A and 6B will be described with reference to FIGS. 5A and 5B described above. "When the preset line length is reached" is changed to "EO
(A) of FIG. 5 except that the LOL (End of Line) determination unit determines that EOL has occurred.
5 and FIG. 5 (b).

The EOL determination unit applied to the description of FIGS. 6A and 6B can be realized as shown in FIG. That is, the JBI shown in FIG.
In the case of the G codec, “FF” represents an escape sequence, and “02” following it represents “normal termination”.

When this "Nomal Terminate" is used as an EOL code, the EOL determination unit may compare the codes in 8-bit units and find out the parts "FF" and "02".

Such an EOL determination section can be easily realized by using a comparator. With the above-described configuration and control, the code sorter 15 merges code data divided vertically, for example, into two from the I / Fs 23b and 24b with the codecs A13 and B14 at the time of decompression, and the first line of the left image , The code on the first line of the right image, the code on the second line of the left image, the code on the second line of the right image, and so on.

Also, at the time of decompression, the code of the first line of the left image sent from the code storage unit 16 → the code of the first line of the right image →
The codes can be rearranged and output in the order of the code on the second line of the left image → the code on the second line of the right image.

Next, the overall operation will be described. At the time of compression, the image sorter 12 first stores the original image in the main scanning length of one.
/ 2, and codecs A13 and B13
A parameter required for the compression operation is set in 14 and a mode is set in which an EOL code is added to each line and compression is performed.

When the operation starts, the CPU or DM
By A, the original image in the image storage unit 11 is read out by raster scan and sent to the image sorter 12. Image sorter 1
2, the original image coming in by the raster scan is divided into n parts, for example, vertically, and the left half image is divided into the codec A13.
Then, the right half image is sent to the codec B14.

Codecs A13 and B14 compress the left and right halves of the original image, respectively. At this time, the codecs A13 and B14 add and code an EOL code for each line.

The codes of the left half and the right half are input to the code sorter 15 from the codecs A13 and B14B. The code sorter 15 determines the EOL, merges the two codes, and codes the first line of the left image → the first line of the right image → the second line of the left image → the second line of the right image. Are sent to the code storage unit 16 in the order of the codes.

At the time of decompression, first, 1/2 of the main scanning length of the original image is set in the image sorter 11, and parameters necessary for the decompression operation are set in the codecs A13 and B14.

When the operation starts, the CPU or DM
The code is sent from the code storage unit 16 to the code sorter 15 by A. At this time, the code in the first row of the left image → 1 in the right image
The code in the order of the code on the line → the code on the second line of the left image → the code on the second line of the right image... Is transmitted to the codec A13 by the code sorter 15 determining the EOL. Code of half image (code of first line of left image → 2 of left image)
The code of the right half image (the code of the first line of the right image → the code of the second line of the right image...) Is rearranged in the codec B14. Output.

The codecs A13 and B14 decompress and decompress the left half image and the right half image, respectively. In this way, the decompressed and restored left half image and right half image are
Each is sent to the image sorter 12.

In this image sorter 12, by counting the line lengths, the two images can be merged to obtain the same image as when the original image was raster-scanned.

That is, from the CPU or the DMA,
A codec having twice the performance can be configured while performing the same control as when compression and decompression are performed using one codec.

In the above embodiment, the case where the original image is vertically divided into two and the left half image is sent to codec A and the right half image is sent to codec B has been described. May be sent in a state where they are overlapped with each other.

In this case, the image source section may control the codecs A and B so that the overlapped images are deleted and merged and output as one image to the image storage section.

Further, in this case, the number of overlap pixels for vertically dividing the image into n is equal to or larger than the number of reference pixels of the codec A and B methods, and the image sorter unit calculates the overlap pixel number × conversion rate. It is only necessary to control to delete only.

According to the first embodiment as described above,
An image storage unit for storing an original image or a decoded image, a plurality of codecs for performing a predetermined encoding process for compression and decompression on the original image or the decoded image stored in the image storage unit, and the image storage unit An image sorter unit that connects the codec to a plurality of codecs; a code storage unit that stores codes from the plurality of codecs; and a code sorter unit that connects the plurality of codecs and the code storage unit. The image sorter unit sometimes divides the image read by the raster scan from the image storage unit into a plurality of pieces and sends the plurality of codecs to the plurality of codecs, and adds a code indicating a code break for each block of a certain size in the plurality of codecs. The code sorter unit merges the codes from the plurality of codecs into one code by using the block as a unit. Output to the code storage unit, and at the time of decompression operation, the code from the code storage unit is divided by the code sorter unit into blocks and sequentially sent to the plurality of codecs, and the plurality of codecs An image processing apparatus for decompressing, the image sorter unit rearranging the images decompressed by the plurality of codecs in units of the blocks, and outputting the same to a raster scan of the original image to the image storage unit. Can be provided.

This image processing apparatus has the above-described predetermined effects based on the following background. That is, when the processing performance is insufficient with one codec, conventionally, as shown in FIG. 3, the number of codecs is increased and the divided image is processed by each codec.

However, in order to increase the speed of the compression / expansion processing by such a method, the number of transfer channels increases.
There were drawbacks such as the PU or DMA processing becoming complicated and the management being complicated due to the multiple code files. In the present embodiment, an image sorter for rearranging image data and a code sorter for rearranging code data are arranged before and after the codec, and the following control is performed.

At the time of compression, first, の of the main scanning length of the original image is set in the image sorter, and codec A
A and B are set with parameters necessary for the compression operation, and a mode is set in which an EOL code is added to each line and compression is performed.

When the operation starts, the CPU or DM
By A, the original image in the image storage unit is read by raster scanning and sent to the image sorter. In the image sorter, the original image coming in by raster scanning is vertically divided into two parts, and the left half image is assigned to codec A and the right half image is assigned to codec B.
Send to

Codecs A and B compress the left and right halves of the original image, respectively. At this time, the codecs A and B add and code an EOL code for each line.

The codes of the left half A and the right half are input from the codecs A and B to the code sorter. In the code sorter, the EOL is determined, the codes of both are merged, and the code of the first line of the left image → the code of the first line of the right image → the code of the second line of the left image → the second line of the right image The codes are rearranged in the order of... And sent to the code storage unit.

At the time of decompression, first, 1/2 of the main scanning length of the original image is set in the image sorter, and parameters necessary for the decompression operation are set in codecs A and B.

When the operation starts, the CPU or DM
A sends the code from the code storage to the code sorter. The code in the order of the code on the first line of the left image → the code on the first line of the right image → the code on the second line of the left image → the code on the second line of the right image By determining the EOL, the code of the left half image (code of the first line of the left image → code of the second line of the left image...) Is provided to the codec A.
And the code of the right half image (1 of the right image)
(The order of the code in the line → the code in the second line of the right image).

In these codecs A and B, the left half image and the right half image are decompressed and restored, respectively. The restored left half image and right half image are each sent to the image sorter.

In the image sorter, by counting the line length, two images can be merged to obtain an image in the same arrangement as when the original image was raster-scanned.
That is, with such a configuration and control, a codec having twice the performance can be configured from the CPU or the DMA while performing the same control as when compression and decompression is performed using one codec.

The present invention is not limited to the above-described first embodiment, but includes the following embodiments and the configuration and operation of the above-described first embodiment. It can be applied accordingly.

FIG. 8A shows, as a second embodiment, an image processing apparatus applied to general image processing. That is, the image processing apparatus according to the second embodiment includes an image storage unit 41 that stores an original image, and a plurality of image processing units 43 that perform a predetermined conversion process on the original image from the image storage unit 41. , 44, an image sorter unit 42 connecting the image storage unit 41 and the plurality of image processing units 43, 44, and a converted image storage unit storing converted image data from the plurality of image processing units 43, 44. 46, and a converted image sorter unit 45 for connecting the plurality of image processing units 43 and 44 and the converted image storage unit 46. The original image read by the raster scan from the image storage unit 41 is stored in the image sorter A plurality of image processing units 43 and 44, which perform image processing on each image and output converted image data; Image sorter unit 45 and outputs the converted image from the plurality of image processing units 43 and 44, the converted image storage unit 46 as one converted image by merging predetermined block units.

In this case, the image sorter unit 42
(B), the original image is vertically divided into n
The image may be divided into a plurality of overlapping images.

FIG. 9A shows, as a third embodiment, an image processing apparatus applied to image enlargement / reduction processing. That is, the image processing apparatus according to the third embodiment includes an image storage unit 51 that stores an original image, and a plurality of image processing units that perform a predetermined conversion process for scaling the original image from the image storage unit 51. Image conversion units 53 and 54, an image sorter unit 52 connecting the image storage unit 51 and the plurality of image processing units 53 and 54, and the converted image data from the plurality of image processing units 53 and 54. And a converted image sorter unit 55 connecting the plurality of image processing units 53 and 54 and the converted image storage unit 56. The original read out by raster scanning from the image storage unit 51 is provided. The image sorter unit 52 divides the image into a plurality of parts and sends the divided images to the plurality of image processing units 53 and 54. The plurality of image processing units 53 and 54 perform image processing on each image and convert the converted image data. The converted image sorter unit 55 merges the converted images from the plurality of image processing units 53 and 54 in units of predetermined blocks and outputs the merged images to the converted image storage unit 54 as one converted image. I do.

In this case, the image sorter unit 52
(B), the original image may be vertically divided into n parts. In this case, the image read by the raster scan from the image storage unit 51 is stored in the image sorter unit 52.
However, as shown in FIG. 9B, the image is divided into a plurality of images overlapping each other and sent to the plurality of image processing units 53 and 54 serving as enlargement / reduction processing units. Image processing units 53 and 54 perform a scaling process on each image to output a converted image,
The converted image sorter unit 55 merges the converted images from the plurality of image processing units 53 and 54 as the enlargement / reduction processing unit while deleting the overlapped converted images to form the converted image as one converted image. It may be characterized in that it is output to the storage unit 56.

Further, in this case, the number of overlapping pixels for vertically dividing the original image into n is equal to or larger than the number of reference pixels of the conversion method of the plurality of image processing units 53 and 54 as the scaling unit. The conversion image sorter 55 may control to delete by the number of overlap pixels × the conversion rate.

FIG. 10A shows, as a fourth embodiment, an image processing apparatus applied to OCR processing of an image. That is, the image processing apparatus according to the fourth embodiment includes an image storage unit 61 that stores an original image and a plurality of units that perform a predetermined conversion process for the OCR process on the original image from the image storage unit 61. Image conversion units 63 and 64, an image sorter unit 62 for connecting the image storage unit 61 and the plurality of image processing units 63 and 64, and conversion image data from the plurality of image processing units 63 and 64. And a converted image sorter unit 65 connecting the plurality of image processing units 63 and 64 to the converted image storage unit 66. The original image read out by raster scanning from the image storage unit 61 is provided. The image sorter unit 62 divides the image into a plurality of pieces and sends them to the plurality of image processing units 63 and 64 that perform OCR processing, and the plurality of image processing units 63 and 64 perform image processing on each image. And it outputs the converted image data, the converted image sorter unit 65 the plurality of image processing unit 63,
It is characterized in that the converted image from the H.64 is merged in units of predetermined blocks and output to the converted image storage unit 66 as one converted image.

In this case, the image sorter unit 62
The original image may be vertically divided into n parts as shown in FIG. In this case, as shown in FIG. 10 (b), the plurality of image processing units 63 and 64
The image is overlapped and divided by the number of pixels (or more) of the maximum character size recognized in the CR process, and one of the characters double-recognized by the converted image sorter 65 as a conversion result sorter at the boundary of the overlapped portion is deleted. Control may be performed.

Further, in this case, by providing a division point determination unit in advance, the division point determination unit finds a division point (such as an all-white vertical line).
It may be characterized in that control is performed so as to be divided according to the information.

[0098]

Therefore, as described in detail above, according to the present invention, by improving the disadvantages of the conventional high-speed processing such as the compression / decompression processing, it is possible to manage the image processing when the high-speed image processing is required. It is possible to provide an image processing apparatus and an image processing method which can be easily dealt with.

[Brief description of the drawings]

FIG. 1 is a block diagram (a) showing a configuration of an image compression / decompression device applied as a first embodiment of an image processing device according to the present invention, and an original image is vertically divided into n parts. FIG. 7B is a diagram illustrating an example of the operation.

FIG. 2 is a block diagram showing a configuration of a conventional device using a single codec.

FIG. 3 is a block diagram showing a configuration of a conventional device using a plurality of codecs.

FIG. 4 is a block diagram showing a configuration when the codec is configured by a JBIG sequential operation.

FIG. 5 is a diagram showing the configuration and operation of the image sorter unit according to the present invention separately for compression (a) and expansion (b).

FIG. 6 is a diagram showing the configuration and operation of the code sorter unit according to the present invention separately for compression (a) and expansion (b).

FIG. 7 is a diagram exemplifying an EOL judging unit in a sequencer of a code sorter unit according to the present invention.

FIG. 8 is a block diagram (a) showing a configuration of a general image processing apparatus applied as a second embodiment of the image processing apparatus according to the present invention, and the original image is vertically overlapped. FIG. 7B is a diagram illustrating an example of n division.

FIG. 9 is a block diagram (a) showing a configuration of an image enlargement / reduction processing apparatus applied as a third embodiment of the image processing apparatus according to the present invention, and the original image is vertically overlapped with n. It is a figure (b) which shows the example of dividing.

FIG. 10 is a block diagram showing the configuration of an image processing apparatus having an OCR processing function applied as a fourth embodiment of the image processing apparatus according to the present invention, and the original image is vertically overlapped. FIG. 7B is a diagram illustrating an example of dividing the data into n.

[Explanation of symbols]

11 image storage unit, 12 image sorter, 13, 14 codecs A and B, 15 code sorter, 16 code storage unit, 31 image I / F, 32 line memory, 33 typical prediction unit 34: context part, 35: encoding / decoding part, 36: code I / F, 22a, 22b, 23a, 23b, 24a, 24b, 5
2a, 52b, 53a, 53b, 54a, 54b ... I /
F, 25a, 55a ... sequencer P, 25b, 55b ... sequencer R, 26a, 56a ... sequencer Q, 26b, 56b ... sequencer S, 41, 51, 61 ... image storage unit, 42, 52, 62 ... image sorter unit, 43, 44, 53, 54, 63, 64: a plurality of image processing units; 45, 55, 65: a converted image sorter unit; 46, 56, 66: a converted image storage unit.

Claims (14)

[Claims] An image storage unit for storing an original image; a plurality of image processing units for performing a predetermined conversion process on the original image from the image storage unit; an image storage unit and the plurality of image processing units An image sorter unit that connects the plurality of image processing units, a converted image storage unit that stores converted image data from the plurality of image processing units, and a converted image sorter unit that connects the plurality of image processing units and the converted image storage unit. The image sorter unit divides the original image read by the raster scan from the image storage unit into a plurality of parts and sends the plurality of images to the plurality of image processing units, and the plurality of image processing units perform image processing for each image. The converted image data is output by performing processing, and the converted image sorter unit merges the converted images from the plurality of image processing units in units of predetermined blocks and outputs the merged images as one converted image to the converted image storage unit. The image processing apparatus according to claim. 2. An image storage unit for storing an original image, a plurality of image processing units for performing a predetermined conversion process on the original image from the image storage unit, the image storage unit and the plurality of image processing units An image sorter unit that connects the plurality of image processing units, a converted image storage unit that stores converted image data from the plurality of image processing units, and a converted image sorter unit that connects the plurality of image processing units and the converted image storage unit. The image sorter unit vertically divides the original image read by raster scan from the image storage unit into n and sends the divided image to the plurality of image processing units, and the plurality of image processing units The converted image data is output by performing image processing, and the converted image sorter unit merges the converted images from the plurality of image processing units in units of predetermined blocks and outputs the merged image as one converted image to the converted image storage unit. This The image processing apparatus according to claim. 3. An image storage unit for storing an original image or a decoded image, and a plurality of codecs for performing a predetermined encoding process for compression and decompression on the original image or the decoded image stored in the image storage unit. An image sorter unit that connects the image storage unit and a plurality of codecs, a code storage unit that stores codes from the plurality of codecs, and a code sorter unit that connects the plurality of codecs and the code storage unit. In the compression operation, the image sorter unit divides the image read by the raster scan from the image storage unit into a plurality of pieces and sends the plurality of codecs to the plurality of codecs. The code sorter unit performs a coding process by adding a code indicating The code is output as one code to the code storage unit.At the time of decompression operation, the code from the code storage unit is divided by the code sorter unit into block units and sequentially transmitted to the plurality of codecs. Decompressing each code, the image sorter unit rearranges the images decompressed by the plurality of codecs into the above-described block units, and outputs the same to the raster scan of the original image to the image storage unit. Image processing device. 4. An image storage unit for storing an original image or a decoded image, and a plurality of codecs for performing a predetermined encoding process for compression and decompression on the original image or the decoded image stored in the image storage unit. An image sorter unit that connects the image storage unit and a plurality of codecs, a code storage unit that stores codes from the plurality of codecs, and a code sorter unit that connects the plurality of codecs and the code storage unit. In the compression operation, the image sorter unit vertically divides an image read by raster scanning from the image storage unit into n segments and sends the divided image to the plurality of codecs. The coding process is performed by adding a code indicating a break, and the code sorter unit merges codes from the plurality of codecs in units of the block. The code is output to the code storage unit as one code, and during the decompression operation, the code from the code storage unit is divided by the code sorter unit into blocks and sequentially transmitted to the plurality of codecs. Decompressing each code, the image sorter unit rearranges the images decompressed by the plurality of codecs into the above-described block units, and outputs the same to the raster scan of the original image to the image storage unit. Image processing device. 5. An image storage unit for storing an original image, a plurality of image processing units for performing a predetermined conversion process for scaling the original image from the image storage unit, An image sorter unit that connects a plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; and a connection between the plurality of image processing units and the converted image storage unit A converted image sorter unit, the image sorter unit divides the original image read by the raster scan from the image storage unit into a plurality of parts, and sends the divided images to the plurality of image processing units. The converted image data is output by performing image processing on the image, and the converted image sorter unit merges the converted images from the plurality of image processing units in units of predetermined blocks and converts the converted image into one converted image. Image processing apparatus and outputs the section. 6. An image storage unit for storing an original image, a plurality of image processing units for performing a predetermined conversion process for scaling the original image from the image storage unit, An image sorter unit that connects a plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; and a connection between the plurality of image processing units and the converted image storage unit A converted image sorter unit, wherein the image sorter unit vertically divides the original image read by raster scanning from the image storage unit into n and sends the original image to the plurality of image processing units, and each of the plurality of image processing units The converted image data is output by performing image processing on the image of the converted image, and the converted image sorter unit merges the converted images from the plurality of image processing units in units of predetermined blocks and converts the converted image into one converted image. Case Image processing apparatus and outputs the section. 7. An image storage unit for storing an original image, a plurality of image processing units for performing a predetermined conversion process for enlargement / reduction on the original image from the image storage unit, An image sorter unit that connects a plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; and a connection between the plurality of image processing units and the converted image storage unit A converted image sorter unit, wherein the original image read by raster scan from the image storage unit is divided into a plurality of images in which the image sorter unit overlaps each other, and the plurality of image processing units as a scaling unit The plurality of image processing units as the enlargement / reduction processing unit perform enlargement / reduction processing on each image to output a converted image, and the converted image sorter unit serves as the enlargement / reduction processing unit. Image processing apparatus and outputs the converted image from the plurality of image processing units, the converted image storage unit as the merge to one transformation image while deleting the converted image of the overlap fraction. 8. An image storage unit for storing an original image, a plurality of image processing units for performing a predetermined conversion process for enlargement / reduction on the original image from the image storage unit, An image sorter unit that connects a plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; and a connection between the plurality of image processing units and the converted image storage unit A converted image sorter unit, wherein the image sorter unit vertically divides the original image read out by raster scanning from the image storage unit into n vertically. The image processing unit is divided into a plurality of overlapping images as the number of reference pixels or more in the conversion method of the image processing unit and sent to the plurality of image processing units as a scaling unit, and the plurality of image processing units as the scaling unit. The image processing unit performs a scaling process on each image and outputs a converted image. The converted image sorter unit converts the converted images from the plurality of image processing units as the scaling process unit into the overlapped image. An image processing apparatus, wherein the converted images are merged while being deleted by the number of overlap pixels × the conversion rate, and output as one converted image to the converted image storage unit. 9. An image storage unit for storing an original image, a plurality of image processing units for performing a predetermined conversion process for an OCR process on the original image from the image storage unit, An image sorter unit that connects a plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; and a connection between the plurality of image processing units and the converted image storage unit A conversion image sorter unit, wherein the image sorter unit divides the original image read by the raster scan from the image storage unit into a plurality of pieces and sends the original image to the plurality of image processing units that perform OCR processing, and the plurality of image processing units The unit performs image processing on each image and outputs converted image data. The converted image sorter unit merges the converted images from the plurality of image processing units in units of predetermined blocks to form one converted image. Image processing apparatus and outputting the converted image storage unit as. 10. An image storage unit for storing an original image, a plurality of image processing units for performing a predetermined conversion process for an OCR process on the original image from the image storage unit, An image sorter unit that connects a plurality of image processing units; a converted image storage unit that stores converted image data from the plurality of image processing units; and a connection between the plurality of image processing units and the converted image storage unit A converted image sorter unit, wherein the original image read out by raster scan from the image storage unit is used by the image sorter unit to perform OC by the plurality of image processing units.
The plurality of image processing units perform the OCR process by dividing into n vertically by overlapping the number of pixels of the maximum character size recognized by the R process, and the plurality of image processing units perform image processing on each image. And outputs converted image data.The converted image sorter unit converts the converted images from the plurality of image processing units while deleting one of the characters double recognized at the boundary of the overlapped portion as a conversion result sorter unit. To
An image processing apparatus wherein a predetermined block is merged and output as one converted image to a converted image storage unit. 11. A step of storing an original image in an image storage unit, and performing a predetermined conversion process on the original image by a plurality of image processing units, wherein the image storage unit and the plurality of image processing units are Connecting by an image sorter unit; storing converted image data from the plurality of image processing units in a converted image storage unit; and connecting the plurality of image processing units and the converted image storage unit by a converted image sorter unit. Connecting the original image read by the raster scan from the image storage unit to the plurality of image processing units by dividing the original image into a plurality of image processing units, wherein the plurality of image processing units The converted image data is output by performing image processing on the converted image, and the converted image sorter unit merges the converted images from the plurality of image processing units in units of predetermined blocks. Image processing method and outputting the converted image storage unit as One of the converted image. 12. A step of storing an original image or a decoded image in an image storage unit, and performing a predetermined encoding process for compressing and expanding the original image or the decoded image stored in the image storage unit by a plurality of codecs. Performing the steps of: connecting the image storage unit and a plurality of codecs by an image sorter unit; storing codes from the plurality of codecs in a code storage unit; and the plurality of codecs and a code storage unit. By a code sorter unit, during a compression operation, the image sorter unit divides an image read by raster scanning from the image storage unit into a plurality of images and sends the plurality of images to the plurality of codecs, Then, a code indicating a code break is added to each block of a certain size to perform an encoding process. The codes from the plurality of codecs are merged in units of the block and output as one code to the code storage unit. During the decompression operation, the code from the code storage unit is used by the code sorter unit in the block unit. And sequentially sends the codecs to the plurality of codecs, the plurality of codecs expands respective codes, and the image sorter rearranges the images expanded by the plurality of codecs in the block units, An image processing method, wherein the image is output to the image storage unit in the same order as the raster scan. 13. A step of storing an original image in an image storage unit, a step of performing a predetermined conversion process for scaling the original image from the image storage unit by a plurality of image processing units, Connecting a storage unit and the plurality of image processing units by an image sorter unit; storing converted image data from the plurality of image processing units in a converted image storage unit; Connecting the converted image storage unit with a converted image sorter unit, and the image sorter unit divides the original image read by the raster scan from the image storage unit into a plurality of parts and sends the original image to the plurality of image processing units. The plurality of image processing units perform image processing on each image and output converted image data, and the converted image sorter unit converts images from the plurality of image processing units. An image processing method and outputting the converted image storing unit as one of the converted image by merging predetermined block units. 14. A step of storing an original image in an image storage unit, performing a predetermined conversion process for OCR processing on the original image from the image storage unit by a plurality of image processing units, Connecting a storage unit and the plurality of image processing units by an image sorter unit; storing converted image data from the plurality of image processing units in a converted image storage unit; Connecting the converted image storage unit with a converted image sorter unit, wherein the image sorter unit divides the original image read by the raster scan from the image storage unit into a plurality of images and performs an OCR process. The plurality of image processing units perform image processing on each image and output converted image data, and the converted image sorter unit outputs the converted image data to the plurality of image processing units. Image processing method and outputting the converted image from the parts, the converted image storing unit as one of the converted image by merging predetermined block units.