JP4031442B2 - Image processing apparatus and image forming apparatus having the same - Google Patents

Description

  The present invention relates to an image processing apparatus that performs compression processing and decompression processing of image data, and an image forming apparatus including the image processing apparatus.

  As is well known, in digital image forming apparatuses such as copiers, printers, and facsimile machines, when image data is input, the image data is subjected to various image processing and then the image data is electrophotographic or inkjet. The image indicated by the image data is recorded on the output engine.

  As an image processing unit that performs various types of image processing, an ASIC (Application Specific Integrated Circuit) or a DSP (Digital Signal Processor) is often used for high-speed processing. Depending on the contents of the image processing, a large capacity memory is required, and the image data is once written in the memory in the ASIC or DSP, or in a larger external memory, etc., and the image data is stored in the memory when necessary. Read from.

  When writing to and reading from such a memory, compression processing is usually not performed. This means that each time image data is input / output to / from the memory at each step in image processing, the image data is compressed and expanded, and an average high compression ratio is achieved for all of the various image data. This is because the compression / decompression algorithm becomes complicated, and it is necessary to provide a compression / decompression processing circuit having a throughput corresponding to the memory bandwidth, and the circuit scale becomes enormous.

On the other hand, in Patent Document 1, a histogram relating to the density of each pixel of an image obtained by pre-scanning is formed, and one of a plurality of predetermined compression processing methods is selected based on the histogram, and this selection is performed. A technique of compressing an image obtained by the main scan using the above-described compression processing method is disclosed.
JP 11-41472 A

  The image processing speed is closely related to a memory that temporarily stores image data. In order to achieve a desired image processing speed, a certain memory bandwidth required for inputting / outputting image data to / from the memory during image processing is required. If the required memory bandwidth is not within the memory bandwidth of the actual memory, the image processing speed cannot be guaranteed.

  If the image data is not compressed and expanded when writing to and reading from the memory, the image data is large, the memory bandwidth must be increased, and the memory capacity must be large.

  On the other hand, if the image data is compressed / decompressed, the image data to be written and read becomes smaller than the original image data, and the memory bandwidth can be reduced (the apparent memory bandwidth is larger than the actual memory bandwidth). Memory capacity can be reduced. However, if a compression / decompression circuit having a throughput corresponding to the memory bandwidth is provided as described above, the circuit scale increases. For this reason, a simple compression / decompression algorithm may be employed. In this case, however, a high compression rate cannot be obtained on average for all of various image data.

  Furthermore, in Patent Document 1, although any one of a plurality of compression processing methods is selected based on a histogram relating to the density of each pixel of an image obtained by pre-scanning, all of various image data are averaged. In order to obtain a high compression rate, it is necessary to prepare many compression processing methods, and the circuit scale also increases.

  Therefore, the present invention has been made in view of the above problems, and can achieve a high compression ratio on average for various image data while reducing the circuit scale, and reducing the required memory bandwidth and the required memory capacity. An object of the present invention is to provide an image processing apparatus capable of realizing the above and an image forming apparatus including the same.

In order to solve the above-described problem, the present invention provides an image processing apparatus that acquires first and second image data indicating the same image and compresses and expands the second image data using the first image data. Histogram creation means for creating histogram data of one image data, histogram correction means for creating correction histogram data by performing correction processing on the histogram data created by the histogram creation means, and histogram correction means A compression / expansion unit that performs reversible compression / expansion of the second image data at a compression / expansion rate according to the correction histogram data , and a storage unit that writes and reads the compressed second image data are provided.

  In the present invention, correction processing is performed on the histogram data to generate correction histogram data, and the second image data is compressed and expanded based on the correction histogram data.

  Furthermore, in the present invention, the image is divided into a plurality of blocks, histogram data is created for each block, and the second image data is compressed and expanded.

  In the present invention, when the data size of the compressed second image data is equal to or larger than a certain size, the storage unit performs writing / reading of the uncompressed second image data.

  On the other hand, an image forming apparatus of the present invention includes the image processing apparatus of the present invention.

  According to the present invention, first and second image data representing the same image is acquired, histogram data of the first image data is created, and the second image data is compressed and decompressed at a compression / decompression ratio according to the histogram data. It is carried out. Therefore, even for various images, the histogram data (image density histogram) can be grasped, and the second image data can be efficiently compressed based on the histogram data, and the storage capacity of the storage means can be reduced. An apparent increase in memory bandwidth is possible.

  In addition, since the correction processing is performed on the histogram data and the second image data is compressed and expanded based on the correction histogram data, a change in the density histogram occurs between the first and second image data due to noise or the like. However, it is possible to suppress a decrease in the data compression rate.

  Furthermore, since the image is divided into a plurality of blocks, histogram data is created for each block, and the second image data is compressed and expanded, the local density change in the image is reflected in the histogram data, and the second Image data can be compressed and expanded, and a reduction in data compression rate can be suppressed.

  In addition, when the data size of the compressed second image data is equal to or larger than a certain size, the storage means is used to write and read the uncompressed second image data, so the storage capacity of the storage means is wasted. Not used.

  The image forming apparatus of the present invention can achieve the same operations and effects as the image processing apparatus of the present invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 is a block diagram showing Embodiment 1 of the image forming apparatus of the present invention. The image forming apparatus 10 of this embodiment is a digital copying machine, and includes a scanner unit (image input device) 11, an image processing unit (image processing device) 12, an engine unit (image output device) 13, a console 14, a controller 15, And a memory 16.

  The scanner unit 11 includes a CCD (Charge Coupled Device) line image sensor unit (not shown) and a sub-scanning direction drive system that reciprocates the CCD line image sensor unit in the sub-scanning direction. The CCD line image sensor unit repeatedly scans the document image in the main scanning direction while scanning the document image in the sub-scanning direction, and performs RGB (R: red, G: green, B: blue) for each main scanning line. Generate and output a color signal. Only the color signal corresponding to the effective range (document reading range) of the output of the CCD line image sensor unit is selected, and this color signal is A / D converted to generate RGB image data to the image processing unit 12. Is output.

  The image processing unit 12 is configured by an ASIC, receives the RGB image data from the scanner unit 11, performs various image processing on the RGB image data, and then converts the RGB image data into CMYK (C: blue-green, M: crimson, Y: yellow, K: black) image data, and the CMYK image data is output to the engine unit 13.

  The controller 15 is composed of a general-purpose CPU (Central Processing Unit), controls the scanner unit 11, the image processing unit 12, and the engine unit 13, and controls a series of copying operations. Although not shown, the console 14 includes a liquid crystal display device for performing an interface with a user, buttons for operation instructions, and the like.

  The engine unit 13 is a tandem type electrophotographic color output engine using CMYK four-color toner (not shown). The engine unit 13 receives CMYK image data for one page from the image processing unit 12 and displays the color indicated by the CMYK image data. The original image is recorded on the recording paper.

  The memory 16 is a storage device for storing a large amount of data composed of, for example, a DRAM (Dynamic Random Access Memory) or an SRAM (Static Random Access Memory), and temporarily stores the image data from the image processing unit 12. Remember.

  Next, the image processing unit 12 in the image forming apparatus 10 of the present embodiment will be described in more detail.

  FIG. 2 is a block diagram illustrating a configuration of the image processing unit 12. The image processing unit 12 includes a memory interface 20, a shading correction processing unit 21, an input gamma processing unit 22, a color image determination unit 23, a resolution conversion processing unit 24, a region separation processing unit 25, a filter processing unit 26, and a color correction processing unit. 27, a halftone processing unit 28, a histogram calculation unit A29, a histogram correction unit A30, a compression processing unit A31, an expansion processing unit A32, a histogram calculation unit B33, a histogram correction unit B34, a compression processing unit B35, and an expansion processing unit B36. Yes.

  The memory interface 20 is an interface that performs input / output control of image data between the compression processing unit A31, the expansion processing unit A32, the compression processing unit B35, and the expansion processing unit B36 and the memory 16.

  The shading correction processing unit 21 corrects the RGB image data in a state where the received light amount distribution of the CCD line image sensor unit (main scanning direction) is not uniform, and the RGB image data is converted into RGB in a state where the received light amount distribution is uniform. Convert to image data. That is, the RGB image data from the scanner unit 11 is subjected to processing for removing various distortions generated in the illumination system, imaging system, and imaging system of the read document.

  The input gamma processing unit 22 corrects the input characteristics of the scanner unit 11 so that the RGB image data from the scanner unit 11 is linear with the sensitivity of the CCD line image sensor unit, and the RGB image data is subjected to subsequent image processing. Make it easy to handle.

  The color image determination unit 23 determines whether the document image indicated by the RGB image data from the input gamma processing unit 22 is color or monochrome.

  The histogram calculation unit A29 refers to the color and gradation of each pixel constituting the image indicated by the RGB image data from the input gamma processing unit 22, and creates frequency data (histogram data) for 256 gradations of each color. . For this reason, the histogram calculation unit A29 incorporates 256 frequency counters for counting the appearance frequency of 256 gradations for each color of RGB, and a total of 768 frequency counters.

  The histogram correction unit A30 calculates corrected histogram data in which the appearance frequency distribution is corrected in consideration of fluctuations in the histogram data calculated by the histogram calculation unit A29, and creates a static Huffman coding table based on the corrected histogram data To do.

  The compression processing unit A31 performs Huffman coding on the RGB image data from the input gamma processing unit 22 using the Huffman coding table created by the histogram correction unit A30, and converts the compressed RGB image data to the memory interface 20 Through the memory 16.

  The decompression processing unit A32 reads out the RGB image data compressed by the compression processing unit A31 from the memory 16 through the memory interface 20, decodes and decompresses the RGB image data using the Huffman coding table, and performs original uncompression. RGB image data of the state is obtained.

  The resolution conversion processing unit 24 uses the RGB image data from the expansion processing unit A32 to generate and output RGB image data indicating the original image scaled in the main scanning direction or the sub-scanning direction. When enlarging by resolution conversion, the amount of image data increases, so it is necessary to keep the processing speed constant by temporarily storing the image data in the memory and then reading the image data from the memory again.

  The area separation processing unit 25 adds 1 bit of an attribute flag indicating whether a pixel represents a character or a pixel other than that to each pixel of the image indicated by the RGB image data.

  The filter processing unit 26 applies RGB image data to a two-dimensional FIR (Finite Impulse Response) filter for each pixel, uses an attribute flag of the pixel, and if the pixel represents a character, the filter processing unit 26 emphasizes the RGB image data. In the process, the character edge is emphasized, and if the pixel represents other than that, the RGB image data is subjected to a smoothing filter to remove noise.

  The color correction processing unit 27 converts RGB image data into CMYK image data. That is, in order to realize color reproduction fidelity, the RGB characteristics are directly converted into CMYK values by referring to the spectral characteristics of the CMYK color material including unnecessary absorption components in a table. This CMYK image data represents 256 gradations for each color of CMYK.

  When the halftone processing unit 28 receives CMYK image data representing 256 gradations for each color of CMYK from the color correction processing unit 26, the halftone processing unit 28 performs multi-value dither processing to obtain 256 gradations for each color of CMYK of the engine unit 13. CMYK image data expressed in 16 gradations for each color of CMYK in accordance with the gradation expression capability is generated and output.

  The histogram calculation unit B33 receives the CMYK image data from the halftone processing unit 28, refers to the color and gradation of each pixel constituting the image indicated by the CMYK image data, and stores frequency data for 16 gradations of each color ( Histogram data). For this reason, the histogram calculation unit B33 includes 16 frequency counters for counting the appearance frequency of 16 gradations for each color of CMYK, and a total of 64 frequency counters.

  The histogram correction unit B34 calculates corrected histogram data in which the appearance frequency distribution is corrected in consideration of fluctuations in the histogram data calculated by the histogram calculation unit B33, and creates a static Huffman coding table based on the corrected histogram data. .

  The compression processing unit B35 performs Huffman coding on the CMYK image data from the halftone processing unit 28 using the Huffman coding table created by the histogram correction unit B34, and the CMYK image data compressed thereby is stored in the memory interface 20. Through the memory 16.

  The decompression processing unit B36 reads the CMYK image data compressed by the compression processing unit B35 from the memory 16 through the memory interface 20, decodes and decompresses the CMYK image data using the Huffman coding table, and performs original uncompression CMYK image data of the state is obtained.

The CMYK image data from the expansion processing unit B36 is output to the engine unit 13 for each color of CMYK in accordance with the document image recording timing by the engine unit 13.
Next, an outline of processing centered on the image processing unit 12 will be described with reference to the flowchart of FIG.

  Here, after performing the pre-scan process of step S1, the main scan process of step S2 is performed.

  In the pre-scan process in step S1, a document image is read and scanned by the scanner unit 11, and a process by the image processing unit 12 is performed to determine whether the document image is color or monochrome, and RGB image data is obtained. A Huffman coding table necessary for compressing / decompressing is created, and a Huffman coding table necessary for compressing / decompressing CMYK image data is also created. However, the CMYK image data is not output to the engine unit 13.

  In the main scan process of step S2, the original image is scanned by the scanner unit 11 and the process by the image processing unit 12 is performed, and the RGB image data is generated using the Huffman coding table created by the prescan process. Compression / decompression, reading / writing of compressed RGB image data to / from the memory 16, compression / decompression of CMYK image data, and reading / writing of compressed CMYK image data to / from the memory 16 are performed. Then, the CMYK image data is output to the engine unit 13.

  Next, the prescan process in step S1 will be described with reference to the flowchart of FIG.

  First, in step S <b> 10, the image processing unit 12 is initialized, and various operation parameters of the image processing unit 12 are set by the controller 15. At this time, the frequency counters in the histogram calculation unit A29 and the histogram calculation unit B33 are also initialized.

  In step S11, it is determined whether the CCD line image sensor unit of the scanner unit 11 has been moved in the sub-scanning direction and has come to the reading position. When the input resolution of the scanner unit 11 is 600 dpi (dot per inch), the reading position is a position obtained by dividing a 1-inch section into 600 parts.

  In step S12, when moved to the reading position, the scanner unit 11 reads each pixel value for one main scanning line for each color of RGB, and converts each pixel value into a 10-bit digital signal, This digital signal is output to the image processing unit 12 as RGB image data.

  In step S13, when the shading correction processing unit 21 receives RGB image data for one main scanning line, the shading correction processing unit 21 performs a shading correction process on the RGB image data. Along with this, a 10-bit digital signal indicating the value of each pixel is converted into an 8-bit digital signal.

  In step S14, the input gamma correction processing unit 22 corrects the input characteristics of the scanner unit 11 so that the RGB image data can be easily handled in subsequent image processing.

  In step S15, the histogram calculation unit A29 refers to the color and gradation of each pixel for one main scanning line indicated by the RGB image data for one main scanning line from the input gamma processing unit 22, and 256 for each color. The 256 frequency counters in the histogram calculation unit A29 are incremented according to the frequency of appearance of gradations, and the frequency of appearance of 256 gradations is counted for each color.

  In step S <b> 16, RGB image data for one main scanning line from the input gamma correction processing unit 22 is temporarily written in the memory 16. At this time, the compression processing unit A31 does not perform the Huffman coding process, outputs the RGB image data for one main scanning line to the memory interface 20 in the uncompressed state as it is, and writes it to the memory 16.

  In step S <b> 17, the decompression processing unit A <b> 32 reads the uncompressed RGB image data from the memory 16 via the memory interface 20 and supplies this RGB image data to the resolution conversion processing unit 24.

  In step S18, the resolution conversion processing unit 24 performs resolution conversion processing on the RGB image data from the expansion processing unit A32 in both the main scanning direction and the sub-scanning direction, and changes the scaled RGB in the main scanning direction and the sub-scanning direction. Generate and output image data.

  In step S <b> 19, the region separation processing unit 25 adds 1 bit of an attribute flag indicating whether the pixel represents a character or the other pixel to each pixel indicated by the RGB image data.

  In step S20, the filter processing unit 26 uses a pixel attribute flag for each pixel. If the pixel represents a character, the RGB image data is subjected to enhancement filter processing to enhance the character edge, and if the pixel represents the other. The RGB image data is subjected to a smoothing filter to remove noise.

  In step S <b> 21, the color correction processing unit 27 refers to the spectral characteristics of the CMYK color material in the table, and converts 256 gradation RGB image data for each color into 256 gradation CMYK image data for each color.

  In step S22, the halftone processing unit 28 performs multi-value dither processing to generate and output CMYK image data in which 256 gradations for each color of CMYK are expressed by 16 gradations for each color of CMYK.

  In step S23, the histogram calculation unit B33 refers to the color and gradation of each pixel indicated by the CMYK image data from the halftone processing unit 28, and the histogram calculation unit according to the appearance frequency of 16 gradations for each color. Each of the 64 frequency counters in B33 is incremented, and the appearance frequency of 16 gradations is counted for each color.

  Through the processing in steps S11 to S23, one main scanning line of the original image is read, and RGB image data and CMYK image data corresponding to the main scanning one line are generated and processed. Separately, the frequency of 256 gradations is counted, and the frequency of 16 gradations is counted for each color of CMYK.

  Subsequently, in step S24, it is confirmed whether or not one page of the document image has been read and processed. If the process has not ended, the processes in steps S11 to S23 are repeated. Thereby, RGB image data and CMYK image data corresponding to one page of the original image are generated and processed, and the frequency of 256 gradations is counted for each color of RGB, and the frequency of 16 gradations is counted for each color of CMYK. Counted.

  Next, in step S25, the histogram correction unit A30 corrects the histogram for each color based on the appearance frequency (histogram data) of 256 gradations for each color counted by the 256 frequency counters in the histogram calculation unit A29. Data is created, and a Huffman coding table for the main scan process is created based on the corrected histogram data.

  In addition, the histogram correction unit B34 generates correction histogram data for each color based on the appearance frequency (histogram data) of 16 gradations for each color counted by the 64 frequency counters in the histogram calculation unit B33. Based on the corrected histogram data, a Huffman coding table for the main scan process is created.

  The correction histogram data creation process uses averaging correction and normalization correction, which will be described later. The reason for creating this correction histogram data is that the same image data indicating the same document image should be generated by the scanner unit 11 in both the pre-scan process and the main scan process. Due to a change in the state of the analog circuit around the sensor unit, a drift of each gradation occurs between the pre-scan process and the main scan process, and the appearance frequency distribution of each gradation collected during the main scan is the gradation collected during the pre-scan. This is to deal with this because there is a slight change from the appearance frequency distribution of.

  Further, in the final step S26 of the prescan process, the color image determination unit 23 determines whether the document image indicated by the RGB image data from the input gamma processing unit 22 is color or monochrome. Since it is not possible to determine whether it is color or monochrome unless data of the entire document image is received, this processing is executed last.

  Next, the process of creating the corrected histogram data and the Huffman coding table in step S25 will be described with reference to the flowchart of FIG.

  First, in step S30, averaging correction is performed as correction processing. In this averaging correction, as shown in the graph of FIG. 6, there is a large variation in the frequency distribution of each gradation at the time of the pre-scan process, and after that, as shown in the same graph, When the appearance frequency distribution is changed, a Huffman coding table to be described later is created using the appearance frequency distribution at the time of pre-scan processing as it is, and when the image data is compressed using this Huffman coding table, This is done to prevent this because there is a risk that the data size will be larger than the original data size or that the data size after compression will exceed a certain size.

この様な補正出現度数Ｃ‘（Ｘ）の決定は、ヒストグラム算出部Ａ２９及びヒストグラム算出部Ｂ３３が生成した各色のヒストグラムデータについてそれぞれ実行される。 As the averaging correction method, the averaging coefficient α is determined in advance, and the corrected appearance frequency C ′ (X) is determined by the following equation, with the appearance frequency C (X) for the gradation X and CA as the average appearance frequency. To do.

Such determination of the corrected appearance frequency C ′ (X) is executed for the histogram data of each color generated by the histogram calculation unit A29 and the histogram calculation unit B33.

正規化出現度数Ｃ‘’（Ｘ）の総和が４０９６以下にならなければ、上式と同様の操作を繰り返し、４０９６以下に収まるまで繰り返す。ヒストグラム補正部Ａ３０においてはハフマン符号長を１２ｂｉｔとして処理を実行し、ヒストグラム補正部Ｂ３４においては階調が１６階調であるため、ハフマン符号長を８ｂｉｔとして処理を実行する。 In step S31, normalization correction is performed on the histogram data subjected to the average correction. This is a process for aligning the Huffman code length. When the Huffman code length is defined as 12 bits, the total CA ′ of the entire corrected appearance frequency with respect to the corrected appearance frequency C ′ (X) calculated in step S30. Is used to calculate the normalized appearance frequency C ″ (X) according to the following equation. The decimal part is rounded down.

If the sum of the normalized appearance frequencies C ″ (X) does not become 4096 or less, the same operation as the above equation is repeated, and is repeated until it falls within 4096. The histogram correction unit A30 executes processing with a Huffman code length of 12 bits, and the histogram correction unit B34 executes processing with a Huffman code length of 8 bits because the gradation is 16 gradations.

  In step S32, a Huffman coding table is created. The Huffman coding table is created by creating a Huffman tree by a conventionally known method using the normalized appearance frequency C ″ (X), and generating a code table for each gradation from the created Huffman tree. Do that. As a result, the Huffman coding tables 71 and 72 as shown in FIG. 7 are created in the histogram correction unit A30, and the Huffman coding tables 81 and 82 as shown in FIG. 8 are created in the histogram correction unit B34. Is done. Each of the Huffman coding tables 71 and 81 is used as a forward Huffman coding table and used for compression in order to obtain a coded value and a bit length from the gradation. The Huffman encoding tables 72 and 83 are used as decompression Huffman encoding tables when decompressing to obtain a gradation and a bit length from the encoded bit pattern. Each Huffman encoding table 71, 72, 81, 82 is prepared for each color.

  The pre-scan process is performed in the above procedure.

  Next, the main scanning process in step S2 will be described with reference to the flowchart of FIG.

  First, in step S40, as in step S10, the image processing unit 12 is initialized, and various operation parameters of the image processing unit 12 are set by the controller 15. At this time, the Huffman encoding tables 71, 72, 81, and 82 are set in the compression processing unit A31, the decompression processing unit A32, the compression processing unit B35, and the decompression processing unit B36.

  Steps S41 to S44 are the same as steps S11 to S14 of FIG. 4 described above.

  In step S45, the compression processing unit A31 receives the RGB image data from the input gamma processing unit 22, and Huffman-codes the image data according to a Huffman encoding table 71 as shown in FIG. To implement. This data compression is performed in units of main scanning lines, and packet data as shown in FIG. 10 is created. When the image data compressed by Huffman coding is not in byte units, dummy bits are added to the image data compressed to be in byte units.

  In step S46, the packet data is stored in the memory 16 through the memory interface 20 for each main scanning line.

  In step S47, necessary packet data of the main scanning line is read from the memory 16 through the memory interface 20.

  In step S48, the decompression processing unit A32 decompresses the compressed image data according to the Huffman coding table 72 as shown in FIG. 7 for each of the RGB colors to obtain the original image data.

  Steps S49 to S51 are the same as steps S18 to S20 of FIG. 4 described above.

  In step S52, the color correction processing unit 27 converts RGB image data into CMYK image data. At this time, if the document image is determined to be color by the pre-scan process, color correction processing for color is performed, and if it is determined to be monochrome, monochrome color correction processing is performed.

  Step S53 performs the same process as S22.

  In step S54, one of the following two processes is selected depending on whether the document image is color or monochrome.

  If the document image is monochrome, steps S61 and S62 are selected. In steps S61 and S62, the CMYK image data is input / output to / from the memory 16 without being compressed or expanded. This is based on the CMYK image data generated by performing the color correction process for color on the histogram data acquired during the pre-scan by the histogram calculation unit B33, and this histogram data performs the color correction process for monochrome. This is because it is different from the histogram data based on the generated image data.

  In step S55, the compression processing unit B35 receives the CMYK image data from the halftone processing unit 28, encodes the image data according to the Huffman encoding table 81 as shown in FIG. 8 for each color of CMYK, and compresses the data. To implement. This data compression is performed in units of main scanning lines, and packet data as shown in FIG. 11 is created. When the image data compressed by Huffman coding is not in byte units, dummy bits are added to the image data compressed to be in byte units.

  In step S56, the packet data is stored in the memory 16 through the memory interface 20 for each main scanning line.

  In step S57, necessary main scanning line data is read from the memory 16 through the memory interface 20 for each color of CMYK.

  In step S58, the decompression processing unit B36 decompresses the compressed image data according to the CMYK color according to the Huffman coding table 82 as shown in FIG. 8, and obtains the original image data.

  In step S <b> 59, the expanded original CMYK image data is output to the engine unit 13 in response to a request from the engine unit 13.

  In step S60, it is confirmed whether or not all the image data of one page of the document image has been processed. If the processing is not completed, the processing in steps S41 to S59 is repeated.

  The main scanning process is performed in the above procedure.

  As described above, in the present embodiment, the histogram data of the document image is created and corrected by the pre-scan process, the Huffman coding table is created based on the corrected histogram data, and the image data is compressed using the Huffman coding table. Since the expansion is performed, the image data can be efficiently compressed and expanded according to the density histogram of the document image, and the data transfer amount to the memory 16 can be reduced. In addition, since the encoding of the compression process in the main scan process is determined using the data acquired at the time of the pre-scan, the compression / decompression process is light and the circuit of the compression / decompression process unit that can follow the memory bandwidth The scale is small.

  In the above-described embodiment, the averaging correction is performed in step S30. However, the scanner unit 11 and the like have a storage device that can store all image data of the original image. If the same image data can be used, the averaging correction need not be performed.

  Even if the data size after compression becomes larger than the data size before compression, the compressed data is stored as it is. However, as shown in FIG. If the data size after compression becomes smaller than the data size before compression, the compressed data is stored in the memory 16, and the data size after compression is larger than the data size before compression. At this time, uncompressed data is stored in the memory 16, and when reading the data, it is possible to determine whether or not data expansion is necessary by referring to the flag of the packet data. As a result, it is possible to prevent a local memory usage increase and a necessary memory bandwidth increase.

  In addition, only one type is created as the histogram data for all image data of the document image. However, the histogram data is created in units of main scanning lines or blocks divided by region separation processing. Alternatively, a Huffman coding table used in compression processing and decompression processing may be created for each block, and each Huffman coding table may be switched and used.

  Further, although static Huffman coding is used as the compression method and decompression method, a method may be used in which only a value having a large appearance frequency is run-length compressed from the histogram data. An arithmetic code may be used instead of the static Huffman coding.

1 is a block diagram illustrating Embodiment 1 of an image forming apparatus of the present invention. FIG. 2 is a block diagram illustrating a configuration of an image processing unit in the image forming apparatus of FIG. 1. 2 is a flowchart illustrating an outline of processing centering on an image processing unit in FIG. 1. It is a flowchart which shows the prescan process of FIG.3 S1. FIG. 5 is a flowchart showing a process of creating correction histogram data and a Huffman coding table in step S25 of FIG. It is a graph which shows the appearance frequency distribution of each gradation at the time of a pre-scan process, and the appearance frequency distribution of each gradation at the time of this scanning process. It is a chart which illustrates the Huffman coding table used for compression and expansion of RGB image data. It is a chart which illustrates the Huffman coding table used for compression and expansion of CMYK image data. It is a flowchart which shows the main scan process of step S2 of FIG. It is a figure which illustrates the packet data for transferring the compressed RGB image data. It is a figure which illustrates the packet data for transferring the compressed CMYK image data.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 Scanner part 12 Image processing part 13 Engine part 16 Memory 20 Memory interface 29 Histogram calculation part A
30 Histogram correction part A
31 Compression processing part A
32 Decompression processing part A
33 Histogram calculator B
34 Histogram correction part B
35 Compression processing part B
36 Decompression processing part B

Claims (4)

In an image processing apparatus that acquires first and second image data indicating the same image, and compresses and expands the second image data using the first image data.
A histogram creating means for creating histogram data of the first image data;
Histogram correction means for performing correction processing on the histogram data created by the histogram creation means to create corrected histogram data; and
Compression / expansion means for performing reversible compression / expansion of the second image data at a compression / expansion rate corresponding to the corrected histogram data created by the histogram correction means ;
An image processing apparatus comprising: storage means for writing and reading compressed second image data. The image processing apparatus according to claim 1, wherein the image is divided into a plurality of blocks, histogram data is created for each block, and the second image data is compressed and expanded. 2. The image processing apparatus according to claim 1, wherein when the data size of the compressed second image data is equal to or larger than a predetermined size, the storage unit performs writing / reading of the second image data in an uncompressed state. . An image forming apparatus comprising the image processing apparatus according to claim 1.

Images