JP2013197984A - Jpeg compression device and image reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent an increase in circuit scale.SOLUTION: A JPEG compression section 40 includes an input data switching section 44 which acquires block data according to identification information and outputs the block data and identification data; a configuration up to a zigzag sequence section 50 that executes processes which can treat the block data output from the input data switching section 44 independently by the block data among a plurality of treatment processes constituting a JPEG compression process, and outputs the block data having been processed and the identification information; and an entropy encoding section 58 which executes other processes among the plurality of processes constituting the JPEG compression process using block data related to the same identification information as the identification information of the block data, and processed by the configuration up to the zigzag sequence section 50.

Description

  The invention disclosed in this specification relates to a technique used in a JPEG compression apparatus.

  2. Description of the Related Art Conventionally, an apparatus that performs compression processing on image data and generates compressed image data is known (for example, Patent Document 1). This apparatus may further include a front-side image reading unit and a back-side image reading unit for reading image data on both sides of the document, and the image data on the front and back sides read by using these image reading units. On the other hand, it may be required to perform compression processing. In the prior art, a technique is disclosed in which image data obtained by reading the front side of a document and image data obtained by reading the back side of the document are compressed using separate processing circuits to generate compressed image data simultaneously. .

JP 2009-296352 A

  As described above, an apparatus that generates compressed image data may be required to generate a plurality of compressed image data. However, in order to generate a plurality of compressed image data as in the prior art, for example, if it is necessary to prepare a plurality of processing circuits to generate the compressed image data of the front surface and the back surface, the circuit scale of the processing circuit This increases the problem.

  The invention disclosed in this specification discloses a technique for suppressing an increase in circuit scale in a JPEG compression apparatus that performs JPEG compression processing on image data.

  The JPEG compression apparatus disclosed in this specification is a JPEG compression apparatus that performs JPEG compression processing including a plurality of processes on one or a plurality of input image data, and at least any of the input image data. The input image data is acquired according to identification information having information for identifying whether or not, and image data consisting of one or a plurality of block-unit image data, which is a minimum processing unit of JPEG compression processing, and the identification information An acquisition output unit that outputs, and processing that can be performed independently of the plurality of processes on the block-unit image data that is output from the acquisition output unit. The independent compression processing unit that outputs the executed block-unit image data and the identification information, and the independent compression processing unit that is the same identification information are output Among the plurality of processes for each of the image data, other processing is executed, and a non-independent compression processing section for outputting as a JPEG compressed image data corresponding to the identification information.

  In this JPEG compression apparatus, an independent compression processing unit performs processing on block-unit image data for processing that can be independently performed on a block-by-block basis among a plurality of processes constituting the JPEG compression processing. Run. Therefore, when performing JPEG compression processing on one or a plurality of image data to generate a plurality of compressed image data, any compressed image data can be processed independently for each block. Can be executed using a common independent compression processing unit, and all the processes constituting the JPEG compression process are executed using different processing circuits. Compared to the prior art, an increase in circuit scale can be suppressed.

  In this JPEG compression apparatus, the independent compression processing unit includes a DCT transform unit that performs discrete cosine transform in units of blocks, and a quantization that performs quantization on each component in units of blocks generated by the discrete cosine transform And a first output switching unit that switches and outputs each quantized component in accordance with the associated identification information, and the non-independent compression processing units have the same identification A plurality of entropy coding output units that perform entropy coding on each component of the block unit associated with information and output each component of the block unit that has been entropy coded according to the identification information; and It is good also as a structure provided. According to this JPEG compression apparatus, the configuration other than the entropy encoding output unit can be used in common among the configurations that execute a plurality of processes constituting the JPEG compression process, and an increase in circuit scale is suppressed. be able to.

  In this JPEG compression apparatus, the independent compression processing unit includes a DCT transform unit that performs discrete cosine transform in units of blocks, and a quantization that performs quantization on each component in units of blocks generated by the discrete cosine transform An entropy encoding unit that performs entropy encoding on each component of the block unit quantized by the quantization unit according to the associated identification information, and the non-independent compression processing unit A second output switching unit that outputs each component of the block unit according to the identification information, and the entropy encoding unit outputs each of the block units output one before for each of the identification information Corresponding to the identification information associated with the component group of the block unit from a plurality of holding units that hold the head component of the component and the head component of the component unit of the block unit A difference acquisition unit that subtracts the head component held in the storage unit and outputs a difference between the head components, and the difference between the head component and the head component other than the head component of the block unit component group It is good also as a structure which performs entropy encoding with respect to these components. According to this JPEG compression apparatus, it is possible to use in common a configuration other than the holding unit included in the entropy encoding unit among the configurations that execute a plurality of processes constituting the JPEG compression process. The increase can be suppressed.

  In this JPEG compression apparatus, the independent compression processing unit further switches a plurality of quantization tables according to the identification information and the quantization table according to the identification information, and information of the switched quantization table The quantization unit may be configured to perform the quantization using information of a quantization table output by the first selector. According to this JPEG compression apparatus, it is possible to suitably quantize the input image data using the information in the quantization table output by the first selector.

  In this JPEG compression apparatus, the non-independent compression processing unit further switches between a plurality of entropy encoding tables corresponding to the identification information and the entropy encoding tables corresponding to the identification information, and the switched entropy codes. A second selector that outputs the information of the encoding table, and the entropy encoding unit may perform the entropy encoding using the information of the entropy encoding table output by the second selector. According to this JPEG compression apparatus, the input image data can be suitably entropy-encoded using the information in the entropy encoding table output by the second selector.

  In this JPEG compression apparatus, the identification information further includes information related to a compression rate of JPEG compression processing, and the quantization table includes a number corresponding to information related to the compression rate of the JPEG compression processing, and the first selector Outputs the information of the switched quantization table according to the information on the compression rate of the JPEG compression process, and the quantization unit uses the information of the quantization table output by the first selector to It is good also as a structure which performs quantization. According to this JPEG compression apparatus, the first selector switches the quantization table information to be output according to the information on the compression rate of the JPEG compression process, so that the compression rate of the input image data can be increased efficiently.

  In this JPEG compression apparatus, the identification information further includes information related to a compression rate of JPEG compression processing, and the entropy encoding table has a number corresponding to information related to the compression rate of the JPEG compression processing, and the second The selector outputs information of the switched entropy coding table according to the information regarding the compression rate of the JPEG compression process, and the entropy coding unit outputs information of the entropy coding table output by the second selector. It is good also as a structure which performs the said entropy encoding using. According to this JPEG compression apparatus, since the second selector switches the information of the entropy coding table to be output according to the information regarding the compression rate of the JPEG compression process, it is possible to set the compression rate according to the input image data.

  In this JPEG compression apparatus, the identification information may have a processing condition for the JPEG compression process, and a plurality of JPEG compression processes with different processing conditions may be performed on one input image data. According to this JPEG compression apparatus, it is possible to suppress an increase in circuit scale when performing a plurality of JPEG compression processes with different processing conditions for one input image data.

  The invention disclosed in this specification is also embodied in an image reading apparatus including the above-described JPEG compression apparatus. An image processing apparatus disclosed in the present specification includes the JPEG compression apparatus described above, and an input image data acquisition unit that acquires the input image data and sequentially stores the acquired input image data in the storage unit The acquisition output unit includes at least the block unit image data together with the identification information in response to the input image data stored in the storage unit being stored in the block unit image data. To get. According to this image processing apparatus, the storage unit only needs a storage capacity for storing about the amount of image data for each block, and the storage capacity required for the storage unit can be suppressed.

  In this image reading apparatus, the input image data acquisition unit includes at least an image reading unit that reads the object to be read and generates the image data, and the JPEG compression device compresses the input image data using JPEG compression. Processing may be performed so that the at least one input image data is the image data read by the image reading unit. According to this image reading apparatus, in an image reading apparatus that performs JPEG compression processing on a plurality of image data including image data generated using an image reading unit, the circuit of the JPEG compression apparatus included in the image reading apparatus An increase in scale can be suppressed.

  According to the invention disclosed in this specification, an increase in the circuit scale of the JPEG compression apparatus can be suppressed.

2 is a cross-sectional view of the image reading apparatus 10. FIG. FIG. 2 is a circuit block diagram showing an electrical configuration of the image reading apparatus 10. The figure which shows the structure of CPU30, RAM33, and the JPEG compression part 40 in 1st Embodiment. The flowchart which shows a process surface flag switching process. The figure which shows the relationship between 8x8 block data and image data. The figure which shows the structure of CPU30, RAM33, and the JPEG compression part 40 in 2nd Embodiment. The figure explaining DC component and AC component. The figure which shows the structure of the entropy encoding part 58 in 2nd Embodiment. The figure which shows the structure of CPU30, RAM33, and the JPEG compression part 40 in other embodiment. The figure which expands and shows the structure of the quantization part 48 vicinity of the JPEG compression part 40 in other embodiment. The figure which expands and shows the structure of the entropy encoding part in other embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. The embodiments of the present invention are not limited to the following embodiments, and various forms can be adopted as long as they belong to the technical scope of the present invention. This embodiment will be described using an image reading apparatus 10 that reads an image of a document as an example of a JPEG compression apparatus.

1. Configuration of Image Reading Device FIG. 1 is a cross-sectional view of an image reading device 10. The image reading apparatus 10 uses the conveyance rollers 121 to 129 to convey the document placed on the paper feed tray 110 to the paper discharge tray 130 along the conveyance path (chain line). A CIS (Contact Image Sensor) 231 is provided on the upstream side of the conveyance path, and a CIS 230 is provided on the downstream side of the conveyance path. The CIS 231 reads the upper surface of the original in FIG. 1 conveyed on the conveyance path at the upstream position. In addition, the CIS 230 reads the document surface in the downward direction in FIG. 1 of the document conveyed on the conveyance path.

  In the following description, when an original is placed on the paper feed tray 110, the upper side of the original in FIG. 1 is used as the front side of the original, and the lower side of the original in FIG. And

  FIG. 2 is a circuit block diagram showing an electrical configuration of the image reading apparatus 10. The image reading apparatus 10 includes a CPU 30, a ROM 32, a RAM 33, a device driving unit 34, a reading correction / image processing unit 36, a JPEG compression unit 40, and an image reading unit 22 (an input image data acquisition unit, an image reading unit). 1), a USB interface (hereinafter referred to as I / F) 24, and an operation unit 28 as main parts.

  The CPU 30, ROM 32, RAM 33, device drive unit 34, reading correction / image processing unit 36, JPEG compression unit 40, USB I / F 24 (another example of the input image data acquisition unit), operation unit 28 is connected via a bus 29.

  The ROM 32 stores various programs for controlling the operation of the image reading apparatus 10, and the CPU 30 controls each unit according to the program read from the ROM 32.

  The device driving unit 34 is connected to the image reading unit 22 and transmits a signal for controlling reading to the image reading unit 22 based on a command from the CPU 30. The image reading unit 22 reads a document based on a signal from the device driving unit 34 and outputs the read data read to the reading correction / image processing unit 36.

  The image reading unit 22 includes a CIS 231 and a CIS 230. Specifically, the image reading unit 22 reads the surface of the document using the CIS 230 and generates read data of the surface of the analog signal. The image reading unit 22 reads the back side of the document using the CIS 231 and generates read data of the back side of the analog signal.

  The reading correction / image processing unit 36 is connected to the image reading unit 22 and converts read data that is an analog signal output from the image reading unit 22 into image data that is a digital signal. The read correction / image processing unit 36 performs various read corrections such as shading correction and image processing such as enlargement, reduction, and enhancement processing on the converted image data. Then, the various corrected image data is stored in the RAM 33 as image data of each color component including luminance (Y) and color difference (Cb, Cr).

  The USB I / F 24 connects the image reading apparatus 10 to an external PC via a network line. The RAM 33 temporarily stores image data that has been variously corrected by the reading correction / image processing unit 36 and input image data that is image data input from the outside of the image reading apparatus 10 via the USB I / F 24. .

  The JPEG compression unit 40 JPEG-compresses the input image data (front-side image data, back-side image data, and image data input via the USB I / F 24) stored in the RAM 33 to generate JPEG-compressed image data. . Then, the JPEG compressed image data is stored in the RAM 33 again.

  The operation unit 28 includes a group of buttons including a start key, a stop key, a numeric keypad, and the like, and receives an operation instruction from a user.

2. Next, the input image data (front-side image data and back-side image data) obtained by reading both sides of the document by the image reading device 10 are sequentially stored in the RAM 33 and stored in the RAM 33. An embodiment in which the input image data thus compressed is JPEG-compressed will be described.

  First, a description will be given of how the input image data (front side image data and back side image data) obtained by reading both sides of the document is sequentially stored in the RAM 33.

  When the original placed on the paper feed tray 110 reaches the reading position of the CIS 231, the image reading apparatus 10 starts reading the back side of the original. Then, the back surface image data read by the CIS 231 and variously corrected by the reading correction / image processing unit 36 is sequentially stored in the RAM 33 for each line of the CIS 231.

  Next, when the original reaches the reading position of the CIS 230, the image reading apparatus 10 starts reading the surface of the original. At this time, the trailing edge of the document has not passed the reading position of the CIS 231. In the RAM 33, the image data of the front surface read by the CIS 230 and variously corrected by the reading correction / image processing unit 36 and the image data of the back surface described above are sequentially stored for each line.

  When the trailing edge of the document passes through the reading position of the CIS 231, the reading of the back side of the document of the CIS 231 ends. The RAM 33 stores image data on the surface in order for each line. Then, the reading of the surface of the document of the CIS 230 is performed until the trailing edge of the document passes through the reading position of the CIS 230.

3. Next, JPEG compression of input image data stored in the RAM 33 using the JPEG compression unit 40 will be described with reference to FIGS. 3 and 6. FIGS. 3 and 6 are diagrams showing configurations of the CPU 30, the RAM 33, and the JPEG compression unit 40. 3 and FIG. 6 have the same configuration in the input data switching unit 44, the thinning unit 45, and the configuration group included in the dotted line 56 (from the DCT conversion unit 46 to the zigzag sequence unit 50, an example of an independent compression unit). However, an entropy encoding unit 58 (an example of a non-independent compression unit) is different as an embodiment.

  In the following description, JPEG compression is performed in the form of 4: 2: 0 as a sampling factor that represents spatial thinning of each of the three color components of luminance (Y) and color difference (Cb, Cr). explain.

  FIG. 3 shows the configurations of the CPU 30, RAM 33, and JPEG compression unit 40, respectively. The RAM 33 stores image data on the front side and image data on the back side by double-sided reading of the image reading device 10. FIG. 3 shows a configuration in which image data stored in the RAM 33 is JPEG-compressed by the JPEG compression unit 40 and image data that has been JPEG-compressed again is stored in the RAM 33.

  The JPEG compression unit 40 includes a plurality of components that perform JPEG compression processing including a plurality of processes. The JPEG compression unit 40 includes an input data switching unit 44, a decimation unit 45, a discrete cosine transform unit (hereinafter referred to as DCT transform unit) 46, a quantization unit 48, a selector 48A, a zigzag sequence unit 50, and an entropy code. And a conversion unit 58. The input data switching unit 44 and the thinning unit 45 are examples of the acquisition output unit.

  The input data switching unit 44 acquires the image data of the front surface and the image data of the back surface stored in the RAM 33 according to an instruction from the CPU 30, and will be described later for each block data composed of 16 × 16 pixel data. Together with the identification information to be output to the thinning unit 45.

  FIG. 4 is a flowchart showing the processing surface flag switching process of the CPU 30. As described above in “2. Double-sided scanning of image reading apparatus”, the image data on the front side and the image data on the back side are stored for each line.

  The CPU 30 determines whether or not 16 lines of surface image data are stored in the RAM 33 (S2). Specifically, the CPU 30 can determine by counting how many lines of the surface image data are stored in the RAM 33.

  If the front surface image data for 16 lines is not stored in the RAM 33 (S2: NO), the CPU 30 determines whether the back surface image data for 16 lines is stored in the RAM 33 (S4). When the back surface image data for 16 lines is not stored in the RAM 33 (S4: NO), the CPU 30 returns to S2.

  When the surface image data for 16 lines is stored in the RAM 33 (S2: YES), the CPU 30 sets the processing surface flag stored in the RAM 33 on the surface (S6). And CPU30 inputs the identification information which shows that it is the surface with respect to the input data switching part 44 (S8).

  When the identification information indicating that the surface is the front surface is input from the CPU 30, the input data switching unit 44 starts to acquire the front surface image data from the RAM 33. Then, the input data switching unit 44 outputs the obtained surface image data to the thinning unit 45 together with identification information indicating the surface for every 16 × 16 pixel data.

  FIG. 5 shows image data of a luminance (Y) component among three of luminance (Y) and color difference (Cb, Cr). As shown in FIG. 5, the luminance (Y) component image data is composed of a plurality of pixel data, and the remaining two components, ie, the color differences (Cb, Cr), are similarly constructed.

  In the case of 4: 2: 0 as the sampling factor for JPEG compression, the thinning-out unit 45 has 8 × 8 pixel data block data (minimum of JPEG compression processing) shown in FIG. The image data is output to the DCT conversion unit 46 every block). On the other hand, with regard to the color difference (Cb, Cr) component, the thinning unit 45 reduces the 16 × 16 pixel data to 1/2 each of the vertical and horizontal directions, thereby converting the DCT conversion unit 46 into block data of 8 × 8 pixel data. Output to.

  By the input data switching unit 44 and the thinning-out unit 45, a total of 6 block data of block data Y1, Y2, Y3, Y4 of luminance (Y) and 2 block data of color difference (Cb, Cr) The input data switching unit 44 outputs in order for each data. A total of six pieces of block data Y1, Y2, Y3, Y4 of this luminance (Y) and two thinned out block data of color differences (Cb, Cr) constitute one unit (1 MCU) of JPEG compression processing.

  When the input data switching unit 44 and the thinning unit 45 generate 1 MCU block data and output the block data to the DCT conversion unit 46 for each block data, the next 16 × 16 pixel data is processed by 1 MCU. Are generated and output to the DCT converter 46 for each block data.

  The CPU 30 determines whether or not the input data switching unit 44 has acquired the image data of the front surface for 16 lines (S10). The CPU 30 waits for the reading of 16 × 16 pixel data of the input data switching unit 44 to be completed (S10: NO), and when reading of 16 × 16 pixel data is completed (S10: YES), S4 Proceed to the process.

  The CPU 30 determines whether or not 16 lines of back side image data are stored in the RAM 33 (S4). When the back surface image data for 16 lines is stored in the RAM 33 (S4: YES), the CPU 30 sets the processing surface flag stored in the RAM 33 on the front surface (S16). And CPU30 inputs the identification information which shows that it is a back surface with respect to the input data switching part 44 (S18).

  The input data switching unit 44 and the thinning-out unit 45 generate 6 block data of 1 MCU as in the above-described surface image data, and output each block data to the DCT conversion unit 46 in order. Go.

  As described in “2. Double-Sided Reading of Image Reading Device”, the front side image data and the back side image data are stored in the storage unit 33 line by line in the order of reading. In the conventional image reading apparatus, the JPEG compression process is not performed on the back side image data until the JPEG compression process is completed on the front side image data. Therefore, the JPEG compression process on the front side image data is completed. Until then, the image data of the back surface must be stored in the storage unit 33. Therefore, there is a problem that the storage capacity of the storage unit 33 is increased.

  On the other hand, the image reading apparatus 10 of the present invention performs the processing of the flowcharts of S2 and S4 described above, and the input data switching unit 44 and the thinning-out unit 45 are the front surface or the back surface corresponding to the processing of S2 and S4. The identification information indicating these and the block data corresponding to the identification information are sequentially output to the DCT conversion unit 46. For this reason, since the image reading apparatus 10 processes each block data, the storage unit 33 does not need a storage capacity for the image data of one original.

  Next, in JPEG compression, in addition to the thinning process of the thinning unit 45 described above, JPEG compression is finally performed by several processes such as a DCT conversion process, a quantization process, a zigzag sequence process, and an entropy encoding process. The created image data is created. Among them, DCT transformation, quantization processing, and zigzag sequence processing can be performed independently for each block data. On the other hand, the entropy encoding process is performed according to a dependency relationship between block data, which will be described later, and therefore cannot be performed independently for each block data.

  The DCT conversion unit 46 performs DCT conversion on the block data output from the thinning unit 45 in units of the block data and converts the block data into 8 × 8 frequency data.

  The DCT conversion unit 46 outputs 8 × 8 frequency data to the quantization unit 48, and outputs the identification information (whether it is the front side or the back side) to the quantization unit 48 and the selector 48A, respectively. To do.

  Quantization means that 8 × 8 frequency data subjected to DCT conversion is divided by 8 × 8 constant table (quantization table), and the decimal point is rounded down or rounded off. The data is rounded down. The DCT-transformed 8 × 8 frequency data are arranged in the form of low frequency components to high frequency components. When the compression rate of JPEG compression is increased, the compression rate of the image data is increased by setting the quantization table so that the frequency data of this high frequency component is cut off as much as possible.

  The JPEG compression unit 40 includes a first quantization table 48B and a second quantization table 48C. The CPU 30 is configured to be able to change the first quantization table 48B and the second quantization table 48C, respectively. The CPU 30 can change the compression rate of the image data according to each surface such as the front surface or the back surface by changing the constant of the quantization table for each surface.

  The selector 48A switches from the identification information output from the DCT conversion unit 46 to the first quantization table 48B for the front surface and the second quantization table 48C for the back surface. Output to 48.

  Then, the quantization unit 48 performs quantization processing on 8 × 8 frequency data according to the constants of the quantization tables 48B and 48C corresponding to either the front surface or the back surface. Then, the quantizing unit 48 outputs the 8 × 8 quantized data and the identification information to the zigzag sequence unit 50.

  The zigzag sequence unit 50 performs zigzag scanning on the quantized data, and each frequency component of the 8 × 8 data is in the order of DC component (hereinafter DC component), AC component (hereinafter AC component), and AC. In the component, each frequency component is rearranged so as to be arranged in the order of low frequency to high frequency. Then, the zigzag sequence unit 50 outputs the data rearranged by the zigzag scan (an example of block-unit image data for which the processing has been performed) and the identification information to the entropy encoding unit 58.

4). Entropy encoding process -first embodiment-
First, the entropy encoding unit 58 in FIG. 3 will be described. The entropy encoding unit 58 of FIG. 3 includes an output data switching unit 52, a first entropy encoder 54A, and a second entropy encoder 54B.

  The output data switching unit 52 selects any one of the entropy encoders 54 </ b> A and 54 </ b> B corresponding to the 8 × 8 pieces of data that are zigzag scanned according to the identification information input from the zigzag sequence unit 50. Output to. In the first embodiment, when the identification information is on the front side, it is output to the first entropy encoder 54A, and when the identification information is on the back side, it is output to the second entropy encoder 54B.

  The entropy encoders 54A and 54B perform a well-known entropy encoding process on 8 × 8 data including a DC component and an AC component. However, the entropy encoding processing method differs between the DC component and the AC component.

  FIG. 7 is a diagram showing the DC component and the AC component rearranged by the zigzag scan. The entropy encoders 54A and 54B are configured so that the DC components (FIG. 7: DC (i−1), FIG. 7) input from the DC components (FIG. 7: DC (i)) input to the entropy encoders 54A and 54B Hereinafter, the entropy encoding process is performed on the difference value from the previous DC component).

  The entropy encoders 54A and 54B have holding units 55A and 55B, respectively. The holding units 55A and 55B are input immediately before the previous DC component, that is, the DC component input to the entropy encoders 54A and 54B, and have the same identification information (front and back surfaces) as the input DC component. DC component having () is held in the holding portions 55A and 55B.

  The entropy encoders 54A and 54B subtract the previous DC component stored in the holding units 55A and 55B from the input DC component to calculate a DC difference value. After calculating the DC difference value, the entropy encoders 54A and 54B update the previous DC component of the holding units 55A and 55B to the input DC component.

  The entropy encoders 54A and 54B perform entropy encoding processing on the calculated DC difference value and AC component, respectively. The entropy encoder 54A outputs the data generated by the entropy encoding process to the RAM 33 as JPEG compressed image data of the surface. Further, the entropy encoder 54B outputs the data generated by the entropy encoding process to the RAM 33 as JPEG compressed image data on the back surface.

5. Entropy encoding process -second embodiment-
First, the entropy encoding unit 58 in FIG. 6 will be described. The entropy encoding unit 58 in FIG. 6 includes a third entropy encoder 70 and an output unit 72.

  FIG. 8 shows the configuration of the entropy encoding unit 58. The entropy encoding unit 58 includes a plurality of configurations that perform entropy encoding processing including the plurality of processing described in the first embodiment. In addition to the holding units 55A and 55B described in the first embodiment, the entropy encoder 70 includes a first selection unit 60, a second selection unit 62, a difference acquisition unit 64, an update unit 66, and an encoding unit. Part 68 and selector 68A. The output unit 72 includes code data output units 76A and 76B in addition to the output data switching unit 52 described in the first embodiment.

  The first selection unit 60 outputs a DC component from 8 × 8 data that is output from the zigzag sequence unit 50 and includes a DC component and an AC component, and outputs the DC component to the difference acquisition unit 64 and the update unit 66. Further, the remaining AC component is output to the encoding unit 68.

  The second selection unit 62 reads the previous DC component stored in the holding units 55 </ b> A and 55 </ b> B corresponding to the corresponding identification information in accordance with the identification information input from the zigzag sequence unit 50, and outputs it to the difference acquisition unit 64. .

  The difference acquisition unit 64 subtracts the previous DC component output from the second selection unit 62 from the DC component output from the first selection unit 60 to calculate a DC difference value, and the calculated DC difference value is an encoding unit. Output to 68.

  Encoding is to replace 8 × 8 binary data arranged in a line using a Huffman table. In the Huffman table, a correspondence relationship is set in which a pattern indicated by “1” and “0” included in binary data is replaced with a different pattern indicated by “1” and “0”. In the Huffman table, the number of digits of the pattern after replacement is determined to be smaller than the number of digits of the pattern before replacement. When increasing the compression rate of JPEG compression, by setting the correspondence so that many patterns contained in 8 × 8 binary data are replaced with patterns with as few digits as possible, the input image data The compression rate increases.

  The JPEG compression unit 40 includes a first Huffman table 68B and a second Huffman table 68C. The CPU 30 can change the first Huffman table 68B and the second Huffman table 68C. The CPU 30 can change the compression ratio of the input image data according to each surface such as the front surface and the back surface by changing the correspondence relationship of the Huffman table for each surface.

  The selector 68A switches from the identification information output from the zigzag sequence unit 50 to the encoding unit 68 by switching between the first Huffman table 68B for the front side and the second Huffman table 68C for the back side. Output.

  Then, the encoding unit 68 performs the 8 × 8 binary number data in which the DC difference value output from the difference acquisition unit 64 and the AC component output from the first selection unit 60 are arranged in this order. Thus, the Huffman encoding process is performed according to the correspondence between the Huffman tables 68B and 68C corresponding to either the front surface or the back surface. Then, the encoding unit 68 outputs the encoded data and the identification information to the output unit 72.

  The update unit 66 outputs the previous DC components of the holding units 55A and 55B corresponding to the corresponding identification information according to the identification information input from the zigzag sequence unit 50 after the encoding unit 68 outputs the encoded data. The DC component output from the first selection unit 60 is updated.

  The output data switching unit 52 converts the encoded data according to the identification information input from the encoding unit 68 into a first code data output unit 76A and a second code data output unit 76B corresponding to the corresponding identification information. Output to either of these. In the second embodiment, when the identification information is on the front surface, it is output to the first code data output unit 76A, and when the identification information is the back surface, it is output to the second code data output unit 76B.

  The first code data output unit 76A outputs the data generated by the entropy encoding process to the RAM 33 as JPEG compressed image data of the surface. The second code data output unit 76B outputs the data generated by the entropy encoding process to the RAM 33 as JPEG compressed image data on the back surface.

3. Advantages of the present embodiment (1) In the image reading apparatus 10 of the first and second embodiments, some of the plurality of processes constituting the JPEG compression process can be performed independently for each block data. Process data using a common configuration. Therefore, an increase in the circuit scale of the JPEG compression unit 40 can be suppressed as compared with a case where processing is performed using a different configuration for all the processes that constitute the JPEG compression process.

(2) Specifically, in the image reading apparatus 10 according to the first and second embodiments, a zigzag is performed from a DCT conversion unit that can perform processing independently for each block data among a plurality of processes that constitute JPEG compression processing. The processing up to the sequence unit 50 is performed using a common configuration, and the entropy encoding processing in the entropy encoding unit 58 needs to be performed according to the dependency between block data having common identification information. For this reason, processing is performed using an individual configuration for each piece of identification information associated with block data. By making the configuration up to the zigzag sequence unit common, an increase in the circuit scale of the JPEG compression unit 40 can be suppressed.

(3) Furthermore, in the image reading apparatus 10 according to the second embodiment, a part of the configuration is shared by the entropy encoder 70 included in the entropy encoder 58. Specifically, holding units 55A and 55B for each identification information are prepared for the previous DC component that needs to be switched depending on the dependency relationship between the block data, and other configurations included in the entropy encoder 70 are shared. As a result, an increase in the circuit scale of the JPEG compression unit 40 can be further suppressed.

<Other embodiments>
The present invention is not limited to the embodiments described with reference to the above description and the drawings, and for example, the following various aspects are also included in the technical scope of the present invention.
(1) Although the above embodiment has been described using the image reading apparatus 10 having a reading function, the present invention is not limited to this. For example, in addition to the reading function, it may be a multi-function device further provided with at least one of a printer function, a copy function, a facsimile function, and the like.

(2) The reading function may not be provided. The present invention has a function of performing only JPEG compression processing on image data. For example, JPEG compression processing is performed on image data input from a PC outside the image reading apparatus 10 via the USB I / F 24. In this case, the reading function is not necessarily required.

(3) Although the above embodiment has been described using JPEG compression as the compression format of image data, the present invention is not limited to this. The compression process consists of multiple processes, and the multiple processes can be processed independently for each data, and the independent process for each data cannot be performed independently, depending on the dependency between the data Any compression format having a non-independent compression process that performs processing can be applied.

(4) In the above embodiment, as an example of JPEG compression of a plurality of image data, a case where JPEG compression is performed on the image data on the front side and the image data on the back side read by double-sided reading of the document has been described. Is not limited to this. For example, as shown in FIG. 9, only the surface of the document may be read, and the image data on the surface and the image data input via the USB I / F 24 may be JPEG compressed. Also, a plurality of image data input via the USB I / F 24 may be JPEG compressed.

(5) In this case, the identification information may be information indicating the image type (text, photograph, etc.) of the corresponding input image data instead of the information indicating the front surface or the information indicating the back surface. Good. When the identification information is information indicating the image type, the compression rate is determined by the image type. Therefore, the compression rate of the image data can be increased efficiently by switching the Huffman table based on the identification information. it can.

(6) Further, the identification information may be information relating to compression conditions such as the compression rate of the corresponding input image data. When the identification information is information related to the compression condition, the compression rate corresponding to the input image data can be set by switching the quantization table based on the identification information.

(7) Although the case where the JPEG compression unit 40 includes the first quantization table 48B and the second quantization table 48C has been described in the above embodiment, the present invention is not limited to this. For example, as shown in FIG. 9, the JPEG compression unit 40 may have m quantization tables 48B to 48N. The identification information includes information about compression conditions such as the compression rate of the corresponding input image data, and quantization tables suitable for each compression rate are stored in the quantization tables 48B to 48N. Then, the selector 48A switches the quantization tables 48B to 48N to be used from the information regarding the compression condition of the input image data of the identification information. As a result, the quantization unit 48 can set the input image data to a compression rate corresponding to the input image data.

(8) Although the case where the JPEG compression unit 40 includes the first Huffman table 68B and the second Huffman table 68C has been described in the above embodiment, the present invention is not limited to this. For example, as illustrated in FIG. 10, the JPEG compression unit 40 may include m Huffman tables 68B to 68N. The identification information includes information indicating the image type of the corresponding input image data, and the Huffman tables 68B to 68N store Huffman tables suitable for the information indicating each image type. Then, the selector 68A switches the Huffman tables 68B to 68N to be used from information indicating the image type of the input image data of the identification information. As a result, the encoding unit 68 can efficiently increase the compression rate of the input image data.

(9) Although the above embodiment has been described using an example in which a plurality of image data is JPEG compressed, JPEG compression may be performed by setting a plurality of different compression conditions for one input image data. Even in this case, an increase in the circuit scale of the JPEG compression unit 40 can be suppressed as compared with a case where JPEG compression under different compression conditions is performed using a different configuration.

10: Image reading device, 22: Image reading unit, 30: CPU, 33: RAM, 40: JPEG compression unit, 44: Input data switching unit, 45: Decimation unit, 46: DCT converter, 48: Quantization unit, 50: Zigzag sequence unit, 52: Output data switching unit, 54: Entropy encoder, 55: Holding unit, 58: Entropy encoding unit, 64: Difference acquisition unit, 66: Update unit, 68: Encoding unit, 70 : Entropy encoder 72: Output unit 76: Code data output unit

Claims (10)

A JPEG compression apparatus that performs JPEG compression processing consisting of a plurality of processes on one or a plurality of input image data,
The input image data is acquired in accordance with identification information having at least information for identifying which of the input image data, and one or a plurality of block-unit image data that is a minimum processing unit of JPEG compression processing An output unit for outputting the image data and the identification information, and
For the block-unit image data output from the acquisition output unit, a process that can be performed independently of the plurality of processes is executed, and the block unit that has been subjected to the process is executed. An independent compression processing unit for outputting image data and the identification information;
For each of the image data output from the independent compression processing unit having the same identification information, the other processing is executed among the plurality of processes and output as JPEG compressed image data corresponding to the identification information. An independent compression processing unit;
Comprising
JPEG compression device. The JPEG compression apparatus according to claim 1,
The independent compression processing unit
A DCT transform unit for performing discrete cosine transform in units of blocks;
A quantization unit that performs quantization on each block unit generated by the discrete cosine transform;
A first output switching unit that switches and outputs each quantized component in accordance with the associated identification information;
With
The non-independent compression processing unit is:
A plurality of entropy-encoded outputs that perform entropy coding on each component of the block unit associated with the same identification information, and output the entropy-coded component of the block unit according to the identification information And
Comprising
JPEG compression device. The JPEG compression apparatus according to claim 1,
The independent compression processing unit
A DCT transform unit for performing discrete cosine transform in units of blocks;
A quantization unit that performs quantization on each block unit generated by the discrete cosine transform;
With
The non-independent compression processing unit is:
An entropy encoding unit that performs entropy encoding on each component of the block unit quantized by the quantization unit according to the associated identification information;
A second output switching unit that outputs each component of the block unit according to the identification information;
With
The entropy encoder is
A plurality of holding units that hold the leading component of each component of the block unit that was output immediately before each of the identification information;
Subtracting the head component held in the holding unit corresponding to the identification information associated with the block unit component group from the head component of the block unit component group, and outputting the difference of the head component An acquisition unit,
Entropy encoding is performed on the difference between the head component and a component other than the head component of the block-based component group.
JPEG compression device. The JPEG compression apparatus according to claim 2 or 3,
The independent compression processing unit
Furthermore,
A plurality of quantization tables according to the identification information;
A first selector that switches the quantization table according to the identification information and outputs information of the switched quantization table;
With
The quantization unit is
Performing the quantization using information in the quantization table output by the first selector;
JPEG compression device. A JPEG compression apparatus according to any one of claims 2 to 4,
The non-independent compression processing unit is:
Furthermore,
A plurality of entropy encoding tables according to the identification information;
A second selector that switches the entropy coding table according to the identification information and outputs information of the switched entropy coding table;
With
The entropy encoder is
Performing the entropy encoding using information in the entropy encoding table output by the second selector;
JPEG compression device. The JPEG compression apparatus according to claim 4,
The identification information further includes information on the compression rate of the JPEG compression process,
The quantization table has a number corresponding to information related to the compression rate of the JPEG compression process,
The first selector outputs information of the switched quantization table according to information on the compression rate of the JPEG compression process,
The quantization unit is
Performing the quantization using information in the quantization table output by the first selector;
JPEG compression device. The JPEG compression device according to claim 5,
The identification information further includes information on the compression rate of the JPEG compression process,
The entropy encoding table has a number corresponding to information related to the compression rate of the JPEG compression process,
The second selector outputs information of the switched entropy coding table according to the information regarding the compression rate of the JPEG compression process,
The entropy encoder is
Performing the entropy encoding using information in the entropy encoding table output by the second selector;
JPEG compression device. The JPEG compression apparatus according to any one of claims 1 to 7,
The identification information has processing conditions for the JPEG compression processing,
A plurality of JPEG compression processes with different processing conditions are performed on one input image data.
JPEG compression device. A JPEG compression device according to any one of claims 1 to 8,
Furthermore,
An input image data acquisition unit for acquiring the input image data and sequentially storing the acquired input image data in the storage unit;
With
The acquisition output unit
The input image data stored in the storage unit acquires image data for the block unit together with the identification information in response to the image data for at least the block unit being stored.
Image reading device. The image reading apparatus according to claim 9, wherein
The input image data acquisition unit
An image reading unit that reads at least a reading object and generates the input image data;
With
The JPEG compression device performs JPEG compression processing on the input image data,
The at least one input image data is the image data read by the image reading unit,
Image reading device.

Images