JP2013121154A - Image processing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve a problem that, in the case that a code amount exceeds an allowable range in encoding of attribute information related to each pixel, because a part of information is converted and encoded independently of characteristics of each attribute flag to improve a compression rate, reversibility is lost even in data that can maintain reversibility, and an effect of image processing based on the attribute information cannot be obtained.SOLUTION: By arranging for each attribute flag, encoding, and changing an encoding method of each attribute flag on the basis of a code amount of each attribute flag group, encoding with the minimum loss in reversibility can be executed.

Description

  The present invention relates to an image processing apparatus having a function for compressing and encoding image data within a predetermined code amount accompanying the image data.

  Conventionally, the JPEG method using discrete cosine transform and the method using Wavelet transform are often used as still image compression methods. Since this type of encoding method is a variable length encoding method, the amount of code changes for each image to be encoded.

  In the JPEG method, which is an international standardization method, only one set of quantization matrices can be defined for an image. Therefore, without pre-scanning, the code amount cannot be adjusted and used in a system that stores in a limited memory. Could cause memory over.

  In order to prevent this, in order to adjust the code amount by changing the compression rate and re-reading the original when the code amount exceeds the planned code amount, or by estimating the code amount by pre-scanning in advance , The method of resetting the quantization parameter was taken.

  On the other hand, image information includes attribute information associated with the image data in addition to the original image data. The attribute information is mainly used for color processing in the image output unit and adjustment of the number of gradations in order to improve the appearance at the time of image output. For natural images that contain both chromatic and achromatic colors and black characters that are often found in the original, changing the type of ink used in the same black makes the natural image look like a natural image while outputting clear characters. I can do it.

  As described above, each pixel has 1-bit attribute flag data indicating whether it is a chromatic color or an achromatic color and whether it is a character portion, thereby improving the image quality of the output image at the time of image output, particularly at the time of printout. I can do it. The attribute information includes information other than the above.

  In order to compress the image information, it is necessary to compress the attribute information as well as the image data. The attribute information is a collection of binary data, and in order to compress it, it is basically necessary to use a reversible encoding method. Conventionally, Packbits and JBIG encoding methods have been used for compression of attribute information.

  Here, if the attribute information is simply compressed by these encoding methods, the code amount may be larger than the code amount before encoding, and the code amount cannot be adjusted and stored in a limited memory. In the case of using it in the memory, there is a possibility that the memory is over, which is a big problem.

  With respect to this problem, when the code amount is large in the first encoding step, there is a method that converts the attribute information to be partially fixed and increases the compression rate by performing the recompression step to avoid memory over. (For example, refer to Patent Document 1).

Japanese Patent No. 3840076

  However, as in the prior art, in the method of converting so that the attribute flag is partially fixed when the code amount of all the attribute information exceeds a certain allowable amount, the compression rate is originally different for each attribute flag. By controlling the data compression method, the reversibility is lost even in data that can maintain reversibility. As a result, the areas that have been separated based on the attribute flags are regarded as the same, and image processing is performed. Therefore, when the image area separation is performed, the processing expected to improve the image quality cannot be performed, and when a re-encoding operation occurs, there is a problem that the desired image quality improvement cannot be expected by the original image area separation.

  An image processing apparatus (100) for inputting attribute information (305 to 311) of each pixel of multi-valued image data and compressing and encoding the dividing means (131, 132) for dividing the image information into arbitrary pixel blocks Attribute information arrangement means (601, 602) for arranging attribute information corresponding to each pixel of the pixel block for each attribute information, and encoding means (605) for lossless encoding using the arranged attribute information group ) And the amount of encoded data of each attribute information group obtained by the encoding means, and a code amount monitoring means (606) for determining whether or not the amount of encoded data of each attribute information group satisfies a predetermined standard ), And if all the encoded data amounts of each attribute information group satisfy the first criterion in the code amount monitoring means, whether each of the attribute information groups satisfies the second criterion Judgment and criteria The image processing apparatus characterized by encoding by changing the conditions for the attribute information group for which meets.

  Attribute information for each pixel block is arranged and encoded together for each attribute information, and the encoded data amount of all attribute information for each pixel block is monitored, and all encoded data amounts of attribute information are predetermined codes. When the encoded data amount is exceeded, the encoded data amount of each attribute information group is confirmed. As a result, by outputting the attribute information group that is larger than the pre-encoding data amount without encoding, the encoded data can be separated for each attribute information. Attribute information is used by maintaining reversibility without encoding into attribute information groups that have a larger amount of data after encoding, which is also a characteristic of lossless encoding, while using encoding technology. Therefore, it is possible to improve the original image quality.

1 is a block diagram showing an overall configuration of an image processing system according to the present embodiment. It is a block diagram which shows the software module which concerns on this Embodiment. It is an example which shows the data structure of 1 pixel which concerns on this Embodiment. It is a block diagram which shows the packet data which concern on this Embodiment. It is an internal block diagram of the image compression part 103 which concerns on this Embodiment. It is an internal block diagram of the attribute flag encoding unit 505 according to the present embodiment. It is an internal block diagram of the code amount calculating part 606 which concerns on this Embodiment. It is a flowchart of the encoding operation | movement of the attribute flag data which concerns on this Embodiment.

  The best mode for carrying out the present invention will be described below with reference to the drawings.

[Example]
FIG. 1 is a block diagram showing the overall configuration of the image processing system according to the present embodiment. In FIG. 1, an image forming apparatus 100 receives a scanner 101 that is an image input device and a printer that is an image output device via a printer image processing unit 119 that performs image processing on the printer side for received and processed data. An engine 102 is connected internally. Then, control for reading image data and printing is performed. Further, the image forming apparatus 100 is connected to the LAN 10 or the public line 104 to perform control for inputting / outputting image information and device information via the LAN 10.

  A CPU 105 is a central processing unit for controlling the image forming apparatus 100. A RAM 106 is a system work memory for the CPU 105 to operate, and is also an image memory for temporarily storing input image data. A ROM 107 is a boot ROM that stores a system boot program. An HDD 108 is a hard disk drive, and stores system software for various processes, input image data, and the like. The operation unit I / F 109 is an interface unit for the operation unit 110 having a display screen capable of displaying image data and the like, and outputs operation screen data to the operation unit 110. The operation unit I / F 109 serves to transmit information input by the operator from the operation unit 110 to the CPU 105. The network I / F 111 is realized by a LAN card, for example, and is connected to the LAN 10 to input / output information to / from an external device. Furthermore, the modem 112 is connected to the public line 104 and inputs / outputs information to / from an external device. The above units are arranged on the system bus 113.

  The image bus I / F 114 is an interface for connecting the system bus 113 and an image bus 115 that transfers image data at high speed, and is a bus bridge that converts a data structure. On the image bus 115, a RIP packet generation unit 131, which will be described in detail below, and a raster image processor (RIP) unit 116, a device I / F 117, a scanner image processing unit 118, and image processing for image editing are provided. A unit 120, an image compression unit 103, an image expansion unit 121, and a color management module (CMM) 130 are connected.

  A raster image processor (RIP) unit 116 expands a page description language (PDL) code into bitmap image data. The RIP packet generation unit 131 divides the bitmap image data output from the RIP unit 116 into rectangular tiles of a predetermined size, adds a header or the like, converts the data into a format described later, and outputs the packet as a packet. . The device I / F 117 connects the scanner 101 and the printer engine 102 via the scanner image processing unit 118 and the printer image processing unit 119, and performs synchronous / asynchronous conversion of image data.

  The scanner image processing unit 118 adds a header or the like from the scanner 101 to the scan packet generation unit 132, converts it to a predetermined format, and performs various corrections, processing, editing, and the like on the packetized input image data. Process. The image editing image processing unit 120 performs various types of image processing such as image data rotation, scaling, color processing, trimming / masking, binary conversion, multi-value conversion, and blank page determination. The image compression unit 103 encodes image data processed by the RIP unit 116, the scanner image processing unit 118, and the image editing image processing unit 120 by a predetermined compression method when the HDD 108 stores the image data once. The image decompression unit 121 once processes the image data compressed by the HDD 108 in the image editing image processing unit 120 or the image processing by the printer image processing unit 119 as necessary, and outputs it by the printer engine 102. Data that has been compressed and encoded is decoded and expanded.

  The printer image processing unit 119 performs image processing correction corresponding to the printer engine and video data count processing for toner control according to the present invention on image data to be printed out. The CMM 130 is a dedicated hardware module that performs color conversion processing (also referred to as color space conversion processing) on image data based on a profile or calibration data. A profile is information such as a function for converting color image data expressed in a device-dependent color space into a device-independent color space (eg, Lab). The calibration data is data for correcting the color reproduction characteristics of the scanner 101 and the printer engine 102.

  Each software module shown in FIG. 2 mainly operates on the CPU 105.

  A job control process 201 shown in FIG. 2 controls and controls each software module (not shown), and controls all jobs generated in the image forming apparatus 100 (not shown) such as copying, printing, scanning, and FAX transmission / reception.

  The network processing 202 is a module that controls communication with the outside, which is mainly performed via the network I / F 111, and performs communication control with each device on the LAN 10. When the network process 202 receives a control command or data from each device of the LAN 10, it notifies the job control process 201 of the contents. Also, based on an instruction from the job control process 201, a control command and data are transmitted to each device of the LAN 10.

  The UI process 203 mainly performs control related to the operation unit 110 and the operation unit I / F 109. The content of the operation of the operation unit 110 by the operator is notified to the job control process 201, and the display content of the display screen on the operation unit 110 is controlled based on an instruction from the job control process 201.

  The FAX process 204 controls the FAX function. The FAX processing 204 performs FAX reception via the modem 112, performs image processing unique to the FAX image, and notifies the job control processing 201 of the received image. Also, the designated image from the job control process 201 is faxed to the designated notification destination.

  The print processing 207 controls the image editing image processing unit 120, the printer image processing unit 119, and the printer engine 102 based on an instruction from the job control processing 201, and performs a specified image printing process. The print processing 207 receives image data, image information (image data size, color mode, resolution, etc.), layout information (offset, enlargement / reduction, imposition, etc.) and output paper information (size, print direction) from the job control processing 201. Etc.). Then, the image compression unit 103, the image expansion unit 121, the image editing image processing unit 120, and the printer image processing unit 119 are controlled to perform appropriate image processing on the image data, and the printer engine 102 is controlled and designated. Lets you print on paper.

  The scan process 210 controls the scanner 101 and the scanner image process 118 based on an instruction from the job control process 201 to read a document on the scanner 101. The instruction of the job control process 201 includes a color mode, and the scan process 210 performs a process according to the color mode. That is, if the color mode is color, the document is input as a color image, and if the color mode is monochrome, the document is input as a monochrome image. If the color mode is Auto, the color / monochrome determination of the document is performed by pre-scanning, and then the document is scanned again as an image based on the determination result and the image is input. A scan process 210 scans a document on a document table of the scanner 101 and inputs an image as digital data. The color information of the input image is notified to the job control process 201. Further, the scan processing 210 controls the scanner image processing 218 to perform appropriate image processing such as image compression on the input image, and then notifies the job control processing 201 of the input image that has undergone image processing.

  The color conversion process 209 performs a color conversion process on the instruction image based on an instruction from the job control process 201 and notifies the job control process 201 of the image after the color conversion process.

  The job control process 201 notifies the color conversion process 209 of input color space information, output color space information, and an image to which color conversion is applied. When the output color space notified to the color conversion process 209 is a color space (for example, Lab space) that does not depend on the input device, the input color space (for example, RGB) that depends on the input device is converted to Lab. The input profile information, which is the information of, is also notified. In this case, the color conversion processing 209 creates a lookup table (LUT) that maps from the input color space to the Lab space from the input profile, and performs color conversion of the input image using this LUT.

  When the input color space notified to the color conversion process 209 is a Lab space, output profile information for conversion from the Lab space to an output color space depending on the output device is also notified. In this case, the color conversion processing 209 creates an LUT that maps from the Lab color space to the output color space from the output profile, and performs color conversion of the input image using the LUT.

  If both the input color space and the output color space notified to the color conversion processing 209 are device-dependent color spaces, both the input profile and the output profile are notified. In this case, the color conversion processing 209 creates an LUT that directly maps from the input color space to the output color space from the input profile and the output profile, and performs color conversion of the input image using the LUT.

  In the color conversion process 209, if the CMM 130 is in the device, color conversion is performed using the CMM 130 by setting the generated LUT in the CMM 130. On the other hand, when there is no CMM 130, the CPU 205 performs color conversion processing in software.

  The RIP processing 211 interprets (interprets) a PDL (page description language) based on an instruction from the job control processing 201, controls the RIP unit 116, and renders it to develop a bitmap image. .

  With the above configuration, the image processing system receives a print job from the LAN 10, encodes image data and related attribute flag data, compresses the data, and sequentially stores the data in the HDD 108. To do.

  First, as described above, a PDL (page description original language) transmitted from a host computer (not shown) via the LAN 10 is received by the network I / F 111 and input to the RIP unit 116 from the image bus I / F 114. The The PDL data includes a line drawing command, a character drawing command composed of a typeface (font name) and size, a character code, and the like. The RIP unit 116 interprets (interprets) the transmitted PDL, converts it into code data that can be processed by the RIP unit 116, and executes rendering based on the code data.

  At this time, since the attribute of each pixel is determined by the rendering process, it is added to each pixel data as an attribute flag.

  FIG. 3 shows an example of the data structure of one pixel. In the present embodiment, one pixel is obtained by adding 8 bits using an attribute flag as a flag indicating the attribute of the pixel to image data of the three primary colors R (red), G (green), and B (blue). Yes.

  305 is a character flag indicating character / non-character, 306 is a chromatic color flag indicating chromatic / achromatic color, 307 is a line drawing flag indicating a line drawing, 308 is a job flag indicating whether it is a scan job or a PDL job, Reference numeral 309 denotes a flat flag indicating the type of a flat object such as a photograph or graphic with two bits of flat 0 and 1, and a processing 1 flag 310 and a processing 2 flag 311 as necessary.

  Next, the image data rendered by the RIP unit 116 and added with the attribute flag is divided into pieces of rectangular data, which will be described later, by the RIP packet generation unit 131, and further packet data is realized by adding header data to each image data. Is generated.

  FIG. 4 is a configuration diagram showing packet data in the present embodiment.

  The entire packet data configuration includes a header data portion 401, an image data portion 402, and an attribute flag data portion 403.

  The header data portion 401 includes page information in the job file, identification information of image types such as RGB, YMCK, and YUV, identification information whether or not encoded data, and the data size of the packet itself as header information. It is stored.

  The image data unit 402 is obtained by packing image data divided into arbitrary rectangular sizes as data. In the present embodiment, it is the RGB portion shown in FIG. 3 and stores data of 32 × 32 pixels as a rectangular size.

  The attribute flag data unit 403 stores the attribute flags 305 to 311 added for each pixel by the RIP unit 116 as described above.

  The packet data packetized by the RIP packet generation unit 131 in this way is then compression encoded by the image compression unit 103 via the image bus 115.

  FIG. 5 is a block diagram showing the configuration of the image compression unit 103.

  The input packet data is divided by the packet data dividing unit 501 into header data, image data, and attribute flag data. The image data is once buffered by the image data tile buffer 502 and then compressed and encoded by the image data encoding unit 504 using an irreversible compression method (JPEG method in the present embodiment). Since it is a JPEG compression method, the minimum encoding window size is 8x8 pixels, so in the case of a 32x32 pixel tile, there must be 4x4 (16) 8x8 pixels in it. Therefore, when encoding one tile, data is read from the image data tile buffer 502 in units of 8 × 8 pixels, and JPEG compression is performed 16 times. In this manner, the image data encoding unit 504 quantizes the 16 8 × 8 pixel windows included in the 32 × 32 pixel tile image data by performing known DCT (discrete cosine) transformation. Then, a Huffman code word is assigned to the quantized data, and the encoding of the image data is finished.

  On the other hand, the attribute flag data is buffered once in the attribute flag data tile buffer 503 and then compressed and encoded by the attribute flag data encoding unit 505 using a lossless compression method (PackBits method in this embodiment).

  FIG. 6 is a block diagram showing the configuration of the attribute flag data encoding unit 505 in the present embodiment.

  The attribute flag data encoding flow in the attribute flag data encoding unit 505 will be described below with reference to the flowchart of FIG.

  The attribute flag data (FlagDataIn [7: 0] in the figure) composed of the input 8 bits of each pixel is divided into attribute flags by the attribute information dividing unit 601 and each attribute flag buffer 602 [0] to [6]. ] (S1001). In this embodiment, there are seven types of attribute flags as described with reference to FIG. 3, and the attribute flags are stored in the attribute flag buffers 602 [0] to [6] in pixel order. A flat flag indicating the type of one flat object with 2 bits is extracted for each 2 bits of the flat flag portion, and stored in the corresponding attribute flag buffer 602 [4]. Therefore, in the present embodiment, the corresponding attribute flag buffer 602 [4] requires twice as much capacity as other attribute flag buffers.

  After all the attribute flag data is stored in the attribute flag buffer 602 [0] to [6], the attribute flag buffer read control unit 603 first reads the attribute flag data sequentially from the attribute flag buffer 602 [0]. Go. In this embodiment, the attribute flag data is read in the order of the attribute flag buffer 602 [0] to [6], so that the next-stage lossless compression unit 605 stores character flag data, chromatic color flag data, and line drawing flag data for one packet. , Job flag data, flat flag data, processing 1 flag data, and processing 2 flag data are input and encoded in this order (S1002).

  The attribute flag buffer read control unit 603 has a 7-bit signal width of the read status signal [6: 0] as an output signal, and which buffer of the attribute flag buffer 602 [0] to [6] is being read. Is output as a signal. In other words, for example, when reading from the attribute flag buffer [0], “0b0000001” is output with the first bit set to “1”, and when reading from the attribute flag buffer [6], the seventh bit is output. “0b1000000” is output as “1”.

  The lossless compression unit 605 is sequentially compressed by the PackBits encoding method as described above, and is temporarily stored in the subsequent-stage encoded data buffer 607 via the code amount control unit 606 described later.

  It is assumed that the encoded data buffer 607 has a capacity for attribute flag data before encoding in the attribute flag data portion in one packet data. Since the PackBits encoding method is one of the lossless compression methods as is well known, it may be larger than the data amount before encoding due to the attribute flag data. Therefore, when the code amount of the encoded attribute flag data from the lossless compression unit 605 exceeds the capacity of the encoded data buffer, that is, exceeds the data amount before encoding, the encoded data is not stored any more. And

  Next, the operation of the code amount control unit 606 will be described in detail.

  FIG. 7 is a block diagram showing an internal configuration of the code amount control unit 606.

  Each attribute flag data that is PackBits-encoded by the lossless compression unit 605 is counted by the code amount counter 701 for each attribute flag data (S1003). In the present embodiment, the attribute flag buffer 602 [0] to [6] in the order of character flag data, chromatic color flag data, line drawing flag data, job flag data, flat flag data, processing 1 flag data, and processing 2 flag data. Is read out, encoded, and the code amount is counted. Each time the count of the encoded data of each attribute flag data is completed, the result is sequentially stored in the corresponding code amount size registers 702 [0] to [6] in the figure. As described above, after all the attribute flag data in one packet data is encoded and stored in the encoded data buffer 607, the encoded data buffer 607 transmits the encoded buffer write end signal to the code amount calculation unit 705. To do. The code amount calculation unit 705 receives the encoding buffer write end signal, and detects that encoding has been completed (S1004). First, the code amount of the attribute flag data of one packet data is calculated. Specifically, the sum of the code amounts of the attribute flags of the code amount size registers 702 [0] to [6] is calculated (S1005). Now, assuming that the total value of the code amount is SUM, SUM is expressed by the following equation.

SUM = character flag code amount + chromatic color flag code amount + line drawing flag code amount + job flag code amount + flat flag code amount + processing 1 flag code amount + processing 2 flag code amount Next, the sum SUM of this code amount is actually It is compared with the value of the code amount threshold register 703 whether or not the compression rate is within the target.

  The code amount threshold register 703 stores a value corresponding to the target compression rate of the attribute flag data in one packet data. For example, the data amount before encoding of all attribute flag data of one packet data is 1024 bytes in total because there are 1024 pixels of 8-bit data per pixel. If the target compression rate is 1/2, a value indicating 512 bytes may be set in the register. This register is a register that can be set in advance by the CPU 105 via a register bus (not shown).

  The code amount threshold register 703 is compared with the sum SUM of the code amounts of the attribute flags as described above (S1006). If the sum SUM is smaller than the threshold value set in the code amount threshold register 703 (Yes in S1006), that is, if it falls within the preset target compression rate, one of the output signals of the code amount calculation unit 705 The encoded buffer read start signal is transmitted to the encoded data buffer 607. The encoded data buffer 607 receives the encoded buffer read start signal from the code amount calculation unit 705, reads the encoded attribute flag data from the buffer, and outputs it to the encoded packet data generation unit 506 at the subsequent stage. .

  Next, the code amount threshold register 703 is compared with the code amount total sum SUM of the previous attribute flag. If the sum SUM is larger than the threshold set in the code amount threshold register 703 (No in S1006), that is, set in advance. A case where the compression rate is not within the target will be described. First, each register of the code amount size register 702 is checked sequentially (S1007), and it is determined whether or not the code amount of each attribute flag data is larger than that before the code. For example, the code amount size register 702 [0] stores the encoded code amount of character flag data, and the character flag data amount of one packet data before encoding is a fixed value of 128 bytes. It can be judged by comparing whether it is larger. More specifically, the lossless compression unit 605 performs PackBits encoding by comparing the value of each register of the code amount size registers 702 [0] to [6] with the data amount of each flag data before encoding. The attribute flag that has become larger can be identified (S1008).

  The code amount calculation unit 705 converts the attribute flag whose code amount has increased by this encoding into the data amount before encoding (S1009), and calculates the sum SUM as described above.

  For example, if it is determined that the processing 1 flag data 310 and the processing 2 flag data 311 are larger than before the code amount, the expression of the sum SUM of the code amounts of the attribute flag data is converted as follows. (S1010).

SUM = character flag code amount + chromatic color flag code amount + line drawing flag code amount + job flag code amount + flat flag code amount + 128 bytes (data amount before processing 1 flag encoding)
+128 bytes (data amount before encoding of processing 2 flag)
In this way, the value of the calculated sum SUM is compared with the value of the code amount threshold register 703 as before (S1011).

  In this case, if the value of the sum SUM is smaller than the threshold value set in the code amount threshold register 703 (Yes in S1011), that is, if it falls within the preset target compression rate, the code amount calculation unit 705 The encoding mask signal [6: 0], which is one of the output signals, is changed and output. The encoding mask signal [6: 0] is a signal having a 7-bit width, and each corresponds to each attribute flag. As in the previous example, if it is determined that the processing 1 flag 310 and the processing 2 flag 311 have a large code amount after encoding, the sixth and seventh bits of the encoding mask signal are set to “1”. , “0b1100000” is output to the compression control unit 604. Further, an input buffer read start signal is output to the attribute flag buffer read control unit 603, the attribute flag data is read again from each attribute flag buffer 602 [0] to [6], and the attribute is sent to the lossless compression unit 605. Send flag data.

  At this time, the compression control unit 604 receives the two signals of the read status signal [6: 0] and the encoding mask signal [6: 0] described above as inputs, and controls the encoding enable signal which is an output signal. Then, whether or not the reversible compression unit 605 can perform the compression operation is determined. Specifically, when the encoding mask signal [6: 0] = “0b1100000” and the read status signal [6: 0] = “0b0100000” or “0b1000000” (the processing 1 flag 310 or the processing 2 flag 311 is read) The encoding enable signal is set to '0'. Therefore, when the lossless compression unit 605 reads the attribute flag buffers 602 [5] and [6], that is, when the processing 1 flag 310 or the processing 2 flag 311 is read and input to the lossless compression unit 605, the PackBits code The data is stored in the subsequent encoded data buffer 607 without performing the conversion operation (S1012). When the attribute flag data for one packet is stored in the encoded data buffer 607, the encoding of the attribute flag data is finished (S1013).

  As a result, attribute flag data that has become larger than before encoding due to partial encoding can be output as it is without being compressed, so that the target compression rate can be achieved.

  Next, the code amount control unit 606 compares the code amount threshold register 703 with the code amount total sum SUM of the attribute flag, and the sum SUM is larger than the threshold set in the code amount threshold register 703. As described above, a case will be described in which the amount of code of attribute flag data that has become larger than that before encoding is large even if it is replaced with the amount of data before encoding (No in S1011).

  In this case, a method of increasing the compression rate by fixing some attribute flag data to any value is adopted. The code amount calculation unit 705 is provided with priority setting registers 704 [0] to [6] that can set the priority of attribute flag data to be fixed in this method.

  This register can set the priority of the attribute flag to be fixed depending on the degree of influence on the image quality such as image processing. The priority setting register 704 [0] contains the identification code of the attribute flag having the highest priority. It is assumed that the identification code of the attribute flag with the lowest priority is set in the priority setting register 704 [6].

  For example, when the identification code indicating the processing 2 flag 311 is set in the attribute flag priority register 704 [6] indicating that the priority is the lowest in the priority setting register 704, all the processing 2 flags 311 are “0”. or Converted as a code amount fixed to '1'. Therefore, in this case, the processing 2 flag 311 is 1024 bits = 128 bytes for 1024 pixels. This becomes all “0”, and when PackBits is encoded, it becomes 2 bytes (S1014).

  Therefore, the sum SUM of the code amount when the above method is used can be calculated as follows, assuming that the processing 1 flag is already replaced with the data amount before encoding by the above method (S1015).

SUM = character flag code amount + chromatic color flag code amount + line drawing flag code amount + job flag code amount + flat flag code amount + 128 bytes (data amount before processing 1 flag encoding)
+2 bytes (data amount encoded with processing 2 flag as a fixed value)
This process is performed in accordance with the priorities of the sequential priority setting registers [0] to [6] until the total sum SUM of the code amounts becomes smaller than the value for reaching the target compression rate set in the code amount threshold register 703. The value of the sum SUM is calculated by replacing the corresponding attribute flag data with a fixed code amount.

  The process is terminated when a result that the sum SUM value does not exceed the value of the code amount threshold register 703 is finally calculated by such a process (S1016).

  When the processing is completed, the code amount calculation unit 705 reflects the information on the calculation result in the data mask signal [6: 0]. The data mask signal [6: 0] is an output control signal having a 7-bit width, and each bit corresponds to each attribute flag.

For example, when only the processing 2 flag 311 needs to be fixed, the data mask signal [6: 0] = “0b1000000” is output and the fixing of the data of the processing 2 flag 311 is controlled.
Further, in this case, “0b0100000” is output to the compression control unit 604 in order to mask only the processing 1 flag 310 in the previous encoding mask signal [6: 0].

  In this way, the data mask signal [6: 0] determined by the code amount calculation unit 705 is output to the attribute flag buffer read control unit 603 in the previous stage.

  Further, the code amount calculation unit 705 also outputs an input buffer read start signal to the attribute flag buffer read control unit 603, and again reads attribute flag data from each attribute flag buffer 602 [0] to [6], The attribute flag data is transmitted to the lossless compression unit 605.

  At this time, the attribute flag buffer read control unit 603 receives the data mask signal [6: 0] described above as an input, and when the attribute flag data of the corresponding bit is read from the attribute flag buffer 602, the attribute flag is read. The buffer is replaced with a fixed value and transmitted to the lossless compression unit 605. In the previous example, when the process 2 flag 311 is read from the attribute flag buffer 602 [6], the attribute flag buffer read control unit 603 always replaces the data with “0” or “1” and the lossless compression unit 605. (S1017). The lossless compression unit 605 performs PackBits encoding with this fixed value, and the previously calculated 2-byte code amount data is stored in the encoded data buffer 607 as encoded data of the process 2 flag. When the attribute flag data of one packet is stored in the encoded data buffer 607, the encoding of the attribute flag data is finished (S1018).

  As a result, in addition to the method of outputting attribute flag data that has become larger than before encoding without compression, attribute flag data with a lower priority for some image processing is also fixed and output. By encoding, it becomes possible to achieve the target compression rate.

  In the present embodiment, the code amount total SUM calculation process of the code amount calculation unit 705 has been described as being configured by the hardware of the code amount calculation unit. However, the CPU 105 uses the code amount size registers [0] to [0] to [6] may be confirmed, and the value of the sum SUM of code amounts may be calculated each time to control each control signal.

  Through the processing as described above, the encoded attribute flag data that finally falls within the target compression rate is output from the encoded data buffer 607 to the encoded packet data generation unit 506 in FIG. In the encoded packet data generation unit 506, one packet of encoded image data subjected to JPEG compression from the image data encoding unit 504, and PackBits compression from the attribute flag encoding unit 505 described above. Combines the encoded attribute flag data for one packet. Further, a part of the header data divided by the packet data division unit 501 at the input stage is rewritten to an appropriate value and added as header data, thereby generating encoded packet data. As a result, the packet data encoded by the image compression unit 103 is sequentially stored in the HDD 108 via the image bus I / F 114, and the compression process ends.

  In this embodiment, the RIP packet generation unit 131 divides the image data into 32 × 32 rectangular data and the image compression unit 103 compresses the image data. However, the image data of one page is compressed. Even in this case, by providing a register that can set the code amount for one page as the attribute flag code amount threshold register that has a large capacity for each buffer and has determined the code amount size according to the compression rate target. This method can be applied.

  As described above, by using the attribute flag data compression method as in the present embodiment, the attribute flag data can be targeted while maintaining the reversibility of the attribute data or with a minimum loss of attribute data information. It is possible to compress within the range of the compression rate.

103 ... Image compression unit 505 ... Attribute flag encoding unit 601 ... Attribute information division unit 602 ... Attribute flag buffer 606 ... Code amount control unit 701 ... Code amount counter 702 ... Code amount size register 703... Code amount threshold register 704... Priority setting register 705.

Claims (6)

  An image processing apparatus (100) for inputting attribute information (305 to 311) of each pixel of multi-valued image data and compressing and encoding the dividing means (131, 132) for dividing the image information into arbitrary pixel blocks Attribute information arrangement means (601, 602) for arranging attribute information corresponding to each pixel of the pixel block for each attribute information, and encoding means (605) for lossless encoding using the arranged attribute information group ) And the amount of encoded data of each attribute information group obtained by the encoding means, and a code amount monitoring means (606) for determining whether or not the amount of encoded data of each attribute information group satisfies a predetermined standard ), And if all the encoded data amounts of each attribute information group satisfy the first criterion in the code amount monitoring means, whether each of the attribute information groups satisfies the second criterion Judgment and criteria The image processing apparatus characterized by encoding by changing the conditions for the attribute information group for which meets.   The first criterion of the code amount monitoring means is a criterion based on a comparison with a threshold value indicating whether or not all code amounts of each encoded attribute information group are within a predetermined data amount. The first criterion for comparing the total amount of data of the attribute information group with the threshold value and determining that the total amount of data of the encoded attribute information groups is greater than the threshold value satisfies the first criterion The image processing apparatus according to claim 1, further comprising a determination unit.   The second criterion of the code amount monitoring means is a criterion based on a comparison with the data amount of each attribute information group provided in each attribute information group as a threshold when not encoded, and when this threshold is exceeded The image processing apparatus according to claim 1, further comprising a second reference determination unit that determines that the reference is satisfied.   The encoding means for the attribute information group that satisfies the first reference in the first reference determination means according to claim 2 and that satisfies the second reference in the second reference determination means according to claim 3, The image processing apparatus according to claim 1, further comprising an encoding control unit that outputs the attribute information group without encoding the attribute information group.   For the attribute information group that satisfies the first criterion of claim 2 and that satisfies the second criterion of claim 3, the encoding means converts the attribute information group into fixed information. The image processing apparatus according to claim 1, further comprising fixed output means for encoding and outputting. 6. The image processing apparatus according to claim 5, further comprising priority setting means for enabling setting of priority of attribute information to be converted into fixed information.