JP2009060593A - Image reader and image reading method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable processing of image data of a joint pixel in a multi-output type image sensor in relatively simple circuit arrangement without newly holding a line memory in an image reader. <P>SOLUTION: The image reader has an image sensor in which a plurality of sensor arrays for reading a document image are connected into a line, and which parallel outputs image data for each sensor array. Of the image data outputted from each sensor array, image data of a pixel at one end and image data of the remaining pixels are compressed by different compression methods. The compressed data of one line are decompressed by a set method in timing of reaching a preset number of pixels. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image reading apparatus that reads an image of a document and an image reading method thereof.

  A general image reading apparatus controls light sources such as LEDs to irradiate a document to be read with light, and photoelectrically converts reflected light from the document by a line-shaped image sensor composed of a plurality of elements to accumulate charges. To read the image. An image is read line by line while the line-shaped image sensor having a length corresponding to the width of the document is scanned with respect to the document. For the image data for one line obtained by reading the image for one line, each process is performed for each pixel from the image data of the pixel read by the element at one end of the sensor.

  FIG. 6 is a flowchart illustrating the flow of each process performed for each pixel of image data. First, image data in an image for one line read by the image sensor 116 is sequentially transferred to the A / D converter 112 serially for each pixel. Thereafter, it is converted into digital data and output to the reading control unit 107. The reading control unit 107 first packs the image data input from the A / D conversion unit 112 in the image data pack unit 201 into 16 bits per pixel image data. Thereafter, the shading correction unit 202 and the γ correction unit 203 perform shading correction and γ correction, respectively. The gamma correction is a process of inputting an image signal and correcting the level so that recording is performed at an appropriate density. Thereafter, the pixel cutout unit 204 cuts out only a specific image data range, and the image data compression unit 205 compresses the cutout image data. Thereafter, the data are sequentially stored in a RAM built in the image reading apparatus via the DMA controller 117.

  In recent years, it has been required to read images at high speed. For this reason, an image sensor capable of reading at high speed is required. This is a line image sensor having a length corresponding to the width of the document, and is composed of a plurality of chips. This sensor is a sensor different from the “single output type” image sensor that sequentially outputs serially the image data of the pixels read by the end element. This type of sensor is referred to as a “multi-output type” image sensor. The “multi-output type” image sensor performs reading at high speed by connecting a plurality of sensor chips in a line and outputting image data in parallel for each sensor chip. For example, by providing three sensor chips and outputting three pieces of image data for one line in parallel, a reading speed equivalent to three times that of a “single output type” image sensor can be achieved.

On the other hand, a technique for performing a different process by determining a compression method of compressed image data is disclosed (see Patent Document 1). Also, a decoding device that selects and outputs desired image data is known (see Patent Document 2). This decoding apparatus is configured to include means for discriminating image data when encoded image data that has been subjected to two different types of processing is sequentially input. In order to discriminate the image data type and change the subsequent processing, the discrimination circuit has a complicated configuration.
JP-A-6-086790 JP-A-7-298266

  Incidentally, the image data compression unit (encoding unit) 205 in FIG. 6 compresses the image data in order to reduce the capacity of the RAM for temporarily storing the image data. As this method, there is a method of compressing the image data compression unit 205 by adopting the DPCM encoding method, that is, a method of compressing the image data of the target pixel from 16 bits to 8 bits based on the previous pixel. Image data compression techniques include a modeling technique for modeling an image signal and an encoding technique that actually assigns a code to a signal sequence converted by this modeling technique. Modeling techniques include a run length model, a Markov model, differential coding (= differential pulse code modulation, hereinafter referred to as DPCM), and the like. On the other hand, Huffman codes, arithmetic codes, and the like are known as encoding techniques. Among these technologies, DPCM uses the property of image data that a certain pixel value has a high probability of taking a value close to the immediately preceding pixel value, and is used as a compression technique for image data. It is often used. However, in this compression method, the image data of the pixel first output from the image sensor for each line of the read image does not include the image data of the previous pixel. For this reason, compression processing by the DPCM encoding method based on the previous pixel is not performed, and compression is performed from 16-bit data to 16-bit data using higher-order 8-bit data from 16-bit data per pixel. That is, the data after compression processing [7: 0] = data before compression processing [15: 8]. The subsequent pixel image data is sequentially compressed by the DPCM encoding method and stored in the RAM sequentially.

  Thereafter, the compressed image data is sequentially read from the RAM via the DMA controller, and is sequentially expanded from 8 bits to 16 bits in the image data expansion unit in the image reading apparatus. Note that, as described above, the image data of the pixel output first from the image sensor is compressed using the upper 8 bits of the image data before compression. For this reason, when decompressing, decompression processing is performed by using the 8-bit data before decompression input to the decompression unit as lower 8-bit data in the decompressed output 16 bits. Since the image data of the pixels that are subsequently input to the decompression unit are compressed based on the previous pixel using the DPCM encoding method, the data is also based on the DPCM decoding method, that is, based on the previous pixel. Is expanded to 16 bits. That is, the expansion process is performed by changing the expansion method at the time of expansion depending on the image data of which pixel.

  However, in the “multi-output type” image sensor, one line is composed of a plurality of sensor chips. In the sensor having such a configuration, in order to compress and expand image data for one line, a line memory for one line must be held. That is, at the time of image reading, image data is output in parallel from each sensor chip constituting one line for the purpose of improving reading speed. Therefore, for the joint pixels of each sensor chip, the image data of the previous pixel is not already output from the sensor at the time of compression. This is because a line memory for holding image data for one line is necessary to perform the differential encoding.

  On the other hand, Patent Document 2 discloses a technology that includes a circuit that determines a compression method of compressed image data, and performs different processing according to the determination result. A circuit is provided that discriminates between inter-frame image data and intra-frame image data when inter-frame image data and intra-frame image data subjected to different compression processes are sequentially input. Here, a method of changing a processing method to be output at a timing determined as intra-frame image data is disclosed. The inter-frame image data is image data compressed by taking a difference between the previous frame and the current frame. The intra-frame image data is image data compressed in the current frame without taking the difference between the previous frame and the current frame. However, it is not preferable to have a means for discriminating between inter-frame image data and intra-frame image data because the circuit becomes complicated as a result and the cost increases.

  As described above, in an image reading apparatus using a multi-output type image sensor, image data for one line can be processed at high speed by a relatively simple circuit configuration without newly holding a line memory during compression. It was difficult. And there was nothing that specifically disclosed this.

  Accordingly, the present invention has been made in view of the above problems, and an object thereof is to provide an image reading apparatus capable of high-speed processing with a relatively simple circuit configuration without newly holding a line memory. To do. This image reading apparatus uses a multi-output type image sensor. Another object of the present invention is to provide an image reading method of the image reading apparatus.

In order to solve the above-described problems, the present invention includes an image sensor in which a plurality of sensors that read a document image are connected in a line, the document image is read line by line, and image data is output in parallel for each sensor. An image reading device,
Among the image data of the original image output from each of the plurality of sensors, the image data of one end side pixel read by each of the plurality of sensors is different from the image data of the other remaining pixels of the pixel. A compression means for compressing by a compression method;
Storage means for storing the image data compressed by the compression means;
Decompression means for decompressing the image data output from the storage means;
Counting means for counting the number of pixels input to the expansion means;
A pixel number setting means included in the decompression means,
When the number of pixels counted by the counting unit coincides with the number of pixels set by the pixel number setting unit, the expansion method of the image data by the expansion unit is changed.

In another aspect of the present invention for solving the above-described problems, a plurality of sensors that read a document image are connected in a line, the document image is read line by line, and image data is output in parallel for each sensor. An image reading method by an image reading apparatus provided with a sensor,
Out of the image data of the original image output from each of the plurality of sensors, the image data of the pixel on one end read by each of the plurality of sensors and the image data of the other remaining pixels of the pixel A compression step of compressing with different compression methods;
A storage step of storing the image data compressed by the compression step in a storage means;
Expanding the image data output from the storage means by expansion means;
A step of changing the expansion method of the image data by the expansion means when the number of pixels set in the expansion means matches the number of pixels counted by the counting means;
It is characterized by reading the original image.

  According to the present invention, it is possible to provide an image reading apparatus and an image reading method using a multi-output type image sensor having a relatively simple circuit configuration without newly holding a line memory for compression.

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

  Image data compression processing according to the present invention when an image of a document is read by a “multi-output type” image sensor in which three sensor chips are provided on one line and outputs from the three sensors are performed in parallel. The decompression process will be described with reference to FIGS. 1, 2A, and 2B.

  First, the one-line image sensor is composed of a leftmost sensor array 206, a central sensor array 207, and a rightmost sensor array 208, and reads an image of a target document. A sensor row means a sensor chip in which read pixels are arranged in a row. After reading, image data is sequentially output from each sensor array to the A / D converter 112 in parallel as shown in FIG. 2A. Image data from each sensor array is output serially. In the compression processing by the image data compression unit 205, the image data 400 of the pixel output first from the leftmost sensor array 206 does not include image data of the previous pixel (adjacent pixel). For this reason, the image data is compressed from 16 bits to 8 bits using the upper 8 bits of the 16 bits of data for one pixel. The image data for the subsequent pixels output from the sensor array 206 after the image data 400 is compressed by the DPCM encoding method.

  On the other hand, the image data 402 of the pixel output first from the central sensor array 207 should be compressed based on the image data 401 of the pixel output last from the sensor array 206 at the left end. Similarly, the pixel image data 404 output first from the rightmost sensor array 208 should be compressed based on the pixel image data 403 output last from the center sensor array 206. However, in order to be compressed in this way in the “multi-output type” image sensor, a line memory for one line must be held. For example, in order to compress the image data 402 of the sensor array 207 by DPCM compression, the image data 401 of the sensor array 206 is necessary. However, since the image data 401 is read earlier than the image data 402, a memory area for holding the image data 402 is required. Therefore, in order to perform the compression process without having such an unnecessary line memory, the image data 402 and 404 of the first pixel in the sensor array 207 and 208 are the same as the image data 400 of the first pixel in the sensor array 206. Perform compression. That is, the image data is compressed from 16 bits to 8 bits using the upper 8 bits of the 16 bits of image data for one pixel. The subsequent image data is compressed by the DPCM encoding method. In FIG. 2A, 405 is the last pixel data output from the sensor array 208.

  Next, the decompression process will be described with reference to FIGS. 1 and 2B.

  From the time the image data is compressed to 8 bits in the image data compression unit 205 to the time when the image data is written to the RAM via the DMA controller 117, the image data output from each sensor array is sequentially parallel as shown in FIG. It is processed. When image data is read from the RAM, for example, image processing performed later such as edge enhancement is considered, and image data for pixels for one line of the original image is associated with the arrangement order of the pixels for one line. Sorted and read. Needless to say, the amount of data is small because it is after compression, so even if the image data for one line of pixels is stored, the memory capacity is small.

  At the time of decompression, as shown in FIG. 2B, the image data for the pixels for one line of the original document image after compression is rearranged corresponding to the arrangement order of the pixels for one line, and then the image data is serially converted in that order. The data is input to the decompression unit and sequentially decompressed. In this decompression process, the image data 400 is the image data of the first pixel output from the sensor array 206. Accordingly, the input 8-bit image data is expanded to 16 bits as lower 8-bit image data in the expanded 16-bit image data. That is, data [15: 0] after decompression processing = data [7: 0] before decompression processing + data [7: 0] before decompression processing. The subsequent pixel image data is sequentially expanded to 16 bits by the DPCM decoding method.

  In the “single output type” image sensor, other than the image data of the first pixel may be expanded by the DPCM decoding method. However, in the “multi-output type” image sensor, the image data 402 of the first pixel output from the sensor array 207 is compressed based on the image data 401 of the last pixel output from the sensor array 206. is not. Similar to the image data 400 of the first pixel output from the sensor array 206, compression is performed based on the upper 8 bits of the 16 bits of image data. For this reason, the image data 402 is subjected to the same expansion process as the image data 400.

  Therefore, at the timing when the image data 402 is input to the decompression unit, the decompression process using the DPCM decoding method that has been performed is temporarily changed, and then the decompression process using the DPCM decoding method is performed again. The same applies to the image data 404. That is, when decompressing image data of pixels for one line after compression, in the decompression process in the middle of the image data, the decompression method of the image data of each pixel is changed depending on the compression method of the image data of each pixel.

  FIG. 3 is an overall configuration diagram of an image reading apparatus 101 to which the present invention can be applied.

  A main CPU 103 controls the entire system, and reads and executes a control program stored in the ROM 106 via the ROM control unit 105. Reference numeral 114 denotes a RAM serving as a storage unit for storing the work area of the CPU 103 and the read image data, and writing and reading are performed via the RAM control unit 110. A motor control unit 108 controls a motor 113 provided in the image reading apparatus. Reference numeral 107 denotes an image data reading control unit. The read control unit 107 sends a light source 115 on / off control signal to the light source current control unit 111, and the light source 115 is turned on / off based on the light on / off control signal at an appropriate current value via the light source current control unit 111. I do. Light is emitted from a light source 115 to a document to be read, and reflected light from the document is photoelectrically converted by a charge storage type image sensor 116 to store an image signal. The reading control unit 107 sends a line synchronization signal to the image sensor 116, and the image sensor 116 sequentially outputs image signals based on the line synchronization signal. The image signal output from the image sensor 116 is converted into digital data by the A / D converter 112 and sequentially transferred to the reading controller 107. The read image data is processed by the reading control unit 107 and stored in the RAM 114 via the DMA controller 117. Thereafter, the image data is sequentially read out from the RAM 114 via the DMA controller 118, subjected to image processing by the image processing unit 109, and sequentially transferred to the host PC 102 via the USB device I / F 104.

  Details of the reading control unit 107 will be described with reference to FIG. Image data read by the image sensor 116 divided into the three regions of the sensor arrays 206, 207, and 208 as shown in FIG. 2A is converted into digital data by the A / D converter 112. Thereafter, it is sent to the reading control unit 107. Further, the image data pack unit 201 packs the digital data serially input from the A / D conversion unit 112 into data of each pixel 16 bits. At the same time, image data output in parallel from each sensor array (sensor chip) is distributed to each data processing channel (0ch, 1ch, 2ch) in the reading control unit 107. The image data distributed to each channel is sequentially input to the shading correction unit 202 and the γ correction unit 203, and subjected to shading correction processing and γ correction processing, respectively. Thereafter, only the desired image data range is cut out by the pixel cutout unit 204, and the cut out image data is compressed by the image data compression unit (encoding unit) 205 in order to reduce the capacity of the RAM 114. The compressed image data is sequentially stored in the RAM 114 via the DMA controller 117.

  The compressed image data stored in the RAM 114 is sequentially read out via the DMA controller 118 in an order corresponding to the pixel arrangement order of one line of the document image, and is serially input to the image processing unit 109. In the image processing unit 109, the image data compressed by the image data compression unit 205 is sequentially expanded and then subjected to various image processing and transferred to the host PC 102.

  Here, the flow of the decompression process in the image processing unit 109 will be described in more detail with reference to FIG.

  First, the number of pixels is set in the internal register 301 of the image data decompression unit 300 in the image processing unit 109 (that is, the image processing unit 109 functions as a pixel number setting unit). Specifically, (1) the number of pixels from the leftmost sensor row 206 starting from the image data 400 of the first output pixel to the image data 402 of the first output pixel from the central sensor row 207 Is set as the first pixel number. Similarly, (2) the image data 400 of the first output pixel of the leftmost sensor array 206 is started, and the number of pixels from the rightmost sensor array 208 to the image data 404 of the first output pixel is set to 2 is set as the number of pixels. At the same time, (3) at the timing when the set first pixel number and second pixel number are reached, either the data expansion method is set to “change” or “do not change”. In this embodiment, the DPCM decoding method is set to be reset at that timing. The image data decompression unit 300 is equipped with a counter 302 that counts the number of serially input image data for each pixel, and counts the number of pixels from sequentially input image data.

  The pixel image data 400 that is first input to the internal register 301 is compressed from 16 bits to 8 bits using the upper 8 bits of the 16-bit image data before compression. Therefore, the 8-bit image data is first serially input from the RAM 114 to the image data expansion unit 300. The decompression calculation processing unit 303 decompresses the image data to 16-bit image data as lower-order 8-bit image data in the decompressed 16-bit image data. The image data inputted thereafter is sequentially decompressed by the decompression arithmetic processing unit 303 by the DPCM decoding method, decompressed to 16 bits, and output. When decompressing image data by the DPCM decoding method, the image data of the previous pixel is stored in the SRAM 304 and read out each time to perform DPCM decoding processing. The counter 302 sequentially counts the number of pixels from the input image data at the same time as the expansion calculation processing unit 303 performs the expansion processing.

  A comparison circuit 305 is mounted in the image data decompression unit 300 and compares the number of pixels counted by the counter 302 with the first number of pixels set in the internal register 301. If the two match, the coincidence detection signal 306 is output at the coincidence timing. When the internal register 301 is set to reset the DPCM decoding, a logical operation with the coincidence detection signal 306 is performed. Thereafter, a reset signal 307 for resetting decoding by the DPCM decoding method is output to the decompression arithmetic processing unit 303. At the timing when the reset signal 307 is output, the expansion arithmetic processing unit 303 resets the decoding by the DPCM decoding method. That is, decompression processing that is not based on the DPCM decoding method is performed on the image data 402 input to the image data decompression unit 300 at that timing. Specifically, the decompression is performed by a method of decompressing the 16-bit image data as the lower 8-bit image data in the decompressed 16-bit image data. The image data for the subsequent pixels is sequentially expanded again by the DPCM decoding method and expanded to 16 bits.

  Similarly, the number of pixels counted by the counter 302 is compared with the second number of pixels set in the internal register 301. If the two match, the coincidence detection signal 308 is output at the coincidence timing. When the internal register 301 is set to reset the decoding by the DPCM decoding method, a logical operation with the coincidence detection signal 308 is performed. Thereafter, a reset signal 307 for resetting decoding by the DPCM decoding method is output to the decompression arithmetic processing unit 303. At the timing when the reset signal 307 is output, the expansion arithmetic processing unit 303 resets the decoding by the DPCM decoding method. The image data 404 input to the image data expansion unit 300 at that timing is subjected to expansion processing that is not based on the DPCM decoding method. Specifically, the decompression is performed by a method of decompressing the 16-bit image data as the lower 8-bit image data in the decompressed 16-bit image data. The image data for the subsequent pixels is sequentially expanded again by the DPCM decoding method and expanded to 16 bits.

  When all the compressed image data of pixels for one line is expanded and the next compressed image data of pixels for one line is input to the image data expansion unit 300, the counter 302 is cleared. Newly input image data of pixels for one line is also sequentially expanded as described above.

  Hereinafter, an example of the image reading method of the present invention will be described with reference to the flowchart of FIG.

  First, in step S110, the image data of one line of document image output from each sensor array is compressed by different compression methods depending on the arrangement order of pixels of one line of the document image. Specifically, the image data of the pixel at one end (one end side) read by each sensor array and the image data of the remaining pixels other than the pixel are compressed by different compression methods. For example, the image data of the pixel on one end side of the sensor array is compressed based on the image data of the pixel, and the image data of the remaining pixel is compressed based on the image data of the adjacent pixel read first. Next, in step S120, the compressed image data is stored in the memory. Next, in step S130, the image data is read out serially from the memory corresponding to the arrangement order of pixels for one line of the original image with respect to the compressed image data. Next, in step S140, the number of pixels corresponding to the image data read in step S130 is counted. In step S150, the read image data is decompressed by a plurality of decompression methods corresponding to different compression methods in accordance with the number of pixels counted in step S140. This is repeated until the reading of the document image to be read is completed (step S160).

  As described above, in the configuration of the image reading apparatus according to the present invention, it is possible to change the expansion method by a designated method at a designated timing using a multi-output type image sensor. It is a simple circuit configuration in which only counting by the counter and selection of the set processing method is selected, and compression processing and decompression processing can be performed without newly providing a data holding line memory at the time of data compression. Further, by making it possible to arbitrarily set the number of pixels to be reset by decoding using the DPCM decoding method, it is possible to apply regardless of the specifications of the image sensor. Further, the present invention can be applied even when the number of pixels to be processed is different each time, for example, when a process of extracting a desired image range from the original image data is performed. Furthermore, by making it possible to set a plurality of pixels to be reset by decoding using the DPCM method, the present invention can be applied to an image sensor that performs more than three parallel outputs. (An image sensor or the like that performs 5 parallel outputs and 6 parallel outputs). In addition, by making it possible for the user to set whether or not to reset the expansion method (expansion method can be selected), for image sensors that perform fewer than three parallel outputs (image sensors that perform two parallel outputs, etc.) However, it can be applied.

It is a figure which shows the flow of the image data in the reading control part in the image reading apparatus of this invention. It is a figure which shows the flow of the compression process of the image data by a multitype output sensor. It is a figure which shows the flow of the expansion process of the image data by a multitype output sensor. 1 is an overall configuration diagram of an image reading apparatus that is an embodiment of the present invention. 1 is a configuration diagram of an image data expansion unit inside an image reading apparatus that is an embodiment of the present invention. FIG. It is a flowchart which shows an example of the image reading method of this invention. It is a figure which shows the flow of the image data in the reading control part in the conventional image reading apparatus.

Explanation of symbols

101 Image Reading Device 107 Reading Control Unit 109 Image Processing Unit 114 RAM
116 Image sensor 206 Sensor array 207 Sensor array 208 Sensor array

Claims (8)

A plurality of sensors that read a document image are connected in a line shape, the image reading device includes an image sensor that reads the document image line by line and outputs image data in parallel for each sensor,
Among the image data of the original image output from each of the plurality of sensors, the image data of one end side pixel read by each of the plurality of sensors is different from the image data of the other remaining pixels of the pixel. A compression means for compressing by a compression method;
Storage means for storing the image data compressed by the compression means;
Decompression means for decompressing the image data output from the storage means;
Counting means for counting the number of pixels input to the expansion means;
A pixel number setting means included in the decompression means,
The image reading apparatus according to claim 1, wherein when the number of pixels counted by the counting unit matches the number of pixels set by the pixel number setting unit, the expansion method of the image data by the expansion unit is changed. The image processing apparatus further includes a control unit that reads out image data for each of the plurality of sensors stored in the storage unit in an order corresponding to an arrangement order of pixels for one line of the document image and serially inputs the data to the compression unit. The image reading apparatus according to claim 1, wherein the image reading apparatus is an image reading apparatus. The compression means compresses image data of one end side of each of the plurality of sensors based on the image data of the pixel, and image data of the remaining pixel is based on image data of an adjacent pixel that is read first. The image reading apparatus according to claim 1, wherein the image reading apparatus is compressed. The image reading apparatus according to claim 2, wherein the decompression unit includes a decompression method selection unit that can set a plurality of decompression units and can select whether or not to select each decompression method. The image reading apparatus according to claim 1, wherein the number of pixels set in the pixel number setting unit is a number of pixels based on a first pixel read by each of the plurality of sensors. The expansion means expands the image data of the pixel on one end side of each of the plurality of sensors based on the image data of the pixel, and the image data of the remaining pixel is based on the image data of the adjacent pixel that is read first. The image reading apparatus according to claim 1, wherein the image reading apparatus is expanded. When the number of pixels counted by the counting unit matches the number of pixels set by the pixel number setting unit, it is possible to select whether or not to change the expansion method of the image data by the expansion unit. The image reading apparatus according to claim 1, wherein the image reading apparatus is an image reading apparatus. An image reading method by an image reading apparatus including an image sensor that includes a plurality of sensors that read document images connected in a line, reads the document image line by line, and outputs image data in parallel for each sensor,
Among the image data of the original image output from each of the plurality of sensors, the image data of one end side pixel read by each of the plurality of sensors is different from the image data of the other remaining pixels of the pixel. A compression step of compressing by a compression method;
A storage step of storing the image data compressed by the compression step in a storage means;
Expanding the image data output from the storage means by expansion means;
A step of changing the expansion method of the image data by the expansion means when the number of pixels set in the expansion means matches the number of pixels counted by the counting means;
An image reading method comprising reading an original image.

Images