JP2009284190A - Image reading apparatus and image reading system - Google Patents

Abstract

Image data obtained by correcting an inclination of a document image area and extracting only an area including an image of a document without mounting a large-capacity memory on the image reading apparatus side is input.
A line image sensor for reading an image of a document, a first image buffer memory for temporarily storing image data for each predetermined line corresponding to a predetermined length in the sub-scanning direction, and a first image buffer memory 11 detects the position, angle, and width of one side of the document image from image data of a predetermined length in the sub-scanning direction at the leading edge of the document stored in 11, and the inclination of the image data based on the detected data. And the second image buffer memory 14 that temporarily stores the image data processed by the ASIC 13 until the image data processed by the ASIC 13 is transferred to the PC 2. Have.
[Selection] Figure 1

Description

  The present invention relates to an image reading apparatus and an image reading system for reading and inputting a document image.

  When a plurality of originals are continuously supplied to the reading device by the auto sheet feeder and read, the originals may be supplied obliquely to the reading device. In this case, since the document image portion included in the read image data is tilted, it is necessary to correct the document image portion so that it is not tilted. Conventionally, this inclination correction corrects the inclination by extracting an edge (contour) of a document image included in image data, or exists in image information as described in, for example, (Patent Document 1). A feature in the horizontal direction is extracted, the inclination of the image information is automatically detected from the feature, and the image information is read based on a value for correcting the detected inclination.

  However, the above-described conventional method can be used only when the edge of the document image is clear or the document image has characteristics, and the document to be read is the color behind the document reading unit of the reading device. If there is not much difference from the above, or if the image has no features, the inclination of the document cannot be extracted.

  Therefore, as described in (Patent Document 2), the present applicant inputs multivalued image data including an original image read by an image reading apparatus to a computer, and the contour enhancement of the multivalued image data is performed on this computer. Processing, binarizing the multivalued image data after the edge emphasis processing, detecting the inclination of the binarized image data, and from the binarized image data based on the detected inclination of the document image area An image input system that corrects the inclination of the original image area and detects the original image area based on the histograms of the vertical and horizontal directions of the image data after the inclination correction has been proposed.

However, if the multi-value image data read by the image reading device is processed on a general-purpose computer in this way, the amount of data read by the image reading device becomes very large, and therefore, the calculation processing on the computer takes a very long time. It will hang. Therefore, as described in, for example, (Patent Document 3), it is considered that high-speed processing can be performed if image data processing is performed on the image reading apparatus side instead of on a computer.
Japanese Patent Laid-Open No. 2-199959 JP 2007-1889577 A JP 2003-219085 A (paragraph 0036)

  However, since the amount of multi-value image data is enormous, a large-capacity memory is mounted on the image reading apparatus for arithmetic processing. Since such a large-capacity memory is very expensive, if it is installed in an image reading apparatus, it causes a high cost.

  Therefore, in the present invention, image data obtained by correcting the inclination of the document image area on the image reading device and extracting only the region including the image of the document without mounting a large-capacity memory on the image reading device side. An object of the present invention is to provide an image reading apparatus and an image reading system capable of inputting at high speed.

  In order to solve the above-described problems, the present invention provides a line image sensor that reads an image of a document in a predetermined line unit in the main scanning direction, and a predetermined line that is read by the line image sensor while moving the document or the line image sensor in the sub-scanning direction. First buffer means for temporarily storing image data corresponding to a predetermined length in the sub-scanning direction, and an image for a predetermined length in the sub-scanning direction at the leading edge of the document stored in the first buffer means Image analysis means for detecting the position, angle and width of one side of the document image from the data, and the first buffer based on the position, angle and width of one side of the document image detected by the image analysis means The image processing means for correcting the inclination of the image data stored in the means and extracting only the area of the image data containing the original image; Second buffer means for temporarily storing the image data processed by the processing means until the image data is transferred to the host computer, and the first buffer means is located at one end side of the original image by the image analysis means. After detecting the angle and width, the image data for each predetermined line read by the line image sensor is overwritten in the area of the image data processed by the image processing means and stored in the second buffer means. It is characterized by.

  According to the present invention, image data obtained by correcting the inclination of the document image area on the image reading device and extracting only the region including the image of the document without mounting a large-capacity memory on the image reading device side can be obtained. Can be entered.

  According to a first aspect of the present invention, a line image sensor that reads an image of a document in units of a predetermined line in the main scanning direction, and an image for each predetermined line that is read by the line image sensor while moving the document or the line image sensor in the sub-scanning direction. A first buffer means for temporarily storing data corresponding to a predetermined length in the sub-scanning direction, and an original from image data for a predetermined length in the sub-scanning direction at the leading edge of the original stored in the first buffer means The image analysis means for detecting the position, angle and width of one side of the image and the first buffer means based on the position, angle and width of the one side of the original image detected by the image analysis means. The image processing means for correcting the inclination of the image data and extracting only the area containing the original image from the image data, and the image processing means Second buffer means for temporarily storing the processed image data until the processed image data is transferred to the host computer. The first buffer means has a position, an angle and a position of one end side of the image of the document by the image analysis means. After the width is detected, the image data is processed by the image processing means, and the image data for each predetermined line read by the line image sensor is overwritten in the image data area stored in the second buffer means. An image reading apparatus. According to the present invention, the position, angle, and width of one end of an image of a document are detected from image data for a predetermined length in the sub-scanning direction at the leading edge of the document, and the position of one end of the detected image of the document is detected. The image data read by the line image sensor is corrected based on the angle and width, and can be transferred to the host computer sequentially while extracting only the area containing the original image. Without mounting a memory of a capacity, it is possible to correct the inclination of the original image area on the image reading apparatus and input image data obtained by extracting only the area including the original image at high speed.

  According to a second invention of the present application, the image analysis unit is configured to cause each position of all pixels in the main scanning direction of the image data corresponding to a predetermined length in the sub-scanning direction at the front end of the document stored in the first buffer unit (hereinafter referred to as the following). , The position where the change point of the pixel in the sub-scanning direction from the leading edge of the document (hereinafter referred to as “y coordinate”) is detected, and the range where the change point is detected. Is divided into a plurality of sections, and the angle of each section is calculated. When the sections having substantially the same angle continue for a predetermined number of sections or more, the average value of the substantially same angles is calculated as the angle of one end of the document image. An image reading apparatus. According to the present invention, one end side of an image of a document is divided into a plurality of sections and the angle is detected. This is a case where a part of one end side of an image of a document is missing or rough. However, the inclination of the document image area can be corrected more accurately.

  According to a third aspect of the present invention, the image processing means stores the image data in the second buffer means without processing the image data when the sections having substantially the same angle do not continue for a predetermined number of sections or more. This is an image reading apparatus. According to the present invention, image data is not processed when an angle cannot be detected from one end of an original image because a part of one end of the original image is missing or rough. Since the image data is transferred to the host computer as it is, it is possible to prevent the image reading apparatus from erroneously correcting the inclination and erroneously extracting the document image area.

  According to a fourth invention of the present application, the image analyzing means detects the x coordinate of the change point of each pixel at the front end portion and the rear end portion in the main scanning direction with respect to all the y coordinates, and the front end in the main scanning direction. Detected when the angles of the leading edge and the trailing edge in the main scanning direction are substantially perpendicular to the angle of one edge of the detected document image. This is an image reading apparatus in which the distance between the change points of the respective pixels at the front end portion and the rear end portion in the main scanning direction is set as the width of one end side of the image of the document. According to the present invention, a document image region can be extracted only when the document is not bent or the like and is substantially perpendicular to the angle of one side of the document image.

  According to a fifth aspect of the present application, the image processing means calculates the image data when the calculated angles of the leading edge and the trailing edge in the main scanning direction are not substantially perpendicular to the angle of one edge of the detected document image. This is an image reading device for storing image data in the second buffer means without performing processing. According to the present invention, if the detected angle of the original image is not substantially perpendicular to the angle of one end of the original image, it is determined that the original is broken, and the image data is processed without being processed. Since the data is transferred to the host computer, it is possible to prevent the document image area from being erroneously extracted by the image reading apparatus.

  A sixth invention of the present application is an image reading system including any of the image reading devices described above and a host computer to which the image reading device is connected. According to the present invention, the position, angle, and width of one end of an image of a document are detected from image data for a predetermined length in the sub-scanning direction at the leading edge of the document, and the position of one end of the detected image of the document is detected. The image data read by the line image sensor is corrected based on the angle and width, and can be transferred to the host computer sequentially while extracting only the area containing the original image. Without mounting a memory of a capacity, it is possible to correct the inclination of the original image area on the image reading apparatus and input image data obtained by extracting only the area including the original image at high speed.

  A seventh invention of the present application includes the image reading device according to the third or fifth invention and a host computer to which the image reading device is connected. The host computer does not process image data. When the image data is transferred to the host computer, the image data is processed on the host computer. According to the present invention, when image data is not processed and transferred to the host computer as it is, the angle and width of the original image data can be changed on the host computer having a large-capacity memory. By detecting and processing the image data, it is possible to reliably correct the inclination of the document image area and extract only the area including the image of the document.

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

(Embodiment)
FIG. 1 is a configuration diagram of an image reading system according to an embodiment of the present invention. 1, an image reading system according to an embodiment of the present invention includes a scanner 1 as an image reading apparatus, and a personal computer (hereinafter referred to as “PC”) 2 as a host computer to which the scanner 1 is connected. Consists of

  The scanner 1 includes a line image sensor 10 that reads an image of a document in full color in units of predetermined lines in the main scanning direction, and a first image buffer memory that serves as a first buffer unit that temporarily stores image data read by the line image sensor 10. 11, a CPU (Central Processing Unit) 12 as image analysis means for detecting the position, angle and width of one side of the original image, and correction of the inclination of the image data (hereinafter referred to as “descue”). ASIC (Application Specific Integrated Circuit) 13 as image processing means for extracting only the area including the image of the original document (hereinafter referred to as “crop”) from the image data, and the image data processed by the ASIC 13 to the PC 2. Roll The second image buffer memory 14 temporarily stores the data until it is sent.

  The line image sensor 10 includes a general solid-state imaging device such as an optical reduction type CCD (Charge Coupled Device) and a contact sensor type CIS (Contact Image Sensor). Although not shown, a reading table colored in black is arranged facing the line image sensor 10. The reason why the reading table is colored black is that the original to be read is generally a white original in many cases, so that the difference between the white original and the background color appears.

  When the scanner 1 is a flatbed type, the document is fixed on a transparent document table, so that the line image sensor 10 moves in the sub-scanning direction while the image of the document is moved in a predetermined line unit in the main scanning direction. Read with. When the scanner 1 is a sheet feed type, the line image sensor 10 is fixed, so that the line image sensor 10 moves the document image in the main scanning direction in the main scanning direction while moving the document in the sub scanning direction by the transport roller. Read line by line. In the line image sensor 10 according to the present embodiment, reading is performed in units of one line in the main scanning direction, but a configuration in which reading is performed in units of a plurality of lines, for example, in units of three lines may be possible.

  The first image buffer memory 11 corresponds to a predetermined length in the sub-scanning direction for the image data for each line read by the line image sensor 10 while moving the document or the line image sensor 10 in the sub-scanning direction as described above. Temporarily store as much as you want.

  FIG. 2 is an explanatory diagram showing a predetermined length in the sub-scanning direction of image data stored in the first image buffer memory. For example, as shown in FIG. 2, the first image buffer memory 11 has a capacity for storing data for a line corresponding to 50 mm at the front end of the document image F as a predetermined length in the sub-scanning direction.

  For example, when the width in the main scanning direction read by the line image sensor 10 is 302 mm and the reading resolution of the line image sensor 10 is 600 dpi × 600 dpi, the number of pixels in the main scanning direction is 302 ÷ 25.4 × 600 = 7133 dots, sub-scanning Since the length in the direction (number of lines) is 50 / 25.4 × 600 = 11181, the capacity of the first image buffer memory 11 is 7133 dots × 1181 lines × 3 (each RGB color) ≈24 Mbytes. Conventionally, when reading the entire surface of the JIS standard A4 size (length: 297 mm), the upper and lower margins are each 18 mm, and 7133 dots × 7866 lines × 3 (each RGB color) ≈160 Mbytes are required.

  The CPU 12 detects the position, angle, and width of one side of the document image from image data of a predetermined length in the sub-scanning direction at the leading edge of the document stored in the first image buffer memory 11 described above.

  FIG. 3 is an explanatory diagram for detecting the angle of one end of the original image, and FIG. 4 is an explanatory diagram for detecting the width of one end of the original image. First, regarding the angle detection, the CPU 12 detects all the pixels in the main scanning direction of image data for a predetermined length in the sub-scanning direction at the leading edge of the document stored in the first image buffer memory 11 as shown in FIG. For each position (x coordinate), a position (y coordinate) from the leading edge of the document at a pixel change point (a point changing from black to white in the illustrated example) in the sub-scanning direction is detected. Next, the range W where the change point is detected is divided into a plurality of sections (for example, equally divided into 10), and the angle of each section is calculated. At this time, the angle is obtained by linearly approximating the coordinates of the changing point by the method of least squares.

  When the angle between adjacent sections of the plurality of sections is within ± 1 °, for example, it is determined that the sections have the same inclination. When the sections having the same inclination continue for five or more sections, the average value is calculated as one end of the document. The angle of the side. That is, when the sections having substantially the same angle continue for a predetermined number of sections or more, the CPU 12 sets the average value of the substantially same angles as the angle of one end side of the document image. Note that when the sections having substantially the same angle do not continue for a predetermined number of sections or more, the ASIC 13 stores the image data as it is in the second image buffer memory 14 without performing image data processing described later.

  Next, with respect to the width detection, the CPU 12 performs the left direction with respect to all the y coordinates of the image data for a predetermined length at the leading edge of the document stored in the first image buffer memory 11 as shown in FIG. And the x coordinate of the change point of the pixel from the right direction is detected, that is, the x coordinate of the change point of each pixel at the front end portion and the rear end portion in the main scanning direction. Is detected. And the change point detected here is rotated by the angle calculated | required by the detection of the above-mentioned angle (refer FIG. 4).

Then, the CPU 12 obtains the respective angles of the front end portion and the rear end portion in the main scanning direction by linear approximation of the coordinates of the change point by the least square method. B ₁ and B _{2 in} FIG. 4 indicate ranges of change points of the front end and the rear end in the main scanning direction. The CPU 12 detects the angle of the calculated front end and rear end in the main scanning direction at substantially right angles (for example, within 90 ° ± 1 °) with respect to the main scanning direction, that is, by detecting the aforementioned angle. When the angle is substantially perpendicular to the angle of one end of the original image, the difference between the x-coordinates of the detected change points of the pixels at the front and rear ends in the main scanning direction, that is, the change points of the respective pixels. The distance between them is defined as the width of one side of the original image. If the calculated angles of the leading end and the trailing end in the main scanning direction are not substantially right angles, the ASIC 13 stores the image data as it is in the second image buffer memory 14 without processing image data described later.

  Note that the position of one side of the original image can be easily detected if these angles and widths are detected.

  The ASIC 13 corrects (descends) the inclination of the image data stored in the first image buffer memory 11 based on the position, angle, and width of one end of the original image detected by the CPU 12, and Only the area including the image of the original is extracted (crop). Since the descuing process and the cropping process can be performed by a known method, detailed description thereof is omitted here.

  The second image buffer memory 14 temporarily stores the image data processed by the ASIC 13 until it is transferred to the PC 2. Since the second image buffer memory 14 is only used for communication processing when image data is transferred from the scanner 1 in response to a transfer request from the PC 2, the second image buffer memory 14 may have a very small capacity relative to the first image buffer memory 11. .

  FIG. 5A is a diagram showing a stored image in the first image buffer memory, and FIG. 5B is a diagram showing a stored image in the second image buffer memory. As shown in FIG. 5B, the lines processed by the ASIC 13 are sequentially stored in the second image buffer memory 14, and the image data stored in the second image buffer memory 14 is sequentially transferred to the PC 2. It will be done. In FIG. 5B, the image data to be stored seems to gradually increase, but in reality, it is transferred to the PC 2 almost simultaneously with the storage, and the amount to be stored is small.

  On the other hand, in the first image buffer memory 11, the CPU 12 detects the position, angle, and width of one side of the original image, and after the processing by the ASIC 13, the processed image data is not necessary. As shown in a), the image data for each predetermined line read by the line image sensor 10 is overwritten one after another on the original area of the image data that has been processed and stored in the second image buffer memory 14.

  Next, image reading processing by the image reading system having the above-described configuration will be described. FIG. 6 is a sequence diagram of the image reading system according to the embodiment of the present invention, FIG. 7A is a diagram showing an example in which the leading end of the document is bent, and FIG. 7B is an example in which the leading end of the document is rough. FIG.

  When the scan conditions of the scanner 1 are set from the PC 2 and the start of scanning is instructed, the scanner 1 starts scanning processing.

  Here, the scanner 1 temporarily stores the image data captured by the line image sensor 10 in the first image buffer memory 11 as described above, and the CPU 12 analyzes the image data to determine the position, angle, and width of the document. Is detected and the information is set in the ASIC 13. Then, the ASIC 13 executes a deskew process and a crop process based on the set information. The image data processed by the ASIC 13 is stored in the second image buffer memory 14.

  The PC 2 acquires the size (width) of the document image from the scanner 1 and information related to the deskew process and the crop process, and sequentially acquires image data from the second image buffer memory 14. On the other hand, in the scanner 1, when the CPU 12 continues to analyze the image data in the first image buffer memory 11 and detects the trailing edge of the document, information on the size (length) of the document image is set in the ASIC 13. Finally, the PC 2 acquires the size (length) of the document image.

  It should be noted that the scanner 1 has an angle of one side of an image of the document, as shown in FIG. 7A, where the leading edge of the document is bent or rough as shown in FIG. 7B. When the width cannot be detected, the image data is stored in the second image buffer memory 11 as it is without performing the deskew process and the crop process as described above. When the deskew process and the crop process are not performed in this way, the PC 2 performs the deskew process and the crop process by the PC 2 itself.

  As described above, in the image reading system according to the present embodiment, the position, angle, and width of one side of the document image are detected from the image data for a predetermined length in the sub-scanning direction at the leading edge of the document, and this detection is performed. The image data read by the line image sensor 10 is corrected (descued) based on the position, angle and width of one side of the original image, and only the area containing the original image is extracted (crop) sequentially. Since it is transferred to the PC 2, it is not necessary to mount a large-capacity memory on the scanner 1 side, and the image data obtained by correcting the inclination of the original image area at high speed by the ASIC 13 of the scanner 1 and extracting only the area including the original image. Can be input to the PC 1.

  Also, in this image reading system, one side of the original image is divided into a plurality of sections and the angle is detected, so that a part of one side of the original image is missing or rough However, since the average value of the substantially identical angles is processed as the angle of one side of the document image only when there are more than a predetermined number of sections having substantially the same angle, it is more accurate. It is possible to correct the inclination of the document image area.

  If the angle and width cannot be detected from one side of the original image, the image data is transferred from the scanner 1 to the PC 2 as it is without processing the image data, and the desk 2 processing and cropping are performed on the PC 2 side. Since processing is performed, it is not necessary to mount a large amount of memory in the scanner 1. In addition, when the angle and width cannot be detected in this way, the scanner 1 does not process image data, so that it is possible to prevent the scanner 1 from erroneously correcting the tilt or erroneously extracting the document image area. can do.

  In addition, when the image data is not processed by the scanner 1 and is transferred to the PC 2 as it is, the angle and width of the original image data are changed from the original image data on the PC 2 having a large-capacity memory. By detecting and processing the image data, it is possible to reliably correct the inclination of the original image area and extract only the area including the original image.

  The image reading apparatus and the image reading system of the present invention are useful for reading and inputting a document image. In particular, the present invention can read an image without mounting a large-capacity memory on the image reading apparatus side. This is suitable as an image reading apparatus and an image reading system capable of correcting the inclination of the original image area on the apparatus and inputting image data obtained by extracting only the area including the original image at high speed.

Configuration diagram of an image reading system according to an embodiment of the present invention Explanatory drawing which shows the predetermined length of the subscanning direction of the image data stored in a 1st image buffer memory Explanatory drawing about the detection of the angle of one side of the image of the document Explanatory drawing about the detection of the width of one side of the image of the document (A) The figure which shows the storing image of the 1st picture buffer memory, (b) The figure which shows the storing image of the 2nd picture buffer memory Sequence diagram of an image reading system in an embodiment of the present invention (A) The figure which shows the example in which the front-end | tip of the original is bent, (b) The figure which shows the example in which the front-end | tip of the original is rough

Explanation of symbols

1 Scanner 2 PC (Personal Computer)
10 line image sensor 11 first image buffer memory 12 CPU
13 ASIC
14 Second image buffer memory

Claims (7)

A line image sensor for reading a document image in a predetermined line unit in the main scanning direction;
First buffer means for temporarily storing image data for each predetermined line read by the line image sensor while moving the original or the line image sensor in the sub-scanning direction, corresponding to a predetermined length in the sub-scanning direction;
Image analysis means for detecting the position, angle and width of one end side of the image of the document from image data for a predetermined length in the sub-scanning direction of the leading edge of the document stored in the first buffer means;
The inclination of the image data stored in the first buffer means is corrected based on the position, angle and width of one side of the image of the original detected by the image analyzing means, and the original of the image data is corrected. Image processing means for extracting only an area including the image of
Second buffer means for temporarily storing the image data processed by the image processing means until the image data is transferred to the host computer,
The first buffer means detects the position, angle and width of one side of the original image by the image analysis means, and then the image data processed by the image processing means and stored in the second buffer means. An image reading apparatus, wherein the image data for each predetermined line read by the line image sensor is overwritten in the area. The image analysis means includes
With respect to each position (hereinafter referred to as “x coordinate”) of all pixels in the main scanning direction of image data for a predetermined length in the sub-scanning direction at the leading edge of the document stored in the first buffer means. The position of the change point of the pixel in the sub-scanning direction from the leading edge of the original (hereinafter referred to as “y coordinate”) is detected
Dividing the range where the change point is detected into a plurality of sections, calculating the angle of each section,
The image reading apparatus according to claim 1, wherein when the sections having substantially the same angle continue for a predetermined number of sections or more, an average value of the substantially same angles is set as an angle of one end side of the image of the document. The image processing means is configured to store the image data in a second buffer means without processing the image data when sections having substantially the same angle do not continue for a predetermined number of sections or more. Item 3. The image reading apparatus according to Item 2. The image analysis means includes
Detecting the x coordinate of the change point of each pixel at the front end and the rear end in the main scanning direction for all of the y coordinates;
Calculate the respective angles of the front end and the rear end in the main scanning direction,
When the calculated angles of the leading edge and the trailing edge in the main scanning direction are substantially perpendicular to the angle of one end of the detected original image, the detected leading edge and trailing edge in the main scanning direction are detected. 4. The image reading apparatus according to claim 2, wherein a distance between change points of each pixel is a width of one side of the image of the document. The image processing means does not process the image data when the calculated angles of the leading edge and the trailing edge in the main scanning direction are not substantially perpendicular to the angle of one end of the detected document image. 5. The image reading apparatus according to claim 4, wherein the image data is stored in the second buffer means. An image reading system comprising the image reading device according to claim 1 and the host computer to which the image reading device is connected. The image reading device according to claim 3 and the host computer to which the image reading device is connected,
The image reading system, wherein the host computer performs processing of the image data on the host computer when the image data is transferred without being processed.

Images