JP2008311963A - Image reader, and shading correcting method in image reader - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image reader capable of obtaining an excellent read result even when dust and dirt are attached to a white reference plate to be used for shading correction. <P>SOLUTION: The image reader for using white reference data read from the white reference plate to perform shading correction in each pixel is provided with a detecting means for comparing image data of an edge part of an original to be a reading object with the white reference data in each pixel to detect an abnormal state of the white reference plate, and a correcting means for using a white reference correction data calculated from a white reference data group of peripheral pixels of an abnormality detection pixel, an image data group of the edge part of the original and the image data of the edge part of the original in the abnormality detection pixel in place of the white reference data to perform shading correction of a pixel whose abnormality of the white reference plate is detected. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image reading apparatus, and more particularly to an image reading apparatus and a shading correction method in an image reading apparatus that can obtain a good reading result even when dust or dirt adheres to a white reference plate used for shading correction.

  The image reading apparatus performs shading correction prior to reading a document in order to prevent unevenness of a read image due to non-uniformity of light amount of a light source lamp and sensitivity variation of an image reading element. In the shading correction, a white reference plate provided in the image reading device is read, and a correction formula is created for each reading element based on the read data. In practice, the shading correction is performed by reading the black reference as well as the white reference, but the black reference reading is not handled in this specification.

  FIG. 10 is a diagram showing an outline of shading correction. In this figure, the result of reading the white reference plate by the reading elements arranged in the main scanning direction is shown as white reference read data. The read data is represented by 8-bit gradation, and is represented by a value from 0 to 255 from the dark side to the bright side. In the shading correction, for example, if the white reference read data read by a certain reading element is 235, a correction formula for converting the read data 235 at the time of reading the document into 255 of maximum brightness is applied to the reading element. create.

  However, if dust, dirt, or the like adheres to the white reference plate, the corresponding reading element outputs low brightness (for example, 70) as read data, as shown in FIG. In this case, since a correction formula is created so that the read data 70 at the time of reading the original is converted to the maximum brightness of 255, the read data of the original is excessively corrected to the white side. As a result, white streaks appear in the sub-scanning direction in the read image of the document.

Patent Document 1 describes the following technique for preventing image defects due to dust and dirt adhering to the white reference plate. In other words, the non-uniformity of light quantity of the light source lamp and the sensitivity variation of the image reading element affect the read data even when reading the original, while the dust and dirt adhering to the white reference plate are Since the reading data is sometimes not affected, the difference between the reading data of the white reference plate and the reading data of the front end portion of the document is calculated, and if the difference exceeds a predetermined value, the white reference plate A judgment is made that dirt or the like is attached, and a warning for prompting the user to clean or replace the white reference plate is output.
JP 11-252362 A

  Since the technique described in Patent Document 1 detects that dust or dirt has adhered to the white reference plate and notifies the user of the warning, the user ignores the warning or does not know how to deal with it. When the image reading process is performed, white streaks in the sub-scanning direction are still generated in the read image.

  The present invention has been made in view of such a situation. In the image reading apparatus and the image reading apparatus, a good reading result can be obtained even when dust or dirt adheres to a white reference plate used for shading correction. An object is to provide a shading correction method.

  In order to solve the above-mentioned problem, an image reading apparatus according to a first aspect of the present invention is an image reading apparatus that performs shading correction for each pixel using white reference data read from a white reference plate. Detecting means for detecting the abnormal state of the white reference plate by comparing the read data of the edge of the original and the white reference data for each pixel, and replacing the white reference data for the pixel in which the abnormality of the white reference plate is detected And correction means for performing shading correction using the read data of the edge portion of the document.

  If dust or dirt is attached to the white reference plate and the white reference data values are used for shading correction as they are, white streaks may occur in the scanned image of the original. Thus, reading data at the edge of the document is employed as shading correction data to prevent white streaks.

  In order to solve the above-mentioned problem, an image reading apparatus according to a first aspect of the present invention is an image reading apparatus that performs shading correction for each pixel using white reference data read from a white reference plate. Detecting means for detecting the abnormal state of the white reference plate by comparing the read data of the edge of the original and the white reference data for each pixel, and replacing the white reference data for the pixel in which the abnormality of the white reference plate is detected Then, shading correction is performed using white reference correction data calculated from the white reference data group of the peripheral pixels of the abnormality detection pixel, the read data group of the document edge, and the read data of the document edge of the abnormality detection pixel. And a correcting means.

  Use white reference correction data calculated from white reference read data at the peripheral pixels of the pixel affected by dust / dirt on the white reference plate, document edge read data, and document edge read data of the affected pixel. As a result, a good reading result can be obtained even if the user does not deal with dust, dirt, etc. adhering to the white reference plate.

  More specifically, the white reference correction data is the ratio of the white reference data average value of the peripheral pixels of the abnormality detection pixel to the read data average value of the edge of the document, and the reading data of the edge of the document in the abnormality detection pixel. Can be multiplied by.

  Further, a storage unit for storing the reading data of the document edge and a background color removing unit for removing the background color of the document from the document reading data using the reading data of the document edge stored in the storage unit. You may do it. When performing the background color removal processing, it is not necessary to separately read and read the background color removal data and store it in the storage means, so that it is possible to reduce the storage area and the processing time.

In order to solve the above problems, a shading correction method according to a third aspect of the present invention is a shading correction method in an image reading apparatus that performs shading correction for each pixel using white reference data read from a white reference plate. Reading a white reference plate to obtain white reference data; obtaining read data of an edge of a document to be read;
A detection step for detecting the abnormal state of the white reference plate by comparing the read data of the edge of the document and the white reference data for each pixel, and replacing the white reference data for the pixel in which the abnormality of the white reference plate is detected. And a white reference data correction step using the read data at the edge of the document as data used for shading correction.

  In order to solve the above problem, a shading correction method according to a fourth aspect of the present invention is a shading correction method in an image reading apparatus that performs shading correction for each pixel using white reference data read from a white reference plate. The step of acquiring the white reference data by reading the white reference plate, the step of acquiring the read data of the edge of the document to be read, and the read data of the edge of the document and the white reference data are compared for each pixel. A detection step for detecting an abnormal state of the white reference plate, and for a pixel in which an abnormality of the white reference plate is detected, instead of the white reference data, the white reference data group of the peripheral pixels of the abnormality detection pixel and the edge of the document White reference data in which correction data calculated from the read data group and the read data at the edge of the original in the abnormality detection pixel are used for shading correction And a positive step.

  According to the present invention, there are provided an image reading apparatus and a shading correction method in an image reading apparatus that can obtain a good reading result even when dust or dirt adheres to a white reference plate used for shading correction.

  Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram illustrating a configuration of an image reading apparatus 1 according to the present embodiment. The image reading apparatus 1 includes a CPU 10, an image reading mechanism unit 20, a ROM 30 in which processing programs are stored in a nonvolatile manner, and a RAM 40 used as a work area. The CPU 10 virtually constructs the reading control unit 110 and the image processing unit 120 by executing a processing program stored in the ROM 30.

  The reading control unit 110 controls the image reading mechanism unit 20 based on an instruction from the user, and performs image reading processing. The image processing unit 120 converts image data read by the image reading mechanism unit 20 into digital data, and performs image processing such as shading correction and background color cut.

  The image reading mechanism 20 includes a document transport mechanism, a CCD as an image reading element, a light source, a white reference plate, and the like. In the present embodiment, a description will be given by taking as an example an original reading type image reading apparatus in which CCDs are arranged in a line to form a line sensor, and an original is transferred to the line sensor to read image data. However, the present invention is not limited to the document conveying system, and can be applied to a flat bed type image reading apparatus that moves a carriage on which a CCD is mounted with respect to the document, and other types of image reading apparatuses.

  FIG. 2 is a diagram schematically illustrating the configuration of the image reading mechanism unit 20. As shown in the figure, the image reading mechanism section 20 includes a line sensor 21 in which CCDs are arranged in a line, a light source lamp 22 such as an LED, a white reference plate 23 disposed on the opposite side of the line sensor 21, and a plurality of rollers. A document transport mechanism 24 composed of, for example, is provided. With this configuration, the document G is conveyed in a direction (sub-scanning direction) orthogonal to the direction of the line sensor 21 (main scanning direction) and read line by line. In the following description, the main scanning direction is described as the X-axis direction.

  FIG. 3 is a diagram schematically illustrating a reading state of the image reading apparatus 1. 3A shows a reading state of shading correction data performed before reading the original G, and FIG. 3B shows a reading state of the leading end portion of the original G. FIG. As shown in FIG. 3A, when reading the shading correction data, the light source lamp 22 is turned on while the original G is not being conveyed, and the line sensor 21 reads the reflected light from the white reference plate 23. For this reason, when dust, dirt, or the like adheres to the white reference plate 23, the read data is affected, and the brightness of the place is lowered. As shown in FIG. 3B, when reading the leading end portion of the document G, the document G is conveyed to the reading position, and the reflected light of the document G is read by the line sensor 21. In this state, even if dust or dirt adheres to the white reference plate 23, the read data is not affected.

  FIG. 4 is a diagram illustrating an example of a reading result. 4A shows the white reference reading data that is the reading result of the white reference plate 23, and FIG. 4B shows the document leading edge reading data that is the reading result of the leading edge portion of the document G. FIG. c) shows a diagram in which both are superimposed.

  In the white reference read data in FIG. 4A, the brightness is remarkably lowered at two locations X1 and X2. As a cause of this, it is generally considered that the sensitivity of the reading element CCD is low, or dust, dirt, etc. are adhered to the white reference plate 23. In the document leading edge reading data of FIG. 4B, the brightness is slightly lowered in the area corresponding to the document width. This drop is noticeable when the background color of the original G is darker than that of the white reference plate 23. On the contrary, when the background color of the original G is as white as the white reference plate 23, almost no drop is observed. Further, in the document leading edge reading data, the lightness is remarkably lowered at one position X2.

  That is, as shown in FIG. 4C, the brightness of both the white reference read data and the document leading edge read data is lowered at the position X2, whereas only the white reference read data has the brightness at the position X1. I'm depressed. From this, it can be seen that the sensitivity of the CCD corresponding to the location X2 is low and that dust and dirt are attached to the white reference plate 23 at the location X1. Of these, when the sensitivity of the CCD is low, it can be dealt with by conventional shading correction. However, if dust or dirt adheres to the white reference plate 23, it cannot be dealt with by the conventional correction, and white streaks occur in the original read image.

  Therefore, in the present embodiment, the following correction processing is performed. FIG. 5 is a flowchart for explaining the shading correction processing in the first embodiment of the present invention. Here, an example in which the line sensor 21 reads 3000 pixels will be described. As described above, the description of reading the black reference in the shading correction is omitted.

  First, the light source lamp 22 is turned on (S101), and when the output of the light source lamp 22 is stabilized, the reflected light from the white reference plate 23 is read to obtain white reference data (S102). The white reference data is stored in the work area of the RAM 40 as W (1) to W (3000) for each pixel. The white reference data may be acquired by reading the same line a plurality of times and calculating the average. Further, in the case of the flat bed system, an average may be calculated by reading a plurality of lines.

  Next, the document G is transported using the document transport mechanism 24 until the leading end portion reaches the reading position (S103). Although it may be conveyed to the rear end portion, the front end portion is desirable because it coincides with the image reading start position. In this state, the reflected light of the document G is read to acquire document tip read data (S104). The document leading edge read data is stored in another work area of the RAM 40 as G (1) to G (3000) for each pixel. The document leading edge read data may be acquired by calculating the average by reading the same line a plurality of times or by reading a plurality of lines.

  Then, the initial value of the variable X indicating the reading element to be processed is set to 1 (S105), and the value of the variable X is increased until the variable X reaches 3000 (S111) (S110). The shading correction formula creation process (S106 to S109) is repeated.

  In the shading correction formula creation process, is the difference (G (X) −W (X)) between the document leading edge reading data G (X) and the white reference data W (X) greater than a predetermined threshold S (> 0)? It is determined whether or not (S106). When dust or dirt is adhered to the white reference plate 23, the brightness of the document leading edge read data G (X) does not decrease, and only the white reference read data W (X) decreases, so (G ( When X) −W (X)) is larger than the threshold value S, it can be determined that dust or dirt is attached to the white reference plate 23. Note that the threshold S is set to a value that takes into account normal reading variations (tolerances). In particular, if the threshold value S is too small, an erroneous determination is likely to occur when the background color of the original G is close to the white reference plate 23, so it is desirable that the threshold S has a certain size.

  As a result, when (G (X) -W (X)) is equal to or less than the threshold value S (S106: No), that is, when it is determined that no dust or dirt is attached to the white reference plate 23, the reading is performed. The white reference read data W (X) is employed as it is as shading correction data (S107).

  On the other hand, when (G (X) -W (X)) is larger than the threshold value S (S106: Yes), that is, when it is determined that dust or dirt is attached to the white reference plate 23, the white reference read data. Document leading edge reading data G (X) is employed as shading correction data instead of W (X) (S108).

  Then, a shading correction formula is created for the pixel to be processed (S109). When the above processing is performed for all pixels (S111: Yes), the shading correction formula creation processing in the first embodiment is finished (S112).

  That is, in the first embodiment, dust and dirt are attached to the white reference plate 23, and if the value of the white reference read data W (X) is used as it is for shading correction, white streaks appear in the read image of the document. In the case where there is a risk of occurrence of white streaks, the leading edge reading data G (X) is used as shading correction data instead of the white reference reading data W (X), thereby preventing white streaks. . This is because dust and dirt adhering to the white reference plate 23 do not affect the document leading edge reading data G (X) and the background color of the document is white. As a result, even if the user does not deal with dust or dirt on the white reference plate 23, a good reading result can be obtained.

  Next, a second embodiment of the present invention will be described. FIG. 6 is a flowchart showing a shading correction formula creation process according to the second embodiment. As shown in the figure, in the second embodiment, the processing (S208) is performed instead of the processing (S108) in the shading correction formula creation processing (S106 to S109) of the first embodiment shown in FIG. Unlike the first embodiment, the other points can be the same as those of the first embodiment.

  That is, in the first embodiment, dust and dirt adhere to the white reference plate 23, and if the value of the white reference read data W (X) is used as it is for shading correction, white streaks appear in the read image of the document. If there is a fear, the document leading edge reading data G (X) is adopted as shading correction data instead of the white reference reading data W (X) (S108). In contrast, in the second embodiment, white reference data that has become an abnormal value due to dust or dirt is corrected by creating white reference data interpolation data (S208). As a result, a good reading result can be obtained even when the background color of the document is not white.

  With reference to FIG. 7, the white reference data interpolation data creation processing (S208) in the second embodiment will be described. Here, it is assumed that the pixel determined to have dust / dirt adhered to the white reference plate 23 is the mth pixel.

  In the white reference data interpolation of the present embodiment, the mth document leading edge reading data G (m) is corrected using the white reference reading data W and the document leading edge reading data G in N pixels around the mth pixel. By doing.

Specifically, the (m−N / 2 to m−1) th N / 2 pixel group and the (m + 1 to m + N / 2) th N / 2 pixel group centered on m. For a total of N pixel groups, the ratio of the average Wav of the white reference read data W and the average GAv of the document leading edge read data G is calculated, and multiplied by G (m) to obtain white reference data interpolation data WCp (m) Is calculated. That is,


From

Can be obtained. In the second embodiment, the white reference data interpolation data WCp (m) is used as shading correction data instead of the white reference read data W (m). If another pixel that is determined to have dust / dirt adhered to the white reference plate 23 is included in the average value calculation range, that pixel is either the white reference data W or the document leading edge reading data G. It is desirable to omit it from the average value calculation.

  Note that the N pixels used for the average calculation are preferably continuous around m, but may be discrete. In addition, when m is close to the end, it is not always necessary to center m. Although the N specific values are not limited, for example, values suitable for experiment or design are determined.

  As described above, in the second embodiment, the average value of the white reference read data W and the average value of the document leading edge read data G in the N pixels around the mth pixel affected by dust and dirt on the white reference plate 23. And correcting the m-th document leading edge reading data G (m), a good reading result can be obtained even if the user does not deal with dust, dirt, etc. adhering to the white reference plate 23. .

  Next, a third embodiment of the present invention will be described. In the third embodiment, a case where the background color of the original G is cut in addition to the shading correction will be described. Generally, when a text document is read by an image reading apparatus, it is desirable that the character portion is clearly reproduced and the background color of the original G is not reproduced in order to make the character easy to read. In particular, when reading an original G such as rough paper or colored paper, the background color of the original G is dark, so that a process of cutting the background color is necessary to obtain a read image that can easily recognize a text document.

  In order to cut the background color of the original G, as shown in the flowchart of FIG. 8, after the creation of the shading correction formula shown in the first embodiment is completed (S112), the background color cut is performed. Data is acquired (S213), and the background color cut processing (S214) is performed.

  Here, in the background cut data acquisition (S213), in order to accurately recognize the background color of the document G, it is necessary to acquire only the background portion of the document G, that is, the data of the portion not including text or the like. There is. In general, a margin is provided in the peripheral portion of the document G and text or the like is often not included. Therefore, the background color cut data may be acquired by acquiring image data of the peripheral portion of the original G.

  Therefore, in the third embodiment, the following processing is performed. Referring to the flowchart of FIG. 5 again, when the white reference data is acquired in the process (S102), W (1) to W (3000) are stored in the area A40a of the RAM 40 as shown in FIG. Further, when the document leading edge reading data is acquired in the process (S104), G (1) to G (3000) are stored in the area B40b of the RAM 40 as shown in FIG. Shading correction is performed using these stored data.

  In the present embodiment, the document leading edge reading data stored for shading correction in the area B40b is also used as background color cut data. This is because the document leading edge reading data stored for shading correction is the blank portion reading data, and therefore can also be used as background color cutting data.

  For this reason, in the third embodiment, when the background color cut processing is performed, it is not necessary to separately read and read the background color cut data and store it in another area of the RAM 40, thereby reducing the memory area and the processing time. Can be realized.

It is a block diagram which shows the structure of an image reading apparatus. It is a figure which shows typically the structure of an image reading mechanism part. It is a figure which shows typically the reading state of an image reading apparatus. It is a figure which shows an example of a reading result. It is a flowchart explaining the process of the shading correction | amendment in 1st Example. It is a flowchart which shows the shading correction type | formula creation process in 2nd Example. It is a figure for demonstrating white reference data interpolation data creation processing. FIG. 10 is a flowchart for explaining background color cut processing of a document. It is a figure which shows the storage area of RAM. It is a figure which shows the outline | summary of a shading correction.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Image reading device, 20 ... Image reading mechanism part, 21 ... Line sensor, 22 ... Light source lamp, 23 ... White reference board, 24 ... Original conveyance mechanism, 30 ... ROM, 40 ... RAM, 110 ... Reading control part, 120 ... Image processor

Claims (6)

An image reading apparatus that performs shading correction for each pixel using white reference data read from a white reference plate,
Detecting means for detecting the abnormal state of the white reference plate by comparing the read data of the edge of the document to be read with the white reference data for each pixel;
An image reading apparatus, comprising: a correcting unit that performs shading correction using read data at an edge of the original document instead of white reference data for a pixel in which an abnormality of a white reference plate is detected. An image reading apparatus that performs shading correction for each pixel using white reference data read from a white reference plate,
Detecting means for detecting the abnormal state of the white reference plate by comparing the read data of the edge of the document to be read with the white reference data for each pixel;
For a pixel in which an abnormality of the white reference plate is detected, instead of white reference data, a white reference data group of peripheral pixels of the abnormality detection pixel, a read data group at the edge of the original, and the original of the original in the abnormality detection pixel An image reading apparatus comprising: correction means for performing shading correction using white reference correction data calculated from read data at an end. The image reading apparatus according to claim 2,
The white reference correction data includes the read data at the edge of the document in the abnormality detection pixel based on the ratio of the white reference data average value of the peripheral pixels of the abnormality detection pixel and the read data average value of the edge of the document. An image reading apparatus characterized by being multiplied. The image reading apparatus according to any one of claims 1 to 3,
Storage means for storing read data of an edge of the document;
An image reading apparatus, further comprising: a background color removing unit that removes the background color of the document from the document read data using the read data of the edge of the document stored in the storage unit. A shading correction method in an image reading apparatus that performs shading correction for each pixel using white reference data read from a white reference plate,
Reading the white reference plate to obtain white reference data;
Obtaining read data of an edge of a document to be read;
A detection step of detecting an abnormal state of the white reference plate by comparing the read data of the edge of the document and the white reference data for each pixel;
A shading correction method comprising: a white reference data correction step that uses, as a data to be used for shading correction, the read data at the edge of the document, instead of white reference data, for pixels in which an abnormality of the white reference plate is detected . A shading correction method in an image reading apparatus that performs shading correction for each pixel using white reference data read from a white reference plate,
Reading the white reference plate to obtain white reference data;
Obtaining read data of an edge of a document to be read;
A detection step of detecting an abnormal state of the white reference plate by comparing the read data of the edge of the document and the white reference data for each pixel;
For a pixel in which an abnormality of the white reference plate is detected, instead of white reference data, a white reference data group of peripheral pixels of the abnormality detection pixel, a read data group at the edge of the original, and the original of the original in the abnormality detection pixel A shading correction method comprising: a white reference data correction step using correction data calculated from the read data at the end as data used for shading correction.