JP4787685B2 - Seam image data correction method, image reading apparatus, and image forming apparatus - Google Patents

Description

  The present invention relates to a seam image data correction method for performing seam processing, an image reading apparatus for executing the seam image data correction method, and an image forming apparatus including the image reading apparatus.

  In an image reading apparatus that reads a document image, shading correction, which is a generally known technique, is performed to correct variations in main scanning direction data in order to guarantee reading characteristics. In an image reading apparatus in which a plurality of image sensors are arranged in a staggered manner, output characteristics of each image sensor vary, and it is difficult to obtain uniform density output over the entire surface. In addition, the side of the document reading ahead of the image sensors arranged in a staggered pattern temporarily adjusts the image data to be stored in the memory and later adjusted to the data of the image sensor on the side of reading the document. However, there may be a case where the data is shifted in the main and sub-scanning directions due to feeding unevenness during document transportation, displacement of the image sensor, or the like.

  Therefore, for example, the inventions described in Patent Documents 1 to 3 are known as techniques for dealing with such problems. Among them, Japanese Patent Application Laid-Open No. 2004-228561 reduces the relative deviation of the read divided images due to the position difference L in the sub-scanning direction of the image pickup apparatus when a single document is divided and read in the main scan direction by a plurality of image pickup apparatuses. In addition, a plurality of imaging devices arranged in the main scanning direction for dividing and reading a wide area in the main scanning direction, and sub-scanning means for relatively driving at least one of the imaging device and the document in the sub-scanning direction A speed detection means for detecting the speed of the sub-scanning, and a timing shift of image data between the imaging devices due to a position difference in the sub-scanning direction of the reading field of the plurality of imaging devices as a detection speed of the speed detection means. Disclosed is an image reading apparatus comprising delay means for correcting based on the image forming apparatus, and collecting means for continuously arranging the image data in which the timing deviation of the image data is corrected in correspondence with the image distribution in the main scanning direction. To have.

  Japanese Patent Laid-Open No. 2004-268531 joins image data read by a plurality of reading sensors that are displaced in the sub-scanning direction and extend in the main scanning direction with a predetermined amount of end portions overlapping. In order to perform accurately, the image pattern of the joint portion of the image data read by the adjacent reading sensors among the respective reading sensors is recognized, and if the image pattern is a predetermined image pattern, Based on the image data, a displacement amount detection process for detecting the displacement amount in the sub-scanning direction of each image data read by the adjacent reading sensor and storing the displacement amount is performed. Image reading provided with joint image recognition correction means for performing delay time correction processing for correcting the delay time of the image data read by the reading sensor upstream of the document conveying direction by the delay means. Ri apparatus is disclosed.

Further, in Patent Document 3, in order to prevent the occurrence of a gray level boundary or image shift appearing at the joint portion of each sensor in the output image, among the plurality of image sensors arranged in a staggered pattern, the sub-scanning direction is disclosed. The image data obtained from the read pixels not used for creating image data of the entire original image in the image sensor arranged on one side is stored, and the image arranged on the other side in the sub-scanning direction based on the stored data An image reading apparatus that corrects image data obtained from a reading pixel that overlaps in the main scanning direction with a reading pixel that has obtained stored image data in a sensor is disclosed.
Japanese Patent Application Laid-Open No. 2004-215011 JP 2004-214769 A JP 2003-046736 A

  However, the above-described known technique is good for images such as characters and lines and images having a uniform halftone density. However, the image sensor arranged in a zigzag pattern is misaligned or constituted by halftone dots. In the case of an image, the density of the joint portion may decrease or the image may be blurred, and it is difficult to say that the image has been appropriately processed.

  Therefore, the problem to be solved by the present invention is that when performing the seam correction process, it is possible to divide the process into a character or halftone density image and a halftone dot portion to prevent a decrease in density or blurring. Is to make it.

To achieve the object, the first means includes reading means in which reading portions of adjacent image sensors are arranged in a staggered manner in the main scanning direction by a predetermined number of pixels, and image data read by each image sensor. comprising obtaining means for obtaining, and an image data correction method for correcting image data obtained, divided into blocks comprising the image data of the overlaid disposed portion of the image data of a plurality of pixels, wherein If the position of the image sensor is deviation occurs for the ideal location, in the disposed overlapping the according to the size of the shift area, so that the amount of shifting as outer block increases , changing the cubic function convolution coefficient for each said block, a first step of the three-dimensional convolution on the image data, the first Engineering Against the output data, the second step of multiplying the weighting coefficient corresponding to the position of the main scanning direction of the adjacent image sensor, is outputted from the pre-Symbol second step, overlapping of the image sensor to the adjacent the image data of the arrangement portion, and a third step of adding according to the position of the main scanning direction, and wherein the obtaining Bei a.
The second means includes a third step of determining whether the acquired image data is a halftone dot in the first means, and the image data determined to be a halftone dot in the third step. The third step is performed through the first step and the second step.
Third means includes reading means in which reading portions of adjacent image sensors are arranged in a staggered manner in the main scanning direction by a predetermined number of pixels, acquisition means for acquiring image data read by each of the image sensors, An image reading device that corrects the acquired image data, wherein the image data of the overlapped portion of the image data is divided into blocks composed of a plurality of pixels, and the position of the image sensor is an ideal position If there is a shift with respect to the third block, a cubic function is set for each block so that the shift amount increases toward the outer block in the overlapped region according to the shift size. First correction means for performing three-dimensional convolution on the image data by changing a convolution coefficient, and output data from the first correction means On the other hand, a second correction unit that multiplies a weighting coefficient corresponding to the position of the adjacent image sensor in the main scanning direction, and a portion that is output from the second unit and that is disposed so as to overlap the adjacent image sensor. And a third correction unit that adds the image data according to the position in the main scanning direction.
The fourth means comprises halftone dot detection means for recognizing whether the image data of the overlapping portion is a halftone dot image or a non-halftone dot image in the third means. A line for performing the joint correction process and a line not to be performed are combined for each line.
The fifth means is a halftone dot attention area setting means for setting a halftone dot determination area for judging whether the image data of the overlapping portion is a halftone dot image or a non-halftone dot image in the third or fourth means. When the determined region of interest is determined to be a halftone dot, a line for performing the seam correction processing of claim 1 and a line not to be performed are combined for each sub-scanning line.
The sixth means is characterized in that, in the fourth or fifth means, there is provided means for arbitrarily setting the number of lines to be subjected to the joint correction processing and the number of lines not to be performed.
Seventh means is the third means for performing halftone dot determination for each sub-scanning line, not performing seam correction processing on the halftone image line, and performing seam correction processing on the non-halftone image line. It is characterized by that.
The eighth means is characterized in that in any of the fourth to seventh means, there is provided means for reducing the number of pixels for performing seam correction processing performed on the overlapping portion in the case of halftone images. .
A ninth means is characterized in that the image forming apparatus includes an image reading apparatus according to any one of the third to eighth means.

According to the present invention, even when weighted and added, the boundary can be made inconspicuous .

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

  FIG. 1 is a block diagram showing an outline of an image reading apparatus according to Embodiment 1 of the present invention. The image reading apparatus 100 includes first to third CIS 1, CIS 2, CIS 3 ((contact image sensor), first to third A / D (analog / digital) converters 101, 102, 103, first and second memories 111 and 112, a line composition processing block 120, and a memory 130.

  In the image reading apparatus 100, the first to third image sensors CIS1 to CIS3 are arranged so as to overlap with adjacent sensors by a predetermined pixel in the main scanning direction. The second sensor CIS2 is arranged in a staggered manner at a predetermined interval on the upstream side in the sub-scanning direction, and the first sensor CIS1 and the third sensor CIS3 are arranged on the downstream side in a predetermined interval.

  The image data output from the first image sensor CIS1 is converted into a digital signal by the first A / D converter 101, input to the line synthesis processing unit 120, and output from the second and third image sensors CIS2 and 3. The output image data is converted into a digital signal by the second and third A / D converters 102 and 103, and the first and second memories 111 and 112 are used for delaying the timing in the sub-scanning direction. And then transferred to the line composition processing unit 120 at the same timing. Note that the first CIS 1 is located on the most downstream side in the sub-scanning direction and does not require a delay, and therefore is directly stored in the line composition processing block 120 without being stored in the memory. On the other hand, the third CIS 3 is on the same line as the first CIS 1, but is arranged on the upstream side for several lines for easy adjustment, and therefore stores data in the second memory 112. The desired line delayed data in each of the first to third CISs 1 to 3 is sent to the line composition processing unit 120 as described above. Here, the correction processing of the joint portion and the one-line processing circuit 123 perform the one-line processing on the data of the first to third CISs 1 to 3 that have flown in parallel, and pass them to the subsequent stage. .

FIG. 2 is a block diagram showing details of the line composition processing unit.
As shown in FIG. 2, the line composition processing unit 120 includes a joint correction processing circuit 201 and a halftone detection circuit 202. The joint correction processing circuit 201 includes first to third cubic function convolutions 230, 231, 232, first to third weighting coefficient units 210, 211, 212, and first to third multiplication circuits 220, 221. , 222, first to third selectors 240, 241, 242, and an adder circuit 250. In each of these units, a first cubic function convolution 230, a first multiplication circuit 220, and a first selector 240 are arranged in series, and the first selector 240 has an input from the first multiplication circuit 220. In addition, data that has passed through the first cubic function convolution 230 and the first multiplication circuit 220 is input. Therefore, the first selector 240 selects either the data from the first multiplication circuit 220 or the data that has passed through the data. Further, the weighting coefficient and data that has passed through the first cubic function convolution 230 are further input to the first multiplication circuit 220 from the weighting coefficient unit 210.

  Such a circuit is also arranged in parallel on the second and third cubic function convolutions 231 and 232 side, and image data D1, D2, and D3 of the overlapping portions of the first to third CIS1 to CIS3 are input respectively. Is done. An adder circuit 250 is provided in the subsequent stage of the first to third selectors 240, 241, and 242. Data from the selectors 240, 241, and 242 is input, subjected to addition processing, and then output as corrected data. Is done.

  Further, the halftone dot detection circuit 202 provided separately from the joint correction processing circuit 201 is inputted with the image data D1, D2, D3 of the overlapping portions of the first to third CIS1 to CIS3. Whether it is a dot or a non-halftone dot (character) is determined, and data to be selected by the selectors 240, 241, and 242 is determined based on the determination result. In this embodiment, “0” is selected when it is determined as a non-halftone (character) region, and “1” is selected when it is determined as a halftone dot. The former is the data processed by the cubic function convolution, and the latter is the data passing through the cubic function convolution. In the case of a halftone dot, the input data is output as it is. . However, for example, in the case of a solid image, the processing by the cubic function convolution is not required, so that the solid image data that has passed through the first cubic function convolution 230 is input to the first multiplication circuit 220. Correction processing is executed by the multiplier circuit 220. The halftone dot detection itself is a known technique, and various methods are known. Therefore, although not specifically described here, for example, as described with reference to FIGS. A halftone dot can be detected by the processing.

  In the line image processing unit 120 configured as described above, the pixel data D1, D2, and D3 including the overlapping pixel data of the first to third CIS1 to 3 are simultaneously input to the joint correction processing circuit 201 to check the network. The corrected data is output through the path determined by the result A (non-halftone dot or halftone dot) from the output circuit 202. When the result A from the halftone dot detection circuit 202 is a character (non-halftone dot) (0), the joint correction process is selected. When the result A from the halftone dot detection circuit 202 is a halftone dot (1), the through process is selected. Details of the first to third cubic function convolutions 230, 231 and 232 and the first to third weighting coefficients 210, 211 and 212 will be described later.

  As described above, in the first embodiment, when the image information is recognized and the image is a character image, a predicted pixel value using a cubic function convolution method is calculated for the overlapping portion, and the weighting coefficient unit 210, Arbitrary weighting coefficients are selected from 211 and 212 and multiplied, and addition processing is performed by first to third adders 220, 221 and 222. On the other hand, if the image is a halftone dot, it is selected whether correction processing is to be performed on the overlapped portion for each line or through, and addition processing is performed to obtain corrected data.

  FIG. 3 is a diagram showing tone calculation of an expected pixel in the cubic function convolution method, FIG. 4 is a diagram showing the relationship between the inter-pixel distance and the correction coefficient, and FIG. 5 is a diagram showing a correction coefficient table. Hereinafter, with reference to these drawings, a method of calculating an expected pixel by cubic function convolution will be described.

FIG. 3 uses four points of data: a point of interest (S _i ), a point after one line (S _{i + 1} ), a point after two lines (S _{i + 2} ), and a point before one line (S _i-1 ). Then, an expected point (σ) between the point S _i and the point S _{i + 1} is obtained and set as the value of the target point (S _i ). An arithmetic expression for calculating the predicted point (σ) is an expression shown in FIG. 3 .sigma. = {S _i-1 * h (1 + r) + S _i * h (r) + S _i + 1 * h (1-r) + S _i + 2 * h (2-r)} / {h (1 + r) + h (r) + h (1-r) + h (2-r)}
... (1)
Is used.

H (r) in equation (1) in FIG. 3 is an equation representing the relationship between the inter-pixel distance and the correction coefficient, and the function shown in FIG.
1-2 | r | ² + | r | ³ (0 ≦ | r | ≦ 1)
h (r) = 4-8 | r | +5 | r | ² − | r | ³ (1 ≦ | r | ≦ 2) (2)
0 (2 ≦ | r |)
It is represented by

  FIG. 5 is a table showing the correction coefficient derived from the equation (2) as a table. FIG. 5 is a correction coefficient table with an accuracy of 1/8 pixel. The coefficients of this table are substituted into equation (1) to obtain the predicted point value.

The method of obtaining the predicted pixel using the cubic function convolution executed by the joint correction processing circuit 201 in this embodiment is as follows.
When the determination result from the halftone dot detection circuit 202 is a character (non-halftone dot), the left and right overlapping portions to be joined are set to 128 pixels, and the data from the first CIS 1 in units of 16 pixels is from the left side. In order, the expected pixels are calculated from the correction coefficients of r (code-decimal display) 0, 1, 2,..., 6 and 7 shown in FIG. The first multiplication circuit 220 performs arithmetic processing on 6/8,. The data from the second CIS2 is calculated from the correction coefficient of r (code-decimal display) 0, 1, 2,..., 6 and 7 shown in FIG. Further, 7/8, 6/8,... 1/8 weighting coefficients are processed by the second multiplication circuit 221. Regarding the correction coefficient, the correction coefficient is 0 to 7 in units of 16 pixels at this time. Thereafter, addition processing is performed by the addition circuit 250 to obtain corrected data. By this method, it is possible to connect the deviation of the overlapping portions of the adjacent image sensors and the density step at the time of reading step by step, and it is possible to absorb the positional deviation and the density step.

  On the other hand, when the determination result from the halftone dot detection circuit 202 is a halftone dot, if the same processing is performed on the halftone dot image as the character image, the halftone dot portion disappears when the image sensor is misaligned. There is a possibility that. Therefore, as shown in FIG. 7, for each sub-scan line, the first CIS1 data is used as it is for the overlapping portion of the image sensor for the first line, and is connected to the second CIS2 data (FIG. 7 * pattern 1 -1). The second line is processed in the same way as the character part, multiplied by the correction coefficient, and then multiplied by the weighting coefficient, and then the data of the first CIS1 and the second CIS2 are joined (see FIG. 8 * pattern 1-2). . In contrast to the first line, the third line uses the second CIS2 data as it is for the overlapping portion of the image sensor, and connects the CIS1 and CIS2 data (see FIG. 9 * Pattern 1-3). The fourth line performs the same processing as the character part in the same way as the second line (FIG. 10 * pattern 1-4). Thereafter, as shown in FIG. 7, the process returns to the pattern 1-1 and the same flow is repeated. As described above, it is possible to prevent the halftone image from being erased by this control method.

  If it is a halftone dot, it can be processed as shown in FIG. That is, as shown in FIG. 11, the correction coefficient is multiplied in the same manner as the character part, and after that, the weighting coefficient is multiplied, and then the first CIS1 and second CIS2 data are added. Reduce the number of pixels. A seam correction process is performed on 16 pixels on the first CIS1 side and the second CIS2 side from the center of the overlapped part, for a total of 32 pixels. At this time, the weighting coefficient is multiplied by 1/2 and added by the adder circuit. With this method, it is possible to prevent the halftone dot image from disappearing due to the joint processing. Here, the number of pixels to be subjected to the joint correction process is 32 pixels, but the number of pixels can be arbitrarily changed.

  In the present embodiment, two methods are used for the processing of halftone dots, but which one to select can be arbitrarily selected as necessary.

  FIG. 12 is a diagram illustrating a configuration of the entire system according to the first embodiment. This system is an image forming system corresponding to a wide sheet of paper including a copying machine main body 500 and a paper folding device 401 connected to the back side of the copying machine main body 500. The paper folding device performs end face folding and bellows folding of the paper. The paper folding device 401 includes a connecting portion 402 to the main body, an ear folding portion 403 for folding paper ears, a bellows folding portion 404 for folding the paper in a bellows shape in the transport direction, a transport switching device 405 for changing the transport direction by 90 °, a cross A folding device (not shown), a reversing device 407 for reversing the front and back, a rotating device 408 for rotating paper by 90 °, for example, changing A4 horizontal to A4 vertical, and a tray 409 for discharging and stacking folded paper are included. .

  An image reading apparatus 100 is disposed in the copying machine main body 500, and a manual sheet feed table 508 is disposed below the image reading apparatus 100. A sheet is set on the manual sheet feeding tray 508, and the sheet is temporarily stopped by a registration roll 507, and is supplied to the image forming unit 506 at a timing. The image forming unit 506 forms a latent image corresponding to image data on a photosensitive member (not shown), develops the latent image with toner, transfers the toner onto a sheet, and fixes the image on a fixing device 510. . The recorded paper on which the toner is fixed by the fixing device 510 is discharged by the recorded paper discharge roll 511 to the paper folding device 401 when paper folding is performed. When paper folding is not performed, the paper is discharged into the main body by the upper paper discharge roller 509 by a switching claw (not shown).

  When the paper is folded, the paper is fed from the recorded paper discharge roll 511 to the paper folding device 401. When the paper is folded, the corner of the paper is folded by the ear fold 403. After the corner of the paper is folded at the ear fold 403, the paper is folded into a bellows at the bellows folding unit 404 and sent to the conveyance switching device 405. In the transport switching device 405, the transport direction is changed, the short side direction of the paper folded by the cross folding device is folded, the front and back sides are reversed by the reversing device 407, and the paper direction is further rotated by the rotating device 408. 409 is stacked. The image reading apparatus 100 is used in such a wide-width image forming apparatus.

  As described above, according to this embodiment, the following effects can be obtained.

1) Acquire image data of the overlapping part of each image sensor, and if there is a deviation in the position of the image sensor, perform correction by multiplying the correction coefficient according to the magnitude of the deviation, and Since the image data of the joint portion is corrected by multiplying the weighting coefficients corresponding to both main scanning directions and adding the both data, the joint portion can be made inconspicuous.

2) Since the correction coefficient and the weighting coefficient are arbitrarily selected from a plurality, the optimum correction process can be performed only by a selection operation.

3) Since halftone dot detection means for recognizing whether the image data of the overlapping portion is a halftone dot image or a non-halftone dot image is provided, the joint correction is performed for each sub-scanning line in the case of a halftone dot image. A line to be processed and a line not to be processed can be combined, and even when the joint is a halftone image, density reduction and blurring are possible.

4) When the image is a halftone dot image, the number of pixels to be subjected to the joint correction processing performed on the overlapping portion can be reduced, and the optimum joint correction processing for the halftone dot portion can be performed.

  Example 2 will be described below. In addition, about the same structure as Example 1, it abbreviate | omits suitably.

  In this embodiment, as described in, for example, Patent Document 3 described above, in the conventional joint correction processing, even if there is a difference in the reading value caused by the deviation, as shown in FIG. Since the addition process is performed, particularly in an image composed of halftone dots, the density of the joint portion may be reduced or blurring may occur. In the present embodiment, in order to prevent such a phenomenon from occurring, correction value calculation is performed as shown in FIG. 13B, and joint correction processing is performed after mitigating the deviation of overlapping pixels.

  FIG. 14 is a block diagram schematically illustrating an image reading apparatus according to the second embodiment of the present invention. In this image reading apparatus 100, an arbiter 121 and a memory computer 122 are provided in the line composition processing unit in the first embodiment. A memory controller 122 is connected to a memory 130, and the memory controller 122 is mutually connected to the arbiter 121. Information can be sent and received. Further, the arbiter 121 includes the first CIS1 read data from the first A / D converter 101, the second CIS2 read data from the first memory, and the second CIS1 read data from the second memory. 3 CIS3 read data is input. The read data stored in the memory 130 is output from the arbiter 121 to the joint correction processing, halftone detection processing circuit 200 and 1-line processing circuit 123.

  That is, the read image data read by the first to third CIS 1 to 3 is simultaneously input to the line composition processing block 120 and temporarily stored in the memory 130 via the arbiter 121. Thereafter, in order to perform the one-line processing, data other than the overlapping portion is sent to the one-line processing circuit 123, and the overlapping portion data is sent to the joint correction processing circuit 200.

  FIG. 15 is a block diagram showing details of the joint correction processing circuit 201 and the halftone detection circuit 202. In the figure, the joint processing circuit 201 includes first and second cubic function convolutions 230 and 231, first and second weighting coefficients 210 and 211, and the first and second cubic function convolutions. First and second selectors 245 and 246 provided before the volumes 230 and 231, the first and second cubic function convolutions 230 and 231, and the first and second weighting coefficient units 210 and 211, respectively. , And third and fourth selectors 247 and 248, respectively. In addition, first and second through paths 251 and 252 are provided in parallel with the first and second cubic function convolutions 230 and 231, and further in parallel with the first and second weighting coefficient units 210 and 211. Are provided with third and fourth through paths. The outputs of the first and second weighting coefficient units 210 and 211 and the outputs of the third and fourth through paths are input to the adder circuit 250, added by the adder circuit 250, and the added value is output.

  In the first and second selectors 245 and 246, D4 and D5, which are data of overlapping portions of the first CIS1 and the second CIS2, or data of overlapping portions of the second CIS2 and the third CIS3, are stored. Entered. These data are simultaneously input to the halftone dot detection circuit 202, and the halftone dot detection circuit determines whether the data D4 and D5 are halftone dots or non-halftone dots. Based on this determination, the first to fourth selectors are selected. Data selection is performed by 245 to 248.

  In the joint correction processing circuit 201 and the halftone detection circuit 202 configured as described above, the processing path of the joint portion image data D4 and D5 is determined by the halftone detection result A of the halftone detection circuit 202. That is, when it is determined that the data D4 is a character image (non-halftone dot), the detection result is “0”, “0” is selected by each of the selectors 245 to 248, and the data D4 and D5 are connected to each other. Correction is performed. In joint correction, cubic function convolution is executed, weighting coefficients are multiplied, and the correction results of data D4 and D5 are added by addition circuit 250 and output as correction data.

  The halftone dot detection can be performed for each line. When the result A from the halftone dot detection circuit 202 is a character (0), the joint correction process is selected. When the result A from the halftone dot detection circuit 202 is a halftone character (1), the through process is selected. However, regardless of the result of halftone processing, it is also possible to select whether or not to perform joint correction processing by the selectors 220, 221, 240, and 241 according to register settings. A cubic function convolution method and a method of selecting and multiplying weighting coefficients are performed in the same manner as in the first embodiment.

  As described above, in the second embodiment, when the image information is recognized and the image of the overlapping portion is a character image, the value of the predicted pixel using the cubic function convolution method is calculated for the overlapping portion, and an arbitrary value is calculated. A weighting coefficient is selected and multiplied, and an addition process is performed. On the other hand, if the image is a halftone dot, it is selected whether correction processing is to be performed on the overlapped portion for each line or through, and addition processing is performed to obtain corrected data. Details will be described later.

  Since the method of calculating the predicted pixel by the cubic function convolution is the same as that described in the first embodiment with reference to FIGS.

  A method of obtaining the predicted pixel using the cubic function convolution in the joint correction process will be described.

  If the determination result from the halftone dot detection circuit 202 is a character (non-halftone dot), the left and right overlapping parts to be joined can be changed to 128 pixels (8, 16, 32, 64, 128, 256). And the same r code coefficient (see FIG. 5) is multiplied to a unit obtained by dividing the number of overlapping pixels into eight (16 pixels in the case of 128 pixels). However, if the same r-code coefficient is multiplied in all the blocks in the overlapping portion, a gap may occur at the boundary between the overlapping portion and the other portion in the same CIS. Also, as shown in the relationship between the coefficients, the value of the r code to be multiplied is also changed for the CIS position and the deviation amount within one pixel.

  Description will be made using a case where the main scanning deviation amount is 4/8 pixels as shown in FIG. The eight blocks of every 16 pixels of the overlapping portion of 128 pixels of CIS1 are multiplied from the left by a coefficient of (r0, r1, r1, r2, r2, r3, r3, r4), and 16 of 16 overlapping portions of CIS2 are obtained. The eight blocks for each pixel are multiplied by coefficients (r4, r3, r3, r2, r2, r1, r1, r0) from the left. By multiplying by such a coefficient, the gap at the boundary between the overlapping portion and the other portion in the same CIS is reduced.

  Next, a weighting coefficient calculation (multiplication) shown in FIG. 18 is performed by the weighting coefficient circuit 210 on the value obtained by multiplying the coefficient of the cubic function convolution as shown in FIG. That is, the block divided into 8 is multiplied by a coefficient of 7/8, 6/8,..., 1/8 from the left of the first CIS1, and 7/7 from the right in the second CIS2. Multiply by a factor of 8, 6/8, ... 1/8. Thus, the closer to the first CIS1 of the overlapping portion, the more weight is placed on the pixel value of the first CIS1, and the closer to the second CIS2, the more weight is placed on the pixel value of the second CIS2 (FIG. 18 (a )), And synthesized (FIG. 18B).

  As shown in FIG. 19, the weighting coefficient can also be combined with a weighting coefficient of 1/2. In this case, the concentrations of the first and second CISs 1 and 2 are averaged and synthesized.

  Further, by selecting “1” with the selectors 247 and 248, the weighting coefficient calculation can be made through, and as shown in FIG. Alternatively, CIS data on only one side can be used for the overlapping portion data (FIGS. 20B and 20C).

  After the weighting coefficient calculation, the adder circuit 250 adds the data of the overlapping portion of the first CIS1 and the second CIS2 or the overlapping portion of the second CIS2 and the third CIS3. Thereafter, the data other than the overlapping portion from the arbiter 121 is connected by the one-line processing circuit 123 to become data for one line.

  By processing as in the present embodiment, it is possible to connect the deviation of the overlapping portions of the adjacent image sensors and the density step at the time of reading step by step, and to absorb the positional deviation and the density step. Become.

If a halftone image is processed in the same way as a character image, the halftone portion may disappear if the image sensor is misaligned. The previous joint correction processing is only the weighting coefficient calculation described above, and as shown in FIG. 21 (in this case, the fine line of one pixel, the positional deviation of one main scanning pixel of CIS), the weighting calculation and addition Since the processing is performed, the image is blurred. The halftone dot image disappears when sub-scanning deviation occurs. Therefore, when the determination result from the halftone dot detection circuit 202 is a halftone dot, the processing is performed as described with reference to FIG. 7 of the first embodiment. If the dot is a halftone dot, it can be processed as shown in FIG. 11 in the first embodiment.
If it is a halftone dot, it can be further processed as follows.
That is, halftone dot detection is performed for each sub-scanning line, and in the case of a halftone image, the selectors 220, 221, 231, and 241 are switched to perform through processing. When the through process is performed, the connection method can be selected from FIG. 20 as before. Further, only in the case of a halftone image, the number of pixels in the joint portion is reduced as described in the previous paragraph, and the joint correction process is performed.

  A method for detecting halftone dots will be described.

  FIG. 22 is a flowchart showing a procedure for detecting halftone dots. In this processing procedure, the input image data (step S1) is filtered (step S2), and simple binarization (step S3) is performed. Thereafter, a determination is made according to a specified determination criterion (described later) (step S4). That is, in step S2, a filter operation is performed on the input image data using the filter shown in FIG. Next, in step S3, simple binarization is performed on the image data after the filter calculation. In this case, the quantization threshold is 128 when the data is 8 bits and 512 when the data is 10 bits. For a simple binarized image, a halftone dot is determined if the number of change points of the image in the attention area is equal to or greater than a certain value, and a non-halftone dot is determined otherwise. However, the region of interest is a total of 41 points of (target pixel) + (20 pixels on the left and right of the target pixel), and the halftone detection circuit 202 sets the region (see FIG. 24). As shown in FIG. 25, the threshold value of the change point is confirmed by the change point threshold values 10 to 12. Note that the quantization threshold and the change point threshold can be changed. In addition, halftone dot detection itself is a known technique, and various methods are known. Methods other than this method can also be applied.

As described above, according to this embodiment, the following effects can be obtained.
1) Acquire image data of the overlapping part of each image sensor, and if there is a deviation in the position of the image sensor, perform correction by multiplying a correction coefficient according to the magnitude of the deviation, Since the image data of the joint portion is corrected by multiplying the weighting coefficients corresponding to both main scanning directions and adding the both data, the joint portion can be made inconspicuous.

2) Since a correction coefficient and a weighting coefficient are arbitrarily selected from a plurality, correction processing according to the application can be performed.

3) Provided with halftone dot detection means capable of arbitrarily setting a halftone dot determination area for recognizing whether the image data of the overlapping portion is a halftone dot image or a non-halftone dot image. Since it is possible to combine the line for which the seam correction processing is performed and the line for which the seam correction processing is not performed every time, it is possible to prevent image loss even when the seam is a halftone image.

4) Since the number of lines to be subjected to joint correction processing and the number of lines not to be performed are arbitrarily set, correction processing according to the application can be performed.

5) Since the number of pixels to be subjected to the seam correction processing performed on the overlapping portion when it is determined to be a halftone image is reduced, the seam correction processing of the halftone portion according to the application can be performed.

6) Halftone dot determination is performed for each sub-scanning line, and the seam correction process is not performed for lines that are halftone images, and the seam correction process is performed only for non-halftone images. Even in such a case, it is possible to prevent image loss.

7) Halftone dot determination is performed for each sub-scanning line, and seam correction processing is performed only when the image is a non-halftone image, and the number of pixels subjected to seam correction processing is reduced for lines that are halftone images. Corresponding halftone dot seam correction processing can be performed.

1 is a block diagram showing a schematic configuration of an image reading apparatus in Embodiment 1 of the present invention. FIG. 3 is a block diagram illustrating a circuit configuration of a circuit that performs line synthesis processing in the first embodiment. It is a figure which shows the relationship between the calculation formula of the prediction pixel gradation calculation in a cubic function convolution method, and a pixel. It is a figure which shows the relationship between the distance between pixels, and a correction coefficient. It is a figure which tabulates and shows the correction coefficient derived | led-out from Formula (2). It is a figure which shows the data derivation image after a seam correction | amendment. It is a figure which shows an example (processing pattern 1-1) of the process pattern of the halftone dot part of the 1st line of a subscanning line. It is a figure which shows an example (processing pattern 1-2) of the processing pattern of the halftone dot part of the 2nd line of a subscanning line. It is a figure which shows an example (processing pattern 1-3) of the halftone dot part processing pattern of the 3rd line of a subscanning line. It is a figure which shows an example (processing pattern 1-4) of the halftone dot part processing pattern of the 1st line of a subscanning line. It is a figure which shows the other example (processing pattern 2) of the processing pattern of a halftone part. 1 is a diagram illustrating a system configuration of an image forming system in which an image reading apparatus according to a first embodiment is mounted on an image forming apparatus. It is a figure which shows the difference of the correction process of a prior art example and Example 2. FIG. It is a block diagram which shows schematic structure of the image reading apparatus in Example 2 of this invention. 10 is a block diagram illustrating a circuit configuration of a circuit that performs joint correction processing and halftone dot detection in Embodiment 2. FIG. It is a figure which shows the correction coefficient with respect to the deviation | shift amount of a main scanning direction. It is a figure which shows the prediction pixel position image of main scanning deviation | shift amount 4/8. It is a figure which shows an example of the weighting coefficient calculation in Example 2. FIG. It is a figure which shows the other example of the weighting coefficient calculation in Example 2. FIG. It is a figure which shows the example through the weighting coefficient calculation in Example 2. FIG. It is a figure which shows the seam correction process in the past. It is a flowchart which shows a halftone dot determination processing procedure. It is a figure which shows an example of the filter used by the halftone dot determination process of FIG. It is a figure which shows the attention area | region in the case of performing a halftone dot determination. It is a figure which shows the relationship between a change point threshold value and the number of reference pixels.

Explanation of symbols

100 Image reading apparatus 101, 102, 103 A / D
DESCRIPTION OF SYMBOLS 111,112,130 Memory 120 Line composition processing part 121 Arbiter 122 Memory controller 123 1-line processing circuit 200 Seam correction processing, halftone detection processing part 201 Seam correction processing circuit 202 Halftone detection circuit 210, 211, 212 Weighting coefficient part 220,221,223 Adder circuit 230,231,232 Third order function convolution unit 500 Image forming apparatus CIS1-3 Contact image sensor

Claims (9)

It has reading means in which reading portions of adjacent image sensors are arranged in a staggered manner in the main scanning direction by a predetermined number of pixels, and acquisition means for acquiring image data read by each image sensor . An image data correction method for correcting image data,
Divided into the Cascade composed of a plurality of pixels of the image data of the deployed partial blocks of the image data, wherein when the position of the image sensor is deviation occurs for the ideal position, the magnitude of the deviation Accordingly, the coefficient of the cubic function convolution is changed for each block so that the shift amount increases toward the outer block in the overlapped region, and the three-dimensional convolution is performed on the image data. A first step of performing
The output data of the first step, a second step of multiplying the weighting coefficient corresponding to the main scanning direction of the position of the image sensor for the adjacent,
Is output from the previous SL second step, a third step of the image data of the overlapping portion disposed by adjacent image sensors, adds in accordance with the position in the main scanning direction,
Image data correction method characterized in that obtaining Bei a. A third step of determining whether the acquired image data is a halftone dot;
2. The image according to claim 1, wherein for the image data determined to be halftone dots in the third step, the third step is executed through the first step and the second step. Data correction method. It has reading means in which reading portions of adjacent image sensors are arranged in a staggered manner in the main scanning direction by a predetermined number of pixels, and acquisition means for acquiring image data read by each image sensor . an image reading apparatus for correcting the image data,
Divided into the Cascade composed of a plurality of pixels of the image data of the deployed partial blocks of the image data, wherein when the position of the image sensor is deviation occurs for the ideal position, the magnitude of the deviation Accordingly, the coefficient of the cubic function convolution is changed for each block so that the shift amount increases toward the outer block in the overlapped region, and the three-dimensional convolution is performed on the image data. First correcting means for performing
The output data from said first correcting means, and second correcting means you calculate multiplied by the weighting coefficient corresponding to the position of the main scanning direction of the image sensor to the adjacent,
Is output from the previous SL second means, the image data of the overlapping portion disposed of an image sensor to the adjacent, and third correction means for adding according to the position of the main scanning direction,
An image reading apparatus comprising: Comprising halftone dot detection means for recognizing whether the image data of the overlapping portion is a halftone dot image or a non-halftone dot image;
4. The image reading apparatus according to claim 3, wherein when the image is a halftone image, a line for performing the joint correction processing and a line for not performing the joint correction processing are combined for each sub-scanning line . Comprising halftone dot attention area setting means for setting a halftone dot determination area for determining whether the image data of the overlapping portion is a halftone dot image or a non-halftone dot image;
5. The method according to claim 3 or 4, wherein when it is determined that the set attention area is a halftone dot, a line for performing the seam correction processing according to claim 1 and a line not to be performed are combined for each sub-scanning line. The image reading apparatus described. 6. The image reading apparatus according to claim 4, further comprising means for arbitrarily setting the number of lines to be subjected to the joint correction processing and the number of lines not to be performed . 4. The image according to claim 3, wherein halftone dot determination is performed for each sub-scanning line, and the seam correction processing is not performed for the lines of the halftone image, but the seam correction processing is performed for the lines of the non-halftone image. Reading device. 8. The image reading apparatus according to claim 4, further comprising means for reducing the number of pixels for performing seam correction processing performed on an overlapping portion in the case of a halftone image . An image reading apparatus according to any one of claims 3 to 8 is provided.
An image forming apparatus .

Images