JP3916596B2 - Image position correcting apparatus and method - Google Patents

Description

  The present invention relates to an image position correcting apparatus and method.

  As an inspection system that automatically inspects printed matter (for example, color quality inspection, inspection for presence / absence of stains, inspection for presence / absence of omission), an image to be inspected is obtained by imaging the printed matter with an imaging device. There is one that compares data with reference image data prepared in advance (generally, image data to be inspected that seems to be accurate).

In such an inspection system, it is necessary to align the relative positions of the reference image data to be compared and the image data to be inspected. This is because if the relative positions of the reference image data to be compared and the image data to be inspected deviate, even a normal printed matter is erroneously determined to be defective. After detecting the relative displacement between the reference image data and the inspected image data and correcting the position of the inspected image data, the reference image data and the inspected image data are compared and inspected. (For example, see Patent Document 1). In Patent Document 1, the image data to be inspected is position-corrected based on the obtained positional deviation amount.
Japanese Patent Laid-Open No. 9-330411

  An object of the present invention is to more accurately detect a relative displacement amount between the reference image data and the inspected image data, and to align the reference image data and the inspected image data more precisely.

  An image position correction apparatus (image position deviation amount calculation apparatus) according to the present invention includes a first memory for storing reference image data representing a reference image, and an inspection image obtained by imaging the object to be inspected. A second memory for storing the inspected image data to be represented; the inspected image data stored in the second memory is divided into a plurality of blocks in a direction corresponding to the transport direction; For each of the obtained first dividing means and the plurality of inspected block image data obtained by the first dividing means, the inspected block image data and the reference image data stored in the first memory Based on the data of the portion corresponding to the block image data to be inspected, the amount of displacement of the block image data to be inspected in the direction corresponding to at least the transport direction is calculated. Those having a block position deviation amount calculating means for.

  An image position correction method (image position deviation amount calculation method) according to the present invention is an inspection object representing an inspection image obtained by imaging the object to be inspected, which is compared with reference image data representing the reference image. This is a method for correcting the position of image data. The image data to be inspected is divided in a direction corresponding to the transport direction to obtain a plurality of block image data to be inspected, and each of the plurality of block image data to be inspected is inspected. Based on the block image data and the data of the portion corresponding to the block image data to be inspected in the reference image data, the displacement amount of the block image data to be inspected in the direction corresponding to at least the transport direction is calculated. .

  According to the present invention, the displacement amount is calculated for each of a plurality of block image data to be inspected obtained by dividing the image data to be inspected in a direction corresponding to the transport direction.

  The calculated positional deviation amount includes at least a positional deviation amount in a direction corresponding to the transport direction. In addition to the direction corresponding to the transport direction, the positional deviation amount in two or more directions including other directions (for example, a direction corresponding to a direction orthogonal to the transport direction) may be calculated. When calculating two or more misregistration amounts, the calculated misregistration amount is accompanied by data indicating in which direction the misregistration amount is.

  According to the present invention, for each of the plurality of block image data to be inspected divided in the direction corresponding to the transport direction, at least the position shift amount in the direction corresponding to the transport direction is calculated. Based on this, it is possible to correct the position corresponding to at least the conveyance direction of each block image data to be inspected. It is possible to precisely (finely) align the relative positions of the reference image data and the inspected image data. In particular, it is possible to eliminate a positional shift between the reference image data and the image data to be inspected, which is likely to occur in the direction corresponding to the transport direction.

  In one embodiment, the image position correction apparatus uses the block image data to be inspected, the position of which is corrected based on the position shift amount calculated by the block position shift amount calculation means, in a direction corresponding to a direction orthogonal to the transport direction, and Obtained by a second dividing means for obtaining a plurality of inspected segment image data (inspected section image data or inspected piece image data) by dividing in a direction corresponding to the transport direction, and the second dividing means. For each of the plurality of segment image data to be inspected, based on the segment image data to be inspected and the portion corresponding to the segment image data to be inspected in the reference image data stored in the first memory, A segment positional deviation amount calculating means for calculating the positional deviation amount of the image data is further provided.

  The positional deviation amount is calculated for each of the inspected segment image data obtained by dividing the position-corrected inspected block image data in the direction corresponding to the direction orthogonal to the transport direction and the direction corresponding to the transport direction. Each position of the segment image data to be inspected can be corrected by the amount of misalignment. Further, the relative positions of the reference image data and the image data to be inspected can be matched more precisely.

  In the present invention, the relative alignment between the reference image data and the inspected image data is performed by correcting the position of the inspected block image data (inspected segment image data) based on the calculated displacement amount. Of course, instead of correcting the position of the block image to be inspected (inspected segment image data), the position of the block image data (segment image data in the reference image data) in the reference image data is corrected. By doing so, it is also possible to relatively align the reference image data and the inspected image data.

  FIG. 1 is a perspective view schematically showing a part of a printed matter inspection system.

  The inspection system images a printed material to be inspected by an imaging device including a CCD line sensor, and inspects the printed material based on image data obtained by imaging. In the inspection of printed matter, the inspected image data obtained by imaging is compared with the pre-acquired reference image data (appropriate image data of the printed matter) (through preprocessing of the image data if necessary), and the result is This is done by determining whether a predetermined condition is satisfied. The image processing checks (determines) the quality of the color tone, the presence / absence of dirt, the presence / absence of omission, and the like.

  A predetermined image (design, character, etc.) is printed on the surface of a continuous sheet W such as paper or film by a rotary printing machine (not shown). From the rotary printing machine, a sheet W (hereinafter referred to as a printed matter) on which a predetermined image is printed at a predetermined interval is sent out. The printed material W sent from the rotary printing press is transported by the transport device. The conveying device includes a large number of rollers 4, and the large number of rollers 4 are arranged in parallel at substantially constant intervals on a horizontal plane along the conveying path.

  An imaging device 1 including a CCD line sensor and a linear illumination light source 2 are fixedly provided above the imaging position determined on the conveyance path of the conveyance device. The linear illumination light source 2 extends in the width direction of the conveyance path (direction perpendicular to the conveyance direction) and illuminates the entire imaging position on the conveyance path. The CCD line sensor of the imaging apparatus 1 is arranged so that the width direction of the transport path is the scanning direction, and images the imaging position illuminated by the light source 2 over the entire width direction (at least the entire width of the printed matter W).

  Image data obtained by the imaging device 1 is input to the image position correction device 3.

  As described above, the inspection of the printed matter W in the inspection system is performed by pattern matching the inspected image data obtained by imaging and the reference image data (appropriate image data of the printed matter) acquired in advance, and the degree of conformity is predetermined. This is done by determining whether the condition is met.

  Pattern matching is based on the premise that the relative positions of the reference image data and the inspected image data match, but the print W over time due to variations in the conveyance speed, temperature, humidity, etc. of the conveyance device. The relative position between the reference image data and the image data to be inspected may change due to a change (stretching / shrinking), meandering of the printed matter W, or the like. This leads to an erroneous determination that even a normal printed matter W is defective in the inspection (pattern matching) of the printed matter W. The image position correction device 3 detects a relative position shift (position shift amount (including direction)) between the reference image data and the inspected image data, and detects the relative position shift. The processing of correcting the position of the image data to be inspected (correcting the positional deviation) is performed based on the amount of positional deviation.

  FIG. 2 is a functional block diagram showing a detailed configuration of the image position correction apparatus 3. In FIG. 2, a block showing the imaging device 1 is also shown.

  The reference image data (appropriate image data of the printed matter) obtained by the imaging device 1 is stored in the reference image data memory 5. On the other hand, the image data to be inspected of the printed matter W to be inspected is stored in the inspected image data memory 6.

  The image position correction processing function (means) 7 uses the reference image data stored in the reference image data memory 5 and the inspected image data stored in the inspected image data memory 6 to A relative position shift (including the size and direction) of the image data to be inspected is detected, and the position of the image data to be inspected is corrected based on the detected amount of position shift (details will be described later).

  The corrected image data to be inspected is stored in the corrected image data to be inspected memory 8. The position-corrected inspection image data is read from the corrected inspection image data memory 8 and used for inspection (pattern matching) of the printed matter W.

  The image position correction processing function 7 is realized by a program routine. Each memory (reference image data memory 5, inspected image data memory 6 and corrected inspected image data memory 8) is a storage element such as a RAM, This is realized by a disk memory such as a hard disk. The image position correction processing function 7 can also be realized by a hardware circuit.

  FIG. 3 shows the procedure of the image position correction processing function 7 in the image position correction apparatus 3 while showing a specific image. 4, 5, and 6 are flowcharts showing the procedure of the image position correction processing function 7 in the image correction apparatus 3. As shown in FIGS. 3 and 4, the image position correction processing function 7 of the image correction apparatus 3 is more specifically divided into a block division process (step 30), a block-unit position correction process (step 40, FIG. 5), Segment division processing (step 50) and segment unit position correction processing (step 60, FIG. 6) are performed.

  Prior to the inspection of the printed material W (image position correction processing), the reference image data (appropriate image data of the printed material) is stored in the reference image data memory 5 of the image correction device 3. After the reference image data is stored in the reference image data memory 5, the imaging device 1 starts imaging of the printed matter W to be inspected (taking inspected image data).

  The inspected image data 10 (FIG. 3A) obtained by the imaging device 1 is stored in the inspected image data memory 6 of the image correcting device 3. The inspected image data 10 is two-dimensional image data composed of a plurality of one-dimensional image data obtained by the CCD line sensor of the imaging apparatus 1.

  The inspected image data stored in the inspected image data memory 6 is divided in a direction (vertical direction, vertical direction) corresponding to the transport direction of the printed material W (step 30, FIG. 3B). Each of the plurality of image data obtained by dividing the inspection image data in the direction corresponding to the conveyance direction of the printed matter W is referred to as “inspected block image data”. FIG. 3B shows an example in which the inspected image data 10 is divided into four in the direction corresponding to the conveyance direction of the printed material W. From the inspected image data 10, four inspected block image data 10A, 10B, 10C and 10D are obtained. The division in the direction corresponding to the transport direction is due to the synchronization shift in the transport direction rather than the positional shift in the direction corresponding to the direction orthogonal to the transport direction due to meandering of the printed matter (shift between the reference image data and the image data to be inspected). This is because the positional deviation is larger and becomes an obstacle to the inspection, so that this can be corrected efficiently.

  For each block image data to be inspected obtained by dividing the image data to be inspected, position correction processing in units of blocks (step 40, FIG. 5) is performed.

  In the block-unit position correction process, each of the reference image data stored in the reference image data memory 5 of the image position correction device 3 and a plurality of block image data to be inspected obtained by the block division process (step 30). Is used.

  Block variable i is initialized (step 41). The following processing is performed in order for each of a plurality of block image data i (i = 1, 2, 3,...) Obtained from the image data to be inspected.

  The reference image data includes an image data portion corresponding to the inspected block image data i. For example, referring to FIG. 3C, the reference image data 20 (the reference image represented by the reference image data 20 is indicated by a broken line in FIG. 3C) is an image corresponding to the block image data 10A to be inspected. It includes a data portion 20A, an image data portion 20B corresponding to the inspected block image data 10B, an image data portion 20C corresponding to the inspected block image data 10C, and an image data portion 20D corresponding to the inspected block image data 10D. In FIG. 3C, the block image data 10C to be inspected is superimposed on the image data portion 20C in the reference image data 20 corresponding to the block image data 10C to be inspected (block image data 10C to be inspected). And the position of the image data portion 20C in the reference image data 20 are shifted).

  First, a reference line segment (line address) is selected from the image data portion (hereinafter referred to as reference block image data) in the reference image data 20 corresponding to the block image data i to be processed (step 42). A data string (image (pixel) data string) representing the reference line segment is extracted from the reference block image data (step 43; see FIGS. 7A and 7B). Next, in the inspected block image data i to be processed, an image data string of a plurality of lines is extracted around the position corresponding to the position (line address) of the reference line segment (step 44; FIG. 7). (See (A) and (B)).

  A plurality of images centered on a position corresponding to the position (line address) of the data line representing the reference line segment selected and extracted from the reference block image data and the above-described reference line segment position extracted in the inspected block image data As will be described below, the data string is used for calculating the amount of positional deviation between the reference block image data and the block image data to be inspected. In the selection of the reference line segment (step 42), a data string representing the reference line segment and a data string adjacent to the data string representing the reference line segment (preferably a data string for a plurality of lines adjacent to the data string representing the reference line segment) And a reference line segment at such a position is clearly selected.

  The direction of the selected reference line segment may be a direction along the direction corresponding to the transport direction of the printed material W (top and bottom (up and down) direction), or a direction along the direction orthogonal to the transport direction (left and right direction). There may be. FIG. 7A shows the position (line address Lx) of the image data string representing the reference line segment selected from the reference block image data in the direction (top (upper and lower) direction) along the direction corresponding to the conveyance direction of the printed material W. And the positions (line addresses Lx-2, Lx, Lx + 2: there are intervals of two lines) of the image data sequence of a plurality of lines (three in this example) selected in the inspected block image data. ing. FIG. 7B shows the position (line address Ly) of the image data string representing the reference line segment selected from the reference block image data in the direction (left-right direction) along the direction orthogonal to the conveyance direction of the printed matter W, and The positions (line addresses Ly-2, Ly, Ly + 2: there are intervals of two lines) of a plurality (three in this example) of image data sequences selected in the block image data are shown. In FIGS. 7A and 7B, the widths of the line segments indicated by hatching are drawn large for easy understanding. 7A and 7B, for the sake of clarity, the positions (line addresses) of a plurality of image data sequences in the block image data to be inspected are spaced by two lines, but adjacent to one line. Alternatively, an image data string of the line address may be used.

  For each of an image data sequence representing a reference line segment extracted from the reference block image data and an image data sequence for a plurality of lines extracted from the inspected block image data i, an operation for obtaining an evaluation value of the degree of coincidence is performed. Performed (step 45). In the example shown in FIG. 7A, the image data string representing the reference line segment of the line address x in the reference block image data and the line addresses Lx-2, Lx, Lx + 2 in the block image data i to be inspected. Three evaluation values are obtained for each of the three image data strings. In the example shown in FIG. 7B, the image data string representing the reference line segment of the line address y in the reference block image data and the line addresses Ly-2, Ly, Ly + 2 in the inspected block image data i. Three evaluation values are obtained for each of the three image data strings.

  The evaluation value may be a correlation coefficient, a sum of absolute values of differences, or a sum of squares of differences.

  The correlation coefficient can be obtained by the following equation.

  Here, i is a pixel position (address), f (i) is a pixel value in the reference pixel data string, g (i) is a pixel value in the inspection pixel data string, Rj is a normalized cross-correlation coefficient, j Represents the pixel position in the reference pixel data string, and n represents the number of pixels in the pixel data string.

  The sum of the absolute values of the differences can be obtained by the following formula.

  The sum of the squares of the differences can be obtained by

  When the correlation coefficient (Equation 1) is used as the evaluation value, the closer the evaluation value (correlation coefficient) is to 1, the greater the degree of coincidence. When the sum of absolute values of differences (Equation 2) or the sum of squares of differences (Equation 3) is used as an evaluation value, the closer the evaluation value (the sum of absolute values of differences and the sum of squares of differences) is to 0, the degree of coincidence Is big.

  The position (line address) of the image data string in the inspected block image data i from which the evaluation value having the highest degree of coincidence is obtained by the evaluation values obtained from the reference pixel data string and the plurality of inspection pixel data strings; Based on the position (line address) of the reference line segment, a displacement amount is calculated (step 46). If the direction of the reference line segment is a direction (vertical direction) along the direction corresponding to the conveyance direction of the printed material W (see FIG. 7A) (see FIG. 7A), the calculated positional deviation amount is a direction corresponding to the direction orthogonal to the conveyance direction. If the reference line segment direction is the direction (left-right direction) corresponding to the direction orthogonal to the conveyance direction of the printed material W (see FIG. 7B), the calculated position deviation is Needless to say, the amount is the amount of misalignment in the direction corresponding to the transport direction (vertical direction).

  As described above, the evaluation value calculation (step 45) includes data representing one reference line segment selected and extracted from the reference block image data, and the above-described reference line segment in the inspection target block image data i. It is not limited to using a plurality of lines of image data sequences extracted with a position corresponding to the position (line address) as the center, and may be data using a plurality of reference line segments. (For example, FIGS. 3 (A) and 3 (B) of JP-A-9-330411), a reference line segment in a different direction including an intersection pixel may be used (for example, JP-A-9-330410). Publication).

  Based on the calculated amount of displacement, the position of the block image data i to be inspected is corrected (step 47). For example, based on the calculated displacement amount of the inspected block image data i, the position of the inspected block image data i in the inspected image data memory 6 (or the corrected inspected image data memory 8) (for example, Data (reference position data) representing the pixel position of the upper left corner of the block image to be inspected represented by the block image data to be inspected i is corrected (corrected) Of course, in step 47, the amount of displacement (including direction) is corrected. The data to be represented may be stored in the inspected image data memory 6 (or the corrected inspected image data memory 8) in association with the inspected block image data i.

  The block variable i is incremented (NO in step 48, step 49), and the same processing is repeated for the next block image data to be inspected.

  When the block position correction process is completed for all the block image data to be inspected (YES in step 48), the position correction process for each block is completed.

  The position correction in units of blocks may be correction only in the direction corresponding to the conveyance direction of the printed material W or correction only in the direction corresponding to the direction orthogonal to the conveyance direction of the printed material W, but preferably both of them. Done in

  The block image data to be inspected (block image data to be inspected whose block position has been corrected) that has undergone the block-unit position correction processing is further divided (FIG. 4: step 50). Image data obtained by further dividing the inspected block image data is referred to as inspected segment image data (see FIG. 3D). In FIG. 3 (D), the block image data 10C to be inspected is divided into two in the direction corresponding to the transport direction and into three in the direction corresponding to the orthogonal direction of the transport direction. It is shown that the image data is divided into 10Ca, 10Cb, 10Cc, 10Cd, 10Ce, and 10Cf. For each piece of segment image data to be inspected, a segment unit position correction process (step 60, FIG. 6) is performed.

  In the position correction processing in units of segments, each of the reference image data stored in the reference image data memory 5 of the image position correction device 3 and the plurality of segment image data to be inspected obtained in the segment division processing (step 50). And are used.

  Block variable i is initialized (step 61), and segment variable j is further initialized (step 62). The following processing is performed for each of the plurality of inspected segment image data j obtained from each of the plurality of inspected block image data i.

  The reference block image data includes an image data portion corresponding to the inspected segment image data j. For example, referring to FIGS. 3D and 3E, the reference block image data 20C includes an image data portion 20Ca corresponding to the inspected segment image data 10Ca and an image data portion 20Cb corresponding to the inspected segment image data 10Cb. , Image data portion 20Cc corresponding to the inspected segment image data 10Cc, image data portion 20Cd corresponding to the inspected segment image data 10Cd, image data portion 20Ce corresponding to the inspected segment image data 10Ce, and inspected segment image data 10Cf Corresponding image data portion 20Cf is included.

  The difference calculation is performed using the portion of the reference block image data (hereinafter referred to as reference segment image data) corresponding to the processing target segment image data j to be processed and the inspection segment image data j (step 63). . In the difference calculation, a difference between a predetermined value for each pixel in the reference segment image data (for example, a luminance value, a density value, etc.) and a predetermined value for each pixel in the inspected segment image data j (predetermined between pixels having the same address) Value difference) is calculated. Thereafter, the difference value for each pixel is integrated (step 64).

  Next, the segment image data to be inspected is shifted in 8 directions of one pixel (one line), up, down, left, right, top right, bottom right, top left, bottom left, respectively, and at each shifted position. In the same manner as described above, a difference value for each pixel of the reference segment image data and the inspected segment image data (a difference between predetermined values of pixels having the same address) is calculated, and then the difference values are integrated (in step 65). NO, repeat steps 66, 63 and 64). FIG. 8 shows that the segment image data 10Cc to be inspected is shifted by one pixel in the upper right direction (shifted by (+1, −1)) (shown by a broken line) and shifted by one pixel in the lower left direction. (Shifted by (-1, +1)) (indicated by a one-dot chain line).

  When a total of nine integration results are obtained (YES in step 65), the segment image data to be inspected is located at the position of the segment image data to be inspected that has produced the minimum integration value among the obtained integration results (integration values). The position is corrected (step 67). For example, the inspected segment image stored in the inspected image data memory 6 (or the corrected inspected image data memory 8) so that the inspected segment image data is located at the position where the minimum integrated value is provided. The position of the data is corrected (corrected or rewritten) (for example, data (reference position data) representing the pixel position of the upper left corner of the inspected segment image represented by the inspected segment image data j is corrected (corrected)). Of course, in step 67, data representing the amount of displacement (including the direction) is stored in the inspected image data memory 6 (corrected inspected image data memory 8) in association with the inspected segment image data j. Processing may be performed.

  The segment variable j is incremented (NO in step 68, step 69), and the same processing is performed on the next inspected segment image data.

  When the processing for a plurality of segment image data obtained from one block image data i to be inspected is completed (YES in step 68), the block variable i is incremented (NO in step 70, step 71). Processing is performed on a plurality of inspected segment image data obtained from the inspected block image data.

  When the segment unit position correction processing is completed for all the inspected segment image data obtained from all the inspected block image data (YES in step 70), the segment unit position correction processing is completed. Since the position correction processing for each segment is performed after the position correction processing for each block, the processing load of the position correction processing for each segment can be reduced (the correction range is narrowed).

  The inspected image data after the image position correction processing (after the block unit position correction and the segment unit position correction) is temporarily stored in the corrected inspected image data memory 8. The corrected inspected image data read from the corrected inspected image data memory 8 is used for pattern matching (inspection of the printed matter W).

  When the CCD line sensor of the image pickup apparatus 1 outputs color (RGB) image data, the color reference image data and the color image data to be inspected are converted into monochrome image data, and the monochrome reference image data And the black and white inspected image data are used to obtain data representing the amount of misalignment for each block image data to be inspected and the amount of misalignment for each segment image data to be inspected (the position where the integrated value of the difference values is minimum). The position (block position and segment position) of the color inspected image data may be corrected using the obtained values (amount and position). In this case, the image position correcting device 3 is provided with a conversion function (conversion circuit, conversion means) for converting the color reference image data and the color image data to be inspected into monochrome image data. The conversion function adds the multiplication values obtained by multiplying the R component, the G component, and the B component of the color reference image data and the color inspected image data by predetermined coefficients, respectively, thereby obtaining the color reference image data. The black-and-white reference image data is generated from the image data, and the black-and-white inspection image data is generated from the color inspection image data.

It is a perspective view showing a part of inspection system roughly. The functional block diagram of an image position correction apparatus is shown. (A) to (E) show the procedure of the image position correction process while showing a specific image. It is a flowchart which shows the process sequence of image position correction | amendment. It is a flowchart which shows the process sequence of the position correction of a block unit. It is a flowchart which shows the process sequence of the position correction of a segment unit. (A) shows a line segment in the direction corresponding to the transport direction in which the evaluation value is calculated in the reference block image data and the inspected block image data. (B) shows a line segment in a direction corresponding to the orthogonal direction of the transport direction in which the evaluation value is calculated in the reference block image data and the inspected block image data. A state in which the inspected segment image data is shifted by one pixel is shown.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image pick-up device 2 Linear illumination light source 3 Image position correction apparatus 5 Reference | standard image data memory 6 Inspection image data memory 7 Image position correction processing function 8 Corrected inspection image data memory
10 Image data to be inspected
10A, 10B, 10C, 10D Inspected block image data
10Ca, 10Cb, 10Cc, 10Cd, 10Ce, 10Cf Inspected segment image data
20 Reference image data
20A, 20B, 20C, 20D Reference block image data
20Ca, 20Cb, 20Cc, 20Cd, 20Ce, 20Cf Reference segment image data

Claims (2)

A first memory for storing reference image data representing the reference image;
A second memory for storing inspected image data representing an inspected image obtained by imaging an object to be inspected being conveyed;
First dividing means for dividing the inspected image data stored in the second memory into a plurality of blocks in a direction corresponding to the conveying direction to obtain a plurality of inspected block image data;
Corresponding to the inspected block image data and the inspected block image data in the reference image data stored in the first memory for each of the inspected block image data obtained by the first dividing means A block displacement amount calculation means for calculating a displacement amount in a direction corresponding to at least the conveyance direction of the inspected block image data based on the data of the portion to be inspected
The block image data to be inspected, which has been position-corrected based on the positional deviation amount calculated by the block positional deviation amount calculating means, is divided into a direction corresponding to the orthogonal direction of the conveying direction and a direction corresponding to the conveying direction, Second segmenting means for obtaining a plurality of segment image data to be inspected, and each of the plurality of segment image data to be inspected obtained by the second segmenting unit are inspected segment image data and the first memory. Segment misregistration amount calculating means for calculating the misregistration amount of the segment image data to be inspected based on the portion corresponding to the segment image data to be inspected in the stored reference image data;
An image position correction apparatus comprising: Calculates the amount of misalignment between the position of the inspection image data representing the inspection image obtained by imaging the object to be inspected and the reference image data compared with the reference image data representing the reference image Is a way to
A plurality of block image data to be inspected is obtained by dividing the image data to be inspected in a direction corresponding to the transport direction,
For each of the plurality of block image data to be inspected, at least in the transport direction of the block image data to be inspected based on the block image data to be inspected and the data of the portion corresponding to the block image data to be inspected in the reference image data. Calculate the amount of misalignment in the corresponding direction,
The block image data to be inspected whose position is corrected based on the calculated displacement amount is divided into a direction corresponding to a direction orthogonal to the transport direction and a direction corresponding to the transport direction to obtain a plurality of segment image data to be inspected. ,
For each of the plurality of segment image data to be inspected, based on the segment image data to be inspected and the portion corresponding to the segment image data to be inspected in the reference image data stored in the first memory, Calculate the amount of image data displacement,
A method for calculating the amount of displacement.

Images