JP2011123683A - Device for measuring number of sheets - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately count the number of sheets from a plurality of pickup images, in a device for moving an imaging means in a stacking direction of the sheet, performing a plurality of pieces of photography, and measuring the number of the sheets. <P>SOLUTION: This device for measuring the number of the sheets includes: a reference mark part 26 arranged with identification marks M1, M2 in the stacking direction parallel to the stacked sheets at prescribed intervals; and a processing part including a count means counting the number of the sheets from the pickup images obtained by the imaging means 30. The imaging means 30 photographs the stacked sheets and the reference mark part 26, and performs the photography so that the two identification marks M1, M2 are included in the pickup image by one-time photography and that one identification mark is included in the pickup images so as to be overlapped in preceding photography and succeeding photography. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a sheet number counting device that measures the number of sheets such as stacked paper sheets and recording media such as CDs and DVDs by taking an image with an imaging unit that moves in the stacking direction.

  Conventionally, for example, devices described in Patent Documents 1 to 4 are known as devices that measure the number of sheets by performing imaging using an imaging unit that moves in the stacking direction.

  In Patent Document 1, a CCD camera and a marking device are provided on the side surface of a laminate on which flat paper is laminated so as to be movable up and down, and image data is sent to the control device. The marking device attaches an index to the side surface of the laminated body for each predetermined number of sheets and counts the number of sheets.

  In Patent Document 2, a camera device having a line CCD sensor that is movable in the height direction is arranged on the side of a paper pile on which paper is stacked, and the density data of the boundary portion is added from the output from the camera device 15. Thus, an arithmetic processing device for calculating the number of sheets of paper and a counting auxiliary jig provided with a pressing mechanism for urging the paper pile above and below the paper pile are provided.

  In Patent Document 3, a CCD camera for photographing a laminated printed material obtained by dividing the printed material into a predetermined number of sheets by a slip sheet, and an elevating mechanism for raising and lowering the CCD camera in the vertical direction are provided. 13 is counted, and the interleaving paper and printed matter of the laminated printed matter are counted while discriminating the interleaving paper by comparing the thickness or reflectance from the image waveform obtained based on the image signal and the reference waveform registered in advance. In addition, the number of sheets between the slip sheets is counted.

  In Patent Document 4, a video camera is moved at a constant speed along one side of a stack of many sheets, a still image is taken at a predetermined timing with a predetermined number of sheets as one shooting unit, and the shooting timing is set. Shooting is continued while correcting from the advance or delay of the reference window in the shot image, and the number of sheets is counted from the number of sheets in the shooting unit and the number of shots.

JP 2008-198081 A JP 2004-318430 A Japanese Patent Laid-Open No. 7-141483 Japanese Patent Laid-Open No. 5-6471

  However, in the conventional sheet number counting device, a technique has not been established as to how to join adjacent captured images in a plurality of obtained captured images. For example, in Patent Document 1, two frames with overlapping images are joined together by collating the matching of the images. However, in the stacked sheets, all the images are almost all. Since the images are similar, it is difficult to check the matching of the images.

  Further, when a line CCD camera is used as in Patent Document 2, there is a problem that the cost increases.

  The present invention has been made in view of such a problem, and an object of the present invention is to use a plurality of picked-up images in an apparatus for measuring the number of sheets by shooting a plurality of times by moving an image pickup unit in the sheet stacking direction. An object of the present invention is to provide a sheet number counting device capable of accurately counting the number of sheets.

In order to solve the above-mentioned problem, the invention according to claim 1 of the present invention is configured to reduce the number of sheets by photographing the end face of the laminated sheets a plurality of times by an imaging means that moves in the laminating direction. In the sheet number counting device to measure,
A reference mark portion in which identification marks are arranged at predetermined intervals in the stacking direction of the stacked sheets;
A processing unit comprising counting means for counting the number of sheets from a captured image obtained by the imaging means;
With
The image pickup unit takes a picture of the laminated sheets and the reference mark part together, and at least two identification marks are included in the picked-up image in one shooting, and in the previous shooting and the next shooting. The imaging is performed so that at least one identification mark is overlapped and included in each captured image.

As for invention of Claim 2, the said process part of Claim 1 is,
In each captured image, detection range extraction means for extracting a range between two identification marks as a detection range;
A sheet boundary detection means for detecting a portion corresponding to the boundary between the sheets and the adjacent sheets in the detection range;
An offset amount calculating means for obtaining and recording the distance from the boundary of the last sheet in the detection range to the end of the detection range as an offset amount;
With
The counting means determines whether or not a sheet exists between the distances of the boundary adjacent to the detected boundary, counts the number of sheets in the detection range, and starts from the beginning of the detection range. A distance obtained by adding the offset amount of the previous captured image to the distance to the boundary of the sheet is obtained, and it is determined whether or not the sheet exists between the obtained distances.

The invention according to claim 3 is the sheet boundary detection means according to claim 2,
Differential value calculating means for obtaining a differential value in the stacking direction of the image data in the detection range;
Secondary differential value calculation means for further differentiating the differential value;
With
The peak value of the secondary differential value is a portion corresponding to the boundary between sheets.

  The invention according to claim 4 is characterized in that the differential value calculation means according to claim 3 replaces the value of the portion where the differential value is positive with zero.

  According to a fifth aspect of the present invention, the counting means according to the third or fourth aspect is configured such that an average distance between a boundary between sheets and a boundary between adjacent sheets is a thickness of the sheets, and is adjacent to the boundary between the sheets. The distance between the boundary of the sheets and the thickness of the sheets is compared, and when the distance is within a predetermined range with respect to the thickness, it is counted as one sheet, and the distance is less than the predetermined range with respect to the thickness. In some cases, it is counted as 0 sheets, and when the distance is not less than a predetermined range with respect to the thickness, it is counted as 2 sheets or more according to the distance.

The invention according to claim 6 is the processing unit according to claim 1,
In each captured image, detection range extraction means for extracting a plurality of detection ranges having a predetermined width between two identification marks;
The counting means for obtaining a cumulative value from each count value for each of a plurality of detection ranges;
Error determination means for determining whether or not the cumulative values for a plurality of detection ranges match;
It is characterized by providing.

According to a seventh aspect of the present invention, the processing unit according to any one of the first to sixth aspects includes:
An identification means for setting the image data in the detection range equal to or less than a reference value to 0,
The counting means does not count the number of sheets for a portion where image data is zero.

The invention according to claim 8 is the processing unit according to any one of claims 1 to 7,
Each captured image includes a captured image suitability determination unit that determines whether or not an identification mark is appropriately included.

The invention according to claim 9 is the one according to claim 8, wherein the identification mark is formed by alternately arranging different first marks and second marks in the stacking direction,
The captured image suitability determination means includes
Determining whether there is an identification mark in the first area and the second area of the captured image;
The case where the first mark is present in the first area and the second mark is present in the second area, and the case where the second mark is present in the first area and the first mark is present in the second area are photographed. When it is alternately repeated for each order, it is determined that the captured image is appropriate.
It is characterized by that.

  According to the present invention, the imaging unit photographs the stacked sheets and the reference mark portion together, and the at least one identification mark of the reference mark portion is overlapped and included in each captured image. By doing this, it is ensured that the previous captured image and the next captured image satisfy a certain physical relationship, the adjacent captured images can be reliably joined, and the stacked sheets The number of sheets can be accurately measured.

  In particular, the distance from the boundary of the last sheet in the previous captured image to the end of the detection range is recorded as an offset, and the offset is the distance from the start of the detection range to the boundary of the first sheet in the next captured image And determining whether or not sheets exist between the added distances, it is possible to reliably count sheets that straddle adjacent detection ranges.

  Further, when detecting the boundary between sheets, the boundary between sheets can be detected more clearly by obtaining up to the second order differential value of the image data. In particular, by setting the value of the portion where the differential value is positive to 0, the boundaries of the sheets can be more clearly distinguished.

  In addition, among the laminated sheets, a certain sheet is retracted, so that the brightness may be shot lower than the other sheets and the boundary may not be clear, but even in such a case, By using the average thickness and interpolating on the assumption that sheets exist, counting omission can be prevented and the number of sheets can be counted reliably.

  By performing the same process in multiple detection ranges and counting the number of sheets, if a certain detection range is affected by local dust, scratches, dirt, etc., and the count value is incorrect, an error is notified and measurement is performed. Generation of errors can be prevented.

  Large gaps without sheets are likely to cause image data to fall below the reference value. Therefore, by identifying the reference value and not counting the number of sheets in the gap, measurement errors will occur. Can be prevented.

  By determining whether or not the identification mark is properly included in the captured image, it is possible to prevent occurrence of measurement errors.

  In particular, shooting is performed by determining whether or not identification marks exist in predetermined first and second areas of the captured image and whether or not identification marks are appropriately present in each area. It is possible to determine whether it has been performed correctly.

It is a schematic block diagram showing the structure of the sheet number counting apparatus by this invention. (A) is a principal part perspective view showing the structure of the sheet | seat number counting apparatus by this invention, (b) is the elements on larger scale. It is a flowchart of the measurement process of the sheet number counting apparatus by this invention. It is a flowchart of a sheet count process. It is a flowchart of the suitability determination process of a captured image. It is a flowchart of a sheet | seat kind boundary detection process. It is a flowchart of a result output process. It is a figure showing a certain captured image and the next captured image. It is a figure showing the relationship between a captured image and a detection range. It is a figure showing the relationship between a captured image and a brightness | luminance. It is a figure showing the relationship between a captured image, a brightness | luminance, a 1st derivative, and a 2nd derivative. It is a figure showing the relationship between the offset amount of a certain captured image, and the next captured image.

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

  As shown in FIGS. 1 and 2, the sheet number counting apparatus 10 according to the present invention roughly includes a sheet placement unit 12, an imaging unit 14, and a processing unit 16.

  The sheet placement unit 12 includes a horizontal placement table 20 on which sheets to be counted are placed in a stacked state, and a vertical wall 22 perpendicular to the placement table 20. The vertical wall 22 is provided with a transparent portion 24 made of transparent glass, resin or the like, and a reference mark portion 26 adjacent to the transparent portion 24.

  In the reference mark portion 26, different types of identification marks are arranged at predetermined intervals in opposition to the imaging means 30 described below in the stacking direction. Specifically, as shown in FIG. 2B, square identification marks M1 and circular identification marks M2 are alternately arranged at equal intervals. The interval between the identification mark M1 and the identification mark M2 is such that two identification marks M1 and M2 can enter within one imaging range of the imaging means 30, and there are a plurality of intervals between the two identification marks. The size is such that sheets can be laminated. The identification marks do not have to be arranged at a precise accuracy interval. In this example, two types of identification marks M1 and M2 are used. However, the present invention is not limited to this, and three or more types of identification marks can be arranged in order, or one type of identification marks. It is also possible to configure only with this.

  The imaging unit 14 includes an imaging unit 30 that captures an end face of the stacked sheets, and a moving unit 32 that moves the imaging unit 30 in the stacking direction of the sheets.

  The imaging means 30 includes a CCD camera 40 and illumination means 42 disposed at the tip of the CCD camera 40. The CCD camera 40 is an area CCD camera and includes a plurality of pixels vertically and horizontally. For example, if it is compatible with VGA, a CCD camera having a resolution of 640 pixels wide and 480 pixels high can be used. Usually, since the resolution in the horizontal direction is high, the horizontal direction with high resolution may be arranged so as to match the sheet stacking direction. The photographing range of the CCD camera 40 is set across the transparent portion 24 and the reference mark portion 26. And it is good to adjust so that it may correspond to about 10 micrometers per pixel and has a resolution of 8-10 pixels or more per sheet. In order to take a picture while moving the camera, it is preferable to use a field mode of a progressive camera or a normal interlace camera so that an image does not flow. The illumination means 42 can perform blinking illumination by a strobe light or lighting illumination.

  The moving unit 32 includes a support bracket 50 that supports the CCD camera 40, a toothed belt 52 to which the support bracket 50 is attached and routed in the vertical direction, pulleys 54 and 54 around which the toothed belt 52 is stretched, A motor 56 coupled to one of the pulleys 54. As the toothed belt 52 moves via the pulley 54 by the rotation of the motor 56, the CCD camera 40 moves and scans in the vertical direction.

  The processing unit 16 processes the captured image acquired by the CCD camera 40 and calculates the number of stacked sheets. An A / D converter that A / D converts the image signal from the CCD camera 40. 60, an image memory 62 that records image data that is luminance data for each pixel, a drive unit 64 that drives the motor 56 and the CCD camera 40, and a calculation unit 66 that includes a CPU that counts the number of sheets. Prepare.

  The operation of the sheet number counting apparatus 10 configured as described above will be described.

  In the measurement, the sheet to be measured is placed on the sheet placement unit 12. The sheets are placed horizontally on the mounting table 20 and are stacked in close contact with the transparent portion 24 of the vertical wall 22 as much as possible. At this time, the imaging means 30 is in the initial position. The initial position of the imaging unit 30 is normally the lowest limit position of the imaging unit 30.

  When measurement is started, the processing unit 16 performs processing according to the flowchart shown in FIG.

  First, a motor drive signal is output from the drive unit 64 to start driving of the motor 56, and at the same time, photographing by the imaging means 30 is performed (step S100). That is, the shutter of the CCD camera 40 is opened and photographing is performed, and blinking or lighting by the illumination unit 42 is started.

  The image signal captured by the CCD camera 40 is A / D converted by the A / D converter 60, and image data corresponding to one captured image is stored in the image memory 62 (step S102). Next, the number counting process is performed on the captured image (step S104).

  As a result of counting the number of sheets, it is determined whether or not the obtained number is 0 (step S106). If the number is not 0, determination is performed until the next shooting position is reached according to driving of the motor 56 ( In step S108), when it is determined that the imaging unit 30 has moved a predetermined distance and has reached the next imaging position, the imaging unit performs imaging at that position (step S110), returns to step S102, and repeats the processing. If the obtained number is 0, the motor 56 is rotated in the reverse direction to return the imaging means 30 to the initial position (step S112), and the number measurement result is output (step S114).

  As can be seen from FIG. 3, the number counting process in step S104 is performed in real time from one shooting to the next shooting, and the imaging means 30 moves during the number counting process. At this time, the predetermined distance that the imaging unit 30 moves from the previous shooting to the next shooting is set to be the same as the pitch of the identification marks M1 and M2 of the reference mark portion 26. That is, the two identification marks M1 and M2 of the reference mark portion 26 are included in the captured image in one shooting, and at least one identification mark overlaps between the previous shooting and the next shooting. To be included. More specifically, when the identification mark M1 and the identification mark M2 are photographed in one shooting (see FIG. 8A), the same identification mark M2 and another identification mark are captured in the next shooting. M1 is photographed so as to be photographed at substantially the same position (see FIG. 8B). Thus, each captured image partially overlaps with an adjacent captured image.

  The number counting process performed in step S104 by the processing unit 16 is performed according to the flowchart shown in FIG.

  First, the suitability of the obtained captured image is determined (step S200). As described above, this determination processing corresponds to determining whether or not shooting has been performed partially overlapping with the previous shooting.

  Specifically, the appropriateness determination process of the captured image can be performed according to the flowchart shown in FIG. First, the number of photographing is incremented (step S300). The initial value of the number of photographing is 0, and is reset at the start of measurement.

  Then, two identification marks are extracted from the captured image. The two identification marks should be taken at substantially the same position in each picked-up image as described above when the image pickup means 30 is shooting correctly for each pitch of the identification marks. Therefore, in each captured image, an identification mark is extracted from the determined first area and second area (step S302, FIG. 8C). If the identification mark is not extracted (No in step S304), it is determined that a correct captured image has not been captured (step S314). If an identification mark is extracted (Yes in step S304), it is further determined whether the number of times of shooting is an odd number or an even number (step S306). If the number of times of photographing is an odd number, it is determined whether or not the first identification mark is detected in the first area and the second identification mark is detected in the second area (step S308). Then, it is determined whether the second identification mark is detected in the first area and the first identification mark is detected in the second area (step S310). If the determination result is yes in step S308 and step S310, it is determined that a correct captured image has been captured (step S312). On the other hand, if any of the determination results is no, it is determined that a correct captured image has not been captured (step S314). This is because any of the determination results in step S312 and step S314 is “no”, which means that shooting partially overlapping with the previous shooting has not been performed.

  The above flowchart is an example in which two types of identification marks are used. However, when three or more types of identification marks are used, the order in which the identification marks are arranged matches the order in which the identification marks are arranged. Whether an identification mark appears in the captured image is determined. By using various types of identification marks, it is possible to detect missing images and the like more reliably.

  Next, returning to FIG. 4, a detection range is extracted from the captured image (step S202). The start end of the detection range is the same stacking direction position as the center position of the identification mark detected in the first area, and the end of the detection range is the same stacking direction position as the center position of the identification mark detected in the second area. (See FIG. 9). The width of the detection range is set to an appropriate number of pixels. If the detection range is too wide, the difference in density in the stacking direction will be small when the sheets are stacked diagonally, and if it is too narrow, the detection will be greatly affected by dust, scratches, dirt, etc. Therefore, the area should be appropriate. A plurality of detection ranges can be set in parallel. In the case of this example, four detection ranges are extracted, and the following processing is similarly performed in two detection ranges (first detection range and second detection range) separated in the width direction.

  In general, a captured image has a relatively black portion where a sheet is not present, a relatively white portion where a sheet is present, and a relatively black gap between adjacent sheets. Brightness difference). In rare cases, the gap between the sheets may be brighter than the portion where the sheets are present. In any case, since a difference in density occurs between the portion where the sheets are present and the gap between the sheets, the difference in density is detected and the number of sheets is counted. In the following description, it is assumed that the brightness of the portion where the sheets are present is high and the brightness of the gap where the sheets are not present is low.

  Then, luminance data that is image data within the detection range extracted in step S202 is extracted, and an average value of the luminance data in the width direction of the detection range is calculated (step S204). By taking the average in the width direction, when dust, scratches, dirt, and the like are locally generated in the width direction, it is possible to reduce the influence of image disturbance due to them.

  Next, the obtained average value is compared with a certain reference value. This reference value is set to a value for identifying the presence or absence of sheets. As described above, the brightness of the gap between the sheets is low, and the brightness is very low when the gap is large, but not only that, but the brightness is low when the sheet is present but retracted. As shown in FIG. 10 (a), a portion with very low luminance is determined as a large gap where sheets are not present, and a portion with slightly low luminance is determined as presence of sheets. These are identified by the reference value (FIG. 10B). And the data of the part whose brightness | luminance is lower than a reference value is set to the brightness | luminance zero (black) (step S206, FIG.10 (c)).

  Next, a boundary between sheets is detected from the luminance data (step S208). Arbitrary image processing can be used as the sheet boundary detection processing, but in this example, it can be performed according to the flowchart shown in FIG.

  First, the amount of change (differential value) in the stacking direction of the average value of the luminance data is calculated (step S400, FIG. 11B). Then, the value of the portion where the differential value is positive is set to 0 (step S402, FIG. 11C). This leaves only the region that changes from a bright part to a dark part. By this processing, even if the image may not be uniform depending on the state of the end faces of the sheets, it is possible to reduce erroneous detection due to such changes in the end faces of the sheets.

  Next, the differential value is differentiated again to calculate a secondary differential value (step S404, FIG. 11 (d)). The positive / negative peak positions of the secondary differential values are extracted, and the positive peak position (that is, the darkest part) is detected as the position of the sheet boundary (step S406).

  Although it is possible to detect the position of the boundary between sheets using only the primary differential value, a stable peak can be obtained by using the secondary differential value as described above. In particular, by deleting the positive part in the first derivative, the darkest part in which the first derivative value changes from negative to 0 (that is, positive) is used as the peak of the second derivative value as shown in FIG. It becomes possible to extract clearly and to prevent erroneous detection.

  When the boundary is thus detected, the process returns to FIG. 4, the distance between the boundary and the adjacent boundary is measured, the average value is calculated in the detection range, and the average value is set as the thickness of the sheets (step S <b> 210). .

  Next, each distance between the boundary and the boundary adjacent to the boundary is compared with the thickness of the sheets obtained in step S210 (step S212).

  If the distance is less than 50% of the thickness, the sheet is not counted as one sheet because there is no sheet between the boundaries (step S214). If the distance is greater than 50% and less than 200%, it is assumed that there is one sheet between the boundaries, and the count value is incremented by one (step S216). In the case where the distance is 200% or more and the luminance data is not zero, interpolation is performed assuming that there are two or more sheets between them, and INT (distance / sheets) is used. ) (Here, INT (x) is an integer obtained by rounding down the decimal point of x) is added to the count value (step S218). On the other hand, in step S218, when there is a portion where the luminance data is zero in the range of the target distance, the count value is only 1 because there should be a gap portion where no sheets exist. Increase. In this way, the average thickness and the distance between the boundaries are compared, and if necessary, interpolation is performed assuming that sheets exist, thereby preventing counting omission and counting the number of sheets reliably. it can. In addition, the numerical value range of the thickness compared with the distance is not limited to the above, and can be changed as appropriate.

  The processing from step S212 to step S218 is repeated for all boundaries existing in the detection range. Then, as shown in FIG. 12, the distance from the last boundary to the end of the detection range is obtained and recorded as an offset amount (step S220).

  This offset amount is added to the distance from the detection range start to the first boundary in the process of step S212 of the number counting process (S104) in the next shooting, and is compared with the thickness. That is, in the processing for the next captured image, a distance obtained by adding the distance between the first boundary position and the detection range start to the offset amount obtained for the previous captured image is defined as the boundary adjacent to the boundary. Is compared in step S212. Thereby, the sheet | seats straddling adjacent captured images can be counted reliably, and the process of the boundary part between adjacent captured images can be performed without an error.

  In FIG. 4, the processing for one detection range is described, but the processing from step S204 to step 220 is performed for two or more detection ranges.

  Returning to FIG. 1, when the image pickup device 30 passes through the stacked sheets and photographs a portion where no sheets exist, the number of sheets obtained in the sheet count process (S104) becomes zero. The number measurement result is output (step S114). For example, the result output can be performed in accordance with the flowchart shown in FIG.

  First, in the captured image suitability determination process described with reference to FIG. 5, if the captured image determination result is an error (Yes in step S500), an error is output because the image has not been correctly captured (step S500). S510). If the image is correctly captured (No in step S500), the count values for the plurality of detection ranges (that is, the cumulative value of the count values of the first detection range and the second detection range of each captured image) are equal. Is determined (step S502). If they are equal, it is determined that the measurement reliability is high, and the count value (the sum of the count values in the detection range of each captured image) is output as a result (step S504). If the count values for a plurality of detection ranges are not equal, it is determined that the reliability of measurement is low, and a plurality of detection ranges (third detection ranges) from each captured image already stored in the image memory. And the fourth detection range) are extracted, and the number counting process shown in FIG. 4 is performed (step S506). Then, it is determined whether or not the count values (that is, the cumulative value of the count values of the third detection range and the fourth detection range of each captured image) for another plurality of detection ranges are equal (step S508). If they are equal (Yes in step S508), it is determined that the measurement reliability is high, and the count value (the sum of the count values in the detection range of each captured image) is output as a result (step S504). . If the count values for a plurality of detection ranges are not equal (No in step S508), it is determined that the measurement reliability is low, and an error is output (step S510).

For example, when only a certain detection range is affected by local dust, scratches, dirt, etc., it is conceivable that the count value and the accumulated value thereof are different in a plurality of detection ranges. Therefore, the accuracy of measurement can be improved by performing the same number counting process for a plurality of other detection ranges that are less likely to be affected locally.
Through the above processing, it is possible to accurately count the number of sheets.

  Note that the above-described step S200 constitutes a captured image suitability determination means, step S202 constitutes a detection range extraction means, step S206 constitutes an identification means, step S208 constitutes a sheet type boundary detection means, and step S212. To S218 constitute a counting means, step S220 constitutes an offset amount calculating means, steps S400 and S402 constitute a differential value calculating means, step S404 constitutes a secondary differential value calculating means, and steps S502 or S508. The error determination means is configured by the above.

DESCRIPTION OF SYMBOLS 10 Sheet | seat number measurement apparatus 16 Processing part 26 Reference | standard mark part 30 Imaging means M1, M2 Identification mark S200 Captured image suitability determination means S202 Detection range extraction means S206 Identification means S208 Sheet | seat kind boundary detection means S212-S218 Counting means
S220 Offset amount calculation means S400, S402 Differential value calculation means S404 Secondary differential value calculation means S502, S508 Error determination means

Claims (9)

In the sheet number counting device that measures the number of sheets by photographing the end face of the stacked sheets a plurality of times by imaging means that moves in the stacking direction,
A reference mark portion in which identification marks are arranged at predetermined intervals in the stacking direction of the stacked sheets;
A processing unit comprising counting means for counting the number of sheets from a captured image obtained by the imaging means;
With
The image pickup unit takes a picture of the laminated sheets and the reference mark part together, and at least two identification marks are included in the picked-up image in one shooting, and in the previous shooting and the next shooting. An apparatus for measuring the number of sheets, wherein photographing is performed so that at least one identification mark is included in each captured image in an overlapping manner. The processor is
In each captured image, detection range extraction means for extracting a range between two identification marks as a detection range;
A sheet boundary detection means for detecting a portion corresponding to the boundary between the sheets and the adjacent sheets in the detection range;
An offset amount calculating means for obtaining a distance from the boundary of the last sheet in the detection range to the end of the detection range and recording it as an offset amount;
With
The counting means determines whether or not a sheet exists between the distances of the boundary adjacent to the detected boundary, counts the number of sheets in the detection range, and starts from the beginning of the detection range. The distance obtained by adding the offset amount of the previous captured image to the distance to the boundary of the sheet is obtained, and it is determined whether or not the sheet exists between the obtained distances. The sheet number counting device described. The sheet boundary detection means includes
Differential value calculating means for obtaining a differential value in the stacking direction of the image data in the detection range;
Secondary differential value calculation means for further differentiating the differential value;
With
3. The sheet number counting apparatus according to claim 2, wherein a peak value of the secondary differential value is a portion corresponding to a boundary between sheets.   4. The sheet number counting device according to claim 3, wherein the differential value calculating means replaces a value of a portion where the differential value is positive with zero.   The counting means uses the average distance between the boundary of the sheet and the boundary of the adjacent sheet as the thickness of the sheet, and the distance between the boundary of the sheet and the boundary of the adjacent sheet and the thickness of the sheet When the distance is within a predetermined range with respect to the thickness, the sheet is counted as one sheet. When the distance is within the predetermined range with respect to the thickness, the sheet is counted as zero. The sheet number measuring device according to claim 3 or 4, wherein the number of sheets is counted as two or more according to the distance when the number is within a predetermined range. The processor is
In each captured image, detection range extraction means for extracting a plurality of detection ranges having a predetermined width between two identification marks;
The counting means for obtaining a cumulative value from each count value for each of a plurality of detection ranges;
Error determination means for determining whether or not the cumulative values for a plurality of detection ranges match;
The sheet number counting apparatus according to claim 1, further comprising: The processor is
An identification means for setting the image data in the detection range equal to or less than a reference value to 0,
7. The sheet number counting apparatus according to claim 1, wherein the counting unit does not count the number of sheets for a portion where the image data is zero. The processor is
8. The sheet number counting device according to claim 1, further comprising: a captured image suitability determination unit configured to determine whether or not an identification mark is properly included in each captured image. 9. The identification mark, the first mark and the second mark different in the stacking direction are alternately arranged,
The captured image suitability determination means includes
Determining whether there is an identification mark in the first area and the second area of the captured image;
The case where the first mark is present in the first area and the second mark is present in the second area, and the case where the second mark is present in the first area and the first mark is present in the second area are photographed. When it is alternately repeated for each order, it is determined that the captured image is appropriate.
The sheet number counting device according to claim 8.