Distance measuring method

Abstract

Provided is a distance measuring method for measuring the distance between a plurality of points in the main scanning direction on the basis of waveform data having been read, the distance measuring method comprising: a first step for converting the waveform data into a binarized waveform by binarizing the waveform data on the basis of a predetermined threshold; a second step for comparing the binarized waveform with a black-and-white pattern formed in a distance correction chart, and converting the binarized waveform into edge information that represents a falling edge and a rising edge with noise removed; and a third step for comparing the edge information with the physical length of the black-and-white pattern in the distance correction chart, and deriving the number of pixels that represent light receiving elements corresponding to the physical interval between black patterns adjacent to each other in the main scanning direction.

Description

Distance measurement method

[0001] The present disclosure relates to a distance measurement method.

[0002] Image reading devices, the applications of which have expanded in recent years, form erect, life-size images, and are therefore sometimes used to measure distances between arbitrary positions in the main scanning direction of an original. However, due to the influence of manufacturing variations in image reading devices, manufacturing variations in rod lens arrays, and thermal expansion of the substrate on which the light receiving elements are fixed, there is a limit to the measurement accuracy when calculating distances simply from the read waveform, i.e., the number of elements (number of pixels) of the light receiving elements involved in waveform reading.

[0003] For this reason, corrections are made to improve the accuracy of distance measurement (for example, see Patent Documents 1 and 2). Patent Document 1 describes a method in which a calibration chart with black lines drawn at known intervals is prepared, the chart is read by an image sensor, and correction is performed.

[0004] Patent Document 2 describes a method for correcting the effects of thermal expansion.

[0005] Japanese Patent Application Laid-Open No. 5-172531 International Publication No. 2020 / 129850

[0006] The technique described in Patent Document 1 has a problem in that calibration errors occur due to erroneous detection of the calibration chart caused by scratches, defects, or stains on the calibration chart.

[0007] The technique described in Patent Document 2 does not describe any correction that takes into consideration expansion and contraction due to thermal expansion of the image reading device itself.

[0008] The present disclosure is intended to solve the above-mentioned problems, and has an object to provide a distance measurement method that can improve the accuracy of distance measurement.

[0009] The distance measurement method according to the present disclosure uses an image reading device having pixels, which are light-receiving elements, arranged in the main scanning direction to read a distance correction chart, on which black and white patterns are formed at regular intervals in the main scanning direction, as waveform data of the black and white patterns using the light-receiving elements arranged in the main scanning direction, and measures the distance between multiple points in the main scanning direction based on the read waveform data. The distance measurement method includes a first step of converting the waveform data into a binary waveform that is binarized based on a predetermined threshold value; a second step of comparing the binary waveform with the black and white patterns formed on the distance correction chart and converting the binary waveform into edge information, which is falling edges and rising edges, with noise removed; and a third step of comparing the edge information with the physical length of the black and white patterns on the distance correction chart, and deriving the number of pixels, which are light-receiving elements, that corresponds to the physical interval between adjacent black patterns in the main scanning direction.

[0010] According to the present disclosure, the waveform data obtained by reading the distance correction chart with an image reading device is converted into a binary waveform, and correction is performed based on the results of comparing the edge information of the binary waveform with the physical length of the black and white pattern on the distance correction chart, thereby making it possible to improve the accuracy of distance measurement.

[0011] FIG. 1 is a perspective view of an image reading device according to the first embodiment; FIG. 2 is a view showing a state in which an original is being read by the image reading device according to the first embodiment; FIG. 3 is a view showing an output waveform when a black-and-white chart is read by the image reading device according to the first embodiment; FIG. 4 is an enlarged view of an output waveform when a black-and-white chart is read by the image reading device according to the first embodiment; FIG. 5 is a side view of the image reading device according to the first embodiment with the side panel removed; FIG. 6 is a top view of the image reading device according to the first embodiment as seen from the reading surface; FIG. 7 is a view showing a rod lens array according to the first embodiment; FIG. 8 is a view showing an image projection by the rod lens array according to the first embodiment; a flowchart of processing for checking physical length and line spacing according to the first embodiment; a flowchart of processing for checking physical length and line spacing according to the first embodiment; a flowchart of processing for checking chart position validity according to the first embodiment; a flowchart of processing for checking chart position validity according to the first embodiment; a diagram showing the relationship between the effective reading range and the effective length measurement correction range according to the first embodiment; a diagram showing an example of waveform processing for processing for checking chart position validity (first measurement, 1st side); a diagram showing an example of waveform processing for processing for checking chart position validity (first measurement,END side) Flowchart of the process of combining the first and second measurements according to embodiment 1. Diagram explaining the process of combining the first and second measurements. Diagram showing the temperature change in the body of the image reading device according to embodiment 1. Flowchart of data processing for determining the temperature correction coefficient according to embodiment 1. Diagram showing the amount of positional misalignment on a pixel basis according to embodiment 1. Diagram showing an example of measurement of the change in the amount of positional misalignment over time ... the change in the amount of positional misalignment with respect to temperature. Diagram showing the change in the slope of the amount of positional misalignment with respect to temperature depending on the pixel position. Flowchart of temperature correction procedure using the temperature correction coefficient. Flowchart of data processing of the distance correction chart waveform according to embodiments 2 and 3. Diagram showing the joint of the rod lens array according to embodiment 3. Diagram showing the attachment part of the rod lens array according to embodiment 3. Diagram showing imaging through the joint of the rod lens 16 according to embodiment 3. Diagram showing the temperature correction value at the joint of the rod lens 16 according to embodiment 3.

[0012] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In the following drawings, the same or corresponding parts are designated by the same reference characters, and description thereof will not be repeated.

[0013] Embodiment 1. FIG. 1 is a perspective view of an image reading device 100 according to embodiment 1. The image reading device 100 according to the present disclosure is a contact image sensor (CIS). As shown in FIG. 1, the x direction is the main scanning direction, the y direction is the sub-scanning direction, and the z direction is the reading depth direction. The side plate 2 is a sealing member for ensuring dust protection inside the image reading device 100. It is generally made of a metal or resin plate. The first transparent body 3 can be made of, for example, resin or glass, and extends in the x direction.

[0014] 2 is a schematic diagram showing a state in which an image of an original M is being captured by the image reading device 100 according to the first embodiment. For example, the original M is a medium to be read (an irradiated object) having image information of a banknote, a security, or other general document. Since the image reading device 100 has a row of light receiving elements 15 in the main scanning direction, capturing an image of the original requires either transporting the original in the sub-scanning direction or moving the image reading device 100 in the sub-scanning direction.

[0015] 3 shows an output waveform 5 of the image reading device 100 when the document M has a pattern as shown in chart 4. The output of the white parts of chart 4 is high, and the output of the black parts is low.

[0016] FIG. 4 shows a partially enlarged view of the output waveform 5 of the image reading device 100. The output waveforms of two sample image reading devices 100 are shown by solid and dotted lines. Compared to the deviation between the two waveforms at edge 1 in the upper left of FIG. 4, the deviation between the two waveforms at edge 12 in the lower right of FIG. 4 is greater. In other words, even when the same chart 4 is imaged, the position at which the edge is observed may differ depending on the image reading device 100. For example, if you want to measure the distance between two points using the image reading device 100, you can do so by multiplying the number of pixels between the points by the size of one pixel (42.33 μm for 600 dpi). However, the number of pixels between the points varies depending on the individual image reading device 100, which results in measurement errors.

[0017] Patent Document 1 describes a solution to a similar problem with two-dimensional image sensors. Patent Document 1 proposes a method of preparing a calibration chart with black lines arranged at known intervals, as described below, reading the chart with an image sensor, and performing correction. However, there is no description of a way to avoid calibration errors that occur due to false detection caused by scratches, defects, or dirt on the calibration chart.

[0018] 5 is a side view of the image reading device 100 with the side panel 2 removed. The first transparent body 3 and light source 8 are fixed by a second frame 7. Inside the second frame 7 is a first frame 9. A second transparent body 10, a rod lens array 11, a light receiving element 15, etc. are attached to the first frame 9. The light receiving element 15 is fixed to a substrate 13 with an adhesive. A substrate 14 is fixed to the substrate support plate 12 together with the substrate 13, aligned in the sub-scanning direction with the substrate 13.

[0019] Fig. 6 is a top view of the image reading device 100 as seen from the reading surface. The rod lens array 11 is arranged in the main scanning direction. Fig. 7 is a schematic diagram of the rod lens array 11. Rod lenses 16 are sandwiched between fixed side plates 17, and multiple rod lenses 16 are aligned in the main scanning direction and fixed with adhesive. Fig. 8 is a schematic diagram showing the projection of an image when a document is read with the rod lens array 11.

[0020] Assume that an original 18 with length L is placed at an object distance lo from the rod lens array 11. The rod lens array 11 forms an erect, life-size image 19 at a distance li. When the object distance lo and the image plane distance li are equal and the two are separated by a conjugate length Tc, the position of the light receiving element 15 becomes the focal position, and the rod lens array 11 theoretically forms an image at the same magnification as the original size on the light receiving element side. In other words, when lo = li, the relationship between the original size and the image size is L = L'. However, due to variations in the manufacturing process of the rod lens 11 or the manufacturing process of the image reading device 100, even if the relationship lo = li is satisfied, L = L' may not be strictly satisfied. In other words, the image may be enlarged or reduced relative to the original.

[0021] FIG. 9 shows a top view of the structure, with the light-receiving element 15, substrate 13, and substrate 14 fixed to the substrate support plate 12, removed from the first frame 9. The light-receiving elements 15 (pixels) are arranged continuously across the effective reading length in the main scanning direction. However, the image reading device 100 generates heat and begins to thermally expand over time after power is turned on. Because the substrate 13 is constructed as a single unit in the main scanning direction without being divided, thermal expansion occurs. Because the light-receiving element 15 is fixed to the substrate 13 via adhesive, the position of the light-receiving element 15 also changes in accordance with the thermal expansion of the substrate 13. A commonly used material for the substrate 13 is glass-reinforced epoxy resin, such as FR-4. Glass-reinforced epoxy resin expands as its temperature rises. Therefore, when the image reading device 100 generates heat, the number of pixels decreases, making the document appear smaller, even if the same size document is read.

[0022] The problem of variations in the number of pixels between points due to manufacturing variations in the image reading device 100 or the rod lens array 11 can be solved by creating correction data and applying it to the measurement results. Therefore, a method for creating the correction data will be described below.

[0023] First, the image reading device 100 must be fixed in an appropriate environment. Figure 10 shows a fixing jig used when acquiring distance correction data. The image reading device 100 is fixed to a table 20, the flatness of which is ensured, by fixtures 21 and 22 via a rail 24 that is movable in the main scanning direction. However, the fixture 21 on the first side fixes the image reading device 100 with, for example, a screw. This fixes the image reading device 100 on the first side during measurement. When acquiring distance correction data, a distance correction chart 25 is placed on the table 20.

[0024] When the effective reading length of the image reading device 100 is long, the length of the distance correction chart 25 may be shorter than the effective reading length due to chart manufacturing reasons. In this case, correction data can be obtained by dividing the image reading device 100 into the 1st pixel side and the END pixel side. First, the distance correction chart 25 is placed at the 1st pixel side position shown in FIG. 11, and waveform data is obtained. Next, the distance correction chart 25 is placed at the END pixel side position shown in FIG. 12, and waveform data is obtained. Note that the distance correction chart 25 must be placed on the table 20 parallel to the image reading device 100, and rotation in the θ direction must be suppressed as much as possible. (A structure for suppressing rotation in the θ direction may be provided on the table 20.)

[0025] 12 is a top view of the distance correction chart 25. Black lines 26 are drawn at regular intervals on the distance correction chart 25. The thickness and spacing of the black lines 26 on the chart can be any desired value, but a thickness of approximately 200 μm and spacing of approximately 1 mm are preferable. Furthermore, stainless steel or glass is preferable as the base material for the distance correction chart 25 to reduce the effects of thermal expansion due to changes in the measurement environment (room temperature).

[0026] 13 is a block diagram showing the configuration of a distance measurement system 1000 according to this embodiment. The distance measurement system 1000 reads a distance correction chart, on which black and white patterns are formed at regular intervals in the main scanning direction, as waveform data of the black and white patterns using light-receiving elements arranged in the main scanning direction of the image reading device 100, and processes the data to generate distance correction data. Furthermore, the distance measurement system 1000 measures the distance between multiple points in the main scanning direction using the distance correction data, based on the waveform data acquired by reading an image of a distance measurement target using the image reading device 100. The distance measurement system 1000 includes the image reading device 100 and a data processing device 200 that measures distances based on image data read by the image reading device 100.

[0027] The data processing device 200 includes a processor 210 and a storage unit 220. The processor 210 includes, for example, a CPU (Central Processing Unit), and functions as a correction data generation unit 211 and a distance measurement unit 212 by executing a program stored in the storage unit 220.

[0028] The storage unit 220 includes a RAM (Random Access Memory) and a non-volatile memory such as an EEPROM (Electrically Erasable and Programmable Read Only Memory), a flash memory, etc. The storage unit 220 stores distance correction data 221 used in distance measurement and various programs executed by the processor 210.

[0029] The correction data generating unit 211 of the processor 210 acquires waveform data obtained by the image reading device 100 reading the distance correction chart 25, and processes the data to generate distance correction data 221. The correction data 221 is data used when measuring distance.

[0030] 14 is a flowchart of data processing executed by the correction data generating unit 211 of the processor 210 for the waveform data of the distance correction chart 25. The steps will be explained below in order.

[0031] First, the correction data generating unit 211 of the processor 210 performs a binarization process on the captured waveform data (step 1, first step). The waveform is expressed as 0 or 1 with an arbitrary threshold as the boundary. In Fig. 16, the threshold is set to 128, and the captured waveform data (solid line) is converted into a binarized waveform (dotted line). More specifically, the position where the distance correction chart changes from a white pattern to a black pattern is taken as the falling position where the value of the binarized waveform changes from 1 to 0, and the position where the distance correction chart changes from a black pattern to a white pattern is taken as the rising position where the value of the binarized waveform changes from 0 to 1.

[0032] Next, the correction data generation unit 211 performs noise removal by pattern matching (step 2, second step). Specifically, the correction data generation unit 211 compares the binarized waveform with the black and white pattern formed on the distance correction chart, and converts the binarized waveform into edge information including falling and rising edges with noise removed. Flowcharts of the noise removal process are shown in FIGS. 15A and 15B. Because the width of the black and white pattern on the distance correction chart 25 to be read is predetermined, the arrival positions of rising edges (binarized waveform 0 → 1) and falling edges (binarized waveform 1 → 0) are predictable. The purpose of this process is to remove signals whose arrival positions deviate significantly from the predicted arrival positions, thereby improving the reliability of the distance correction data 221. Examples of events that could impair the reliability of the distance correction data 221 include the events listed in the possible failure modes in FIG. 17 occurring on the distance correction chart 25. To avoid these problems, it is effective to (i) set upper and lower limits for the width of the black line 26, and (ii) restrict the lower limit of the width of the plain white portion 27.

[0033] First, the image reading device 100 scans in the main scanning direction, and the correction data generation unit 211 reads the waveform (step S101). When the measurement pixel reaches a predetermined measurement start pixel (step S102: Yes), detection of the falling edge, which is the starting point of the black pattern of the black line 26 in the main scanning direction, is enabled (step S103, step 21). If a falling edge is detected in the area above the measurement start pixel (area A in Figure 16 ) (step S104: Yes), the received light data from the pixel where the falling edge was detected is invalidated, and the falling edge detection is disabled (step S105, step 22). The correction data generation unit 211 stores the edge position and the fact that it is a falling edge in the memory unit 220 (step S106). Thereafter, detection of the rising edge is not performed until the minimum number of pixels of the width of the black line 26 is reached. This is to prevent false detection when a white scratch or chip occurs in the black line 26 (area B in Figure 16 ). While a falling edge is not detected (step S104: No), the waveform is repeatedly read (step S102).

[0034] After detecting a falling edge, if the pixel width is equal to or greater than the width at which rising edge detection is prohibited (lower limit of black line width) (step S108: Yes), the pixel falls within the range predicted to be the end of the black pattern (area C in FIG. 16 ), and the correction data generation unit 211 validates the received light data, enables rising edge detection, and begins monitoring the width of the black line (step S109, step 23). In other words, when the predicted end pixel of the black pattern in the main scanning direction is reached, rising edge detection is enabled. By enabling rising edge detection at the predicted end pixel, it is possible to prevent erroneous detection of black foreign matter adjacent to the black line and detect edge droop.

[0035] Furthermore, the correction data generator 211 reads the waveform (step S110). If a rising edge is detected as expected in the region below the upper limit of the black line width (step S111: Yes) (step S112: Yes), the received light data from the pixel where the rising edge was detected (region D in FIG. 16) onward is invalidated (step 24). That is, the detection of the rising edge is invalidated, and monitoring of the black line width is reset (step S114). Even if a rising edge does not arrive as expected, if the upper limit of the black line width is exceeded (step S111: No) and scanning has exceeded the range of the predicted end pixel of the black pattern in the main scanning direction, the process is forced to proceed to step S114 (step 24). This is to prevent false detection when black stains or scratches are attached to the plain white portion 27 (region D in FIG. 16). Until a rising edge arrives (step S112: No), if the value of the binarized waveform is "0" (step S115: No), the process returns to step S110, and if the value of the binarized waveform is "1" (step S115: Yes), the process proceeds to step S114.

[0036] The correction data generator 211 then reads the waveform data (step S116). If the pixel is not the final pixel to be measured (step S117: No), or if the pixel is below the lower limit of the solid white width (step S118: Yes), the process returns to step S116. If the pixel exceeds the lower limit of the solid white width (step S118: No), enters the range where the falling edge of the black line 26 is predicted to arrive, and reaches the measurement start pixel for the next black pattern in the main scanning direction, the process returns to step S101 and re-enables falling edge detection (Figure 16, area A'). If the waveform readout reaches the final pixel to be measured (step S117: Yes), or if it is the first measurement (step S119: Yes), the process proceeds to the second waveform acquisition and binarization process (step S120). If it is the second measurement (step S119: No), the process ends, the detected edge information is written, and the process proceeds to the next step.

[0037] If the black lines 26 on the distance correction chart are 211.5 μm wide and the center-to-center spacing of the black lines 26 is 1 mm, the width of the black lines 26 will be approximately 5 pixels, and the width of the plain white areas 27 will be approximately 18.5 pixels. If the lower limit of the width of the black lines 26 is 3.5 pixels, the upper limit is 6.5 pixels, and the lower limit of the plain white areas 27 is 15.5 pixels, the width of the black lines 26 will be ±1.5 pixels, and the width of the plain white areas 27 will be -3 pixels. This will result in an ineffective detection error, which may result in a distance correction error. This range can be changed arbitrarily, but narrowing the range may increase the risk of undetected edges. To compensate for this disadvantage, further data processing is performed in the physical length confirmation and line spacing confirmation processes described below.

[0038] Next, the physical length confirmation and line spacing confirmation process will be described (steps 3 and 3 in FIG. 14). In this process, the edge information is compared with the physical record of the black and white patterns on the distance correction chart, and the number of pixels (number of light receiving elements) corresponding to the physical spacing between adjacent black patterns in the main scanning direction is derived. A processing flowchart for this process is shown in FIG. 18. The waveform data at this time has been converted into edge information by noise removal using pattern matching. The correction data generation unit 211 of the processor 210 reads the edge information (step S201) and performs various processes.

[0039] First, information about a falling edge that is not adjacent to a rising edge is deleted (step S202, step 31). That is, if a rising edge adjacent to a falling edge is not detected, the falling edge is deleted. This is because if a rising edge is not detected by the pattern matching process, the information about the rising edge is missing, and so the process here ultimately leaves only the black line 26 that has been correctly detected.

[0040] Next, the average pixel position of the adjacent rising edge and falling edge is calculated and recorded as the position of the black line 26 (black pattern) (step S203, step 32). More specifically, the pixel where the falling edge is detected is defined as the falling detection pixel, the pixel where the adjacent rising edge corresponding to the falling edge is detected is defined as the rising detection pixel, and the position of the average value of the positions of the falling detection pixel and the rising detection pixel is defined as the pixel position of the black line 26.

[0041] 14 (flowchart in FIG. 15A), the pixel positions of the black lines 26 are derived by performing the process of step S203 on the falling edges and rising edges repeatedly detected in step 2 (flowchart in FIG. 15A) in order to calculate the number of pixels corresponding to the distance between adjacent black lines 26 (step S204).

[0042] Next, the spacing between the pixel positions of the multiple black lines 26 is compared with the physical spacing between the black lines 26 on the distance correction chart 25. If the comparison shows that the spacing between the pixel positions of the multiple black lines 26 is equal to the physical spacing between the black lines 26 on the distance correction chart 25, the number of pixels between the pixel positions of adjacent black lines 26 is set to the number of pixels corresponding to the physical spacing between the black lines 26 on the distance correction chart 25. If the spacing between the pixel positions of the multiple black lines 26 is an integer multiple of two or more of the physical spacing between the black lines 26 on the distance correction chart 25, the number of pixels between the pixel position obtained by interpolating the black line 26 and the pixel position of the adjacent black line 26 is set to the number of pixels corresponding to the physical spacing between the black lines 26 on the distance correction chart 25 (step S205). In other words, the adjacent distance condition is that the spacing between the pixel positions of the multiple black lines 26 is an integer multiple of one or more of the physical spacing between the black lines 26 on the distance correction chart 25, and the detailed number of pixels corresponding to the spacing between adjacent black lines is determined depending on whether or not this condition is met (step S205).

[0043] This will be explained in detail using an example of a black line position 26 shown in FIG. 19. The positions (circled 1), (circled 2), and (circled 3) are spaced 23.5 pixels apart, and if the spacing between the black lines 26 were 1 mm, it would be understood that no black lines 26 would have been missing along the way. On the other hand, there is a gap of 47 pixels between positions (circled 3) and (circled 5), which corresponds to a spacing of 2 mm, and it can be inferred that one black line 26 between them has been missing due to some abnormality. In FIG. 19, position (circled 1) indicates "1 in a circle" in the figure. Similarly, position (circled 2) indicates "2 in a circle" in the figure. The same applies to positions (circled 3) and beyond. In this specification and figures, "numbers in circles" are used in the same way.

[0044] The spacing between the black lines 26 is known, so if the spacing is 1 mm, there should be black lines 26 approximately every 23.5 pixels, and the line spacing is determined based on this. For example, after position (circled 3) in Figure 19, a black line 26 is expected to be present at position (circled 4). If a black line 26 is present within a range of 23.5 ± 1 pixels from position (circled 3), it is recognized that the black lines 26 existed at an interval of 1 mm. However, in the case of Figure 19, there is no positional information for the black line 26 at position (circled 4), so it is ignored.

[0045] Next, a black line 26 is expected to exist at position (circle 5). If a black line exists within a range of 47±1 pixels from position (circle 3), it will be recognized that the black line 26 exists at an interval of 2 mm. In the case of Figure 19, since the black line 26 exists at position (circle 5), the line interval between positions (circle 3) and (circle 5) will be recognized as 2 mm. Similarly, if a black line exists within a range of 70.5±1 pixels from position (circle 3), it will be recognized that the black line 26 exists at an interval of 3 mm.

[0046] In this way, even if the position of the black line 26 is not detected in the previous step of noise removal by pattern matching (step 2 in FIG. 14), data processing continues if the next black line 26 is recognized in the correct position. This means that even if strict conditions are set in noise removal by pattern matching to prevent erroneous detection due to chips or scratches on the black line 26 or scratches or stains on the plain white area 27, and black lines 26 are excessively ignored, the process of generating correction data can continue using the remaining correctly detected black lines 26. Of course, the expected range of the black line 26 can be set arbitrarily (set to ±1 pixel in FIG. 19).

[0047] Furthermore, by performing this process, if a black scratch 28 on a plain white area 27 shown in Figure 17 (an enlarged view of the enlarged portion of Figure 13) is erroneously detected as a black line 26, the next black line 26 will not fit within the specified ranges (23.5±1 pixels, 47±1 pixels, 70.5±1 pixels) and will be ignored. Therefore, this process also serves to remove noise that could not be completely removed by pattern matching.

[0048] 18 illustrates an example in which two consecutive black lines 26 are missing (the expected position range of the black lines 26 is 70.5±1 pixels, corresponding to an interval of 3 mm between the black lines 26). If three or more consecutive black lines 26 are missing, the adjacent distance condition is not met (step S205: No), and an error message is displayed (step S207), the process is forcibly terminated, and the user is prompted to remeasure.

[0049] It is possible to arbitrarily set the number of consecutive missing lines before prompting remeasurement. For example, if a maximum measurement error of 0.5 mm is expected at a reading length of 900 mm, the measurement error expected to occur at 3 mm is 1.7 μm. This is sufficiently small, since the size of one pixel on a 600 dpi light-receiving element 15 is 42.3 μm. (Measurement errors smaller than the size of one pixel cannot be detected.) It is also sufficiently smaller than the maximum expected measurement error.

[0050] In this way, the maximum number of missing black lines 26 that can be recognized can be determined by taking into account the maximum expected length measurement error or the size of the light receiving element 15. If it is determined that the determined adjacent distance condition is met, including the number of missing black lines 26 that can be recognized (step S205: Yes), data is recorded in the form of pixel position - adjacent black line distance (physical length) (step S206). The starting point (0 mm) position is the position where the first black line 26 is detected in each measurement.

[0051] Next, in order to guarantee the effective range of the measurement correction, a process for confirming the validity of the chart position is carried out (steps 4 and 4 in FIG. 14). The process flowcharts are shown in FIGS. 20A and 20B. FIG. 21 shows the effective document reading range and the effective range of the measurement correction value of the image reading device 100.

[0052] The effective document reading range refers to the range determined from the first pixel position to the last pixel position of the light-receiving element 15. On the other hand, the effective measurement correction value range is narrower than the effective document reading range because correction data is generated based on waveform data from the distance correction chart 25 captured by the image reading device 100. However, when used for distance measurement within a customer's process, the effective measurement correction value range must be clearly defined. Measurement correction values ​​can only be generated within the range where the black line 26 on the distance correction chart 25 exists. Therefore, to clearly define the effective measurement correction value range, it is necessary to specify the expected positions of the black line 26 on the first pixel side and the expected positions of the black line 26 on the end side and to confirm that the black line 26 actually exists within that range. Furthermore, it is necessary to confirm that the black line 26 is not an erroneous detection line due to dirt or scratches on the white solid area, and that the distance correction chart 25 is in the desired position in order to perform the first- and second-read data combination process described below.

[0053] Specifically, the correction data generation unit 211 of the processor 210 determines that the black pattern of the distance correction chart 25 has been detected within the range from the first pixel position to the final pixel position in the main scanning direction, within the range of the predetermined expected position of the black pattern on the first pixel side and the expected position of the black pattern on the final pixel side.

[0054] As an example, the first measurement (first-side measurement) will be described. FIG. 22 shows the detection status of the black line 26 up to the 188th pixel on the first side in the first measurement. The position of the black line 26 closest to the first side (circled 1 in FIG. 22) is the 16th pixel. The area surrounded by the black dotted line is the expected presence position range on the first side (0 to 23.5 pixels). It is determined whether the first observation line is within the expected presence position range (step S301). In the example shown in FIG. 22, the black line 26 actually exists within the expected presence position range. Note that if the expected presence position range is set to within 23.5 pixels from the first pixel / end pixel, the guaranteed effective range of the length measurement correction value is 1 mm inside the effective document reading range (approximately 2 mm shorter than the effective document reading range).

[0055] Next, to confirm that the black line 26 in Figure 22 (Circle 1) is not a falsely detected line and that data has been correctly acquired up to the range previously expected for the data connection process with the second measurement data, check whether the distance measurement chart 25 was placed in the correct position when the image was taken.

[0056] FIG. 22 shows an example of the detection status of the black line 26 up to the 188th pixel on the first side in the first measurement. After confirming that the position of the black line 26 closest to the first side (first observation line) is within the expected presence range (step S301: Yes), it is then confirmed whether the second line (circled 2 in FIG. 22) and the third line (circled 3 in FIG. 22) are consecutive within a range of 23.5±1 pixels (corresponding to 1 mm) (steps S302 and S303). This is because if three consecutive black lines 26 are detected, it can be confirmed that the distance correction chart 25 is covering at least the first side. The number of consecutive lines required to determine whether the chart is correctly positioned can be set arbitrarily.

[0057] Similarly, on the END side, it is confirmed that the black line 26 closest to the END (the final observation line, circled 1 in Figure 23) is present within the expected position range (in the case of Figure 23, the range within 23.5 pixels from the first final measurement position of 12,000 pixels, the range surrounded by the black dotted line frame) (step S304).

[0058] Then, it is checked whether the previous observation line (circled 2 in FIG. 23) and the observation line two lines before that (circled 3 in FIG. 23) are continuous within a range of 23.5±1 pixels (corresponding to 1 mm) (steps S305 and S306). In the example of FIG. 23, both conditions are met, and there is no problem with the position on the END side.

[0059] Since it was confirmed that the distance correction chart 25 was applied to both the 1st side and the END side, and since it was confirmed that the position of the first black line 26 on the 1st side and the position of the last black line 26 on the END side were within the expected position range, the image reading device 100 determines that there was no problem with the position of the distance correction chart 25 at the time of the first measurement (steps S301 to S306: Yes).

[0060] On the other hand, if the position of the first black line 26 on the 1st side and the position of the last black line 26 on the END side are not within the expected position range (steps S301 and S304: No), an error message is displayed (step S307) and the process ends. Also, if the distance between the first and second observation lines, the second and third observation lines, the last observation line and the observation line immediately preceding it, or the observation line immediately preceding it and the observation line immediately preceding it are not within the range of 23.5±1 pixels (corresponding to 1 mm) (steps S302, S303, S305, and S306: No), an error message is displayed (step S307) and the process ends.

[0061] If the processing of steps S301 to S306 is being executed for the first data (step S308: No), the process returns to step S301, and the processing of steps S301 to S306 is executed for the second data to determine whether there is a problem with the position of the distance correction chart 25. If the processing of steps S301 to S306 is being executed for the second data (step S308: Yes), the process ends.

[0062] Completion of this process determines (guarantees) the valid range of the length measurement correction values ​​and determines that there will be no problems in the process of combining the first measurement data and the second measurement data, which will be described later.

[0063] Next, when multiple distance correction charts 25 are arranged in the main scanning direction, data of the number of pixels corresponding to the physical spacing between the black patterns on each distance correction chart 25 is combined. For example, the first measurement data and the second measurement data are combined (steps 5 and 5 in FIG. 14). A processing flowchart is shown in FIG. 24. Note that this process can be skipped if the reading length of the image reading device 100 is short and the distance correction chart 25 is sufficiently longer than the reading length, since divided measurement is not necessary.

[0064] This will be explained using the data processing example shown in Figure 25. First, the first measurement result is read (step S401). In the example of Figure 25, the end position of the first measurement data acquisition is set to the 12,000th pixel. Next, the pixel position of the final black line 30 of the first measurement is identified (step S402), and then the first measurement result is copied to the final result file (step S403).

[0065] Next, the cumulative physical length is calculated from the spacing between adjacent black lines calculated and recorded in the physical length confirmation and line spacing confirmation process (step 3 in FIG. 14).The cumulative physical length is then recorded together with the pixel position of the final black line 30 from the first measurement result copied to the final result file (step S404).The cumulative physical length is calculated by adding up the value of the spacing between adjacent black lines up to the pixel position to be calculated.

[0066] Next, the second measurement data is read (step S405), and it is determined whether the black line in the second measurement data has exceeded the pixel position of the final black line in the first measurement (step S406). If there is a first valid start black line 29 that has exceeded the pixel position of the final black line 30 in the first measurement (step S406: Yes), an offset value is calculated (step S407). Note that even if the final black line 30 in the first measurement and the black line in the second measurement coincide in step S406, the process may proceed to step S407. As long as the black line in the second measurement data has not exceeded the pixel position of the final black line 30 in the first measurement (step S406: No), reading of the second measurement data continues (step S405).

[0067] The offset value calculated in step S407 is the value obtained by subtracting the pixel position of the first final black line 30 from the pixel position of the effective start black line 29 that first exceeded the first final black line 30 in the second measurement data. If the offset value exceeds a preset value (step S408: Yes), an error message and a prompt to remeasure are displayed (step S409), and the process ends. This is because there is no distance correction data between the first final black line 30 and the pixel position of the effective start black line 29 that exceeded the first final black line 30 in the second measurement data, and it is desirable to set an upper limit for this offset value in consideration of the accuracy of distance correction.

[0068] The offset tolerance can be set arbitrarily. How the offset tolerance should be set can be considered in the same way as the concept of setting tolerances in the physical length confirmation and line spacing confirmation processes. If the offset value is within the preset tolerance (step S408: No), the actual measurement data merging process is started. The second measurement data is recorded additionally following the first measurement data, as described below.

[0069] If the offset value is 0 (the final black line 30 of the first measurement and the effective start black line 29 of the second measurement completely coincide) (step S410: Yes), the final black line 30 of the first measurement and the effective start black line 29 of the second measurement overlap, and the second measurement data is recorded sequentially, starting from the black line next to the overlapping effective start black line 29, with pixel positions and cumulative physical length data (step S412). If the offset value is not 0 (step S410: No), distance A is calculated by multiplying the pixel position of the first effective start black line 29 of the second measurement by 1 pixel size, adding distance A to the cumulative physical length of the final black line 30 of the first measurement, and recording this as the cumulative physical length of the effective start black line 29 of the second measurement (step S411). Thereafter, the spacing between adjacent black lines is added from the second effective black line onwards to the final black line in the second measurement, and this is recorded as the cumulative physical length for the image reading device 100, along with the black line pixel positions (step S412). By the above process, when a plurality of distance correction charts are arranged in the main scanning direction, data of the number of pixels corresponding to the cumulative physical length, which is the physical distance between the black patterns of each distance correction chart, can be combined and processed.

[0070] So far, we have explained the data processing method for generating distance correction data 221 by capturing an image of distance correction chart 25 with image reading device 100 to address the issue of variations in the number of pixels between points due to manufacturing variations in image reading device 100 or manufacturing variations in rod lens array 11. These processes must be performed individually for each image reading device 100 for which distance correction data 221 is to be generated.

[0071] Next, a correction method for the problem that the position of the light receiving element 15 changes due to thermal expansion when the image reading device 100 is powered on, resulting in a measurement error during length measurement, will be described.

[0072] Patent Document 2 describes a method for correcting the effects of thermal expansion, although the correction chart has a different format. However, this method is related to correcting the thermal expansion of the calibration plate itself, and does not take into account the expansion and contraction due to thermal expansion of the image reading device itself.

[0073] FIG. 26 is a graph showing the relationship between time and temperature after power-on of the image reading device 100 (contact image sensor: CIS). When power is turned on, the temperature of the image reading device 100 begins to rise, and after 120 minutes has passed since power-on, it reaches thermal equilibrium and the temperature change becomes small. The image reading device 100 is incorporated into a device that inspects objects for scratches, chips, etc., and therefore operates continuously for long periods of time. Therefore, it is effective to understand the effects of thermal expansion approximately 120 minutes after power-on, when it reaches thermal equilibrium.

[0074] However, if it is necessary to wait 120 minutes to obtain the distance correction data, a delay will occur in the process of obtaining the distance correction data during the assembly process of the image reading device 100, which will result in an increase in the manufacturing cost of the image reading device 100.

[0075] Therefore, distance correction data at a certain temperature is acquired during the assembly process of the image reading device 100, and distance correction data at temperatures that could not be measured during the assembly process can be estimated using parameters determined experimentally in advance. This makes it possible to correct the temperature dependency of the distance correction data without impairing the productivity of the image reading device 100.

[0076] Therefore, here, a method for experimentally extracting the temperature dependency of distance correction data in advance will be described. First, the measurement environment required for experimentally extracting the temperature dependency of distance correction data will be described. The measurement environment used is the environment shown in FIG. 10.

[0077] The image reading device 100 is fixed to a table 20, the flatness of which is ensured, by fixtures 21 and 22 via a rail 24 that is movable in the main scanning direction. However, the fixture 21 on the 1st side fixes the image reading device 100 with, for example, a screw. This allows the image reading device 100 to be fixed on the 1st side during measurement. The fixture 22 on the END side is configured so that the image reading device 100 can move only in the main scanning direction. When acquiring distance correction data, a distance correction chart 25 is placed on the table 20. If the effective reading length of the image reading device 100 is long, the length of the distance correction chart 25 may be shorter than the effective reading length due to chart manufacturing reasons. In this case, correction data can be acquired by dividing the 1st pixel side and the END pixel side of the image reading device 100.

[0078] An infrared sensor 23 is placed on the table 20 to monitor the temperature of the image reading device 100. While the temperature characteristics of the distance correction data are being measured using the distance correction chart 25, the infrared sensor 23 measures the time-temperature relationship from when the power is turned on for the image reading device 100. In Fig. 10, the infrared sensors 23 are placed in three locations, but the number of locations where they are placed can be changed as desired.

[0079] The temperature correction process will be described with reference to Fig. 27. Fig. 27 is a flowchart of data processing for determining the temperature correction coefficient. First, the image reading device 100 for acquiring temperature characteristics is placed on the table 20, and the data processing device 200 acquires the CIS temperature / distance correction chart waveform using the distance correction chart 25 (step 11).

[0080] Next, the correction data generating unit 211 of the processor 210 performs the same processes as steps 1 to 5 in Fig. 14 on the distance correction chart acquired in step 11, and converts the waveform data of the distance correction chart 25 into the relationship between pixel position and cumulative physical length (step S12). That is, the processes of steps 13 to 18 in Fig. 27 correspond to the temperature correction process of step 6 performed after steps 1 to 5 in Fig. 14.

[0081] Here, image acquisition and conversion into the relationship between pixel position and cumulative physical length in steps 1 to 5 of Figure 14 must be performed immediately after power-on, when the image reading device 100 reaches thermal equilibrium, and at several points in between. In Figure 29 (described later), images are acquired at 0 minutes (immediately after power-on), 5 minutes, 10 minutes, 30 minutes, and 120 minutes (thermal equilibrium state), but any time other than 0 minutes (immediately after power-on) and the thermal equilibrium state can be set. However, shortening the time interval allows for more data points when converting to time-position deviation data in step 14 (described later), which is advantageous when calculating the temperature correction coefficient in steps 16, 17, and 18.

[0082] The correction data generator 211 records the relationship between pixel position and cumulative physical length obtained in step S12 together with the time elapsed since power-on. If the correction data acquisition is divided into two or more times, the data is combined at this stage (step 5 in FIG. 14).

[0083] Next, the correction data generation unit 211 approximates the amount of deviation for each pixel from the discrete waveform data (every 1 mm) at each elapsed time, and generates continuous waveform data (step 13). A more detailed explanation of this process will be given. First, the amount of positional deviation at the pixel position is calculated using the calculation method shown in the following equation (1). Positional deviation amount = cumulative physical length at the pixel position - pixel size of one pixel x (pixel position - 26 pixel position of the first black line) (1)

[0084] If the amount of positional misalignment is a positive value, the captured image is smaller than the original. If the amount of positional misalignment is a negative value, the captured image is larger than the original. The calculation result is data in the form of pixel position - amount of positional misalignment at that pixel position. The pixel position here is the position of the black line 26 on the distance correction chart 25, and is data for every 23.5 pixels (every 1 mm).

[0085] In step 13, when considering temperature-related correction, it is difficult to extract the amount of misalignment at the same pixel position at each temperature when performing step 14 (described later in FIG. 27 ) using only the position information of the black line 26. Therefore, the correction data generation unit 211 calculates the amount of misalignment for pixel positions other than the pixel position where the black line 26 is recorded, and continuously records the amount of misalignment for all pixels of the image reading device 100. An example of the calculation results of the amount of misalignment on a pixel-by-pixel basis is shown in FIG. 28 . There is data for the black line 26 at pixels 11013.5 and 11037.5. There is no data for the black line 26 at pixel positions between these two. For pixel positions between these two, the processor 210 uses linear approximation to calculate the amount of misalignment approximately from the amount of misalignment at the recorded position of the black line 26. Then, data on the approximate value of the amount of misalignment for each pixel is recorded. In this way, the relationship between pixel position and amount of misalignment at each elapsed time of the image reading device 100 is recorded. FIG. 29 shows an example of measurement of the change in the amount of positional deviation over time.

[0086] Next, the correction data generation unit 211 extracts the positional misalignment amount for each pixel position at a certain interval (for example, 1000 pixel intervals) from the pixel position-positional misalignment amount relationship obtained in step 13, and generates data on the relationship between elapsed time and positional misalignment amount (step 14). That is, the pixel position-positional misalignment amount data is converted into positional misalignment amount data with respect to elapsed time at the same pixel position. Figure 30 shows an example of measuring the change in positional misalignment amount with respect to elapsed time.

[0087] In the example of Figure 30, the misalignment amounts at 100 pixels, 1000 pixels, 2000 pixels, 10,000 pixels, 20,000 pixels, 21,000 pixels, and 21,500 pixels are extracted from the data in Figure 29. The more pixel positions extracted, the higher the accuracy of the approximation process described below. Therefore, it is desirable to extract data at as many pixel positions as possible, but the extraction frequency can be set as desired. Since the horizontal axis has been converted to time, the correction data generation unit 211 uses the relationship between the body temperature and the elapsed time since the power-on of the image reading device 100, acquired by the infrared sensor 23, to convert the data obtained in step 14 into the relationship between the body temperature of the image reading device 100 and the amount of misalignment (step 15). That is, the correction data generation unit 211 calculates the change in the amount of misalignment with temperature shown in Figure 31 based on the relationship between elapsed time and temperature shown in Figure 26 and the relationship between elapsed time and amount of misalignment shown in Figure 30.

[0088] 31 has been calculated, the correction data generating unit 211 then performs a linear approximation calculation for each pixel position using the least squares method (step 16). The amount of linear expansion is defined by the following equation (2): Amount of linear expansion (change in positional deviation) ΔL=α×L×ΔT (2) where α is the linear expansion coefficient, L is the pixel position from the first pixel, and ΔT is the temperature difference.

[0089] The linear expansion coefficient α is a value determined by the material. However, since the image reading device 100 is a combination of various components, it is not easy to estimate. If L is considered a constant, the relationship between temperature and misalignment amount can be linearly approximated from the linear expansion equation. For the graph in FIG. 31, if a linear approximation is performed using, for example, the least squares method, the misalignment amount is expressed by the following equation (3). In equation (3), b is a term that is not temperature-dependent, i.e., it can be ignored when determining the amount of change in misalignment amount. (Misalignment amount) = a × (Temperature) + b (3)

[0090] Since Figure 31 creates a graph for each pixel position L, which corresponds to the distance from the starting point (pixel position) of the distance correction data, focusing on each line representing the misalignment amount at each pixel position L, the pixel position L can be considered a constant. Each line for each pixel position L in Figure 31 indicates that the misalignment amount changes as the temperature changes. The slope a of the linear approximation in the temperature range shown in the graph in Figure 31 increases the farther away from the starting point (pixel position) of the distance correction data. This means that for a given temperature T, the misalignment amount increases as the pixel position L increases. While the linear expansion coefficient α generally depends on temperature, as shown in Figure 26, the ΔT of the image reading device 100 is at most about 20°C, and the operating environment is room temperature (about 24°C), the linear expansion coefficient α can be considered constant. Therefore, the slope a of equation (3) when linearly approximating the change in misalignment amount with temperature corresponds to α × L in equation (2) for linear expansion ΔL, but only the pixel position L can be treated as a variable. The correction data generator 211 extracts the gradient a (a1, a2, a3, ... a7) of the linear approximation result at each pixel position L to create the relationship between pixel position and gradient a as shown in Fig. 32 (step S17). Fig. 32 is a diagram showing how the gradient a of the positional misalignment amount with respect to temperature changes depending on the pixel position.

[0091] Since the gradient a has a linear response to the pixel position L, the correction data generating unit 211 again performs linear approximation on the graph using the least squares method (step S18). When the linear approximation is performed, the following equation (4) is derived: (gradient) a = c × (pixel position from the 1st pixel) + d (4)

[0092] Therefore, the amount of positional misalignment is expressed by the following equation (5). In equation (5), b is a term that does not depend on temperature, that is, it can be ignored when calculating the amount of change in positional misalignment. (Amount of positional misalignment) = (c × (pixel position from the 1st pixel) + d) × (temperature) + b (5)

[0093] Furthermore, the necessary information is the positional misalignment amount change ΔL when the body temperature of the image reading device 100 reaches Tcis, which is calculated based on the positional misalignment amount L' measured on the mass production line when the body temperature of the image reading device 100 is at a certain temperature T'. ΔL is calculated using the following equation (6): Positional misalignment amount change ΔL=(c×(pixel position from the 1st pixel)+d)×(Tcis-T') (6)

[0094] In the above formula (6), c and d are items related to the linear expansion coefficient and are temperature correction coefficients that should be determined in advance through experiments from the graph shown in Fig. 32. Since these coefficients are constant if the materials or combination of parts do not change, the values ​​can be determined experimentally in advance for each model.

[0095] Furthermore, by monitoring Tcis and determining the pixel position for which the amount of misalignment is to be calculated, the change in the amount of misalignment can be calculated using equation (6). Regarding Tcis, if the image reading device 100 is warmed up at startup and used in a thermal equilibrium state, there is no need to monitor it (however, it is necessary to know the temperature at which the thermal equilibrium state is reached). Furthermore, since T' is the body temperature of the image reading device 100 measured on the mass production line, it is sufficient to record this in the storage unit 220 within the image reading device 100.

[0096] Furthermore, when the cumulative physical length at a certain pixel position is P', and the body temperature of the image reading device 100 is Tcis, the cumulative physical length Ptemp taking into account the temperature dependency at that pixel position is expressed by the following equation (7): Ptemp=P'+ΔL (7)

[0097] In equation (7), P' and ΔL depend on the pixel position from the first pixel, so they can be calculated using the pixel position for which the positional deviation is desired. Here, ΔL also depends on Tcis, so ΔL is calculated by substituting the pixel position and Tcis into equation (6). In this way, the influence of temperature can be corrected using the temperature correction coefficient shown in equation (6) without measuring the temperature dependency of the positional deviation for each of the image reading devices 100.

[0098] Fig. 33 is a flowchart of the temperature correction process (step 6 in Fig. 14) using a temperature correction coefficient. More specifically, Fig. 33 shows a processing method for applying correction of the positional misalignment amount due to linear expansion to the relationship between pixel position and cumulative physical length derived in steps 1 to 5 in Fig. 14, based on data on changes in the positional misalignment amount with respect to temperature derived by the process shown in the flowchart in Fig. 27.

[0099] As mentioned above, the change in the amount of positional misalignment depends on the pixel position. The temperature correction process shown in FIG. 33 first requires the temperature of the image reading device 100 when the distance correction data was measured, the temperature at which temperature correction is desired, and the temperature of the image reading device 100 (CIS) body when the distance correction data was measured. The temperature when the distance correction data was measured is pre-recorded in the memory unit 220 of the data processing device 200. The correction data generation unit 211 of the processor 210 then adds or subtracts the amount of positional misalignment change ΔL calculated using equation (6) above to or from the cumulative physical length (step S501), thereby completing the process. This process allows the number of pixels corresponding to the cumulative physical length, which is the physical spacing between black patterns on the distance correction chart, to be temperature-dependently corrected. If this temperature correction process is not required, step 6 in FIG. 14 can be skipped.

[0100] In the flowchart of FIG. 14 , the correction data generation unit 211 of the processor 210 finally processes the output data (step 7 in FIG. 14 ). Because steps 1 through 6 were processed in the pixel position-cumulative physical length data format, skipping step 7 would result in the distance correction data 221 being the format output after step 6. However, since the pixel positions recorded in the distance correction data 221 are values ​​for each interval between the black lines 26 on the distance correction chart 25, the cumulative physical length is not recorded for every pixel position, which may make the data less user-friendly. In such cases, the data can be converted to continuous data by converting the cumulative physical length into the positional misalignment amount and performing an approximate calculation for the pixels between adjacent lines, as performed in step 13 of FIG. 27 . In this way, the correction data generation unit 211 of the data processing device 200 converts the data into a data format convenient for the end user of the image reading device 100 and outputs the data (step 7 in FIG. 14 ).

[0101] The distance measurement unit 212 of the data processing device 200 uses the distance correction data 221 generated by the processing of the flowchart in Figure 14 to measure the distance between multiple points in the main scanning direction based on the waveform data obtained by reading the image of the measurement object using the image reading device 100.

[0102] Second Embodiment A second embodiment according to the present disclosure will be described with reference to the drawings.

[0103] Fig. 34 is a processing flowchart of a distance correction chart waveform according to embodiment 2. The processing flowchart of Fig. 34 includes a step of checking white output (step 101), a step of checking whether output has been reduced due to foreign matter such as dust (step 102), and a step of removing dust (step 103) before the steps from waveform capture to binarization (step 1) in the flowchart of Fig. 14.

[0104] In the first embodiment, the data processing device 200 generates distance correction data 221 for correcting the physical length in the main scanning direction of the image reading device 100 based on output data from the image reading device 100 that reads black and white patterns formed at equal intervals in the main scanning direction on the distance correction chart 25. However, if the distance correction chart 25 is read when the surface of the first transparent body 3 is dirty or has foreign matter attached thereto, the dirt or foreign matter on the surface of the first transparent body 3 will be erroneously recognized as black lines 26 on the distance correction chart 25.

[0105] Therefore, in the second embodiment, a white output confirmation step (step 101) is provided before the waveform capture to binarization process (step 1), and a white chart for white luminance correction (for correcting the amount of light received by the light receiving element 15) is read in advance by the image reading device 100 to confirm the white output (white correction process).

[0106] If the surface of the first transparent body 3 is dirty or has foreign matter attached thereto, the light receiving element 15 corresponding to that location will receive a reduced amount of light. Therefore, the state of the surface of the first transparent body 3 is determined in a step (step 102) for checking for a decrease in output due to foreign matter such as dust. If there is data from the light receiving element 15 indicating a decrease in the amount of light received (light receiving data indicating a decrease in white output) (step 102: Yes), it is determined that the surface of the first transparent body 3 is dirty or that foreign matter is attached to the surface, and the image reading device 100 proceeds to a dust removal step (step 103). In step 103, the first transparent body 3 is cleaned, and the process returns to the white output confirmation step (step 101). If there is no light receiving data from the light receiving element 15 indicating a decrease in the amount of light received (step 102: No), it is determined that the surface of the first transparent body 3 is normal, free from dirt or foreign matter, and the process proceeds to waveform capture and binarization processing (step 1).

[0107] By carrying out the above steps (steps 101 to 103), reading errors of the distance correction chart 25 can be prevented.

[0108] In the above, we have explained a flow in which, when it is determined that the surface of the first transparent body 3 is dirty or that foreign matter is attached, the image reading device 100 proceeds to the dust removal step (step 103) and cleans the first transparent body 3. However, in the output reduction confirmation step (step 102), data from the light receiving element 15 where the amount of received light has decreased may be set as invalid data, and the process may proceed to waveform capture and binarization processing (step 1). In this case, the data from the light receiving element 15 is treated as invalid data in the subsequent steps.

[0109] Third Embodiment A third embodiment of the present disclosure will be described with reference to the drawings.

[0110] Fig. 34 is a processing flowchart of a distance correction chart waveform according to embodiment 3. In the processing flowchart of Fig. 34, a process of determining whether or not a lens joint is present (step 301) and a process of invalidating the temperature correction value near the lens joint (step 302) are provided between the temperature correction process (step 6) and the output data process (step 7) in the flowchart of Fig. 14 .

[0111] In the case of an image reading device 100 that is long in the main scanning direction, one rod lens array 11 cannot cover the image reading range, so several rod lens arrays 11 may be joined together in the main scanning direction as shown in Fig. 35. Fig. 35 shows a case in which a rod lens array 11A and a rod lens array 11B are joined together in the main scanning direction.

[0112] When multiple rod lens arrays 11 (11A, 11B) are joined in the main scanning direction, the rod lenses 11A and 11B are joined with a sealant 111. At the joint between the rod lenses 11A and 11B, which includes the sealant 111, a joining error may occur between the two joined rod lens arrays 11 (11A, 11B). As shown in FIG. 36 , in the image reading device 100, the rod lens arrays 11 (11A, 11B) are pressed against the plate 31 in the sub-scanning direction by the lens plate 112 and the adjustment screw 113, so that no misalignment occurs in the sub-scanning direction (Y direction) at the joint. For the same reason, no misalignment occurs in the main scanning direction (X direction).

[0113] However, in the direction perpendicular to the main scanning direction and the sub-scanning direction, a joining error may occur in the rod lens array 11 (11A, 11B) at the joint. Specifically, as shown in Figure 37, in adjacent rod lenses 16 sandwiching a joint, the distance between the optical axes may differ between one end (e.g., the light entrance portion) and the other end (e.g., the light exit portion).

[0114] In the rod lens 16 at the joint, if the distance between the optical axes is the same at one end (e.g., the light entrance end) and the other end (e.g., the light exit end) (Figure 37(a)), measurement correction of the image reading device 100 is possible using steps 1 to 6 in Figure 34, as in embodiment 1.

[0115] However, if the distance between the optical axes of one end (e.g., the light entrance portion) and the other end (e.g., the light exit portion) of the rod lens 16 at the joint is different (FIG. 37(b)), the temperature correction value at the joint may show a singular point (discontinuity point), as shown in FIG. 38.

[0116] Therefore, data near the junction of the rod lens array 11 (11A, 11B) where the temperature correction value may show a singularity is invalidated (for example, FAULT-DATA is forcibly input), and the process moves to output data processing (step 7).

[0117] Since the positions of the joints between the rod lens arrays 11 (11A, 11B) are known in advance, the light-receiving data from a predetermined number of light-receiving elements 15, including the joints, is invalidated. For example, the light-receiving data from 50 light-receiving elements 15 in the front and rear of the main scanning direction (i.e., 100 light-receiving elements in total) is invalidated (for example, FAULT-DATA is forcibly input) and the process proceeds to output data processing (step 7).

[0118] That is, if the image reading device 100 is configured with a plurality of rod lens arrays 11 connected in the main scanning direction, the lens joint presence / absence determination process determines "present" (step S301: Yes), and the data processing device 200 performs a process of invalidating the temperature correction values ​​near the lens joints (step S302). Specifically, the data from a predetermined number of light receiving elements 15 including the joints of the rod lens array 11 is invalidated, and the process proceeds to output data processing (step S7), where the measurement results of the distances between a plurality of points in the main scanning direction are output.

[0119] If the image reading device 100 is composed of one rod lens array 11, the data processing device 200 determines that there is no lens joint in the lens joint presence / absence processing (step 301: No), and proceeds to output data processing (step 7), where the measurement results of the distance between multiple points in the main scanning direction are output using the results obtained in step 6.

[0120] By performing the above processing, it is possible to perform length measurement correction for the image reading device 100 even when the image reading device 100 is configured with a plurality of rod lens arrays 11 connected in the main scanning direction.

[0121] [Supplementary Note 1] A distance measurement method in which an image reading device having pixels which are light-receiving elements arranged in the main scanning direction reads a distance correction chart on which black and white patterns are formed at regular intervals in the main scanning direction as waveform data of the black and white patterns using the light-receiving elements arranged in the main scanning direction, and measures the distance between a plurality of points in the main scanning direction based on the read waveform data, the distance measurement method comprising: a first step of converting the waveform data into a binary waveform that is binarized based on a predetermined threshold value; a second step of comparing the binary waveform with the black and white patterns formed on the distance correction chart, and converting the binary waveform into edge information which is falling edges and rising edges with noise removed; and a third step of comparing the edge information with the physical length of the black and white patterns on the distance correction chart, and deriving the number of pixels which are the light-receiving elements corresponding to the physical interval between adjacent black patterns in the main scanning direction. [Supplementary Note 2] The distance measurement method according to Supplementary Note 1, wherein in the first step, a position where the distance correction chart changes from a white pattern to a black pattern is set as a falling position where the value of the binarized waveform changes from 1 to 0, and a position where the distance correction chart changes from a black pattern to a white pattern is set as a rising position where the value of the binarized waveform changes from 0 to 1.[Supplementary Note 3] The second step includes: a 21st step of scanning the image reading device in the main scanning direction and validating detection of the falling edge, which is a starting point of the black pattern in the main scanning direction, when a predetermined measurement start pixel is reached; a 22nd step of invalidating light-receiving data from the pixel at which the falling edge is detected, when the falling edge is detected; a 23rd step of validating the light-receiving data and validating detection of the rising edge when a predicted end pixel of the black pattern in the main scanning direction is reached after detecting the falling edge; and a 24th step of invalidating the light-receiving data from the pixel at which the rising edge is detected, when the rising edge is detected, or invalidating the light-receiving data at a point when scanning beyond the range of the predicted end pixel of the black pattern in the main scanning direction even if the rising edge is not detected; and the distance measurement method described in Supplementary Note 1 or Supplementary Note 2, wherein the 21st step, the 22nd step, the 23rd step, and the 24th step are repeated when a measurement start pixel for a next black pattern located in the scanning direction is reached.[Supplementary Note 4] The third step includes: a 31 step of deleting the falling edge for which the adjacent rising edge corresponding to the falling edge has not been detected; a 32 step of setting the pixel for which the falling edge has been detected as a falling detection pixel and the pixel for which the adjacent rising edge corresponding to the falling edge has been detected as a rising detection pixel, and setting the average position of the position of the falling detection pixel and the position of the rising detection pixel as the pixel position of the black pattern; and a 33 step of deriving the pixel positions of the black pattern by performing the process of the 32 step on the plurality of falling edges and the rising edges repeatedly detected in the second step. a 34th step of comparing an interval between pixel positions of the plurality of black patterns with a physical interval between the black patterns on the distance correction chart, and, if the interval between pixel positions of the plurality of black patterns is equal to the physical interval between the black patterns on the distance correction chart, setting the number of pixels between the pixel positions of adjacent black patterns to the number of pixels corresponding to the physical interval between the black patterns on the distance correction chart, and, if the interval between pixel positions of the plurality of black patterns is an integer multiple of two or more of the physical interval between the black patterns on the distance correction chart, setting the number of pixels between the pixel position obtained by interpolating the black pattern and the pixel position of the adjacent black pattern to the number of pixels corresponding to the physical interval between the black patterns on the distance correction chart. [Supplementary Note 5] The distance measurement method according to any one of Supplementary Note 1 to Supplementary Note 4, further comprising a 4th step of determining that the black pattern on the distance correction chart has been detected within a predetermined range between an expected position of the black pattern on the first pixel side and an expected position of the black pattern on the final pixel side within a range from a first pixel position to a final pixel position located in the main scanning direction. [Supplementary Note 6] The distance measurement method according to any one of Supplementary Note 1 to Supplementary Note 5, further comprising a fifth step of arranging a plurality of the distance correction charts in the main scanning direction and combining data of a number of pixels corresponding to the intervals between the black patterns of each of the distance correction charts.[Supplementary Note 7] The distance measurement method according to any one of Supplementary Note 1 to Supplementary Note 5, further comprising a fifth step of arranging a plurality of the distance correction charts in a main scanning direction and combining data on the number of pixels corresponding to the physical spacing between black patterns on each of the distance correction charts. [Supplementary Note 8] The distance measurement method according to any one of Supplementary Note 1 to Supplementary Note 7, further comprising a sixth step of processing to correct the number of pixels corresponding to the spacing between black patterns on the distance correction chart based on temperature dependency data on pixel positional deviation in the main scanning direction of the image reading device, which has been acquired in advance. [Appendix 9] A distance measurement method according to any one of Appendices 1 to 8, further comprising: a white output confirmation step, prior to the first step, of reading a white chart with the image reading device and confirming the white output of the image reading device; and an output reduction confirmation step, in which, if the received light data indicates a decrease in the white output output by the light receiving element, it is determined that dirt or foreign matter is attached to a first transparent body of the image reading device, and the first transparent body is cleaned, and the method branches to a dust removal step in which the white output confirmation step is performed again; and, if there is no decrease in the white output output by the light receiving element, it is determined that the first transparent body is normal, and the method proceeds directly to the first step. [Supplementary Note 10] The distance measurement method according to Supplementary Note 8, further comprising, after the sixth step, a lens joint presence / absence determination step in which the image reading device determines whether or not a lens joint where a plurality of rod lens arrays are joined in the main scanning direction is present, and if a lens joint is present in the lens joint presence / absence determination step, the received light data from a predetermined number of the light receiving elements including the lens joint is invalidated and a result of measuring the distance between a plurality of points in the main scanning direction is output, and if a lens joint is not present in the lens joint presence / absence determination step, a result of processing in the sixth step is used to output a result of measuring the distance between a plurality of points in the main scanning direction. [Supplementary Note 11] The distance measurement method according to Supplementary Note 10, wherein the presence or absence of the lens joint is input in advance.

[0122] The present disclosure allows various embodiments and modifications without departing from the broad spirit and scope of the present disclosure. Furthermore, the above-described embodiments are intended to explain the present disclosure and do not limit the scope of the present disclosure. That is, the scope of the present disclosure is defined by the claims, not the embodiments. Various modifications made within the scope of the claims and the meaning of equivalent disclosures are considered to be within the scope of the present disclosure.

[0123] This application is based on Japanese Patent Application No. 2023-064812 filed on April 12, 2023, and Japanese Patent Application No. 2024-009960 filed on January 26, 2024. The entire specifications, claims, and drawings of Japanese Patent Application No. 2023-064812 and Japanese Patent Application No. 2024-009960 are incorporated herein by reference.

[0124] 100 Image reading device, 2 Side plate, 3 First transparent body, 4 Chart, 5 Output waveform of chart 4, 7 Second frame, 8 Light source, 9 First frame, 10 Second transparent body, 11, 11A, 11B Rod lens array, 12 Substrate support plate, 13 Substrate, 14 Substrate, 15 Light receiving element, 16 Rod lens, 17 Fixed side plate, 18 Original, 19 Erect life-size image of original 18, 20 Table with ensured flatness, 21 Fixture, 22 Fixture, 23 Infrared sensor, 24 Rail, 25 Distance correction chart, 26 Black line, 27 White plain area, 28 Black scratch, 29 Second effective start black line, 30 First final black line, 31 Plate, 111 Sealant, 112 Lens plate, 113 Adjustment screw, 200 Data processing device, 210 Processor, 211 Correction data generation unit, 212 distance measurement unit, 220 storage unit, 221 distance correction data, 1000 distance measurement system.

Claims

1. In a distance measurement method in which a distance correction chart having a black-and-white pattern formed at regular intervals in the main scanning direction is read as waveform data of the black-and-white pattern by a light-receiving element pixel arrayed in the main scanning direction by an image reading apparatus, and distances between a plurality of points in the main scanning direction are measured based on the read waveform data, a first step of converting the waveform data into a binarized waveform obtained by binarizing the waveform data based on a determined threshold value; a second step of comparing the binarized waveform with the black-and-white pattern formed on the distance correction chart and converting the binarized waveform into edge information including a falling edge and a rising edge from which noise has been removed; a third step of comparing the edge information with the physical length of the black-and-white pattern of the distance correction chart and deriving the number of the pixels which are the light-receiving elements corresponding to the physical interval between adjacent black patterns in the main scanning direction; A distance measurement method comprising the above steps.

2. The distance measurement method according to claim 1, wherein in the first step, a position where the distance correction chart changes from a white pattern to a black pattern is set as a falling position where the value of the binarized waveform changes from 1 to 0, and a position where the distance correction chart changes from a black pattern to a white pattern is set as a rising position where the value of the binarized waveform changes from 0 to 1.

3. The second step includes: a 21st step of scanning the image reading apparatus in the main scanning direction and enabling detection of the falling edge which is the starting point of the black pattern in the main scanning direction when reaching a predetermined measurement start pixel; a 22nd step of invalidating received light data after the pixel at which the falling edge is detected when the falling edge is detected; a 23rd step of validating the received light data and enabling detection of the rising edge when reaching a predicted end pixel of the black pattern in the main scanning direction after detection of the falling edge; a 24th step of invalidating the received light data after the pixel at which the rising edge is detected when the rising edge is detected, or invalidating the received light data when scanning exceeds the range of the predicted end pixel of the black pattern in the main scanning direction even if the rising edge is not detected. When reaching the measurement start pixel for the next black pattern located in the scanning direction, the steps of the 21st, 22nd, 23rd, and 24th are repeated, the distance measurement method according to claim 1.

4. The third step is a 31st step of deleting the falling edge where the adjacent rising edge corresponding to the falling edge is not detected; a 32nd step of designating the pixel where the falling edge is detected as a falling detection pixel and the pixel where the adjacent rising edge corresponding to the falling edge is detected as a rising detection pixel, and designating the position of the average value of the position of the falling detection pixel and the position of the rising detection pixel as the pixel position of the black pattern; a 33rd step of deriving the pixel positions of a plurality of the black patterns by performing the process of the 32nd step on the plurality of falling edges and rising edges repeatedly detected in the 2nd step; comparing the interval between the pixel positions of the plurality of black patterns with the physical interval of the black patterns in the distance correction chart, and when the interval between the pixel positions of the plurality of black patterns is equal to the physical interval of the black patterns in the distance correction chart, setting the number of pixels between the pixel positions of the adjacent black patterns as the number of pixels corresponding to the physical interval of the black patterns in the distance correction chart, and when the interval between the pixel positions of the plurality of black patterns is an integer multiple of 2 or more of the physical interval of the black patterns in the distance correction chart, setting the number of pixels between the pixel position of the black pattern after interpolation completion and the pixel position of the adjacent black pattern as the number of pixels corresponding to the physical interval of the black patterns in the distance correction chart, the distance measurement method according to claim 1, comprising a 34th step.

5. The distance measurement method according to claim 1, further comprising a 4th step of determining that the black pattern in the distance correction chart is detected within the range of the expected position of the black pattern on the first pixel side and the expected position of the black pattern on the final pixel side, which are predetermined, in the range from the first pixel position to the final pixel position located in the main scanning direction.

6. The distance measurement method according to claim 1, further comprising a 5th step of arranging a plurality of the distance correction charts in the main scanning direction and performing a combining process on data of the number of pixels corresponding to the interval between the black patterns of each distance correction chart.

7. The distance measurement method according to claim 1, further comprising a fifth step of arranging a plurality of the distance correction charts in the main scanning direction and performing a combining process on data of the number of pixels corresponding to the physical distance between the black patterns of each of the distance correction charts.

8. The distance measurement method according to any one of claims 1 to 7, comprising a sixth step of correcting the number of pixels corresponding to the physical distance between the black patterns of the distance correction chart based on the temperature dependence data of the positional deviation of the pixels in the main scanning direction of the image reading device acquired in advance.

9. Before the first step, a white output confirmation step of reading a white chart with the image reading device and confirming the white output of the image reading device, and When there is received light data in which the white output output by the light receiving element decreases, it is determined that there is dirt or foreign matter adhesion on the first transparent body of the image reading device, and when the first transparent body is cleaned and then the white output confirmation step is performed again, it branches to a dust removal step, and when there is no decrease in the white output output by the light receiving element, it is determined that the first transparent body is normal and directly proceeds to the first step. The distance measurement method according to any one of claims 1 to 7, further comprising an output decrease presence / absence confirmation step of proceeding.

10. After the sixth step, the image reading device further has a lens joint presence / absence determination processing step of determining the presence or absence of a lens joint portion to which a plurality of rod lens arrays are joined in the main scanning direction, When there is a lens joint portion in the lens joint presence / absence determination processing step, the received light data from a predetermined number of the light receiving elements including the lens joint portion is invalidated, and the measurement result of the distance between a plurality of points in the main scanning direction is output, When there is no lens joint portion in the lens joint presence / absence determination processing step, the measurement result of the distance between a plurality of points in the main scanning direction is output using the result processed in step 6. The distance measurement method according to claim 8.

11. The distance measurement method according to claim 10, wherein the presence or absence of the lens joint portion is input in advance.

12. In a distance measurement method of arranging a distance correction chart in which black and white patterns are formed at regular intervals in the x direction, which is the measurement direction, and reading the black and white pattern as waveform data with a light receiving element arranged in the x direction, and measuring the distance between a plurality of points in the x direction based on the read waveform data, A first step of converting the waveform data into a binarized waveform obtained by binarizing the waveform data based on a defined threshold value; A second step of comparing the binarized waveform with the black-and-white pattern formed on the distance correction chart and converting the binarized waveform into edge information including a falling edge and a rising edge from which noise has been removed; A third step of comparing the edge information with the physical length of the black-and-white pattern of the distance correction chart and deriving the number of the light receiving elements corresponding to the physical interval between adjacent black patterns in the x direction; A distance measurement method comprising the above.