Element angle identification device and method for identifying element angle

Abstract

To provide an element angle identification device and a method for identifying an element angle for rapidly and accurately performing processing of identifying angles of a plurality of elements on a substrate.SOLUTION: A processing unit 40 is configured to: acquire a reference image 1 in which one semiconductor element 210 is captured, on the basis of a captured image captured by an imaging unit 20; generate a plurality of rotated element images 1a to 1d in which an element part 2 appearing in the reference image 1 is rotated at a plurality of preset angles; acquire correlation values each indicating the degree of correlation between a semiconductor element 210, which is a target for specifying a rotation angle, and each of the plurality of rotated element images 1a to 1d; acquire correlation relation information indicating a relationship between the angles corresponding to the plurality of rotated element images 1a to 1d and the correlation values; and specify a rotation angle of the semiconductor element 210 based on the correlation relation information.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present invention relates to an element angle specifying device and an element angle specifying method, and more particularly to an element angle specifying device and an element angle specifying method that specify the rotation angle of an element based on a captured image. [Background technology]

[0002] BACKGROUND ART Conventionally, an element angle specifying device that specifies the rotation angle of an element based on a captured image is known (see, for example, Patent Document 1).

[0003] The above-mentioned Patent Document 1 discloses an inspection device (element angle identification device) for mounted components, which includes a storage means for storing chip (element) component patterns, an imaging unit for capturing images of chip components mounted on a jig substrate, and an inspection means for inspecting the position or angle of the chip components based on the captured image of the chip components and the generated chip component patterns.

[0004] In the inspection device of Patent Document 1, the inspection means selects, rotates, and overlaps multiple chip component patterns generated by expanding or contracting chip component patterns stored in a storage unit for each chip component imaged by an imaging unit, thereby selecting two chip component patterns with a high degree of matching through pattern matching. The inspection device of Patent Document 1 also generates multiple additional chip component patterns with a degree of expansion or contraction between the two extracted chip component patterns, and performs a process of again determining the degree of matching for each of the imaged chip components. The inspection device of Patent Document 1 repeats a series of processes, including the process of selecting two chip component patterns with a high degree of matching, the process of generating multiple additional chip component patterns based on the selected chip component patterns, and the process of obtaining the degree of matching of the chip component patterns for each of the imaged chip components, any number of times, and obtains the position and angle of the chip component based on the chip component pattern with the highest degree of matching. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Laid-Open No. 2007-103660 Summary of the Invention [Problem to be solved by the invention]

[0006] As described above, Patent Document 1 discloses an inspection device that repeats a series of processes, including selecting two chip component patterns with high matching degrees, generating multiple other chip component patterns based on the selected chip component patterns, and obtaining matching degrees of the chip component patterns for each of the captured chip components, any number of times, and obtains the positions and angles of chip components based on the chip component pattern with the highest matching degree. However, if the number of times the series of processes is repeated is small, it may be impossible to obtain a chip component pattern with a sufficiently high matching degree, resulting in low accuracy in the angle of the detected element (chip component). However, if the series of processes is repeated multiple times for one element whose angle is to be determined, it may take a very long time to determine the rotation angles of so many elements, especially when a large number of elements are mounted on a single board. Therefore, there is a need for an element angle determination device and element angle determination method that can determine the rotation angles of multiple elements on a board in a short time and with high accuracy.

[0007] The present invention has been made to solve the above-mentioned problems, and one object of the present invention is to provide an element angle identification device and an element angle identification method that can perform processing to identify the rotation angles of multiple elements on a substrate in a short time and with high accuracy. [Means for solving the problem]

[0008] In order to achieve the above object, an element angle identification device in a first aspect of the present invention includes an imaging unit that images a substrate on which a plurality of elements are arranged, and a processing unit that performs processing to identify the rotation angle of each of the plurality of elements, and the processing unit is configured to obtain a reference image in which one element is captured based on the captured image captured by the imaging unit, and to generate a plurality of rotated element images in which the element captured in the reference image is rotated at a plurality of preset angles, to obtain a correlation value that indicates the degree of correlation between the element for which the rotation angle is to be identified and each of the plurality of rotated element images, and to obtain correlation information that indicates the relationship between the angle corresponding to each of the plurality of rotated element images and the correlation value, and to identify the rotation angle of the element based on the correlation information.

[0009] In one aspect of the present invention, the element angle identification device includes a processing unit that acquires a correlation value indicating the degree of correlation between an element whose rotation angle is to be identified and each of a plurality of rotation element images, acquires correlation information indicating the relationship between the angle corresponding to each of the plurality of rotation element images and the correlation value, and identifies the rotation angle of the element based on the correlation information. This allows the processing unit to identify the rotation angle of the element based on the correlation information indicating the relationship between the angle corresponding to each of the plurality of rotation element images and the correlation value. Therefore, the rotation speed of the element can be identified in a short time without repeatedly performing pattern matching between each of the rotation element images and the element whose rotation angle is to be identified until the correlation value is maximized. Furthermore, even when the number of pattern matching operations is small and the number of rotation element images to be compared is small, the rotation angle of the element can be identified by interpolating the angle between the angle conditions of discrete rotation element images based on the correlation information. This allows the rotation angle of the element to be identified with higher accuracy than when the angle conditions of discrete rotation element images are directly used to identify the rotation angle of the element. As a result, the process of identifying the rotation angles of multiple elements on a substrate can be performed in a short time with high accuracy.

[0010] In the element angle identifying device according to the first aspect, the processing unit is preferably configured to acquire, as the correlation information, information identifying an approximation line representing the relationship between the angle corresponding to each of the plurality of rotational element images and the correlation value. With this configuration, the rotation angle of the element can be identified using information identifying an approximation line representing the relationship between the angle corresponding to each of the plurality of rotational element images and the correlation value. As a result, the rotation angle of the element can be identified with higher accuracy than when the rotation angle of the element is identified based on correlation information that does not use approximation, such as a graph in which the angle corresponding to each of the plurality of rotational element images and the correlation value are connected by a broken line.

[0011] In this case, the processing unit is preferably configured to specify, as the rotation angle of the element, the angle at which the correlation value on the approximation line obtained based on the information specifying the approximation line is maximized. With this configuration, the rotation angle of the element can be accurately specified based on the angle at which the correlation value on the approximation line is maximized.

[0012] In the element angle identification device in which the processing unit acquires information identifying an approximation line as correlation information, the processing unit is preferably configured to generate an approximation line using some of the correlation values ​​corresponding to angles corresponding to the plurality of rotation element images, the correlation values ​​having values ​​equal to or greater than a predetermined threshold, when the difference between two correlation values ​​corresponding to the maximum and minimum angles among the plurality of preset angles is equal to or greater than a predetermined difference. Here, when the difference between two correlation values ​​corresponding to the maximum and minimum angles among the plurality of preset angles is equal to or greater than a predetermined difference, the angle corresponding to the maximum of the plurality of correlation values ​​is biased toward either the maximum or minimum angle. Therefore, the generated approximation line will be distorted to one side, and the angle corresponding to the maximum correlation value on the approximation line will also be shifted. However, with the above configuration, correlation values ​​that cause distortion of the approximation line can be eliminated using a predetermined threshold. As a result, the rotation angle of the element can be identified with higher accuracy based on the maximum correlation value on the approximation line using only the necessary correlation values.

[0013] In this case, the processing unit is preferably configured to set the threshold value based on a reference correlation value, which is the largest correlation value among two correlation values ​​corresponding to the maximum and minimum angles among a plurality of preset angles, and to generate the approximation line using some of the correlation values ​​corresponding to angles corresponding to each of the plurality of rotation element images that have values ​​equal to or greater than the threshold. With this configuration, the threshold value can be set based on the reference correlation value, which is the value closest to the maximum value of the approximation line used to identify the rotation angle of the element, among two correlation values ​​corresponding to the maximum and minimum angles among a plurality of preset angles. As a result, the accuracy of the approximation line can be further improved.

[0014] In the element angle identification device in which the processing unit generates an approximation line using some of the correlation values ​​corresponding to angles corresponding to each of the plurality of rotation element images, which have values ​​equal to or greater than a threshold, the processing unit is preferably configured to set a reference correlation value as the threshold. This configuration allows the threshold to be set based on an easily set reference correlation value, which is the magnitude relationship between two correlation values ​​corresponding to the maximum and minimum angles among a plurality of angles set in advance. As a result, an approximation line can be easily generated using only the necessary correlation values.

[0015] In the element angle identification device in which the processing unit generates an approximation line using some correlation values ​​having values ​​equal to or greater than a threshold value among a plurality of correlation values ​​corresponding to angles corresponding to each of a plurality of rotation element images, the processing unit is preferably configured to set the threshold value based on a reference correlation value and the maximum value of the plurality of correlation values. This configuration allows the threshold value to be set taking into consideration not only the reference correlation value but also the maximum value of the plurality of correlation values. As a result, an appropriate threshold value can be set according to the maximum value of the plurality of correlation values, thereby improving the accuracy of the approximation line.

[0016] In the element angle identification device, in which the processing unit generates an approximation line using some correlation values ​​having values ​​equal to or greater than a threshold value among a plurality of correlation values ​​corresponding to angles corresponding to each of a plurality of rotation element images, the processing unit is preferably configured to set a correlation value equal to or less than a reference correlation value and closest to the reference correlation value as the threshold value. This configuration can prevent correlation values ​​whose values ​​are almost the same as the reference correlation value from being excluded. As a result, an approximation line can be generated that can be used to further improve the accuracy of identifying the rotation angle of the element.

[0017] In the element angle identification device in which the processing unit acquires information for identifying an approximation line as correlation information, the processing unit is preferably configured to: generate multiple rotational element images rotated at angles greater than the first angle if the angle corresponding to the value at which the correlation value is maximized matches a first angle, which is the largest angle among a plurality of preset angles; generate multiple rotational element images rotated at angles smaller than the second angle if the angle corresponding to the value at which the correlation value is maximized matches a second angle, which is the smallest angle among a plurality of preset angles; and generate an approximation line based on multiple correlation values, including the correlation values ​​between the element for which the rotation angle is to be identified and each of the newly acquired multiple rotational element images. With this configuration, even if the correlation value between the angle of the set rotational element image and the element for which the angle is to be identified is low, additional rotational element images rotated only in the direction necessary to obtain a higher correlation value can be generated. As a result, even if the rotational angle of the element cannot be identified by a single pattern matching, the rotational angle of the element can be identified with a small number of attempts.

[0018] The element angle identification method according to this second aspect includes an imaging step of imaging a substrate on which a plurality of elements are arranged; an image generation step of acquiring a reference image in which one element is captured based on the captured image, and generating a plurality of rotated element images in which the element captured in the reference image is rotated at a plurality of preset angles; a correlation information acquisition step of calculating a correlation value between the element for which the rotation angle is to be identified and each of the plurality of rotated element images, and acquiring correlation information indicating the relationship of the correlation value with the angle corresponding to each of the plurality of rotated element images; and an angle identification step of identifying the rotation angle of the element based on the correlation information.

[0019] The element angle identification method according to the second aspect includes, as described above, a correlation information acquisition step of acquiring correlation information indicating the relationship between the correlation value and the angle corresponding to each of the plurality of rotational element images, and an angle identification step of identifying the rotation angle of the element based on the correlation information. This allows the processing unit to identify the rotation angle of the element based on the correlation information indicating the relationship between the angle corresponding to each of the plurality of rotational element images and the correlation value. Therefore, the rotation speed of the element can be identified in a short time without repeatedly performing pattern matching between each of the rotational element images and the element whose rotation angle is to be identified until the correlation value is maximized. Furthermore, even when the number of pattern matching operations is small and the number of rotational element images to be compared is small, the rotation angle of the element can be identified by interpolating the angle between the angle conditions of the discrete rotational element images based on the correlation information. This allows the rotation angle of the element to be identified with higher accuracy than when the angle conditions of the discrete rotational element images are directly used to identify the rotation angle of the element. As a result, a device angle identification method can be provided that can perform the process of identifying the rotation angles of multiple elements on a substrate in a short time with high accuracy. [Effects of the Invention]

[0020] According to the present invention, as described above, it is possible to provide an element angle specifying device and an element angle specifying method that can perform the process of specifying the rotation angles of a plurality of elements on a substrate in a short time and with high accuracy. [Brief explanation of the drawings]

[0021] [Figure 1] 1 is a diagram showing the overall configuration of an element angle identifying device according to a first embodiment of the present invention. [Figure 2] FIG. 1 is a plan view illustrating a substrate according to a first embodiment of the present invention. [Figure 3] FIG. 3 is an enlarged plan view illustrating the inclination of elements on a substrate according to the first embodiment of the present invention. [Figure 4] FIG. 2 is a diagram for explaining a reference image according to the first embodiment of the present invention. [Figure 5] FIG. 2 is a diagram for explaining a rotation element image according to the first embodiment of the present invention. [Figure 6] 5 is a flowchart illustrating the operation of a processing unit according to the first embodiment of the present invention. [Figure 7] 5 is a flowchart illustrating detailed processing of a correlation information acquisition step according to the first embodiment of the present invention. [Figure 8] 4 is a graph for explaining the relationship between the angle and the correlation value corresponding to each of the rotation element images according to the first embodiment of the present invention. [Figure 9] 5 is a graph for explaining an approximation line using a corrected correlation value according to the first embodiment of the present invention. [Figure 10] 10 is a graph for explaining the correlation value when the angle corresponding to the maximum value of the correlation value according to the first embodiment of the present invention matches the maximum angle in the generation conditions of the rotation element image. [Figure 11] 10 is a graph for explaining an approximation line using additional correlation values ​​according to the first embodiment of the present invention. [Figure 12] 10 is a graph illustrating an approximation line between the angle and the correlation value corresponding to each of the rotation element images according to the comparative example. [Figure 13] 10 is a graph for explaining an approximation line using a corrected correlation value according to the second embodiment of the present invention. [Figure 14] 11 is a graph for explaining an approximation line using a corrected correlation value according to the third embodiment of the present invention. [Figure 15] 10 is a graph for explaining an approximation line between the angle and the correlation value corresponding to each of the rotation element images according to the modifications of the first to third embodiments of the present invention. [Figure 16] 10 is a graph for explaining the correlation value when the angle corresponding to the maximum value of the correlation value according to the modifications of the first to third embodiments of the present invention matches the maximum angle in the generation conditions of the rotation element image. [Figure 17] 10 is a graph for explaining an approximation line using additional correlation values ​​according to the modifications of the first to third embodiments of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0022] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0023] [First embodiment] The configuration of an element angle identifying device 100 according to a first embodiment of the present invention will be described with reference to FIGS.

[0024] (Overall configuration of element angle identification device) 1, the element angle identifying device 100 includes a substrate holding unit 10, an imaging unit 20, a moving mechanism 30, and a processing unit 40. The element angle identifying device 100 is a device for identifying the positions of a plurality of semiconductor elements 210 arranged in a predetermined region A on a substrate 200.

[0025] 2, a plurality of semiconductor elements 210 are arranged spaced apart from one another in a predetermined region A on the substrate 200. The plurality of semiconductor elements 210 are aligned in the X direction and the Y direction perpendicular to the X direction. The XY plane corresponds to the horizontal plane. The direction in which the plurality of semiconductor elements 210 are placed on the substrate 200 is defined as the Z direction.

[0026] A thin rectangular element with sides of about several hundred microns to several tens of mm, such as a memory, is used as the semiconductor element 210. The semiconductor element 210 is an example of the "element" in the claims.

[0027] The substrate holding unit 10 is configured to be able to hold the substrate 200. Specifically, the substrate holding unit 10 includes a substrate mounting table 11 on which the substrate 200 is placed in a horizontal position. The substrate mounting table 11 has a groove or hole formed therein that is connected to a negative pressure generating means such as a vacuum pump (not shown). The substrate holding unit 10 holds the substrate 200 by creating a negative pressure in the groove or hole formed in the substrate mounting table 11, and releases the substrate 200 from its held state by opening the groove or hole formed in the substrate mounting table 11 to the atmosphere.

[0028] 1, the imaging unit 20 is configured to be able to capture an image of the substrate 200. Specifically, the imaging unit 20 is disposed above, in the Z direction, the substrate mounting table 11 on which the substrate 200 is placed, so as to capture an image of the substrate 200 from above in the Z direction. The imaging unit 20 includes an imaging camera 21 for capturing an image of the substrate 200. Image data of the substrate 200 captured by the imaging camera 21 is output to the processing unit 40.

[0029] The movement mechanism 30 is configured to be able to move the substrate holding unit 10 relative to the imaging unit 20 within the XY plane. Specifically, the movement mechanism 30 includes an X-direction movement mechanism 31 that moves the substrate holding unit 10 along the X direction, and a Y-direction movement mechanism 32 that moves the substrate holding unit 10 along the Y direction. The X-direction movement mechanism 31 and the Y-direction movement mechanism 32 are, for example, linear motion mechanisms that combine a servo motor or a pulse motor with a ball screw, or linear motion mechanisms that use a linear motor. The movement mechanism 30 is disposed below the substrate holding unit 10 in the Z direction.

[0030] The processing unit 40 is configured to be able to control each unit of the element angle identifying device 100. Specifically, the processing unit 40 controls a negative pressure generating means (not shown) to create a negative pressure in the grooves or holes formed in the substrate mounting table 11 to hold the substrate 200, and to open the grooves or holes formed in the substrate mounting table 11 to the atmosphere to release the substrate 200. The processing unit 40 also controls the imaging unit 20 to capture an image of the substrate 200. The processing unit 40 also controls the moving mechanism 30 to move the substrate holding unit 10 relative to the imaging unit 20. The processing unit 40 is a computer that includes a processor such as a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and a GPU (Graphics Processing Unit) and is capable of image processing. The processing unit 40 is also configured to perform a process of identifying the rotation angles of multiple semiconductor elements 210 on the substrate 200 by performing pattern matching based on the captured image captured by the imaging unit 20.

[0031] (Regarding the reference image and the rotated image) Here, the reference image 1 and the rotated element images 1a to 1d acquired by the processing unit 40 will be described with reference to FIGS. 3 to 5. FIG. 3 is an enlarged view of a portion of the substrate 200 shown in FIG. 2, showing that a plurality of semiconductor elements 210 are arranged spaced apart from one another in a predetermined region A. In this case, as shown in the semiconductor element 210a, the side of each semiconductor element 210a along the X-axis direction is arranged parallel to the side of the substrate 200 along the X-axis direction, and the side of each semiconductor element 210a along the Y-axis direction is arranged parallel to the side of the substrate 200 along the Y-axis direction. In this state, the rotation angle of the semiconductor element 210a is set to 0°. On the other hand, as shown in the semiconductor element 210b, for example, there are cases where the semiconductor elements 210b are arranged at an angle such that the side of the semiconductor element 210b along the X-axis direction and the side of the substrate 200 along the X-axis direction are not parallel (the side of the semiconductor element 210b along the Y-axis direction and the side of the substrate 200 along the Y-axis direction are not parallel).

[0032] As described above, when semiconductor element 210 is disposed at an angle, the direction in which semiconductor element 210 is rotated counterclockwise from a rotation angle of 0° is defined as a positive rotation angle, and the direction in which the element is rotated clockwise from a rotation angle of 0° is defined as a negative rotation angle. For example, in Figure 3, semiconductor element 210b is rotated 3° clockwise, and therefore is disposed at a rotation angle of 3.0° (-3.0°) in the negative direction.

[0033] 4 is an image generated by the processing unit 40 based on one semiconductor element 210 selected by an operator or the like from captured images of multiple semiconductor elements 210 arranged on a substrate 200 captured by the imaging unit 20. The reference image 1 is an image consisting of an element portion 2 and a background portion 3, and is also one of the rotation element images described below. A detailed method for generating the reference image 1 will be described later.

[0034] Rotation element images 1a-1d shown in FIG. 5 are images generated by the processing unit 40 by rotating only the element portion 2 in the reference image 1 in the positive and negative directions within an angle range set by an operator or the like. Like the reference image 1 described above, rotation element images 1a-1d are images consisting of element portions 2a-2d and background portions 3a-3d. In this first embodiment, for example, eight patterns of rotation element images 1a-1d are generated at 2.5° intervals within a range from -10° to 10° (0° is generated as the reference image 1). For example, rotation element image 1a in FIG. 4 is an image obtained by rotating the element portion 2 in the reference image 1 by -10.0°. Note that the angle of 10° is the "first angle, which is the maximum angle among the plurality of preset angles" described in the claims, and the angle of -10° is the "second angle, which is the minimum angle among the plurality of preset angles" described in the claims. A detailed method for generating rotation element images 1a-1d will be described later.

[0035] (Processing flow of element angle identification method) Here, an element angle identifying method in which the processing unit 40 identifies the rotation angles of all semiconductor elements 210 arranged in a predetermined area A on the substrate 200 in the element angle identifying device 100 shown in FIG. 1 will be described with reference to FIGS. 6 to 11.

[0036] First, as an imaging step of step S1, the processing unit 40 controls the imaging unit 20 to image a predetermined area A on the substrate 200 in order to detect the rotation angle of each of the plurality of semiconductor elements 210, as shown in Figures 2 and 3. Specifically, with the substrate 200 held by the substrate holding unit 10 (see Figure 1), the processing unit 40 divides the predetermined area A on the substrate 200 into several parts and images them. Therefore, the processing unit 40 alternately controls the movement mechanism 30 (see Figure 1) that moves the substrate holding unit 10 holding the substrate 200 by a predetermined movement amount, and controls the imaging unit 20 that captures the image. Then, the processing proceeds to step S2.

[0037] As a selected element receiving step of step S2, the processing unit 40 (see FIG. 1) receives that a worker or the like has selected one semiconductor element 210 from the captured image of the multiple semiconductor elements 210 arranged on the substrate 200 captured by the imaging unit 20. At this time, the worker or the like selects a semiconductor element 210 with a rotation angle close to 0° from the captured image of the multiple semiconductor elements 210 arranged on the substrate 200, and inputs the selected information to the processing unit 40 using an operation unit (not shown) of the element angle identifying device 100 in FIG. 1. Then, the process proceeds to step S3.

[0038] As a generation condition receiving step in step S3, the processing unit 40 receives input of generation conditions for generating the rotation element images 1a-1d (see FIG. 5) from an operator or the like. At this time, the operator or the like arbitrarily determines the maximum and minimum values ​​and angle increments for generating the rotation element images 1a-1d, and inputs the selected information to the processing unit 40 using an operation unit (not shown) in the element angle identification device 100 of FIG. 1. Note that the maximum angle in the following description refers to the value at which the positive angle of the element portions 2a-2d (see FIG. 5) in the rotation element images 1a-1d is greatest. Also, the minimum angle in the following description refers to the value at which the negative angle of the element portions 2a-2d in the rotation element images 1a-1d is greatest. Note that a "small angle" in the claims refers to an angle at which the absolute value of the negative angle is large. For example, the -10° angle is smaller than the -5° angle. Then, the processing proceeds to step S4.

[0039] As the image generation process of step S4, the processing unit 40 (see FIG. 1) generates a reference image 1 (see FIG. 4) based on one semiconductor element 210 selected by an operator or the like in step S2, and generates rotation element images 1a to 1d (see FIG. 5) based on the reference image 1 and the generation conditions input by the operator or the like in step S3. Then, the process proceeds to step S5.

[0040] In step S5, a correlation value acquisition process, processing unit 40 performs pattern matching on each of semiconductor elements 210 in the captured image whose rotation angle is to be identified, with reference image 1 (see FIG. 4) generated in step S4 and rotation element images 1a-1d (see FIG. 5), and acquires a correlation value indicating the degree of correlation between the two compared images. Here, the correlation value is not a value having a specific unit, but is expressed as a normalized value (normalized value) with a maximum value of "1" representing a perfect match between the two compared images. For example, since semiconductor element 210b in FIG. 3 has a rotation angle of -3.0° as described above, as shown in FIG. 8, the correlation value for semiconductor element 210b is the maximum value in rotation element images 1a-1d in which the angles of element portions 2a-2d in rotation element images 1a-1d are closest to -3.0°. Furthermore, as the angles of element portions 2a-2d in rotation element images 1a-1d increase from -3.0°, the correlation value for semiconductor element 210b decreases. Then, the process proceeds to step S6.

[0041] In the correlation information acquisition process of step S6, the processing unit 40 acquires correlation information indicating the relationship between the angles corresponding to the reference image 1 and each of the plurality of rotation element images 1a to 1d and the correlation value. Here, the correlation information acquisition process of step S6 will be described in detail with reference to FIG.

[0042] In step S6a of Fig. 7, it is determined whether the angle corresponding to the maximum correlation value matches the maximum or minimum angle in the generation conditions. In this first embodiment, as described above, in rotation element images 1a to 1d in which the angles of element portions 2a to 2d in the rotation element images 1a to 1d are closest to -3.0°, the correlation value is maximum, and therefore does not match the maximum angle of 10° or the minimum angle of -10° in the generation conditions (see Fig. 8). Therefore, the process proceeds to step S6d. Note that if the angle corresponding to the maximum correlation value matches the maximum or minimum angle in the generation conditions, the process proceeds to step S6b, which will be described later.

[0043] In step S6d, the processing unit 40 determines whether the difference between the large reference correlation value Pa and the small end correlation value Pb, among the correlation values ​​corresponding to the maximum and minimum angles in the generation conditions, is equal to or greater than a predetermined amount. In this first embodiment, the correlation value at the minimum angle corresponds to the reference correlation value Pa, and the correlation value at the maximum angle corresponds to the end correlation value Pb. For example, in a standard value where the maximum value is "1," if the predetermined amount is set to "0.05" and the difference between the reference correlation value Pa and the end correlation value Pb is "0.08," the difference between the reference correlation value Pa and the end correlation value Pb is deemed to be equal to or greater than the predetermined amount, and the processing proceeds to step S6e. Note that if the difference between the reference correlation value Pa and the end correlation value Pb is less than the predetermined amount, the processing proceeds to step S6f.

[0044] In step S6e, a corrected correlation value acquisition step, the processing unit 40 acquires corrected correlation values ​​as new correlation values ​​obtained by eliminating unnecessary correlation values. In this first embodiment, the processing unit 40 sets the reference correlation value Pa, which is the larger of the reference correlation value Pa and the end correlation value Pb, as a threshold value, and acquires corrected correlation values ​​by eliminating correlation values ​​less than the threshold value. For example, in FIG. 8, the correlation value Pc corresponding to an angle of 2.5° is equal to or greater than the reference correlation value Pa, but the correlation value Pd corresponding to an angle of 5.0° is less than the reference correlation value Pa and is therefore excluded. As a result, the plot data of the correlation values ​​as shown in FIG. 8 is replaced with the plot data of the corrected correlation values ​​shown in FIG. 9. Then, the processing proceeds to step S6f.

[0045] In the approximate equation acquisition step of step S6f, processing unit 40 acquires an approximate equation that indicates the relationship between the angle corresponding to each of rotational element images 1a-1d and reference image 1 and the correlation value. If the difference between reference correlation value Pa and end correlation value Pb is equal to or greater than a predetermined amount in step S6d, plot data of the angle corresponding to each of rotational element images 1a-1d and reference image 1 and the correlation value, as shown in FIG. 9, is obtained, and processing unit 40 derives an approximate equation and approximate line L2 based on this plot data. This approximate equation and approximate line L1 are an example of correlation information described in the claims. This completes the correlation information acquisition step of step S6, and processing proceeds to step S7.

[0046] In the angle identification process of step S7, the processing unit 40 identifies the rotation angle of the semiconductor element 210 to be identified based on the approximation formula acquired in step S6. Specifically, the processing unit 40 identifies the rotation angle of the semiconductor element 210b to be identified as the rotation angle at which the approximation formula, which is a function having a peak acquired in step S6, is maximized. For example, in FIG. 9, the angle corresponding to the maximum value of the approximation formula (approximation line L1) is -3.2°, so the rotation angle of the semiconductor element 210b to be identified as -3.2°. Then, the processing proceeds to step S8.

[0047] In step S8, the processing unit 40 determines whether or not the rotation angle determination has been completed for all of the semiconductor elements 210 that are the target of angle determination. If the rotation angle determination has not been completed for all of the semiconductor elements 210 that are the target of angle determination, the processing returns to step S5, and processing is performed to determine the rotation angle for another semiconductor element 210 that has not yet been determined for its rotation angle. If the rotation angle determination has been completed for all of the semiconductor elements 210 that are the target of angle determination, the processing ends.

[0048] (Acquisition of additional correlation values) Here, the processing performed when, in step S6a, the processing unit 40 acquires correlation values ​​as shown in FIG. 10, determines that 10.0°, the angle corresponding to the maximum correlation value, matches the maximum angle of 10.0° under the generation conditions, and proceeds to step S6b (additional image generation process) will be described. For example, when detecting the rotation angle of a semiconductor element 210c having a rotation angle of 12.5° as shown in FIG. 3, the correlation value with a rotated element image in which the angle of the element portion 2 (see FIG. 4) is 12.5° is highest, but this is outside the range of the generation conditions (-10.0° to 10.0°). Therefore, the plot data of the correlation values ​​rises to the right, and as shown in FIG. 10, the angle of 10.0°, the angle corresponding to the maximum correlation value, matches the maximum angle of 10.0° under the generation conditions. In this case, the plot data of the correlation values ​​does not have a peak, and the processing unit 40 is unable to identify the rotation angle of the semiconductor element 210. Therefore, the processing unit 40 generates additional rotation element images by rotating the element portion 2 in the reference image 1 by a larger angle than the maximum angle of 10.0° set in the generation conditions. For example, the processing unit 40 generates additional rotation element images by rotating the element portion 2 in the reference image 1 by 12.5°, 15.0°, and 17.5°. Then, the processing proceeds to step S6c.

[0049] In step S6c, as an additional correlation value acquisition step, the processing unit 40 performs pattern matching between each of the additionally generated rotational element images and the semiconductor element 210c whose angle is to be identified, as in step S5, to acquire each correlation value. Then, the processing returns to step S6a. The above process is repeated until the angle corresponding to the maximum correlation value no longer matches the maximum or minimum angle in the generation conditions. In this first embodiment, as shown in FIG. 11, by adding correlation values ​​between the rotational element images obtained by rotating the element portion 2 by 12.5°, 15.0°, and 17.5° and the semiconductor element 210c in the captured image, the angle corresponding to the maximum correlation value no longer matches the maximum angle in the generation conditions. Then, the processing proceeds to step S6d, where the process proceeds according to the above-described processing flow. As shown by the approximation line L2 in FIG. 11, the rotation angle of the semiconductor element 210c is identified as 12.5°.

[0050] [Comparative Example] Here, a comparative example will be described in which, for example, in step S6d, an approximate equation is derived based on plot data of all correlation values ​​acquired in step S7, even though the difference between the reference correlation value Pa and the end correlation value Pb is equal to or greater than a predetermined amount. In this case, as shown in FIG. 12, because the maximum acquired correlation value Pe is close to the reference correlation value Pa, the approximate line L3 is distorted toward the end correlation value Pb, and the angle corresponding to the maximum value of the approximate equation (approximate line L3) is −2.7°. Therefore, the rotation angle of the semiconductor element 210b to be identified as the angle identification target is identified as −2.7°, which is 0.5° larger than the rotation angle of −3.2° identified using the corrected correlation value. In this way, the identified rotation angle differs depending on whether or not correction is applied to the generation of the approximate equation.

[0051] [Effects of the first embodiment] In the first embodiment, the following effects can be obtained.

[0052] In the first embodiment, the element angle identifying device 100 includes an imaging unit 20 that captures an image of a substrate 200 on which a plurality of semiconductor elements 210 are arranged, and a processing unit 40 that performs processing to identify the rotation angle of each of the plurality of semiconductor elements 210. The processing unit 40 is configured to acquire a reference image 1 in which one semiconductor element 210 is captured based on the captured image captured by the imaging unit 20, and to generate a plurality of rotated element images 1a to 1d in which the element portion 2 captured in the reference image 1 is rotated by a plurality of preset angles, to acquire correlation values ​​that indicate the degree of correlation between the semiconductor element 210 for which the rotation angle is to be identified and each of the plurality of rotated element images 1a to 1d, and to acquire correlation information (approximation line L1 and approximation formula) that indicates the relationship between the angle corresponding to each of the plurality of rotated element images 1a to 1d and the correlation value, and to identify the rotation angle of the semiconductor element 210 based on the correlation information (approximation line L1 and approximation formula). As a result, the processing unit 40 can identify the rotation angle of the semiconductor element 210 based on correlation information indicating the relationship between the angle corresponding to each of the multiple rotation element images 1a-1d and the correlation value. This eliminates the need to repeatedly perform pattern matching between each of the rotation element images 1a-1d and the semiconductor element 210 for which the rotation angle is to be identified until the correlation value reaches a maximum. This allows the rotation speed of the semiconductor element 210 to be identified in a short time. Furthermore, even when the number of pattern matching operations is small and the number of rotation element images 1a-1d to be compared is small, the rotation angle of the semiconductor element 210 can be identified by interpolating angles between the angle conditions of the discrete rotation element images 1a-1d based on the correlation information. This allows the rotation angle of the semiconductor element 210 to be identified with higher accuracy than when the angle conditions of the discrete rotation element images 1a-1d are directly used to identify the rotation angle of the semiconductor element 210. As a result, the process of identifying the rotation angles of the multiple semiconductor elements 210 on the substrate 200 can be performed in a short time with high accuracy.

[0053] Furthermore, in the first embodiment, the processing unit 40 is configured to acquire, as the correlation information, an approximation formula as information specifying an approximation line L1 that represents the relationship between the angle corresponding to each of the plurality of rotation element images 1a to 1d and the correlation value. This makes it possible to specify the rotation angle of the semiconductor element 210 using information specifying the approximation lines L1 to L3 that represent the relationship between the angle corresponding to each of the plurality of rotation element images 1a to 1d and the correlation value. As a result, it is possible to specify the rotation angle of the semiconductor element 210 with higher accuracy than when the rotation angle of the semiconductor element 210 is specified based on correlation information that does not use approximation, such as a graph in which the angle corresponding to each of the plurality of rotation element images 1a to 1d and the correlation value are connected by a broken line.

[0054] Furthermore, in the first embodiment, the processing unit 40 is configured to specify the angle at which the correlation value on the approximation line L1 or L2 obtained based on the information specifying the approximation line L1 or L2 is maximized as the rotation angle of the semiconductor element 210. This makes it possible to accurately specify the rotation angle of the semiconductor element 210 based on the angle at which the correlation value on the approximation line L1 or L2 is maximized.

[0055] Furthermore, in the first embodiment, the processing unit 40 is configured to generate an approximation line L2 using some of the correlation values ​​corresponding to angles corresponding to the plurality of rotation element images 1a-1d, the correlation values ​​having values ​​equal to or greater than a predetermined threshold, when the difference between the two correlation values ​​Pa and Pb corresponding to the maximum angle of 10° and the minimum angle of −10° among the plurality of preset angles is greater than or equal to a predetermined difference. Here, when the difference between the two correlation values ​​Pa and Pb corresponding to the maximum angle of 10° and the minimum angle of −10° among the plurality of preset angles is greater than or equal to a predetermined difference, the angle corresponding to the maximum of the plurality of correlation values ​​is biased toward either the maximum or minimum angle. Therefore, the generated approximation line L3 is distorted to one side, and the angle corresponding to the maximum correlation value on the approximation line L3 is also shifted. However, by doing the above, it is possible to eliminate correlation values ​​that cause distortion of the approximation line L3 using a predetermined threshold. As a result, the rotation angle of the semiconductor element 210 can be identified with higher accuracy based on the maximum correlation value on the approximation line L2, which uses only the necessary correlation values.

[0056] Furthermore, in the first embodiment, the processing unit 40 is configured to set a threshold value based on a reference correlation value Pa, which is a larger correlation value among two correlation values ​​Pa and Pb corresponding to the maximum angle of 10° and the minimum angle of −10° among a plurality of preset angles, and to generate the approximation line L2 using some of the correlation values ​​corresponding to angles corresponding to each of the plurality of rotation element images 1a to 1d that have values ​​equal to or greater than the threshold value. This makes it possible to set the threshold value based on the reference correlation value Pa, which is a value close to the maximum value of the approximation line L1 used to identify the rotation angle of the semiconductor element 210, among the two correlation values ​​Pa and Pb corresponding to the maximum angle of 10° and the minimum angle of −10° among a plurality of preset angles. As a result, the accuracy of the approximation line L2 can be further improved.

[0057] Furthermore, in the first embodiment, the processing unit 40 is configured to set the reference correlation value Pa as the threshold value. This allows the threshold value to be set based on the easily set reference correlation value Pa, which is the magnitude relationship between two correlation values ​​Pa and Pb corresponding to the maximum angle of 10° and the minimum angle of −10° among a plurality of angles set in advance. As a result, the approximation line L1 can be easily generated using only the necessary correlation values.

[0058] Furthermore, in the first embodiment, the processing unit 40 is configured to generate multiple rotation element images 1a-1d rotated at angles greater than the first angle of 10° when the angle corresponding to the maximum correlation value is equal to the first angle of 10°, which is the largest angle among the multiple preset angles. If the angle corresponding to the maximum correlation value is equal to the second angle of −10°, which is the smallest angle among the multiple preset angles, the processing unit 40 generates multiple rotation element images rotated at angles smaller than the second angle of −10°. The processing unit 40 generates an approximation line L2 based on multiple correlation values, including the correlation values ​​between the semiconductor element 210 for which the rotation angle is to be identified and each of the multiple newly acquired rotation element images. This allows additional rotation element images rotated only in the direction necessary to obtain a higher correlation value, even if the correlation value between the angle of the set rotation element images 1a-1d and the semiconductor element 210 for which the angle is to be identified is low. As a result, even if the rotation angle of the semiconductor element 210 cannot be identified by a single pattern matching, the rotation angle of the semiconductor element 210 can be identified with a small number of attempts.

[0059] [Second embodiment] Next, an element angle specifying device 100a and an element angle specifying method for a semiconductor element 210 according to a second embodiment will be described. The device configuration of the element angle specifying device 100a is the same as that of the element angle specifying device 100 shown in FIG. 1, except for the processing unit 40a. In the second embodiment, a processing unit 40a that executes processing of a correction method different from that in the first embodiment is provided in step S6e. Note that a description of the second embodiment common to the first embodiment will be omitted.

[0060] In the second embodiment, in the corrected correlation value acquisition step of step S6e in FIG. 7, the processing unit 40a acquires corrected correlation values ​​as new multiple correlation values ​​obtained by eliminating unnecessary correlation values. In this second embodiment, the processing unit 40a sets a threshold based on the maximum obtained correlation value Pe, which is the maximum value of all correlation values ​​shown in FIG. 8, and the reference correlation value Pa, which is the larger correlation value between the reference correlation value Pa and the end correlation value Pb, and acquires the corrected correlation values. Specifically, the processing unit 40a sets a value obtained by subtracting a value equivalent to 10% of the maximum obtained correlation value Pe from the reference correlation value Pa as the threshold, and acquires the corrected correlation values ​​by eliminating correlation values ​​less than the threshold. For example, in FIG. 8, if the value equivalent to 10% of the maximum obtained correlation value Pe is "0.098," the processing unit 40a sets the correlation value obtained by subtracting "0.098" from the reference correlation value Pa as the threshold c1. Here, in the first embodiment, as shown in Fig. 9, the correlation value Pd was excluded from the corrected correlation value, but in the second embodiment shown in Fig. 13, the correlation value Pd remains as the corrected correlation value. Then, the process proceeds to step S6f.

[0061] In step S6f, the processing unit 40a acquires an approximate equation indicating the relationship between the angle corresponding to each of the rotational element images 1a-1d and the reference image 1 and the correlation value. In step S6d, if the difference between the reference correlation value Pa and the end correlation value Pb is equal to or greater than a predetermined amount, plot data of the angle corresponding to each of the rotational element images 1a-1d and the reference image 1 and the correlation value, as shown in FIG. 13, is obtained. Based on this plot data, the processing unit 40a derives an approximate equation and an approximate line L4. This completes the correlation information acquisition step of step S6, and the processing proceeds to step S7. The subsequent steps are the same as those in the first embodiment. Comparing FIG. 12, which shows a case where correction by step S6e is not performed, with FIG. 13, which shows a case where correction is performed, the angle of the semiconductor element 210b in the captured image identified in step S8 is off by 0.3°. 12, the angle corresponding to the maximum value of the approximation formula (approximation line L3) is -2.7°, so the processing unit 40a identifies the angle of the semiconductor element 210b to be identified as -2.7°. Also, in FIG. 13, the angle corresponding to the maximum value of the approximation formula (approximation line L4) is -3.0°, which is rotated 0.3° in the negative direction compared to when approximation line L3 is used, so the processing unit 40a identifies the angle of the semiconductor element 210b to be identified as -3.0°.

[0062] (Effects of the second embodiment) Next, the effects of the second embodiment will be described.

[0063] In the second embodiment, the processing unit 40a is configured to set the threshold value c1 based on the reference correlation value Pa and the maximum obtained correlation value Pe, which is the maximum value of the multiple correlation values. This makes it possible to set the threshold value c1 taking into consideration not only the reference correlation value Pa but also the maximum obtained correlation value Pe. As a result, it is possible to set an appropriate threshold value c1 according to the maximum obtained correlation value Pe, thereby improving the accuracy of the approximation line L4.

[0064] The other effects of the second embodiment are the same as those of the first embodiment.

[0065] [Third embodiment] Next, an element angle specifying device 100b and an element angle specifying method for a semiconductor element 210 according to a third embodiment will be described. The device configuration of the element angle specifying device 100b is the same as that of the element angle specifying device 100 shown in FIG. 1, except for the processing unit 40b. In the third embodiment, a processing unit 40b is provided that executes processing of a correction method different from those in the first and second embodiments in step S6e. Note that a description of the third embodiment common to the first and second embodiments will be omitted.

[0066] In the third embodiment, in the corrected correlation value acquisition step of step S6e in Fig. 7, processing unit 40b acquires corrected correlation values ​​as new multiple correlation values ​​obtained by eliminating unnecessary correlation values. In this third embodiment, processing unit 40b sets one correlation value Pd that is equal to or less than and closest to reference correlation value Pa shown in Fig. 8 as threshold value c2, and acquires corrected correlation values ​​by excluding correlation values ​​less than threshold value c2. Then, the process proceeds to step S6f.

[0067] In the approximate expression acquisition step of step S6f, the processing unit 40b acquires an approximate expression that indicates the relationship between the angle corresponding to each of the rotational element images 1a-1d and the reference image 1 and the correlation value. In step S6d, if the difference between the reference correlation value Pa and the end correlation value Pb is equal to or greater than a predetermined amount, plot data of the angle corresponding to each of the rotational element images 1a-1d and the reference image 1 and the correlation value, as shown in FIG. 14, is obtained. Based on this plot data, the processing unit 40b derives an approximate expression and an approximate line L5. This completes the correlation information acquisition step of step S6, and the processing proceeds to step S7. The subsequent steps are the same as those in the first embodiment. Comparing FIG. 12, which shows the case where correction by step S6e is not performed, with FIG. 14, which shows the case where correction is performed, the angle of the semiconductor element 210b in the captured image identified in step S8 is off by 0.3°. 12, the angle corresponding to the maximum value of the approximation formula (approximation line L3) is -2.7°, so the processing unit 40 specifies the angle of the semiconductor element 210b to be specified as -2.7°. Also, in FIG. 15, the angle corresponding to the maximum value of the approximation formula (approximation line L5) is -3.0°, which is rotated 0.3° in the negative direction compared to when approximation line L3 is used, so the processing unit 40 specifies the angle of the semiconductor element 210b to be specified as -3.0°.

[0068] (Effects of the third embodiment) Next, the effects of the third embodiment will be described.

[0069] In the third embodiment, the processing unit 40b is configured to set, as the threshold value c2, a correlation value that is equal to or smaller than the reference correlation value Pa and that is closest to the reference correlation value Pa. This makes it possible to prevent the correlation value Pd that is almost the same as the reference correlation value Pa from being excluded. As a result, it is possible to generate the approximation line L5 that is used to further improve the accuracy of identifying the rotation angle of the semiconductor element 210.

[0070] The other effects of the third embodiment are the same as those of the first and second embodiments.

[0071] [Variations] The embodiments disclosed herein should be considered to be illustrative and not restrictive in all respects. The scope of the present invention is defined by the claims rather than the above description of the embodiments, and further includes all modifications (variations) within the meaning and scope of the claims.

[0072] For example, in the first to third embodiments, examples have been shown in which a thin semiconductor element 210 such as a memory is used as the element of the present invention, but the present invention is not limited to this. Various elements other than the semiconductor element 210 may also be used as the element of the present invention.

[0073] In the above first to third embodiments, the processing unit 40 acquires an approximation formula as information specifying the approximation line L1, L2, L4, or L5 that represents the relationship between the angle corresponding to each of the plurality of rotation element images 1a to 1d and the correlation value, but the present invention is not limited to this. In the present invention, the correlation information acquired by the processing unit 40 does not have to be an approximation line, but may be, for example, a table that represents the relationship between the angle corresponding to each of the plurality of rotation element images 1a to 1d and the correlation value.

[0074] In the first to third embodiments, the processing unit 40 specifies the angle at which the correlation value on the approximation line obtained based on the information specifying the approximation line L1, L2, L4, or L5 is maximum as the rotation angle of the semiconductor element 210. However, the present invention is not limited to this. In the present invention, for example, one angle arbitrarily selected from angles at which the correlation value on the approximation line is equal to or greater than a predetermined value may be specified as the rotation angle of the semiconductor element 210.

[0075] Furthermore, in the above first to third embodiments, the processing unit 40 generates the approximation line L1, L2, L4, or L5 using some of the correlation values ​​corresponding to angles corresponding to each of the plurality of rotation element images 1a to 1d that have values ​​equal to or greater than a predetermined threshold value when the difference between the two correlation values ​​corresponding to the maximum and minimum angles among the plurality of preset angles is equal to or greater than a predetermined difference, but the present invention is not limited to this. In the present invention, even if the difference between the two correlation values ​​corresponding to the maximum and minimum angles among the plurality of preset angles is equal to or greater than a predetermined difference, the approximation line may be generated using all of the correlation values.

[0076] In the above first to third embodiments, the processing unit 40 sets the threshold value c1 or c2 based on the reference correlation value Pa, which is the largest correlation value among two correlation values ​​corresponding to the maximum and minimum angles among a plurality of preset angles, and generates the approximation line L1, L2, L4, or L5 using some of the correlation values ​​corresponding to angles corresponding to each of the plurality of rotation element images 1a to 1d that have values ​​equal to or greater than the threshold value, but the present invention is not limited to this. In the present invention, for example, the threshold value may be set based on the maximum value of all correlation values ​​instead of the reference correlation value Pa.

[0077] In the first to third embodiments, the processing unit 40 generates the approximation line L1, L2, L4, or L5 using some of the correlation values ​​corresponding to the angles corresponding to the plurality of rotation element images 1a to 1d that have values ​​equal to or greater than a predetermined threshold value when the difference between the two correlation values ​​Pa and Pb corresponding to the maximum angle of 10° and the minimum angle of −10° among the plurality of preset angles is equal to or greater than a predetermined difference. However, the present invention is not limited to this. In the present invention, for example, as shown in FIG. 15, if the difference between the two correlation values ​​Pa and Pb corresponding to the maximum angle of 10° and the minimum angle of −10° among the plurality of preset angles is less than a predetermined difference, the approximation line L6 may be generated based on all of the acquired correlation values.

[0078] In the first to third embodiments, when the angle corresponding to the maximum acquired correlation value Pe matches a first angle, which is the largest angle among a plurality of preset angles, the processing unit 40 generates a plurality of rotation element images rotated at an angle greater than the first angle, and generates an approximation line based on a plurality of correlation values ​​including the correlation values ​​between the semiconductor element 210, the target of which rotation angle is to be identified, and each of the plurality of newly acquired rotation element images. However, the present invention is not limited to this. In the present invention, as shown in FIG. 16, when the angle corresponding to the value at which the correlation value is maximized matches a second angle, which is the smallest angle among a plurality of preset angles, the processing unit 40 may generate a plurality of rotation element images rotated at an angle smaller than the second angle. In this case, as shown in FIG. 17, an approximation formula and an approximation line L7 are generated based on a plurality of correlation values ​​including the correlation values ​​between the added rotation element image and the semiconductor element 210, making it possible to identify the rotation angle of the semiconductor element 210.

[0079] In addition, in the above first to third embodiments, an example has been shown in which the processing unit 40 acquires the reference image 1 based on the captured image captured by the imaging unit 20, but the present invention is not limited to this. In the present invention, for example, when detecting the angle of a semiconductor element 210 of the same shape mounted on another substrate 200, the processing unit 40 may acquire the reference image 1 stored in a storage unit (not shown). [Explanation of symbols]

[0080] 1 Reference image 1a~1d Rotating element images 2. Element part shown in the reference image (element shown in the reference image) 10 Board holding part 20 Imaging unit 30 Moving mechanism 40, 40a, 40b Processing section 100, 100a, 100b Element angle identification device 200 boards 210, 210a, 210b semiconductor element (element) L1, L2, L4~L7 approximate line

Claims

1. an imaging unit that images a substrate on which a plurality of elements are arranged; a processing unit that performs processing to identify a rotation angle of each of the plurality of elements, The processing unit A reference image in which one of the elements is captured is obtained based on the captured image captured by the imaging unit, and a plurality of rotated element images are generated by rotating the element captured in the reference image at a plurality of preset angles; Acquire a correlation value indicating the degree of correlation between the element for which the rotation angle is to be specified and each of the plurality of rotation element images, and acquire correlation information indicating the relationship between the angle corresponding to each of the plurality of rotation element images and the correlation value; an element angle identifying device configured to identify a rotation angle of the element based on the correlation information;

2. The element angle identifying device according to claim 1, wherein the processing unit is configured to acquire, as the correlation information, information that identifies an approximation line that represents a relationship between an angle corresponding to each of the plurality of rotation element images and the correlation value.

3. 3. The element angle specifying device according to claim 2, wherein the processing unit is configured to specify, as a rotation angle of the element, an angle at which the correlation value on the approximation line obtained based on information specifying the approximation line is maximized.

4. 3. The element angle identification device according to claim 2, wherein, when a difference between two of the correlation values ​​corresponding to a maximum angle and a minimum angle among the plurality of angles set in advance is a predetermined difference or more, the processing unit is configured to generate the approximation line using some of the correlation values ​​corresponding to angles corresponding to each of the plurality of rotation element images, the some having a value equal to or greater than a predetermined threshold value.

5. The processing unit setting the threshold value based on a reference correlation value that is the larger correlation value from two correlation values ​​corresponding to the maximum angle and the minimum angle of the plurality of angles that have been set in advance; 5. The element angle identification device according to claim 4, wherein the approximation line is generated using some of the correlation values ​​having values ​​equal to or greater than the threshold value, among the plurality of correlation values ​​corresponding to angles corresponding to each of the plurality of rotation element images.

6. The element angle identifying device according to claim 5 , wherein the processing unit is configured to set the reference correlation value as the threshold value.

7. The element angle identifying device according to claim 5 , wherein the processing unit is configured to set the threshold value based on the reference correlation value and a maximum value of the plurality of correlation values.

8. The element angle identifying device according to claim 5 , wherein the processing unit is configured to set the correlation value that is equal to or smaller than the reference correlation value and closest to the reference correlation value as the threshold value.

9. The processing unit If the angle corresponding to the value at which the correlation value is maximized matches a first angle that is the largest angle among the plurality of angles set in advance, a plurality of rotation element images rotated at angles larger than the first angle are generated, If the angle corresponding to the value at which the correlation value is maximized matches a second angle that is the smallest angle among the plurality of angles set in advance, a plurality of rotation element images rotated at angles smaller than the second angle are generated, 3. The element angle identification device according to claim 2, wherein the approximation line is generated based on the plurality of correlation values ​​including the correlation value between the element whose rotation angle is to be identified and each of the plurality of newly acquired rotation element images.

10. an imaging step of imaging a substrate on which a plurality of elements are arranged; an image generation step of acquiring a reference image in which one of the elements is captured based on the captured image, and generating a plurality of rotated element images in which the element captured in the reference image is rotated by a plurality of preset angles; a correlation information acquisition step of calculating a correlation value between the element whose rotation angle is to be specified and each of the plurality of rotation element images, and acquiring correlation information indicating a relationship between the correlation value and the angle corresponding to each of the plurality of rotation element images; and an angle specifying step of specifying a rotation angle of the element based on the correlation information.

Images