Light source position detection method and inspection apparatus

Abstract

To improve the accuracy of light source position detection. [Solution] The light source position detection method includes illuminating the jig 10 with a light source Li such that the shadows 15 of each sphere 14 of the multiple protrusions 12 are projected onto the base 11; capturing multiple images of the jig 10 illuminated by the light source Li with multiple cameras C; calculating a first three-dimensional coordinate X1 of the center of the sphere 14 and a second three-dimensional coordinate X2 of the center of the shadow 15 of the sphere 14 for each of the multiple protrusions 12 based on the multiple images; calculating multiple light source vectors V for the multiple protrusions 12, wherein each of the multiple light source vectors V passes through the first three-dimensional coordinate X1 and the second three-dimensional coordinate X2 for each of the multiple protrusions 12; and calculating the position of the light source Li based on the multiple light source vectors V.

Description

Technical Field

[0001] The present disclosure relates to a method for detecting a light source position and an inspection apparatus.

Background Art

[0002] The illuminance difference stereo method is known as a method for detecting the shape of an object. For example, Patent Document 1 discloses an apparatus for recognizing the three-dimensional shape of an object using the illuminance difference stereo method. This apparatus includes a plurality of light sources that irradiate the object from different directions, and a camera that images the object. The plurality of light sources sequentially irradiate the object. The camera images the object irradiated by each light source. Based on the luminance values of the object in the plurality of images, a normal vector is calculated for each pixel. Thereby, the shape of the object is obtained.

[0003] In the illuminance difference stereo method, in order to accurately detect the shape of an object, it is important to know the exact positions of each light source.

[0004] Non-Patent Document 1 discloses a method for estimating the light source position. In this method, the light source irradiates two spheres, and the camera images the two spheres. Based on the image, the normal of each sphere is calculated. From the normal of each sphere and the optical axis of the camera, the azimuth vector of the light source is estimated. The light source position is estimated from the intersection point between the azimuth vectors estimated from the two spheres.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Non-Patent Documents

[0006]

Non-Patent Document 1

[0007] When using a sphere to estimate the position of a light source, it is necessary to identify the highlighted areas in the image and determine the center position of the sphere. However, if the highlighted areas are made clearer in the image, the outline needed to determine the center position of the sphere becomes blurred. Conversely, if the outline of the sphere is made clearer, the highlighted areas become blurred. Therefore, it is difficult to clearly identify both the highlighted areas and the outline of the sphere.

[0008] This disclosure aims to provide a method for detecting the position of a light source that can improve the accuracy of light source position detection. Furthermore, this disclosure aims to provide related inspection equipment. [Means for solving the problem]

[0009] A method for detecting the position of a light source according to one aspect of the present disclosure includes: preparing a jig including a base and a plurality of protrusions attached to the base, wherein each of the plurality of protrusions includes a shaft attached to the base and a sphere attached to the shaft; illuminating the jig with a light source such that the shadows of each of the spheres of the plurality of protrusions are projected onto the base; capturing a plurality of images of the jig illuminated by the light source with a plurality of cameras; calculating a first three-dimensional coordinate of the center of the sphere and a second three-dimensional coordinate of the center of the shadow of the sphere for each of the plurality of protrusions based on the plurality of images; calculating a plurality of light source vectors for the plurality of protrusions, wherein each of the plurality of light source vectors passes through the first three-dimensional coordinate and the second three-dimensional coordinate in each of the plurality of protrusions; and calculating the position of the light source based on the plurality of light source vectors.

[0010] Another aspect of the present disclosure is an inspection apparatus comprising: a plurality of cameras; a plurality of light sources; and a control device for calculating the positions of the plurality of light sources, the control device being configured to receive a plurality of images captured by the plurality of cameras, each of the plurality of images including a fixture illuminated by a selected light source from the plurality of light sources, the fixture including a base and a plurality of protrusions attached to the base, each of the plurality of protrusions including a shaft attached to the base and a sphere attached to the shaft, each of the plurality of images including a sphere and the shadow of the sphere projected onto the base for each of the plurality of protrusions; calculating a first three-dimensional coordinate of the center of the sphere and a second three-dimensional coordinate of the center of the shadow of the sphere for each of the plurality of protrusions based on the plurality of images; calculating a plurality of light source vectors for the plurality of protrusions, each of the plurality of light source vectors passing through the first three-dimensional coordinate and the second three-dimensional coordinate for each of the plurality of protrusions; and calculating the position of a selected light source based on the plurality of light source vectors. [Effects of the Invention]

[0011] According to this disclosure, the accuracy of detecting the position of the light source can be improved. [Brief explanation of the drawing]

[0012] [Figure 1] Figure 1 is a schematic diagram of an inspection apparatus according to an embodiment. [Figure 2] Figure 2 is a flowchart showing the light source position detection method according to the embodiment. [Figure 3] Figure 3 is a schematic cross-sectional view showing a protrusion illuminated by a light source. [Figure 4] Figure 4 is a schematic perspective view showing a sphere and its shadow. [Figure 5] Figure 5 is a schematic cross-sectional view showing the light source vector and the position of the light source. [Figure 6] Figure 6 shows the simulation results. [Figure 7] Figure 7 shows the experimental results. [Figure 8] Figure 8 shows the relationship between the crossover error and the error in the light source position. [Modes for carrying out the invention]

[0013] Embodiments of this disclosure will be described in detail below with reference to the attached drawings. The specific dimensions, materials, and numerical values ​​shown in these embodiments are merely illustrative for ease of understanding and do not limit this disclosure unless otherwise specified. In this specification and drawings, elements having substantially the same function and configuration are denoted by the same reference numerals to avoid redundant explanations, and elements not directly related to this disclosure are omitted from the illustrations.

[0014] Figure 1 is a schematic diagram showing an inspection device 100 according to an embodiment. The inspection device 100 detects the shape of object 1 using the illuminance difference stereo method. For example, the inspection device 100 is configured to detect defects such as scratches and dents on the surface of object 1. The objects detected by the inspection device 100 are not limited to these.

[0015] For example, the inspection device 100 includes a plurality of light sources L, a plurality of cameras C, a frame 2, a table 3, a jig 10, and a control device 90. The inspection device 100 may further include other components. Also, the inspection device 100 may not include at least one of the above components.

[0016] Four light sources L are shown in FIG. 1. The number of light sources L is not limited to four, and may be two, three, or five or more. For example, the light source L may be LED (Light-Emitting Diode) lighting. The plurality of light sources L are arranged spaced apart from each other and configured to irradiate an object 1 or a jig 10 set on the table 3 from different positions.

[0017] In the present embodiment, the light source L is connected communicably, either wired or wirelessly, to the control device 90, and the control device 90 is configured to control the operation of the light source L. Alternatively, in other embodiments, the light source L may not be connected communicably to the control device 90, and the light source L may be manually controlled by an operator.

[0018] Two cameras C are shown in FIG. 1. The number of cameras is not limited to two, and may be three or more. For example, each of the cameras C may be a CCD (Charge Coupled Device) camera or a CMOS (Complementary Metal Oxide Semiconductor) camera. For example, each of the cameras C includes a macro lens.

[0019] The two cameras C function as a stereo camera during calibration using the jig 10 (details will be described later). Also, one of the two cameras C is the main camera used in the inspection of the object 1, and the other of the two cameras C is an auxiliary camera not used in the inspection of the object 1.

[0020] In this embodiment, camera C is connected to control device 90 via wired or wireless communication and transmits the obtained images to control device 90. Alternatively, control device 90 may be configured to control the operation of camera C. In other embodiments, camera C does not need to be connected to control device 90 for communication purposes, and images may be input to control device 90 via a storage medium.

[0021] In this embodiment, frame 2 supports the light source L and camera C. In other embodiments, at least a portion of the light source L and camera C may be supported separately. For example, the positions of the light source L and camera C may be fixed or adjustable.

[0022] Table 3 is positioned opposite the light source L and camera C. Object 1 and jig 10 are set on table 3. For example, the height from table 3 to frame 2 may be fixed or adjustable.

[0023] The jig 10 is used for calibration to detect the position of each light source L relative to the main camera before inspecting object 1. For example, the jig 10 includes a base 11 and a plurality of protrusions 12.

[0024] The base 11 supports a plurality of protrusions 12. For example, the base 11 may be a flat plate. The base 11 includes a base surface 11a that supports the plurality of protrusions 12. Of the base 11, at least the base surface 11a has a light color such as white so that the shadows of the protrusions 12 projected onto the base surface 11a are clearly visible. In this embodiment, the entire base 11 has a light color.

[0025] Figure 1 shows four protrusions 12. The number of protrusions 12 is not limited to four; there may be two, three, or five or more. Multiple protrusions 12 are attached to the base 11. Multiple protrusions 12 are arranged on the base surface 11a so as to be spaced apart from each other. For example, the multiple protrusions 12 may be arranged in a predetermined pattern or randomly. Also, for example, the heights of the multiple protrusions 12 (the distance from the base surface 11a to the tip of each protrusion 12) may be the same or different from each other.

[0026] Each of the multiple protrusions 12 includes a shaft 13 and a ball 14.

[0027] The shaft 13 is attached to the base 11. The shaft 13 protrudes from the base surface 11a. For example, the shaft 13 may have various shapes, such as a cylindrical shape or a polygonal prism shape. For example, the diameters of multiple shafts 13 may be the same or different.

[0028] The ball 14 is attached to the end of the shaft 13. For example, the diameter of the ball 14 is greater than the thickness of the shaft 13. For example, the diameters of multiple balls 14 may be the same as or different from each other.

[0029] For example, the sphericity of the sphere 14 may be maintained within a predetermined range. For example, in this disclosure, "sphericity" may mean the difference in radius between the smallest circumscribed sphere and the largest inscribed sphere, with the center of the least-squares mean sphere of the sphere's surface as its center, and may be measured according to the evaluation method specified in JIS B1501. For example, the "predetermined range" may be a range that does not appear as an error when calculating the three-dimensional coordinates X1 and X2 described later. For example, the "predetermined range" may be determined according to various factors such as the size of the pixels used in camera C, the angle of view of the lens used in camera C, and the distance from jig 10 to camera C. For example, the sphericity may be maintained at a value smaller than the length of one pixel in the image obtained by camera C.

[0030] Of the protrusions 12, at least the spheres 14 have a dark color, such as black, so that they are clearly distinguishable from the base surface 11a. In this embodiment, the entire protrusion 12 has a dark color.

[0031] The control device 90 controls the inspection device 100, either entirely or partially. For example, the control device 90 includes components such as a processor 90a, a storage device 90b, and a connector 90c, which are connected to each other via a bus. For example, the processor 90a includes a CPU (Central Processing Unit). For example, the storage device 90b includes a hard disk, a ROM (Read Only Memory) for storing programs, and a RAM (Random Access Memory) as a work area. The control device 90 is connected to each component of the inspection device 100 via the connector 90c so as to be able to communicate via wired or wireless means. For example, the control device 90 may further include other components such as a display device such as a liquid crystal display or a touch panel, and an input device such as a keyboard, buttons, or a touch panel. For example, the operation of the control device 90 may be achieved by having the processor 90a execute a program stored in the storage device 90b.

[0032] During the inspection of object 1, multiple light sources L sequentially illuminate object 1. The main camera captures images of object 1 illuminated by each light source. The processor 90a of the control device 90 calculates the normal vectors of each position on the surface of object 1 based on the multiple images captured by the main camera. This allows the shape of object 1 to be obtained.

[0033] Next, we will explain the calibration process for detecting the light source position.

[0034] Figure 2 is a flowchart illustrating a light source position detection method according to an embodiment. The steps shown in Figure 2 (steps S100 to S116) are performed before inspection of object 1. For example, in this embodiment, steps S100 to S116 are performed by the processor 90a of the control device 90. Alternatively, in other embodiments, some of steps S100 to S116 may be performed by an operator.

[0035] The processor 90a performs internal calibration of camera C (step S100). Specifically, the processor 90a corrects (removes) lens distortion for each of the cameras C. Distortion correction can be performed based on various known methods.

[0036] The processor 90a performs external calibration of camera C (step S102). Specifically, the processor 90a calculates the relative distance between cameras C. The calculation of the relative distance between cameras C can be performed based on various known methods.

[0037] The operator prepares the jig 10 and sets it on the table 3.

[0038] The processor 90a lights up one of the light sources L selected from among the multiple light sources L and illuminates the jig 10 (step S104). For example, the processor 90a may light up one of the multiple light sources L based on input from the operator. Alternatively, in other embodiments, one of the multiple light sources L may be manually lit by the operator.

[0039] Figure 3 is a schematic cross-sectional view showing the protrusion 12 illuminated by the light source Li. In Figure 3, only one of the multiple protrusions 12 is shown for better understanding.

[0040] In step S104, the selected light source Li illuminates the jig 10 so that the shadow 15 of the sphere 14 is projected onto the base surface 11a. Although only one protrusion 12 is shown in Figure 3, the light source Li illuminates the jig 10 so that the shadows 15 of the sphere 14 of all the protrusions 12 are projected onto the base surface 11a.

[0041] Returning to Figure 2, the processor 90a captures multiple images of the fixture 10 illuminated by the light source Li using the multiple cameras C (step S106). For example, the processor 90a may control each of the multiple cameras C based on input from the operator. Alternatively, in other embodiments, each of the multiple cameras C may be manually controlled by the operator.

[0042] Referring to Figure 3, for example, in step S106, each camera C takes an image of the jig 10 such that both the sphere 14 and the shadow 15 are included in a single image. Although only one protrusion 12 is shown in Figure 3, each camera C takes an image of the jig 10 such that the sphere 14 and shadow 15 of all the protrusions 12 are included in a single image.

[0043] Returning to Figure 2, the processor 90a receives multiple images of the fixture 10 from multiple cameras C (step S108). For example, in this embodiment, the processor 90a receives two images of the fixture 10 from two cameras C, each image including all the spheres 14 and shadows 15 of the protrusions 12. In this embodiment, the processor 90a receives the images from the cameras C online. Alternatively, in other embodiments, the control device 90 may receive the images via a storage medium.

[0044] The processor 90a calculates the three-dimensional coordinates of the center of the sphere 14 and the three-dimensional coordinates of the center of the shadow 15 for each protrusion 12 based on multiple images (step S110).

[0045] Figure 4 is a schematic perspective view showing the sphere 14 and its shadow 15. In Figure 4, only one of the multiple projections 12 is shown for better understanding.

[0046] In step S110, the processor 90a identifies the center of the sphere 14 and the center of the shadow 15 for each protrusion 12 from each image. For example, the identification of the center of the sphere 14 and the center of the shadow 15 may be performed by various known image recognition techniques.

[0047] The processor 90a uses the forward resection method (stereo measurement) to calculate the three-dimensional coordinates of the center of the sphere 14 (first three-dimensional coordinates) X1 and the three-dimensional coordinates of the center of the shadow 15 (second three-dimensional coordinates) X2 for each projection 12, from multiple images captured by multiple cameras C. Although only one projection 12 is shown in Figure 4, the processor 90a calculates the three-dimensional coordinates X1 and X2 for all projections 12.

[0048] Returning to Figure 2, the processor 90a calculates the light source vector for each protrusion 12 (step S112).

[0049] Referring to Figure 3, in step S112, the processor 90a calculates the light source vector V as the line segment extending from the three-dimensional coordinate X2 of the center of the shadow 15 to the three-dimensional coordinate X1 of the center of the sphere 14. Theoretically, the extension of the light source vector V passes through the center of the light source Li. Although only one protrusion 12 is shown in Figure 3, the processor 90a calculates the light source vector V for all protrusions 12.

[0050] Returning to Figure 2, the processor 90a calculates the position of the light source Li based on the multiple light source vectors V of the multiple protrusions 12 (step S114).

[0051] Figure 5 is a schematic cross-sectional view showing the light source vectors V1, V2, V3 and the position Pi of the light source Li. In Figure 5, for better understanding, only three of the multiple protrusions 12 and their respective light source vectors V1, V2, V3 are shown.

[0052] Theoretically, the extensions of the light source vectors V1, V2, and V3 intersect each other at the center of the light source Li. However, due to various errors, the extensions of the light source vectors V1, V2, and V3 may not intersect each other at a single point. Therefore, for example, the processor 90a uses the least squares method to calculate the point where the extensions of the light source vectors V1, V2, and V3 are closest to each other as the center Pi of the light source Li. In this way, the three-dimensional coordinates of the center Pi of the light source Li are calculated. Note that in Figure 5, only the three light source vectors V1, V2, and V3 are shown, but the processor 90a uses the light source vectors V of all the protrusions 12 to calculate the center Pi of the light source Li. However, if there is a light source vector V that is extremely far from the other light source vectors V, this light source vector may be excluded from the calculation of the position of the light source Li.

[0053] Returning to Figure 2, the processor 90a determines whether or not it has calculated the positions of all light sources L (step S116).

[0054] If the positions of all light sources L have not been calculated in step S116 (NO), the processor 90a repeats steps S104 to S116 until the positions of all light sources L have been calculated.

[0055] If the positions of all light sources L have been calculated in step S116 (YES), the processor 90a terminates the calibration.

[0056] Next, the evaluation results of the inspection apparatus 100 and the light source position detection method according to the embodiment will be described.

[0057] Figure 6 shows the simulation results. The simulation in Figure 6 was performed by a computer in a virtual space.

[0058] In the simulation shown in Figure 6, multiple light sources L, multiple cameras C, and a jig 10 were placed in the virtual space as shown in Figure 1. In the simulation in Figure 6, eight light sources L numbered from No. 1 to No. 8, two cameras C, and eight protrusions 12 were used. Light sources L numbered from No. 1 to No. 8 were placed at the positions indicated in "Illumination Position (True Value)".

[0059] Steps S104 to S116 in Figure 2 were performed using the multiple light sources L, multiple cameras C, and jig 10 in the virtual space described above. The positions of light sources L No. 1 to No. 8 obtained by the simulation are shown in "Illumination Position (Measured)". The difference between "Illumination Position (True Value)" and "Illumination Position (Measured)" is shown in "Difference".

[0060] Furthermore, camera C in the virtual space uses pixels of a predetermined size, just like a real camera. Therefore, a "difference" arises between the "illumination position (true value)" and the "illumination position (measured)" due to the size of the pixels.

[0061] As described above, eight protrusions 12 are used in the simulation. Therefore, eight light source vectors are used to determine the "illumination position (measured)" of one light source L. "Mean Error" represents the average distance (error) between the "illumination position (measured)" of each light source L and the extensions of the eight light source vectors. In other words, "Mean Error" can also be called the "crossover error" between multiple light source vectors.

[0062] As shown by the RMS (root mean square) of the difference d, the detection error of the light source position in the simulation is 0.988 mm.

[0063] Figure 7 shows the experimental results. The experiment in Figure 7 was performed using the actual testing device 100.

[0064] In the experiment shown in Figure 7, eight light sources L numbered No. 1 to No. 8, two cameras C, and eight protrusions 12 were used, similar to the simulation in Figure 6. The conditions for the experiment in Figure 7, including the size and position of the light sources L, cameras C, and fixture 10, as well as the pixels of camera C, were the same as those in the simulation in Figure 6. Note that the true position of the light sources L could not be measured in the experiment, so "Illumination Position (True Value)" is not shown in Figure 7.

[0065] Steps S104 to S116 in Figure 2 were performed using the actual inspection apparatus 100 described above. The positions of light sources L from No. 1 to No. 8 obtained from the experiment are shown in "Illumination Position (Measured)". Note that, as described above, "Illumination Position (True Value)" is not shown in the experiment, so the "Difference" between "Illumination Position (True Value)" and "Illumination Position (Measured)" is also not shown in Figure 7.

[0066] Similar to Figure 6, "Mean Error" represents the average value of the distance (error) between the "illumination position (measured)" of each light source L and the extension lines of the eight light source vectors, i.e., the "crossover error". "Max Error" represents the maximum value of the distance (error) between the "illumination position (measured)" of each light source L and the extension lines of the eight light source vectors.

[0067] Furthermore, as indicated by the RMS (Root Mean Square) of the "Mean Error," the estimation accuracy in the experiment was 0.178 mm.

[0068] Figure 8 shows the relationship between the crossover error and the light source position error. In Figure 8, the horizontal axis represents the crossover error, i.e., "Mean Error," and the vertical axis represents the light source position error (detection error). The plots show the "difference d" for the "Mean Error" of each light source L in Figure 6. The straight lines show the linear approximation of these plots. In the above simulation and experiment, it is assumed that as the crossover error increases, the light source position detection error also increases. The vertical dashed line shows the "Mean Error" in the experiment (=0.178 mm). The intersection of the straight line and the dashed line shows the estimated light source position error in the experiment (=0.88 mm). Therefore, from Figures 6 to 8, it can be seen that the light source position detection error in the simulation (0.988 mm) and the light source position detection error in the experiment (0.88 mm) are approximately the same. Thus, in the actual inspection device 100, it can be seen that the light source position detection error mainly depends on the pixel size. According to several sources, the detection error of the light source position in conventional methods is approximately 10 mm (for example, Non-Patent Literature 1 mentioned above, and H. Santo, et. al., "Light Structure from Pin Motion: Simple and Accurate Point Light Calibration for Physics-Based Modeling", ECCV 2018: 15th European Conference, Munich, Germany, September 8-14, 2018, Proceedings, Part III, September 2018, p. 3-19). Therefore, it can be seen that the detection accuracy of the light source position according to this embodiment is higher than that of conventional methods.

[0069] As described above, the light source position detection method according to the embodiment includes preparing a jig 10 including a base 11 and a plurality of protrusions 12 attached to the base 11, wherein each of the plurality of protrusions 12 includes a shaft 13 attached to the base 11 and a sphere 14 attached to the shaft 13; illuminating the jig 10 with a light source Li such that the shadows 15 of each of the spheres 14 of the plurality of protrusions 12 are projected onto the base 11; capturing a plurality of images of the jig 10 illuminated by the light source Li using a plurality of cameras C; calculating the three-dimensional coordinates X1 of the center of the sphere 14 and the three-dimensional coordinates X2 of the center of the shadow 15 of the sphere 14 for each of the plurality of protrusions 12 based on the plurality of images; calculating a plurality of light source vectors V for the plurality of protrusions 12, wherein each of the plurality of light source vectors V passes through the three-dimensional coordinates X1 and X2 at each of the plurality of protrusions 12; and calculating the position of the light source Li based on the plurality of light source vectors V.

[0070] Furthermore, the inspection apparatus 100 according to this embodiment includes a plurality of cameras C, a plurality of light sources L, and a control device 90 that calculates the positions of the plurality of light sources L. The control device 90 is configured to receive a plurality of images captured by a plurality of cameras C, each of which includes a fixture 10 illuminated by a selected light source Li from a plurality of light sources L, the fixture 10 including a base 11 and a plurality of protrusions 12 attached to the base 11, each of which includes a shaft 13 attached to the base 11 and a sphere 14 attached to the shaft 13, each of which includes, for each of the plurality of protrusions 12, the sphere 14 and the shadow 15 of the sphere 14 projected onto the base 11, and to calculate, based on the plurality of images, the three-dimensional coordinates X1 of the center of the sphere 14 and the three-dimensional coordinates X2 of the center of the shadow 15 of the sphere 14 for each of the plurality of protrusions 12, and to calculate a plurality of light source vectors V for the plurality of protrusions 12, each of which passes through the three-dimensional coordinates X1 and X2 in each of the plurality of protrusions 12, and to calculate the position of the selected light source Li based on the plurality of light source vectors V.

[0071] According to the above-described light source position detection method and inspection apparatus 100, the position of the light source Li is detected based on the three-dimensional coordinate X1 of the center of the sphere 14 and the three-dimensional coordinate X2 of the center of the shadow 15 of the sphere 14. For example, the three-dimensional coordinate X1 of the sphere 14 and the three-dimensional coordinate X2 of the shadow 15 can be identified with higher accuracy compared to identifying the highlight area as in Non-Patent Literature 1. Also, for example, the three-dimensional coordinate X1 of the sphere 14 and the three-dimensional coordinate X2 of the shadow 15 can be identified with higher accuracy compared to identifying the intersection point between multiple lines. As a result, the position of the light source Li can also be detected with high accuracy. Therefore, the detection accuracy of the light source position can be improved. Furthermore, according to the above-described light source position detection method and inspection apparatus 100, it is not necessary to match the dimensions of the multiple protrusions 12. Therefore, the cost of the jig 10 can be reduced. Furthermore, according to the above-described light source position detection method and inspection apparatus 100, only one shot is required for each camera C to calculate the position of one light source L. Therefore, the calibration time can be shortened.

[0072] Although embodiments have been described above with reference to the attached drawings, this disclosure is not limited to the embodiments described above. It will be obvious to those skilled in the art that various modifications or alterations can be conceived within the scope of the claims, and these will naturally also fall within the technical scope of this disclosure. Furthermore, the steps of the methods of the embodiments described above do not have to be performed in the order described above, and may be performed in a different order as long as it does not result in a technical inconsistency. [Explanation of Symbols]

[0073] 10 jigs 11 Bass 12 protrusions 13 shafts 14 balls 15 The Shadow of the Ball 90 Control device 100 Inspection device C Camera L light source Li Selected light source Pi Light source position V Light source vector V1 Light source vector V2 Light Source Vector V3 Light Source Vector X1 is the three-dimensional coordinate of the center of the sphere (first three-dimensional coordinate). X2 Three-dimensional coordinates of the center of the sphere's shadow (second three-dimensional coordinates)

Claims

1. The present invention provides a jig comprising a base and a plurality of projections attached to the base, wherein each of the plurality of projections includes a shaft attached to the base and a ball attached to the shaft. The jig is illuminated by a light source such that the shadow of each of the spheres of the plurality of protrusions is projected onto the base, Multiple cameras capture multiple images of the jig illuminated by the light source, Based on the aforementioned multiple images, for each of the multiple protrusions, the first three-dimensional coordinate of the center of the sphere and the second three-dimensional coordinate of the center of the sphere's shadow are calculated. The process involves calculating multiple light source vectors for the multiple protrusions, wherein each of the multiple light source vectors passes through the first three-dimensional coordinate system and the second three-dimensional coordinate system at each of the multiple protrusions. The position of the light source is calculated based on the plurality of light source vectors, including, Method for detecting the position of a light source.

2. Multiple cameras, Multiple light sources, A control device for calculating the positions of the plurality of light sources, The control device is The system receives multiple images captured by the multiple cameras, each of which includes a fixture illuminated by a selected light source from the multiple light sources, the fixture including a base and a plurality of protrusions attached to the base, each of which includes a shaft attached to the base and a sphere attached to the shaft, and each of which includes, for each of the plurality of protrusions, the sphere and the shadow of the sphere projected onto the base. Based on the aforementioned multiple images, for each of the multiple protrusions, the first three-dimensional coordinate of the center of the sphere and the second three-dimensional coordinate of the center of the sphere's shadow are calculated. The process involves calculating multiple light source vectors for the multiple protrusions, wherein each of the multiple light source vectors passes through the first three-dimensional coordinate system and the second three-dimensional coordinate system at each of the multiple protrusions. Based on the plurality of light source vectors, the position of the selected light source is calculated, A control device configured to perform the following actions: Equipped with, Inspection device.

Images