Inspection equipment, inspection method, and program

Abstract

To provide an inspection device, an inspection method, and a program capable of improving throughput.SOLUTION: An inspection device comprises a processing section for processing an image of a surface of an inspection object illuminated by an illumination section and captured by an imaging section. The surface of the inspection object includes a plurality of areas containing a first area. The illumination section includes first and second illumination that illuminate the surface from a first and second directions. The imaging section acquires information on first and second captured images of the plurality of areas illuminated by the first and second illumination. The processing section calculates a first difference image corresponding to a difference between a first area image and a second area image on the basis of the first area image in the first area based on the first captured image and the second area image in the first area based on the second captured image. The processing section determines a defective area in the first area on the basis of the first difference image.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] Embodiments of the present invention relate to an inspection apparatus, an inspection method, and a program.

Background Art

[0002] There are an inspection apparatus, an inspection method, and a program for inspecting the appearance of an inspection object such as a substrate. By the inspection, defects such as unevenness on the surface of the inspection object are detected.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Embodiments of the present invention provide an inspection apparatus, an inspection method, and a program capable of improving throughput.

Means for Solving the Problems

[0005] The inspection apparatus according to the embodiment includes a processing unit that processes an image of the surface of an object to be inspected, which is illuminated by an illumination unit and imaged by an imaging unit. The surface of the object to be inspected includes a plurality of regions, including a first region. The illumination unit includes a first illumination that illuminates the surface from a first direction and a second illumination that illuminates the surface from a second direction. The imaging unit acquires imaging data which includes information from a first image of the plurality of regions illuminated by at least the first illumination and information from a second image of the plurality of regions illuminated by at least the second illumination. The processing unit calculates a first difference image corresponding to the difference between the first region image and the second region image, based on a first region image of the first region based on the first image and a second region image of the first region based on the second image. The processing unit determines a defective region in the first region based on the first difference image. [Brief explanation of the drawing]

[0006] [Figure 1] This is a schematic diagram illustrating an inspection apparatus according to the first embodiment. [Figure 2] This is a schematic plan view illustrating an example of an object to be inspected. [Figure 3] This is a schematic plan view illustrating a part of the inspection apparatus according to the first embodiment. [Figure 4] This is a flowchart illustrating an inspection method in the inspection apparatus according to the first embodiment. [Figure 5] This is a flowchart illustrating the process of determining a defect area in the inspection apparatus according to the first embodiment. [Figure 6] Figures 6(a) to 6(e) are schematic diagrams illustrating images in the inspection apparatus according to the first embodiment. [Figure 7] This is a schematic diagram illustrating an image in the inspection apparatus according to the first embodiment. [Figure 8] This is a flowchart illustrating the process of calculating multiple region images in an inspection apparatus according to the first embodiment. [Figure 9]Figures 9(a) to 9(c) are schematic diagrams illustrating the process of detecting multiple region images in an inspection apparatus according to the first embodiment. [Figure 10] Figures 10(a) to 10(c) are schematic diagrams illustrating the process of correcting brightness in the inspection apparatus according to the first embodiment. [Figure 11] This is a flowchart illustrating an inspection method in the inspection apparatus according to the first embodiment. [Figure 12] This is a schematic diagram illustrating an inspection apparatus according to the second embodiment. [Figure 13] This is a schematic diagram illustrating an image captured in an inspection apparatus according to the second embodiment. [Figure 14] This table illustrates combinations of captured images in the inspection apparatus according to the second embodiment. [Modes for carrying out the invention]

[0007] The embodiments of the present invention will be described below with reference to the drawings. Drawings are schematic or conceptual, and the relationships between the thickness and width of each part, as well as the ratios of the sizes of different parts, are not necessarily identical to those of reality. Even when representing the same part, the dimensions and ratios may differ between drawings. In this specification and in each figure, elements similar to those already described are denoted by the same reference numerals, and detailed explanations are omitted as appropriate.

[0008] (First embodiment) Figure 1 is a schematic diagram illustrating an inspection apparatus according to the first embodiment. The inspection apparatus 100 (inspection system) according to this embodiment includes a processing unit 30. The inspection apparatus 100 may further include an imaging unit 10, an illumination unit 20, and an output unit 40. The inspection device 100 acquires an image of the surface of the object to be inspected W (workpiece) and detects surface shape defects such as irregularities on the surface of the object to be inspected W based on the obtained image.

[0009] The illumination unit 20 illuminates the surface Wf of the inspection object W. As shown in FIG. 1, for example, the illumination unit 20 is located obliquely above the surface Wf and irradiates the surface Wf with illumination light L from an oblique direction. The illumination unit 20 is, for example, low-angle illumination. As will be described later, in this example, the illumination light L is polarized light.

[0010] The imaging unit 10 images the surface Wf of the inspection object W. For example, the imaging unit 10 is a camera arranged above the inspection object W. As will be described later, in this example, the imaging unit 10 is a polarized camera.

[0011] The processing unit 30 processes the image captured by the imaging unit 10. The processing unit 30 includes, for example, an arithmetic device such as a computer. The processing unit 30 includes a communication interface and is connected to the imaging unit 10 and the illumination unit 20 by wire or wirelessly and can communicate with the imaging unit 10 and the illumination unit 20. For example, the processing unit 30 can control the operations of the imaging unit 10 and the illumination unit 20. However, in the embodiment, the operations of the imaging unit 10 and the illumination unit 20 do not necessarily have to be controlled by the processing unit 30.

[0012] The processing unit 30 includes a storage unit 31 and an arithmetic unit 32. The image (imaging data) captured by the imaging unit 10 is input to the processing unit 30 and is arithmetically processed in the arithmetic unit 32. Specifically, the arithmetic unit 32 includes, for example, an electronic circuit such as a CPU (Central Processing Unit).

[0013] The memory unit 31 stores a program 100p that causes the computer to execute the inspection method performed by the inspection device 100. The memory unit 31 also stores acquired images and their calculation results. The processing unit 30 (arithmetic unit 32) can read the data stored in the memory unit 31 as needed. For example, the processing unit 30 (arithmetic unit 32) sequentially reads the program 100p stored in the memory unit 31 and processes the program 100p sequentially to perform image processing in the inspection device 100 or control the imaging unit 10, etc. The memory unit 31 can use a storage device such as ROM (Read Only Memory) or RAM (Random Access Memory).

[0014] Some or all of the blocks in the processing unit can utilize integrated circuits such as LSIs (Large Scale Integration) or ICs (Integrated Circuit) chipsets. Individual circuits may be used for each block, or integrated circuits may be used for some or all of them. The blocks may be provided as a single unit, or some blocks may be provided separately. Furthermore, parts of each block may be provided separately. Integration is not limited to LSIs; dedicated circuits or general-purpose processors may also be used.

[0015] The processing unit 30 is connected to the output unit 40 by wire or wireless connection and can communicate with the output unit 40. The processing unit 30 can output the inspection results (defect determination results) of the surface Wf to the output unit 40. Specifically, the output unit 40 is, for example, a display device capable of displaying images, etc. This allows the user to confirm the inspection results. The processing unit 30 may output and display acquired images and arbitrary data such as their calculation results to the output unit 40. Alternatively, the output unit 40 may be a device capable of storing acquired images and inspection results. Specifically, the output unit 40 may be a server or a database.

[0016] Figure 2 is a schematic plan view illustrating an example of an object to be inspected. Figure 2 shows an example of an object W to be inspected, with its surface Wf viewed from above. The surface Wf of the object W under inspection contains multiple regions R. For example, the surface Wf has a pattern divided into multiple regions R.

[0017] In this example, each of the multiple regions R has the same pattern (shape). In other words, the surface Wf has a repeating pattern in which multiple regions R are arranged in an array on a plane.

[0018] For example, multiple regions R include regions R1 to R9. One pattern from regions R1 to R9 is the same as another pattern from regions R1 to R9. Note that patterns are considered the same if they are substantially identical, and may differ in terms of manufacturing variations or in the details of the design.

[0019] For example, the object to be inspected W is a substrate on which a battery is formed, and the region R is the battery cell. When viewed from above, the planar shape of the object to be inspected W and the region R is rectangular. For example, the object to be inspected W is a large workpiece with a side length (width) of about 50 cm.

[0020] However, the object to be inspected W is not limited to the above; for example, it may be a set of multiple workpieces arranged together. In this case, each workpiece contains one or more regions R. By arranging multiple workpieces together, multiple regions R are arranged on a plane (on the surface Wf). For example, multiple workpieces may be arranged on a tray, and the tray as a whole may be inspected together.

[0021] Figure 3 is a schematic plan view illustrating a part of the inspection apparatus according to the first embodiment. Figure 3 shows the inspection object W, imaging unit 10, and illumination unit 20 as viewed from above. The illumination unit 20 includes multiple lights 2n. When viewed from above, the multiple lights 2n are arranged around the inspection object W, surrounding it. Each of the multiple lights 2n illuminates the surface Wf of the inspection object W from multiple directions.

[0022] For example, the multiple lights 2n include the first to fourth lights 21 to 24. The first light 21 is positioned on the back side of the object W to be inspected and illuminates the surface Wf from the back (first direction). The second light 22 is positioned on the front side of the object W to be inspected and illuminates the surface Wf from the front (second direction). The third light 23 is positioned on the right side of the object W to be inspected and illuminates the surface Wf from the right (third direction). The fourth light 24 is positioned on the left side of the object W to be inspected and illuminates the surface Wf from the left (fourth direction). In this embodiment, the number of lights 2n is not limited to four and is arbitrary.

[0023] The first direction (back) and the second direction (front) are opposite directions. The third direction (right) is perpendicular to the first direction (back). The third direction (right) and the fourth direction (left) are opposite directions.

[0024] In this example, the illumination unit 20 is polarized illumination. That is, each of the multiple illuminations 2n emits illumination light polarized in each of the multiple polarization directions. The light source of the first illumination 21 is provided with a polarizing film 21f (filter) that transmits light polarized in a first polarization direction. As a result, the first illumination 21 irradiates the surface Wf with first illumination light L1 polarized in the first polarization direction. For example, the first polarization direction is 0°, and the first illumination light L1 is linearly polarized at 0°. The light source of the second illumination 22 is provided with a polarizing film 22f (filter) that transmits light polarized in the second polarization direction. As a result, the second illumination 22 irradiates the surface Wf with second illumination light L2 polarized in the second polarization direction. For example, the second polarization direction is 90°, and the second illumination light L2 is linearly polarized at 90°. The light source of the third illumination 23 is equipped with a polarizing film 23f (filter) that transmits light polarized in the third polarization direction. As a result, the third illumination 23 irradiates the surface Wf with third illumination light L3 polarized in the third polarization direction. For example, the third polarization direction is 135°, and the third illumination light L3 is linearly polarized at 135°. The light source of the fourth illumination 24 is equipped with a polarizing film 24f (filter) that transmits light polarized in the fourth polarization direction. As a result, the fourth illumination 24 irradiates the surface Wf with fourth illumination light L4 polarized in the fourth polarization direction. For example, the fourth polarization direction is 45°, and the fourth illumination light L4 is linearly polarized at 45°.

[0025] The imaging unit 10 images the surface Wf illuminated by the illumination unit 20. As a result, the imaging unit 10 acquires imaging data containing information from multiple imaging images C. Each imaging image C is, for example, an image of the surface Wf illuminated by each illumination, and includes images of multiple regions R. The imaging unit 10 outputs the multiple imaging images C (imaging data) to the processing unit 30.

[0026] More specifically, the imaging data acquired by the imaging unit 10 includes information on the image C1 (first image) of the surface Wf illuminated by at least the first illumination 21, information on the image C2 (second image) of the surface Wf illuminated by at least the second illumination 22, information on the image C3 of the surface Wf illuminated by at least the third illumination 23, and information on the image C4 of the surface illuminated by at least the fourth illumination 24.

[0027] The imaging data containing information from multiple captured images C may consist of multiple data points acquired through multiple imaging attempts, or it may consist of a single data point acquired through a single imaging attempt. Using a polarizing camera can reduce the number of imaging attempts.

[0028] A polarizing camera has a structure in which pixels equipped with polarizing filters for different polarization directions are arranged in an array. For example, a first pixel equipped with a polarizing filter for a first polarization direction, a second pixel equipped with a polarizing filter for a second polarization direction, a third pixel equipped with a polarizing filter for a third polarization direction, and a fourth pixel equipped with a polarizing filter for a fourth polarization direction are arranged periodically in a grid.

[0029] In this embodiment using polarized illumination and a polarized camera, captured image C1 is an image obtained from light polarized in the first polarization direction, captured image C2 is an image obtained from light polarized in the second polarization direction, captured image C3 is an image obtained from light polarized in the third polarization direction, and captured image C4 is an image obtained from light polarized in the fourth polarization direction.

[0030] For example, the imaging unit 10 acquires imaging data by taking a single image while simultaneously irradiating the surface Wf with multiple illumination lights polarized in each of multiple polarization directions. Multiple imaging images C are then created from the imaging data obtained from that single image.

[0031] More specifically, the imaging unit 10 acquires imaging data by imaging the surface Wf in one pass under the condition that the surface Wf is simultaneously irradiated with the first to fourth illumination lights L1 to L4 from each of the first to fourth illuminations 21 to 24. This imaging data contains information about the captured images C1, C2, C3, and C4. By extracting the data of each of the first to fourth pixels from the imaging data, the captured images C1, C2, C3, and C4 can be created. In other words, the imaging unit 10 creates the captured images C1, C2, C3, and C4 from the imaging data obtained from the above single pass. In this way, by using a polarization camera, it is possible to simultaneously capture images of light with multiple polarization directions in a single imaging pass using one camera.

[0032] Figure 4 is a flowchart illustrating an inspection method in the inspection apparatus according to the first embodiment. The processing unit 30 (computer) executes each of the inspection methods shown below, for example, based on program 100p. However, some or all of the inspection methods shown below may be executed based on hardware operations that do not require a program. As shown in Figure 4, the inspection method in the inspection apparatus according to the embodiment includes a step S1 for acquiring imaging data, a step S2 for calculating multiple region images, and a step S3 for determining a defective region.

[0033] In step S1, for example, the imaging unit 10 acquires imaging data by imaging the surface Wf. The imaging unit 10 sends multiple imaging images C included in the imaging data to the processing unit 30. For example, the imaging unit 10 creates imaging images C1, C2, C3, and C4 based on the imaging data obtained in a single imaging, and the processing unit 30 acquires the imaging images C1, C2, C3, and C4 created by the imaging unit 10. Alternatively, the processing unit 30 may create imaging images C1, C2, C3, and C4 based on the imaging data.

[0034] In step S2, the processing unit 30 (calculation unit 32) calculates multiple region images A (see Figure 5) based on at least a portion of the multiple captured images C. A region image A is an image that contains one image of multiple regions R. A region image A is, for example, an image obtained by extracting a portion of one region R from the captured image C (or the polarization image described later). For example, multiple region images A can be obtained by dividing the captured image C (or the polarization image). For example, the processing unit 30 detects images of multiple regions R in the multiple captured images C (or the polarization image described later) and calculates multiple region images A. A region image A contains, for example, an image of at least one region R.

[0035] Specifically, when the first region R1 is the object to be inspected, the processing unit 30 calculates the region image A1 (first region image) of the first region R1 based on the captured image C1, the region image A2 (second region image) of the first region R1 based on the captured image C2, the region image A3 of the first region R1 based on the captured image C3, and the region image A4 of the first region R1 based on the captured image C4. Each of these region images A1, A2, A3, and A4 is an image that includes at least an image of the first region R1.

[0036] In step S3, the processing unit 30 (calculation unit 32) determines the defective area based on the multiple region images A calculated in step S2. For example, as will be described later with respect to Figure 5, the processing unit 30 generates a difference image corresponding to the difference between the multiple region images A, based on the multiple region images A and an average image obtained by averaging at least a portion of the multiple region images A. Then, the processing unit 30 determines the defective area in one of the multiple regions R based on the difference image.

[0037] The processing unit 30 performs steps S2 and S3 described above for each of the multiple regions R that are to be inspected. This allows the entire surface Wf to be inspected.

[0038] Figure 5 is a flowchart illustrating the process of determining a defect area in the inspection apparatus according to the first embodiment. As shown in Figure 5, the processing unit 30 calculates region image A1 based on captured image C1, region image A2 based on captured image C2, region image A3 based on captured image C3, and region image A4 based on captured image C4.

[0039] In step S301, the processing unit 30 calculates an average image obtained by averaging at least region image A1 and region image A2. In this example, the processing unit 30 calculates an average image M obtained by averaging region images A1, A2, A3, and A4.

[0040] In step S302, the processing unit 30 calculates a difference image D1 (first difference image) corresponding to the difference between region image A1 and region image A2. Specifically, the difference image D1 is the difference between the image obtained by dividing the pixel values ​​of region image A1 by the pixel values ​​of average image M and the image obtained by dividing the pixel values ​​of region image A2 by the pixel values ​​of average image M.

[0041] In step S303, the processing unit 30 calculates an image of the difference between the first difference image D1 and the image obtained by smoothing the first difference image D1 (step S303). Then, it calculates an image (binary image) obtained by binarizing the absolute value of the image obtained in step S303 (step S304).

[0042] Similarly, in step S305, the processing unit 30 calculates a difference image D2 corresponding to the difference between region image A3 and region image A4. Specifically, the difference image D2 is the difference between the image obtained by dividing the pixel values ​​of region image A3 by the pixel values ​​of average image M and the image obtained by dividing the pixel values ​​of region image A4 by the pixel values ​​of average image M.

[0043] In step S306, the processing unit 30 calculates an image of the difference between the difference image D2 and the image obtained by smoothing the difference image D2 (step S306). Then, it calculates an image (binary image) obtained by binarizing the absolute value of the image obtained in step S306 (step S307).

[0044] For example, in a binary image, one of the regions where the pixel value is binary (for example, a pixel with a white value) corresponds to a defect. The threshold value for the pixel value when binarizing can be a predetermined value that is appropriately set to detect defects.

[0045] The binary image obtained in step S304 and the binary image obtained in step S307 are integrated (step S308). Noise is removed from the integrated binary image as appropriate. Specifically, in the binary image, the area corresponding to the defect (e.g., the white area) is reduced by a predetermined number of pixels (step S309), and then expanded by a predetermined number of pixels (step S310). The predetermined number of pixels can be any predetermined value that is appropriately determined to remove noise. For example, the predetermined number of pixels is 1 pixel, and the area corresponding to the defect is reduced and expanded by one size.

[0046] In the denoised binary image, regions corresponding to defects that have an area greater than or equal to a predetermined size threshold are determined to be defective regions (step S311). Defective regions are anomaly regions where the pixel value differs from the surrounding area. The processing unit 30 may output the defect region determination result to the output unit 40. The predetermined size threshold may be a value appropriately determined so as to enable the detection of defects of a desired size.

[0047] Figure 5 shows an example of using four captured images (region images), and in this embodiment, the number of captured images used to determine the defect region is not particularly limited. For example, two captured images may be used. In that case, for example, the average image M may be the average of the two images, and steps S302 to S304 and S309 to S311 will be executed, while steps S305 to S308, etc., can be omitted as appropriate.

[0048] Figures 6(a) to 6(e) are schematic diagrams illustrating images in the inspection apparatus according to the first embodiment. Figures 6(a), 6(b), 6(c), and 6(d) show enlarged portions of region images A4, A3, A1, and A2, respectively. Figure 6(e) shows an example of a binary image in which the defective region was determined by the aforementioned process S3, based on the region images in Figures 6(a) to 6(d).

[0049] In the image shown in Figure 6(e), the white areas marked with arrows correspond to the defective areas. An image like Figure 6(e) may be output by the output unit 40. The output unit 40 may also output coordinates on the surface Wf, indicating the location and extent of the defective areas. The output defect area determination result may include information on whether each defective area is concave or convex.

[0050] As described above, the processing unit 30 determines the defect region in the first region R1 based on a difference image (for example, difference image D1) derived from multiple region images A.

[0051] Image calculation (processing) is performed for each pixel, for example, on pixel values ​​such as brightness or grayscale. Calculations between images can be performed for each pixel corresponding to the same coordinate on the surface Wf. The images calculated above (average image, difference image, and binary image, etc.) are simply data obtained by calculating pixel values ​​and do not necessarily have to be output visually as images.

[0052] Figure 7 is a schematic diagram illustrating an image in the inspection apparatus according to the first embodiment. Figure 7 is a schematic diagram illustrating the determination of the unevenness of a defective region in multiple region images (or captured images). The upper part of Figure 7 is a schematic diagram showing each region image when the defective region U is a convex portion. The lower part of Figure 7 is a schematic diagram showing each region image when the defective region U is a concave portion.

[0053] As shown in the upper part of Figure 7, when the defect region U is a convex portion, the right side of the defect region U is brighter than the left side in the image when light is shone from the right, the left side of the defect region U is brighter than the right side in the image when light is shone from the left, the back of the defect region U is brighter than the front of the defect region U in the image when light is shone from behind, and the front of the defect region U is brighter than the back of the defect region U in the image when light is shone from behind.

[0054] On the other hand, as shown in the lower part of Figure 7, when the defect region U is a recess, the left side of the defect region U is brighter than the right side in the image when light is shone from the right, the right side of the defect region U is brighter than the left side in the image when light is shone from the left, the front side of the defect region U is brighter than the back side in the image when light is shone from the back, and the back side of the defect region U is brighter than the front side in the image when light is shone from the front.

[0055] Therefore, the processing unit 30 can determine whether a defective region U is a concave or convex portion in each region image (or captured image) based on the direction of light irradiation and the contrast (distribution of pixel values) within the defective region U. That is, for example, the processing unit 30 compares the pixel value (luminance) of the irradiated region of the defective region U, which is closer to the light source, with the pixel value (luminance) of the opposite region, which is further from the light source than the irradiated region. If the pixel value of the irradiated region is greater than the pixel value of the opposite region, the processing unit 30 determines that the defective region U is a convex portion. If the pixel value of the irradiated region is smaller than the pixel value of the opposite region, the processing unit 30 determines that the defective region U is a concave portion.

[0056] Next, an example of step S2, which calculates multiple region images based on multiple captured images, will be described. Figure 8 is a flowchart illustrating the process of calculating multiple region images in an inspection apparatus according to the first embodiment. As shown in Figure 8, in this example, step S2 includes step S201 for calculating multiple polarization images, step S202 for detecting multiple regions R, step S203 for performing brightness correction, and step S204 for calculating a region image.

[0057] As described above, polarized illumination and a polarized camera are used in this example. Furthermore, captured images C1, C2, C3, and C4 are obtained by imaging the surface Wf at one degree with all first to fourth illuminations 21 to 24 turned on. In such cases, the processing unit 30 calculates multiple polarization images based on the multiple captured images C (step S201). In this example, the first to fourth polarization images are calculated based on the captured images C1, C2, C3, and C4.

[0058] Here, the first to fourth polarized images are images produced by the first to fourth illumination lights L1 to L4, respectively. Let the pixel value of the first polarized image be a, the pixel value of the second polarized image be b, the pixel value of the third polarized image be c, and the pixel value of the fourth polarized image be d. That is, a, b, c, and d are components produced by the first to fourth illumination lights L1 to L4, respectively.

[0059] On the other hand, as already mentioned, the captured images C1, C2, C3, and C4 are images of light polarized in the first to fourth polarization directions, respectively. By decomposing the captured data in each polarization direction, captured images C1, C2, C3, and C4 can be generated. The pixel values ​​of captured image C1 are denoted as IA, the pixel values ​​of captured image C2 as IB, the pixel values ​​of captured image C3 as IC, and the pixel values ​​of captured image C4 as ID.

[0060] Each of the images obtained by decomposing the image in each polarization direction (IA, IB, IC, and ID) contains non-orthogonal polarization components. For example, IA includes not only the component from the first illumination light L1, but also the components from the third illumination light L3 and the fourth illumination light L4. That is, for example, IA = a + (c + d) / 2 1 / 2 That is the case. For example, IB includes not only the component from the second illumination light L2, but also the components from the third illumination light L3 and the fourth illumination light L4. That is, for example, IB = b + (c + d) / 2 1 / 2That is the case. For example, IC includes not only the component from the third illumination light L3, but also the components from the first illumination light L1 and the second illumination light L2. That is, for example, IC = c + (a + b) / 2 1 / 2 That is the case. For example, ID includes not only the component from the fourth illumination light L4, but also the components from the first illumination light L1 and the second illumination light L2. That is, for example, ID = d + (a + b) / 2 1 / 2 That is the case.

[0061] Therefore, the first polarization image is a = (IC + ID) / 2 1 / 2 -Calculated by IB. The second polarization image is b=(IC+ID) / 2 1 / 2 -Calculated by IA. The third polarization image is c = (IA + IB) / 2 1 / 2 -Calculated by ID. The fourth polarization image is d=(IA+IB) / 2 1 / 2 -Calculated by IC. In this way, multiple polarization images can be calculated from multiple captured images. Polarized illumination light from four directions can be acquired simultaneously through image processing.

[0062] In step S202 shown in Figure 8, the processing unit 30 detects multiple regions R in multiple captured images (or polarized images). Figures 9(a) to 9(c) are schematic diagrams illustrating the process of detecting multiple region images in an inspection apparatus according to the first embodiment. For example, the processing unit 30 calculates an average image by averaging all or part of the acquired image capture images, as shown in Figure 9(a).

[0063] Subsequently, the processing unit 30 calculates an image from which edges have been extracted in the average image of Figure 9(a), as shown in Figure 9(b).

[0064] Subsequently, the processing unit 30 detects multiple regions R based on the image in Figure 9(b), as shown in Figure 9(c). In this process, for example, the processing unit 30 compares the image in Figure 9(b) with the pre-stored layouts for each type of object to be inspected. This allows the processing unit 30 to identify the type of object to be inspected in the image in Figure 9(b). For example, the types of objects to be inspected may include types with 9 regions R, as in the example in Figure 9(c), or types with 16, 20, or 25 regions R. For example, the type of object to be inspected can be determined based on the number of regions (cells), etc. The pre-stored layouts of objects to be inspected include the coordinates of multiple regions R. Based on the coordinates of the regions R in the layout of the identified type, the processing unit 30 can perform inspection of each region R.

[0065] However, the method for detecting multiple regions R in step S02 is not limited to the above. For example, instead of the image from which edges were extracted as exemplified in Figure 9(b) in the above example, a binarized image of the average image may be used. Region R may also be detected by extracting regions having a specific area and aspect ratio from the image from which edges were extracted from the average image, or from the binarized image of the average image. Inspection can be performed based on the coordinates of the detected region R. Alternatively, region R may be detected by pattern matching between the average image and the reference image, as illustrated in Figure 9(a). The reference image is a standard image of region R acquired in advance, such as an image of a good product or an image obtained by averaging multiple images of good products. Among the average images illustrated in Figure 9(a), regions with a high similarity to the reference image can be detected as region R.

[0066] In step S203 shown in Figure 8, the processing unit 30 performs brightness correction on multiple polarized images. Figures 10(a) to 10(c) are schematic diagrams illustrating the process of correcting brightness in the inspection apparatus according to the first embodiment. Figure 10(a) shows, for example, a first polarization image. In the image in Figure 10(a), the direction from the top to the bottom of the paper is defined as the y-direction. The example in Figure 10(a) is an image when the first illumination light L1 is irradiated in the -y direction (i.e., from the bottom to the top of the paper).

[0067] Figure 10(b) is a graph illustrating the luminance distribution along the y-direction in the image shown in Figure 10(a). In a polarized image (or captured image), areas relatively close to the light source may have high luminance, while areas relatively far from the light source may have low luminance. That is, as shown in Figures 10(a) and 10(b), the luminance in the +y-side area is higher than the luminance in the -y-side area. In step S203, this luminance distribution is corrected.

[0068] The processing unit 30 divides the polarized image into multiple regions and corrects the brightness of the multiple regions so that the contrast (the difference between the maximum and minimum brightness values ​​within each region) is equal to that of the other regions. In this example, the processing unit 30 divides the first polarized image into first to third regions Z1 to Z3. The processing unit 30 corrects the brightness of the first region Z1, the second region Z2, and the third region Z3 so that they are equal to each other.

[0069] Specifically, the first to third regions Z1 to Z3 are regions delimited in the y-direction. Regions Z1, Z2, and Z3 are in that order, furthest from the first illumination. Regions Z1 to Z3 are regions into which the first polarization image is delimited for each row of multiple regions R arranged in an array. In this example, region Z1 includes the first region R1, the fourth region R4, and the seventh region R7; region Z2 includes the second region R2, the fifth region R5, and the eighth region R8; and region Z3 includes the third region R3, the sixth region R6, and the ninth region R9.

[0070] Figure 10(c) is a graph illustrating the distribution of brightness along the y-direction in a first polarized image corrected so that the contrast (brightness) of multiple regions is equal to each other. Note that "equal contrast" of multiple regions does not mean that the contrasts of multiple regions are exactly the same, but rather that the contrasts of multiple regions are of a similar degree. For example, the brightness in each region is multiplied by a constant so that the difference between the maximum and minimum brightness in the first region Z1, the difference between the maximum and minimum brightness in the second region Z2, and the difference between the maximum and minimum brightness in the third region Z3 are equal to each other. In this way, the parameters are changed for each region, and the brightness is corrected to have the same contrast.

[0071] In step S204 shown in Figure 8, the processing unit 30 calculates a region image A from the polarization image. Based on the first polarization image corrected in step S203, the processing unit 30 calculates multiple region images A using the coordinates of multiple regions R detected in step S202. For example, the processing unit 30 extracts the portion containing the region R to be inspected from the corrected polarization image. More specifically, the processing unit 30 calculates a region image A1 of the first region R1 from the corrected first polarization image.

[0072] The processing unit 30 performs the same processing for each of the multiple polarization images (first to fourth polarization images). Specifically, the processing unit 30 calculates region image A2 of the first region R1 from the second polarization image with corrected brightness, calculates region image A3 of the first region R1 from the third polarization image with corrected brightness, and calculates region image A4 of the first region R1 from the fourth polarization image with corrected brightness.

[0073] Note that the process shown in Figure 8 is just one example, and some parts may be omitted or the order may be changed as appropriate. For example, in this embodiment, polarization is not necessarily required. As will be described later, multiple images may be obtained by irradiating the surface Wf with unpolarized light and imaging the surface Wf with a normal camera. In this case, step S201 may be omitted, and the captured images may be used instead of the polarized images in subsequent calculations. Also, for example, brightness correction may be performed on the captured images rather than on the polarized images. Alternatively, instead of brightness correction, multiple images may be obtained by changing the brightness of the illumination. When brightness correction is performed, it is desirable for the imaging unit 10 to acquire imaging data by imaging in HDR (High Dynamic Range).

[0074] Furthermore, the description of the embodiment mainly uses the inspection of the first region R1 as an example. If necessary, inspections can be performed on each of multiple regions R in the same manner. For example, the processing unit 30 may use the region image of the second region R2 based on the captured image C1 and the region image of the second region R2 based on the captured image C2 to determine the defective region in the second region R2 by the same process as the inspection of the first region R1.

[0075] As described above, in this embodiment, the imaging unit acquires imaging data including information from a first imaging image (imaging image C1) of a plurality of regions R illuminated by at least the first illumination 21, and information from a second imaging image (imaging image C2) of a plurality of regions R illuminated by at least the second illumination 22. The processing unit 30 then calculates a first difference image (difference image D1) corresponding to the difference between the first region image and the second region image based on a first region image (region image A1) of the first region R1 based on the first imaging image, a second region image (region image A2) of the first region R1 based on the second imaging image, and an average image M obtained by averaging at least the first region image and the second region image, and determines the defective region in the first region R1 based on the first difference image. This improves the inspection throughput. There is a reference example method for imaging the surface of an object to be inspected in multiple regions. This reference example method is used, for example, when the workpiece is large. In the reference example method, the stage on which the object to be inspected is placed is moved, or the optical system is moved by a robot arm or the like, according to the region to be imaged. That is, in the reference example, after imaging the first region, the stage or optical system is moved to image the second region. If such movement takes time, it leads to a decrease in throughput. In contrast, in the embodiment, the first and second imaged images each include multiple regions, and the processing unit 30 determines the defective region based on the first region image of the first region R1 based on the first imaged image and the second region image of the first region R1 based on the second imaged image. In the embodiment, it is not necessary to move the stage or optical system during imaging. Therefore, it is possible to improve throughput compared to the reference example. Also, since it is not necessary to provide a mechanism to move the stage or optical system, it is possible to inspect large workpieces with a relatively simple and inexpensive device.

[0076] Furthermore, in this example, polarized illumination and a polarized camera are used as described above. Specifically, the first illumination 21 emits a first illumination light L1 polarized in the first polarization direction, and the second illumination 22 emits a second illumination light L2 polarized in the second polarization direction. The first captured image is an image obtained with light polarized in the first polarization direction, and the second captured image is an image obtained with light polarized in the second polarization direction. This makes it possible to simultaneously acquire an image illuminated with polarization from the first direction and an image illuminated with polarization from the second direction. In the example above, images illuminated from four directions can be acquired simultaneously at once. This reduces the number of imaging cycles, thereby improving the throughput of the examination.

[0077] As explained with respect to Figure 7, the unevenness of the defective area can be determined by illuminating the object to be inspected from multiple directions. However, in this case, as explained with respect to Figure 10(b), uneven illumination (unevenness in brightness in the image) may occur. In particular, when the workpiece is large, one end of the workpiece is close to the light source and the other end is farther away from the light source, so uneven illumination is likely to occur.

[0078] When illumination unevenness occurs, for example, the brightness of a certain area in an image illuminated from the front will differ significantly from the brightness of the same area in an image illuminated from the back. For example, in one of the images, one illuminated from the front and the other from the back, the defective area will be further from the light source, and the brightness around the defective area will be lower. As a result, the contrast between the defective area and its surroundings will also decrease, which may reduce the accuracy of detecting the defective area.

[0079] In contrast, in this embodiment, the processing unit 30 divides the first polarization image into multiple regions, corrects the brightness of the multiple regions so that their brightness levels are equal, and calculates a first region image based on the corrected first polarization image. The contrast in the region image can be adjusted by correcting the brightness (for example, in step S203). Therefore, a decrease in the detection accuracy of the defect region can be suppressed.

[0080] As described later, imaging conditions may be changed for each region R being examined to suppress the decrease in contrast. In this case, multiple imaging sessions will be required due to the change in imaging conditions. On the other hand, when correcting brightness, the contrast can be adjusted without changing the imaging conditions, thus reducing the number of imaging sessions.

[0081] Furthermore, in this example, the first and second captured images are created from imaging data obtained in a single imaging session under conditions where the first illumination light L1 is irradiated onto the surface Wf from the first illumination 21, and the second illumination light L2 is irradiated onto the surface Wf from the second illumination 22. This further reduces the number of imaging sessions, thereby improving the inspection throughput. Specifically, for example, by using polarized illumination and a polarized camera, and correcting the brightness of the polarized image (or captured image), it is possible to acquire images from multiple illumination directions simultaneously in a single imaging session while suppressing a decrease in inspection accuracy.

[0082] Figure 11 is a flowchart illustrating an inspection method in the inspection apparatus according to the first embodiment. Figure 11 shows an example of the overall inspection flow, including the processes S1 to S3 described in Figures 4 to 10, and the judgment process (processes S401 to S404). As shown in Figure 11, process S401 is performed after process S202, which detects multiple regions R. Process S401 determines whether the number N of regions R detected in process S202 is a specified number. The specified number is a predetermined number determined, for example, by the type of object being inspected. If the number N of detected regions R is not the specified number (process S401: No), the inspection ends as an error (process S402). In other words, if the number of detected regions R differs from the number of regions R in the pre-assumed type (e.g., less than the specified number), the inspection ends as an error. Reasons for errors include, for example, inspecting an unsuitable type of object, the workpiece being misaligned and stopping outside the inspection range, or inability to properly image or detect regions due to lighting or camera malfunctions. Process S401 detects such errors that require human verification. If the number N of detected regions R is the specified number, the next process (for example, process S203) is executed.

[0083] Process S403 is performed after process S3, which determines the defective area. As described above, in process S3, a defective area is detected within a single region R. Then, process S403 determines whether that region R is pass or fail. In process S403, it is determined whether each defective area detected in process S3 falls under a predetermined category (condition). Each category is determined by at least one of the following: the size of the defective area (diameter, etc.), the area, and the type of defect (concave or convex, etc.). Process S403 measures the number of defective areas that fall under each category, and if that number exceeds the standard value, that region R is determined to be unacceptable. The standard value is a predetermined value set for each category and is arbitrary.

[0084] For example, if there is one or more defective regions that meet the criteria of "the defect type is convex and the size is L or larger," then that region R is judged to be unacceptable. For example, if there are two or more defective regions that fall under the category of "the defect type is convex and the size is less than L," then that region R will be judged as unacceptable. For example, if there is one or more defective regions that fall under the category of "the defect type is concave and the area is S or greater," then that region R is judged to be unacceptable. For example, if there are five or more defective regions that fall under the category of "the defect type is concave and the area is less than S," then that region R will be judged as unacceptable.

[0085] As shown in Figure 11, steps S204, S3, and S403 are performed for each of the N detected regions R. Subsequently, step S404 (overall pass / fail judgment) is performed. Step S404 makes a pass / fail judgment for the inspected object based on the pass / fail judgment results for each region R determined in step S403. For example, if the number of regions R that failed in step S403 out of the N regions R is greater than or equal to a predetermined value, the inspected object is judged to be a failure (NG). If the number of regions R that failed in step S403 is less than the predetermined value, the inspected object is judged to be a pass (OK).

[0086] After the above inspections are completed, the system proceeds to the next process if the inspection results are OK, NG, or error, or if the inspection results are incorrect. If an error occurs, the system controls the equipment, such as prompting the operator to confirm the result. The inspection results, the judgment results of process 403, and defect information (defect type, size, area, coordinates) are displayed on a monitor or recorded in a database. This allows for timely reference of the inspection results, etc.

[0087] (Second embodiment) Figure 12 is a schematic diagram illustrating an inspection apparatus according to the second embodiment. In this embodiment, imaging conditions (such as exposure time and illumination intensity) are changed for each region R to be inspected, and images are acquired and inspection processing is performed. Polarization does not need to be used in this embodiment. Otherwise, the configuration of the inspection apparatus 101 according to the second embodiment is the same as that of the inspection apparatus 100 described above. That is, the inspection apparatus 101 is also provided with an imaging unit 10, an illumination unit 20 (multiple illuminations 2n), a processing unit 30, and an output unit 40.

[0088] In this embodiment, polarizing films do not need to be provided on the multiple lights 2n (for example, the first to fourth lights 21 to 24). That is, the multiple lights 2n may irradiate the surface Wf of the object W to be inspected with unpolarized light. The imaging unit 10 may be a normal camera instead of a polarizing camera. That is, for example, a polarizing filter may not be provided on the pixel array of the imaging unit 10.

[0089] In this example as well, the imaging unit 10 acquires imaging data that includes information from multiple imaging images C (step S1). For example, the imaging data acquired by the imaging unit 10 includes information from imaging images C11 to C13 of the surface Wf illuminated by the first illumination 21, information from imaging images C21 to C23 of the surface Wf illuminated by the second illumination 22, information from imaging images C31 to C33 of the surface Wf illuminated by the third illumination 23, and information from imaging images C41 to C43 of the surface Wf illuminated by the fourth illumination 24. In this example, the imaging data is obtained through multiple imagings.

[0090] Figure 13 is a schematic diagram illustrating an image captured in an inspection apparatus according to the second embodiment.

[0091] The imaging unit 10 images the surface Wf of the object to be inspected W under three imaging conditions. Specifically, three imaging conditions, the first to third, are used, each with different exposure times. The exposure time for the second imaging condition is longer than that for the first imaging condition. The exposure time for the third imaging condition is longer than that for the first imaging condition and shorter than that for the second imaging condition. For example, the exposure time for the first imaging condition is 78 milliseconds, the exposure time for the second imaging condition is 252 milliseconds, and the exposure time for the third imaging condition is 600 milliseconds.

[0092] The exposure time is the time the surface Wf is exposed (illuminated) during imaging, and corresponds to the illumination time during imaging. The exposure time may be the camera's exposure time, not the illumination time. Also, the imaging conditions do not have to be three conditions; for example, there may be two conditions or four or more conditions. The imaging conditions are not limited to the exposure time; for example, the illumination intensity may be changed.

[0093] The imaging unit 10 performs imaging under each of the three imaging conditions while each of the multiple lights is turned on. In a single imaging cycle, with only one light on and the other lights off, one image is acquired under one of the imaging conditions.

[0094] For example, when the first illumination 21 is turned on and the other illuminations are turned off, the imaging unit 10 images the surface Wf illuminated by the first illumination 21 under the first imaging conditions and acquires an image C11 (first image). The image C11 is an image of the surface Wf illuminated by the first illumination 21 and captured under the first imaging conditions.

[0095] For example, when the second illumination 22 is turned on and the other illuminations are turned off, the imaging unit 10 captures the surface Wf illuminated by the second illumination 22 under the second imaging conditions and acquires the captured image C22 (second captured image). The captured image C22 is an image of the surface Wf illuminated by the second illumination 22 captured under the second imaging conditions.

[0096] Similarly, the imaging unit 10 captures multiple images C. Each of the multiple images C is, for example, an image of the surface Wf illuminated by each of the multiple illuminations 2n under different illumination conditions. In this example, imaging is performed for each of the four illuminations under each of the three imaging conditions, so a total of 12 imaging takes are performed, and 12 images are obtained. That is, images C11-C13, C21-C23, C31-C33, and C41-C44 are captured.

[0097] Each of the captured images C11 to C13 is an image of the surface Wf illuminated by the first illumination 21, captured under the first to third imaging conditions, respectively. Each of the captured images C21 to C23 is an image of the surface Wf illuminated by the second illumination 22, captured under the first to third imaging conditions, respectively. Each of the captured images C31 to C33 is an image of the surface Wf illuminated by the third illumination 23, captured under the first to third imaging conditions, respectively. Each of the captured images C41 to C43 is an image of the surface Wf illuminated by the fourth illumination 24, taken under the first to third imaging conditions, respectively.

[0098] Subsequently, the processing unit 30 calculates multiple region images A based on at least a portion of the multiple captured images C (step S2). In this embodiment, the processing unit 30 inspects by combining different images depending on the region to be inspected.

[0099] Specifically, if the first region R1 is the target of examination, the region image A11 (first region image) of the first region R1 based on the captured image C11, the region image A12 (second region image) of the first region R1 based on the captured image C22, the region image A13 of the first region R1 based on the captured image C31, and the region image A14 of the first region R1 based on the captured image C42 are calculated.

[0100] In the same manner as in step S202 described above with respect to Figure 8, the processing unit 30 detects multiple regions R in each captured image. Then, in the same manner as in step S204 described above with respect to Figure 8, the processing unit 30 calculates multiple region images A11, A12, A13, and A14 based on the captured images C11, C22, C31, and C42, respectively. Figure 13 shows the regions corresponding to the multiple region images A11, A12, A13, and A14. In this example, the steps S201 for calculating multiple polarization images and S203 for brightness correction are omitted.

[0101] The processing unit 30 determines the defect area in the first region R1 based on the calculated region images A11, A12, A13, and A14 (step S3). The method for determining the defect area is the same as the method described above with respect to Figure 5, etc.

[0102] Specifically, the processing unit 30 calculates an average image M1 (corresponding to image M in Figure 5) by averaging region images A11, A12, A13, and A14. Based on region images A11, A12, and the average image M1, the processing unit 30 calculates a difference image D11 (corresponding to image D1 in Figure 5) corresponding to the difference between region image A11 and region image A12. Based on region images A13, A14, and the average image M1, the processing unit 30 calculates a difference image D12 (corresponding to image D2 in Figure 5) corresponding to the difference between region image A13 and region image A14. Based on the difference images D11 and D12, the processing unit 30 determines the defective region in the first region R1.

[0103] The processing unit 30 can also inspect regions R other than the first region R1 in the same manner. However, the processing unit 30 appropriately changes the combined imaging image (imaging conditions) depending on the region R to be inspected.

[0104] Figure 14 is a table illustrating combinations of captured images in the inspection apparatus according to the second embodiment. In Figure 14, the "cell number" represents the region R to be inspected. Cell numbers 1 to 9 correspond to regions R1 to R9, respectively.

[0105] In the case of cell number 1, that is, when the first region R1 is the subject of examination, as already explained, the following images are combined: image C11 taken with "back" illumination (first illumination 21) under the "short" condition (first imaging condition); image C22 taken with "front" illumination (second illumination 22) under the "long" condition (second imaging condition); image C31 taken with "right" illumination (third illumination 23) under the "short" condition; and image C42 taken with "left" illumination (fourth illumination 24) under the "long" condition.

[0106] In the case of cell number 2, that is, when the second region R2 is the subject of examination, the following images are combined: image C13 (third image) taken with "back" lighting under "medium" conditions (third imaging conditions); image C23 (fourth image) taken with "front" lighting under "medium" conditions; image C31 taken with "right" lighting under "short" conditions; and image C42 taken with "left" lighting under "long" conditions. The processing unit 30 calculates the region image A21 (third region image) of the second region R2 based on the captured image C13, the region image A22 (second region image) of the second region R2 based on the captured image C23, the region image A23 of the second region R2 based on the captured image C31, and the region image A24 of the second region R2 based on the captured image C42. The processing unit 30 calculates an average image M2 (corresponding to image M in Figure 5) by averaging region images A21, A22, A23, and A24. Based on region images A21, A22, and the average image M2, the processing unit 30 calculates a difference image D21 (corresponding to image D1 in Figure 5) corresponding to the difference between region image A21 and region image A22. Based on region images A23, A24, and the average image M2, the processing unit 30 calculates a difference image D22 (corresponding to image D2 in Figure 5) corresponding to the difference between region image A23 and region image A24. Based on the difference images D21 and D22, the processing unit 30 determines the defective region in the second region R2.

[0107] The processing unit 30 similarly determines the defective area for cell numbers 3 to 9 based on the combination of captured images shown in Figure 14. For example, the shorter the distance between the illumination 2n and the area R to be inspected, the shorter the exposure time (or lower the illumination intensity). For example, the longer the distance between the illumination 2n and the area R to be inspected, the longer the exposure time (or higher the illumination intensity).

[0108] In the above explanation, in step S1, where imaging data is acquired, images were captured for all combinations of lighting and imaging conditions (12 combinations). Then, depending on the area R to be inspected, a portion of the acquired images was selected and used to determine the defective area. However, this is not limited to this, and in this embodiment, imaging may be performed using only a portion of the lighting and imaging condition combinations as needed.

[0109] As described above, in this embodiment as well, the imaging unit 10 acquires imaging data including information of a first imaging image (imaging image C11) of a plurality of regions R illuminated by at least the first illumination 21, and information of a second imaging image (imaging image C22) of a plurality of regions R illuminated by at least the second illumination 22. The processing unit 30 then calculates a first difference image (difference image D11) corresponding to the difference between the first region image and the second region image based on the first region image (region image A11) of the first region R1 based on the first imaging image, the second region image (region image A12) of the first region R1 based on the second imaging image, and an average image M1 obtained by averaging at least the first region image and the second region image, and determines the defective region in the first region R1 based on the first difference image. This makes it possible to improve the inspection throughput in the same way as in the first embodiment.

[0110] Furthermore, the first image is an image of the surface Wf illuminated by the first illumination 21, captured under the first imaging conditions. The second image is an image of the surface Wf illuminated by the second illumination 22, captured under second imaging conditions different from the first imaging conditions. In this way, by changing the imaging conditions with illumination, the contrast (illumination) in the image of the area R to be inspected can be adjusted. For example, uneven illumination can be suppressed, and a decrease in inspection accuracy can be suppressed.

[0111] Furthermore, in the above example, the brightness correction step S203 described with respect to Figure 8 can be appropriately omitted by appropriately changing the imaging conditions. However, brightness correction may also be performed as needed in this embodiment. That is, for example, the processing unit 30 may divide the first captured image into a plurality of areas, correct the brightness of the plurality of areas so that the brightness of the plurality of areas is equal to that of the plurality of areas, and calculate the first region image based on the corrected first captured image.

[0112] Furthermore, the imaging unit 10 acquires imaging data including a third imaging image (imaging image C13) obtained by imaging the surface Wf irradiated by the first illumination 21 under imaging conditions different from the first imaging conditions, and a fourth imaging image (C23) obtained by imaging the surface Wf irradiated by the second illumination 22 under imaging conditions different from the second imaging conditions.The processing unit 30 then calculates a second difference image (difference image D21) corresponding to the difference between the third and fourth region images based on the third region image (region image A21) of the second region R2 based on the third imaging image, the fourth region image (region image A22) of the second region R2 based on the fourth imaging image, and an average image M2 obtained by averaging at least the first and second region images, and determines the defective region in the second region R2 based on the second difference image.In this way, by changing the imaging conditions for each region to be inspected, the contrast (illumination) in the image of the region R to be inspected can be adjusted.For example, illumination unevenness can be suppressed, and a decrease in inspection accuracy can be suppressed.

[0113] The instructions shown in the processing procedure described in the above-described embodiment can be executed based on a software program. A general-purpose computer system can also obtain the same effects as the device of the above-described embodiment by pre-storing this program and reading it. The instructions described in the above-described embodiment are recorded as a program that can be executed by a computer on a magnetic disk (flexible disk, hard disk, etc.), optical disk (CD-ROM, CD-R, CD-RW, DVD-ROM, DVD±R, DVD±RW, etc.), semiconductor memory, or similar recording medium. Any storage format is acceptable as long as it is a recording medium that can be read by a computer or embedded system. The computer can read the program from this recording medium and have the CPU execute the instructions written in the program based on this program, thereby achieving the same operation as the information processing device of the above-described embodiment. Of course, the computer may acquire or read the program via a network.

[0114] Furthermore, an operating system (OS) running on a computer, database management software, middleware (MW) operating on a network, etc., may execute some of the processes necessary to realize the embodiment based on instructions from a program installed on a computer or embedded system from a recording medium.

[0115] Furthermore, the recording medium in the embodiment is not limited to a recording medium independent of the computer or embedded system, but also includes a recording medium on which a program transmitted via a LAN, the Internet, etc., has been downloaded and stored or temporarily stored. Also, the recording medium is not limited to one; cases where the processing in the embodiment is executed from multiple recording media are also included in the recording medium in the embodiment. The configuration of the recording medium may be any configuration.

[0116] The computer or embedded system in the embodiment is used to execute each process in the embodiment based on a program stored on a recording medium, and may be configured as a single device such as a personal computer or microcomputer, or as a system in which multiple devices are connected via a network.

[0117] Furthermore, the term "computer" in these embodiments is not limited to personal computers, but also includes arithmetic processing units, microcomputers, and the like that found in information processing equipment. It is a general term for any equipment or device that can implement the functions described in these embodiments through a program.

[0118] According to the embodiment, an inspection device, inspection method, and program capable of improving throughput can be provided.

[0119] Although several embodiments of the present invention have been illustrated above, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims of the invention and its equivalents. Furthermore, the embodiments described above can be implemented in combination with each other. [Explanation of symbols]

[0120] 10 Imaging unit, 20 Illumination unit, 21-24 1st-4th illumination, 21f-24f Polarizing film, 30 Processing unit, 31 Memory unit, 32 Calculation unit, 40 Output unit, 100 Inspection device, 100p program, 101 Inspection device, A, A1-A4, A11-A14, A21-A24 Region images, C, C1-C4, C11-C13, C21-C23, C31-C33, C41-C43 Acquired images, D1, D2 Difference images, L Illumination light, L1-L4 1st-4th illumination light, M Average image, R Region, R1-R9 1st-9th regions, S1-S3, S201-S204, S301-S311, S401-S404 Process, U Defect area, W: object to be inspected, Wf: surface, Z1~Z3: 1st~3rd area

Claims

1. An imaging unit including a polarizing camera positioned above the object to be inspected, The system includes a processing unit that controls the imaging unit and processes an image of the surface of the object to be inspected, which is illuminated by the illumination unit and captured by the imaging unit, The surface of the object to be inspected includes a repeating pattern in which a plurality of regions, including a first region, are arranged in an array on a plane, The illumination unit includes a first illumination that illuminates the surface from a first direction, and a second illumination that illuminates the surface from a second direction opposite to the first direction when viewed from above. The imaging unit acquires imaging data including information of a first image of the plurality of regions illuminated by the first illumination and the second illumination, and information of a second image of the plurality of regions illuminated by the first illumination and the second illumination. The aforementioned processing unit, Based on the first region image of the first region based on the first captured image and the second region image of the first region based on the second captured image, a first difference image corresponding to the difference between the first region image and the second region image is calculated. Based on the first difference image, the defective region in the first region is determined. The first illumination emits first illumination light polarized in a first polarization direction, The second illumination emits a second illumination light polarized in a second polarization direction, The first captured image is an image of light polarized in the first polarization direction, The second captured image is an image of light polarized in the second polarization direction, An inspection apparatus in which the first image and the second image are created from imaging data obtained by a single imaging operation under the condition that the first illumination light is irradiated onto the surface from the first illumination and the second illumination light is irradiated onto the surface from the second illumination.

2. The plurality of regions include a second region, The inspection apparatus according to claim 1, wherein the processing unit determines a defective area in the second region based on a difference image corresponding to the difference between a region image of the second region based on the first captured image and a region image of the second region based on the second captured image.

3. An imaging unit including a camera positioned above the object to be inspected, The system includes a processing unit that controls the imaging unit and processes an image of the surface of the object to be inspected, which is illuminated by the illumination unit and captured by the imaging unit, The surface of the object to be inspected includes a plurality of regions, including the first region. The illumination unit includes a first illumination that illuminates the surface from a first direction, and a second illumination that illuminates the surface from a second direction opposite to the first direction when viewed from above. The imaging unit acquires imaging data including information of a first image of the plurality of regions illuminated by the first illumination and information of a second image of the plurality of regions illuminated by the second illumination. The aforementioned processing unit, Based on the first region image of the first region based on the first captured image and the second region image of the first region based on the second captured image, a first difference image corresponding to the difference between the first region image and the second region image is calculated. Based on the first difference image, the defective region in the first region is determined. The first image is an image of the surface irradiated by the first illumination, captured under the first imaging conditions. The inspection apparatus wherein the second image is an image of the surface irradiated by the second illumination, captured under second imaging conditions different from the first imaging conditions.

4. The aforementioned multiple regions include a second region, The imaging unit acquires a third image obtained by imaging the surface irradiated by the first illumination under imaging conditions different from the first imaging conditions, and a fourth image obtained by imaging the surface irradiated by the second illumination under imaging conditions different from the second imaging conditions. The aforementioned processing unit, Based on the third region image of the second region based on the third captured image and the fourth region image of the second region based on the fourth captured image, a second difference image corresponding to the difference between the third region image and the fourth region image is calculated. The inspection apparatus according to claim 3, which determines a defect region in the second region based on the second difference image.

5. The processing unit calculates a first polarization image, which is an image of the first illumination light, based on the first captured image. The inspection apparatus according to claim 1, wherein the processing unit divides one of the first captured image and the first polarized image into a plurality of regions, corrects the brightness of the plurality of regions so that the contrast of the plurality of regions is equal to that of the plurality of regions, and calculates a first region image based on the corrected one of the first captured image and the first polarized image.

6. A method for inspecting the surface of an object to be inspected, which includes a repeating pattern in which multiple regions, including a first region, are arranged in an array on a plane, A first image of the plurality of regions illuminated by a first illumination that shines on the surface from a first direction and a second illumination that shines on the surface from a second direction opposite to the first direction when viewed from above, and a second image of the plurality of regions illuminated by the first illumination and the second illumination are acquired. Based on the first region image of the first region based on the first captured image and the second region image of the first region based on the second captured image, a first difference image corresponding to the difference between the first region image and the second region image is calculated. Based on the first difference image, the defective region in the first region is determined. The first illumination emits first illumination light polarized in a first polarization direction, The second illumination emits a second illumination light polarized in a second polarization direction, The first captured image is an image of light polarized in the first polarization direction, The second captured image is an image of light polarized in the second polarization direction, An inspection method in which the first image and the second image are obtained by taking images in a single imaging operation under the condition that the first illumination light is irradiated onto the surface from the first illumination and the second illumination light is irradiated onto the surface from the second illumination.

7. A method for inspecting the surface of an object to be inspected, which includes a plurality of regions including a first region, A first image of the plurality of regions illuminated by a first illumination from a first direction, and a second image of the plurality of regions illuminated by a second illumination from a second direction opposite to the first direction when viewed from above, are acquired. Based on the first region image of the first region based on the first captured image and the second region image of the first region based on the second captured image, a first difference image corresponding to the difference between the first region image and the second region image is calculated. Based on the first difference image, the defective region in the first region is determined. The first image is an image of the surface irradiated by the first illumination, captured under the first imaging conditions. An inspection method wherein the second image is an image of the surface irradiated by the second illumination, captured under second imaging conditions different from the first imaging conditions.

8. The aforementioned multiple regions include a second region, A third image is obtained by capturing the surface irradiated by the first illumination under imaging conditions different from the first imaging conditions, and a fourth image is obtained by capturing the surface irradiated by the second illumination under imaging conditions different from the second imaging conditions. Based on the third region image of the second region based on the third captured image and the fourth region image of the second region based on the fourth captured image, a second difference image corresponding to the difference between the third region image and the fourth region image is calculated. The inspection method according to claim 7, wherein a defect region in the second region is determined based on the second difference image.

9. Based on the first captured image, a first polarized image which is an image of the first illumination light is calculated, The inspection method according to claim 6, comprising dividing one of the first captured image and the first polarized image into a plurality of regions, correcting the brightness of the plurality of regions so that the contrast of the plurality of regions is equal to that of the plurality of regions, and calculating a first region image based on the corrected one of the first captured image and the first polarized image.

10. A program for inspecting the surface of an object to be inspected, which includes a repeating pattern in which multiple regions, including a first region, are arranged in an array on a plane, On the computer, Based on a first region image of the first region, which is based on first captured images of the plurality of regions illuminated by a first illumination that illuminates the surface from a first direction and a second illumination that illuminates the surface from a second direction opposite to the first direction when viewed from above, and a second region image of the first region, which is based on second captured images of the plurality of regions illuminated by the first illumination and the second illumination, a first difference image corresponding to the difference between the first region image and the second region image is calculated. This is a program that determines the defect region in the first region based on the first difference image. The first illumination emits first illumination light polarized in a first polarization direction, The second illumination emits a second illumination light polarized in a second polarization direction, The first captured image is an image of light polarized in the first polarization direction, The second captured image is an image of light polarized in the second polarization direction, The first captured image and the second captured image are obtained by a single imaging operation in which the first illumination light is shone onto the surface from the first illumination and the second illumination light is shone onto the surface from the second illumination.

11. A program for inspecting the surface of an object to be inspected, which includes a plurality of regions including a first region, On the computer, Based on a first region image of the first region, which is based on first captured images of the plurality of regions illuminated by a first illumination from a first direction, and a second region image of the first region, which is based on second captured images of the plurality of regions illuminated by a second illumination from a second direction opposite to the first direction when viewed from above, a first difference image corresponding to the difference between the first region image and the second region image is calculated. This is a program that determines the defect region in the first region based on the first difference image. The first image is an image of the surface irradiated by the first illumination, captured under the first imaging conditions. The program provides a second image, which is an image of the surface illuminated by the second illumination, captured under second imaging conditions different from the first imaging conditions.

Images