JP2023015418A - Inspection method, program and inspection system - Google Patents

Abstract

An object of the present invention is to provide an inspection method, a program, and an inspection system capable of improving the accuracy of inspection of the surface state of an object. An inspection method includes an acquisition step S11 and an inspection step S12. Acquisition step S11 is a step of acquiring a target image of the surface of the target object. The inspection step S12 inspects the state of the surface of the object based on the object information including the shape information representing the shape of the frequency distribution of the parameters based on the pixel values obtained from the object image. [Selection diagram] Fig. 1

Description

The present disclosure relates to inspection methods, programs, and inspection systems. In particular, the present disclosure relates to an inspection method, program, and inspection system for inspecting the state of the surface of an object using images.

Patent Literature 1 discloses a coloring inspection device. The color inspection apparatus disclosed in Patent Document 1 includes a camera having three spectral sensitivities linearly transformed equivalently to the CIEXYZ color matching function, and an image having the three spectral sensitivities acquired by the camera into three colors in the CIEXYZ color system. It includes an arithmetic processing unit that acquires and calculates coloring data converted into stimulus values, and an illumination unit that illuminates an automobile, which is an example of the object to be measured. The color inspection apparatus inspects colors by calculating a color distribution coincidence index that indicates the overlapping ratio of the two xyz chromaticity histogram distributions of the inspection object and the reference object.

JP 2015-155892 A

The appearance of the color of an inspection object (measuring object, object) can be affected by the shape of the surface of the inspection object. Therefore, even if the color distribution coincidence index obtained in Patent Document 1, which indicates the overlapping ratio of the two xyz chromaticity histogram distributions of the inspection object and the reference object, is high, the visual sensation perceived by people may differ.

An object of the present invention is to provide an inspection method, a program, and an inspection system that can improve the accuracy of inspection of the surface state of an object.

An inspection method of one aspect of the present disclosure includes an acquisition step and an inspection step. The acquisition step is a step of acquiring a target image of the surface of the target. The inspection step is a step of inspecting the state of the surface of the object based on object information including shape information representing the shape of the frequency distribution of parameters based on pixel values obtained from the object image.

A program according to one aspect of the present disclosure is a program for causing one or more processors to execute the inspection method.

An inspection system according to one aspect of the present disclosure includes an acquisition unit that acquires a target image of a surface of an object, and target information including shape information representing the shape of a frequency distribution of parameters based on pixel values obtained from the target image. and an inspection unit that inspects the state of the surface of the object based on the above.

According to the aspect of the present disclosure, it is possible to improve the accuracy of inspection of the state of the surface of the object.

FIG. 1 is a flow chart of an inspection method according to one embodiment. FIG. 2 is a block diagram of an inspection system that implements the above inspection method. FIG. 3 is an explanatory diagram of a method for generating shape information used in the inspection method. FIG. 4 is an explanatory diagram of an example of shape information obtained by the inspection system. FIG. 5 is an explanatory diagram of an example of shape information obtained by the inspection system. FIG. 6 is an explanatory diagram of an example of shape information obtained by the inspection system. FIG. 7 is an explanatory diagram of an example of target information and reference information obtained by the inspection system. FIG. 8 is an explanatory diagram of an example of target information and reference information obtained by the inspection system. FIG. 9 is an explanatory diagram of an example of target information and reference information obtained by the inspection system. FIG. 10 is a graph showing the rate of matching between a plurality of samples prepared as objects and a reference object. FIG. 11 is a graph showing the results of visual evaluation of a plurality of samples prepared as objects. FIG. 12 is a graph showing the relationship between the concordance rate and the visual evaluation results for a plurality of samples prepared as objects. FIG. 13 is an explanatory diagram of a method of generating shape information in a modified example. FIG. 14 is an explanatory diagram of a method of generating shape information in a modified example. FIG. 15 is an explanatory diagram of a method of acquiring a target image in an inspection method of a modified example. FIG. 16 is an explanatory diagram of an example of a target image. FIG. 17 is an explanatory diagram of an example of a target image. FIG. 18 is an explanatory diagram of an example of a histogram obtained from the target image. FIG. 19 is an explanatory diagram of an example of a histogram obtained from the target image. FIG. 20 is an explanatory diagram of an example of an image of the surface of the object. FIG. 21 is an explanatory diagram of a method of acquiring a target image in an inspection method of a modified example. FIG. 22 is an explanatory diagram of a method of generating shape information according to a modified example. FIG. 23 is an explanatory diagram of a method of acquiring a target image in an inspection method of a modified example. FIG. 24 is a diagram showing an example of inspection results presented by the inspection system. FIG. 25 is a diagram showing an example of inspection results presented by the inspection system. FIG. 26 is a diagram showing an example of inspection results presented by the inspection system. FIG. 27 is a diagram showing an example of inspection results presented by the inspection system. FIG. 28 is a diagram showing an example of inspection results presented by the inspection system.

(1) Embodiment (1.1) Overview An inspection method according to one embodiment can be used for inspecting the state of the surface of an object 100 as shown in FIG. The state of the surface of the object 100 may include the color, shape, texture, glossiness, gradation, and paint state of the surface of the object 100 . In this embodiment, gradation can include flip-flop, color travel effect, glitter, and graininess, as well as gradation that occurs depending on the angle of incidence and reflection when the surface is curved. In this embodiment, the target object 100 is assumed to be an automobile. In particular, the surface of the object 100 is part of the outer surface of the vehicle body. Note that the object 100 is not limited to an automobile. For example, the object 100 may be a moving body other than an automobile, or may not be a moving body. Examples of mobile objects include motorcycles, trains, drones, aircraft, construction machinery, and ships. Also, the target object 100 may be electrical equipment, tableware, containers, furniture, clothes, building materials, and the like. In short, the object 100 may be any object having a surface.

FIG. 1 shows a flowchart of the inspection method of this embodiment. The inspection method of this embodiment includes an acquisition step S11 and an inspection step S12. Acquisition step S<b>11 is a step of acquiring a target image of the surface of the target object 100 . The inspection step S12 inspects the state of the surface of the object 100 based on the object information including the shape information representing the shape of the frequency distribution of the parameters based on the pixel values obtained from the object image.

In the inspection method of the present embodiment, the surface state of the object 100 is inspected using shape information representing the shape of the frequency distribution of parameters based on pixel values, which is obtained from the target image of the surface of the object 100. do. In other words, the inspection method of this embodiment utilizes the shape of the frequency distribution itself. As a result, according to the inspection method of this embodiment, it is possible to improve the accuracy of inspection of the state of the surface of the object 100 .

(1.2) Details The inspection method of the present embodiment will be described in more detail below with reference to FIGS. 1 to 12. FIG. The inspection method of this embodiment can be executed by the inspection system 1 shown in FIG. Inspection system 1 is a system for inspection of the surface of object 100 . The inspection system 1 has a function as a coloring inspection device. The inspection system 1 can also paint the object 100 . The inspection system 1 can paint the object 100 according to the result of the inspection, thereby obtaining the object 100 with the desired painting.

The inspection system 1 includes a determination system 10, an illumination system 20, an imaging system 30, and a coating system 40, as shown in FIG.

The illumination system 20 is a system for illuminating the surface of the object 100 with light. Illumination system 20 includes one or more lamps that illuminate object 100 . A lamp is an LED lamp as an example. Also, the lamp emits white light. In addition, in the lighting system 20, the number of lamps is not particularly limited, and the types of lamps may be light sources other than LEDs. Also, the color of light emitted from the lamp is not limited to white. The color of light emitted by the lamp can be appropriately set in consideration of the color of the object 100 and the color detectable by the imaging system 30 . Also, the wavelength of light emitted by illumination system 20 may be variable.

The imaging system 30 is a system for generating an image (digital image) of the surface of the object 100 . In this embodiment, the imaging system 30 images the surface of the object 100 illuminated by the illumination system 20 to generate an image of the surface of the object 100 . Imaging system 30 includes one or more cameras. A camera comprises one or more image sensors. Note that the camera may include one or more line sensors.

The painting system 40 is a system for painting the surface of the object 100 . The painting system 40 includes one or more painting units (painting robots). Since the coating robot may have a conventionally well-known configuration, detailed description thereof will be omitted.

The determination system 10 includes an input/output unit 11, a storage unit 12, and a processing unit 13, as shown in FIG. The determination system 10 can be implemented by a computer system. A computer system may include one or more processors, one or more connectors, one or more communication devices, one or more memories, and the like.

The input/output unit 11 inputs and outputs information to and from the lighting system 20 , imaging system 30 and painting system 40 . In this embodiment, the input/output unit 11 is communicably connected to the lighting system 20, the imaging system 30, and the painting system 40. FIG. The input/output unit 11 includes one or more input/output devices and uses one or more input/output interfaces.

The storage unit 12 is used to store information used by the processing unit 13 . The storage unit 12 includes one or more storage devices. The storage device is, for example, RAM (Random Access Memory) or EEPROM (Electrically Erasable Programmable Read Only Memory).

The processing unit 13 can be implemented by, for example, one or more processors (microprocessors). That is, one or more processors function as the processing unit 13 by executing one or more programs (computer programs) stored in one or more memories. One or more programs may be pre-recorded in one or more memories, or may be provided by being recorded in a non-temporary recording medium such as a memory card or through a telecommunication line such as the Internet.

The processing unit 13 has an acquisition unit F11, an inspection unit F12, a presentation unit F13, and a control unit 14, as shown in FIG. The acquisition unit F11, the inspection unit F12, the presentation unit F13, and the control unit F14 do not represent actual configurations, but represent functions realized by the processing unit 13. FIG.

Acquisition unit F11 executes acquisition step S11 (see FIG. 1) for acquiring a target image of the surface of target object 100 . In this embodiment, the target image is an image obtained by imaging the surface of the target object 100 illuminated by the illumination system 20 with the imaging system 30 . Here, the imaging system 30 images a part of the surface of the object 100 instead of the entire surface. Therefore, from the target object 100, a plurality of target images in which the imaged portions of the surface of the target object 100 are different are acquired. That is, inspection is performed for each portion of the surface of the object 100 . In this embodiment, the acquisition unit F11 acquires an image of the surface of the object 100 from the imaging system 30. FIG. That is, the acquisition unit F11 receives an image from the imaging system 30 via the input/output unit 11. FIG. The image acquired by the acquisition unit F<b>11 from the imaging system 30 is determined by the imaging conditions of the imaging system 30 . The imaging conditions can include the relative positional relationship among the object 100, the illumination system 20, and the imaging system 30 (that is, information on the positional relationship among the object to be imaged, the illumination, and the camera). In this embodiment, the pixel values of the target image may include three color values. The three color values may be an R value corresponding to red, a G value corresponding to green, and a B value corresponding to blue. Thus, in this embodiment, pixel values are expressed in the RGB color system.

The inspection unit F12 executes an inspection step S12 (see FIG. 1) for inspecting the state of the surface of the object 100. FIG. In order to inspect the state of the surface of the object 100, the inspection unit F12 generates object information from the object image (S12a, S12b). The target information includes shape information representing the shape of the frequency distribution of parameters based on pixel values. In this embodiment, the parameter is a color parameter. In particular, as parameters, lightness (L), saturation (C), and hue (h), which are parameters of the LCh color system, are used. Therefore, the inspection unit F12 converts the pixel values of the target image from the RGB color system to the LCh color system. Then, the inspection unit F12 obtains the frequency distribution of the parameters (brightness, saturation, hue) based on the pixel values. Therefore, in the present embodiment, the target information includes shape information regarding each of a plurality of parameters (brightness, saturation, and hue).

In this embodiment, the shape information includes a feature quantity representing the shape of the frequency distribution. For example, FIG. 3 shows a histogram H10 representing the frequency distribution of brightness. In FIG. 3, the horizontal axis represents the brightness, which is a parameter, and the vertical axis represents the frequency. In this embodiment, the shape of the frequency distribution is the shape of a histogram representing the frequency distribution. The feature quantity includes multiple evaluation values and a representative value.

A plurality of evaluation values are values representing the shape of a histogram representing a frequency distribution. In this embodiment, multiple evaluation values are treated as a vector. Hereinafter, a plurality of evaluation values may be referred to as an evaluation vector. The evaluation vector is calculated based on the area of each of a plurality of sections of the parameter of the shape of the frequency distribution. For example, as shown in FIG. 3, the inspection unit F12 obtains areas Si (S1 to S7) of each of a plurality of divisions Di (D1 to D7) of the parameter (brightness) of the shape of the histogram H10. Here, i is an integer of 1 or more. In this embodiment, the multiple divisions Di have the same width (the difference between the upper and lower limits of the division). The inspection unit F12 also obtains the total area S (=ΣSi) of the shape of the histogram H10. The inspection unit F12 sets the value obtained by dividing the area Si of the section by the total area S as the evaluation value of the section. In other words, the evaluation value corresponds to the ratio of each segment to the entire histogram H10. Here, if the evaluation value of the division Di is wi, wi=Si/S is established. In the example of FIG. 3, since there are seven sections D1 to D7, the evaluation vector of the histogram H10 is (w1, w2, w3, w4, w5, w6, w7).

The representative value is the representative value of the frequency distribution. In this embodiment, the representative value corresponds to the value (central value) of the parameter (brightness in FIG. 3) at the center of the frequency distribution. The representative value is a value obtained from the ratio (evaluation vector) of each of the plurality of sections of the parameter of the shape of the frequency distribution to the entire shape of the frequency distribution and the values of the parameters corresponding to the plurality of sections. For example, in FIG. 3, let Li (L1 to L7) be the parameter values corresponding to the divisions Di (D1 to D7). Here, the value Li of the parameter corresponding to the division Di is a representative value of the range of the parameter corresponding to the division Di, and may be any of the average value, the mode value, the median value, the maximum value, and the minimum value. . In this case, if C is the representative value of the frequency distribution, then C=ΣLi*wi.

The inspection unit F12 generates shape information for each parameter based on pixel values. In the present embodiment, the inspection unit F12 generates shape information for each of brightness, saturation, and hue for the target image. In each of the lightness, saturation, and hue shape information, the feature amount includes an evaluation vector and a representative value. For example, in FIG. 4, HT1 indicates a histogram, WT1 indicates an evaluation vector, and CT1 indicates a representative value for lightness. Also, in FIG. 5, HT2 indicates a histogram, WT2 indicates an evaluation vector, and CT2 indicates a representative value regarding saturation. In FIG. 6, regarding hue, HT3 indicates a histogram, WT3 indicates an evaluation vector, and CT3 indicates a representative value. In this way, it can be said that the feature amount of the shape information represents the frequency distribution in a simplified manner to such an extent that the shape and position are not damaged.

The inspection unit F12 compares the object information thus generated with the reference information (S12c). The reference information may be pre-stored in the storage unit 12 . The reference information includes shape information obtained from the reference image regarding the surface of the reference object that serves as the reference of the target object 100 . The reference information is generated in the same manner as the target information. As with the target information, the reference information uses lightness (L), saturation (C), and hue (h), which are parameters of the LCh color system. Therefore, the inspection unit F12 converts the pixel values of the reference image from the RGB color system to the LCh color system as necessary, and obtains the frequency distribution of the parameters (brightness, saturation, hue) based on the pixel values. . Therefore, in the present embodiment, the reference information includes shape information regarding each of a plurality of parameters (brightness, saturation, and hue) in the same manner as the target information. Then, similar to the object information, in each of the lightness, saturation, and hue shape information, the feature amount includes an evaluation vector and a representative value. For example, in FIG. 7, regarding lightness, WR1 indicates an evaluation vector based on the shape of the frequency distribution from the reference image, and CR1 indicates a representative value of the frequency distribution from the reference image. FIG. 8 also shows, with respect to saturation, WR2 indicates an evaluation vector based on the shape of the frequency distribution from the reference image, and CR2 indicates a representative value of the frequency distribution from the reference image. In FIG. 9, regarding hue, WR3 indicates an evaluation vector based on the shape of the frequency distribution from the reference image, and CR3 indicates a representative value of the frequency distribution from the reference image.

In the comparison between the target information and the reference information, the inspection unit F12 obtains the difference in the feature amount of the shape information for each parameter. The feature quantity difference includes the evaluation vector difference and the representative value difference. In the example shown in FIG. 7, the evaluation vector difference Dw1 is |wT1-wR1|, and the representative value difference Dc1 is |CT1-CR1|. Table 1 below shows an example of the difference Dw1 of the evaluation vectors, and Table 2 below shows an example of the difference Dc1 of the representative values.

In the example shown in FIG. 8, the evaluation vector difference Dw2 is |wT2-wR2|, and the representative value difference Dc2 is |CT2-CR2|. Table 3 below shows an example of the difference Dw2 of the evaluation vectors, and Table 4 below shows an example of the difference Dc2 of the representative values.

In the example shown in FIG. 9, the evaluation vector difference Dw3 is |wT3-wR3|, and the representative value difference Dc3 is |CT3-CR3|. Table 5 below shows an example of the difference Dw3 of the evaluation vectors, and Table 6 below shows an example of the difference Dc3 of the representative values.

Then, the inspection unit F12 obtains a determination value (first determination value) of the evaluation vector and a determination value (second determination value) of the representative value from the difference in the feature amount of the shape information for each parameter. Assuming that the first determination value is E1, E1=(Dw1 ² +Dw2 ² +Dw3 ² ) ^1/2 . Assuming that the second determination value is E2, E2=(Dc1 ² +Dc2 ² +Dc3 ² ) ^1/2 .

Then, the inspection unit F12 determines whether the state of the surface of the object 100 is pass or fail based on the result of comparison between the object information and the reference information (S12d). In this embodiment, the inspection unit F12 determines that the state of the surface of the object 100 is acceptable when the first determination value E1 is less than the first specified value and the second determination value E2 is less than the second specified value. Suppose there is On the other hand, the inspection unit F12 determines that the state of the surface of the object 100 is unacceptable when the first determination value E1 is greater than or equal to the first specified value or the second determination value E2 is greater than or equal to the second specified value. and

As an example, the specified values (the first specified value and the second specified value) may be determined using visual evaluation. A brief description will be given of how to determine the specified value using visual evaluation. For example, as the object 100, a plurality of samples d1 to dn with different coating conditions are prepared. Note that n is an arbitrary integer of 2 or more. As an example, the coating conditions may be set so that the colors of the samples d1 to dn become lighter in order.

Each of the plurality of samples d1 to dn is compared with the reference object, and the matching rate is obtained based on the first judgment value and the second judgment value. FIG. 10 shows the relationship between samples d1 to dn and match rates. In FIG. 10, sample dk has the highest match rate. Then, each of the samples d1 to dn is visually evaluated (visual evaluation) by a plurality of persons (for example, 30 persons) to obtain the accuracy rate. Visual evaluation compares one of samples d1-dn with sample dk. The accuracy rate is the ratio of the number of people who selected sample dk to the number of people who performed visual evaluation. FIG. 11 shows the results of visual evaluation. A sample with an accuracy rate of 1.0 may be clearly rejected. A sample with an accuracy rate of 0.5 has the same accuracy rate as random, so it may be accepted as a pass. Further, for samples d2 to dk-2 and dk+2 to dn-1 whose accuracy rate is between 0.5 and 1.0, it is further determined which samples are acceptable.

For example, as shown in FIG. 12, for samples dk+2 to dn-1, sample dl having a large concordance rate gradient (slope) among samples dk+2 to dn-1 and sample dk+2, which is a pass, are visually evaluated. Here, if the accuracy rate is 1.0, the sample dl is rejected. Then, visual evaluation is performed on the sample dl-1 and the sample dl, which have the second highest matching rate after the sample dl. Then, if the accuracy rate is 1.0, the sample dl-1 is rejected. Then, the sample dl-2 and the sample dl-1, which have the second highest coincidence rate after the sample dl-1, are visually evaluated. If the accuracy rate is 0.5, the sample dl-1 is taken as the limit sample of the acceptable range. In this way, the visual evaluation is repeated until the correct answer rate reaches 0.5, and the sample when the correct answer rate reaches 0.5 is defined as the sample at the limit of the allowable range. However, if sample dk+2 is reached before the accuracy rate reaches 0.5, sample dk+2 is regarded as the limit sample of the allowable range and the visual evaluation is terminated.

On the other hand, in the visual evaluation of sample dl and sample dk+2, which is acceptable, if the accuracy rate is 0.5, sample dl is accepted. Then, visual evaluation is performed on the sample dl+1 and the sample dl, which have the second lowest coincidence rate after the sample dl. Then, if the accuracy rate is 0.5, the sample dl+1 is accepted. Then, visual evaluation is performed on samples dl+2 and dl+1, which have the second lowest matching rates after sample dl+1. If the accuracy rate is 1.0, the sample dl+1 is taken as the limit sample of the acceptable range. In this way, the visual evaluation is repeated until the correct answer rate reaches 1.0, and the sample when the correct answer rate reaches 1.0 is defined as the sample at the limit of the allowable range. However, if the sample dn-1 is reached before the accuracy rate reaches 1.0, the sample dn-2 immediately before the sample dn-1 is regarded as the sample at the limit of the allowable range, and visual evaluation is terminated.

Then, for the samples d2 to dk-2, samples at the limit of the allowable range may be similarly determined. Then, a prescribed value ( a first specified value and a second specified value). For example, and the greater of the percent match of an acceptable limit sample selected from samples d2 to dk-2 and the percent match of an acceptable limit sample selected from samples dk+2 to dn-1, The specified values (the first specified value and the second specified value) may be set based on the smaller one or the average value.

The presentation unit F13 executes a presentation step S13 (see FIG. 1) for presentation based on the result of the inspection by the inspection unit F12 (the result of the inspection in the inspection step). In other words, the presentation unit F13 makes a presentation based on the result of the inspection by the inspection unit F12. The presentation based on the result of the inspection may also include the presentation of the result of determination by the inspection part F12 using the result of comparison between the target information and the reference information. Therefore, the presentation unit F13 may present the result of determination by the inspection unit F12. In the present embodiment, the presentation unit F13 outputs the determination result of the inspection unit F12 to an external device through the input/output unit 11. FIG. The external device may present the result of the determination by the inspection unit F<b>12 , that is, the result of the inspection by the inspection system 1 .

The control unit F14 executes the control step S14 (see FIG. 1) for outputting control information based on the inspection result of the inspection step S12 to the painting system 40 that paints the surface of the object 100 . In this embodiment, when the inspection unit F12 determines that the surface condition of the object 100 is unsatisfactory, the control unit F14 outputs control information to the painting system 40 to repaint the object 100. Let As an example, the control information may include data in the format shown in Table 7 below. The control information includes multiple items. The multiple items are "model", "camera", "site", "result", "E1", "E2", "color space", and "number of bins". Here, the “model” indicates information for specifying the inspection system 1 . "Camera" indicates the camera number of the imaging system 30 used to generate the target image. "Part" indicates information for specifying the part of the object 100 shown in the target image. Here, "site" indicates the number assigned to the site. "Result" indicates the result of the inspection in the inspection unit F12. Here, the "result" is "NG", indicating that the inspection result is unsatisfactory. "E1" indicates the first determination value corresponding to the target image. "E2" indicates the second determination value corresponding to the target image. "Color space" indicates information about the color system of the frequency distribution of the target image. Table 7 indicates that the "color space" is the LCh color system. "Number of bins" indicates the number of elements of the evaluation vector for each color system. Table 7 shows that there are 64 elements in each evaluation vector for lightness, saturation, and hue. By providing such control information, it is possible to transmit to the painting system 40 which part of the object 100 should be repainted and how.

In repainting, the control unit F14 controls the painting system 40 based on the result of comparison between the target information obtained from the control information and the reference information (the difference in the feature amount of the shape information for each parameter). That is, the control unit F14 controls the coating system 40 that coats the surface of the object 100 based on the inspection result of the inspection unit F12. As a result, the state of the surface of the object 100 can be brought closer to the desired state. When the inspection unit F12 determines that the surface condition of the object 100 is acceptable, the control unit F14 may output control information to the painting system 40 to finish painting the object 100 .

(1.3) Operation Next, the inspection method performed by the inspection system 1 described above will be briefly described with reference to the flowchart of FIG. In the inspection system 1, the acquisition unit F11 acquires a target image of the surface of the target object 100 from the imaging system 30 (S11). Next, the inspection unit F12 inspects the state of the surface of the object 100 (S12). Here, the inspection unit F12 obtains a frequency distribution for each parameter based on pixel values from the target image (S12a). In this embodiment, since the parameters are lightness, saturation, and hue, the frequency distribution of lightness, the frequency distribution of saturation, and the frequency distribution of hue are obtained. Next, the inspection unit F12 generates target information based on the frequency distribution (S12b). The target information includes shape information regarding lightness, saturation, and hue. The shape information includes a feature quantity representing the shape of the frequency distribution, and includes an evaluation vector (a plurality of evaluation values) and a representative value. Next, the inspection unit F12 compares the target information and the reference information (S12c). In the comparison between the target information and the reference information, the inspection unit F12 obtains the difference in the feature amount of the shape information for each parameter (here, brightness, saturation, and hue). The feature quantity difference includes the evaluation vector difference and the representative value difference. Then, a determination value (first determination value) of the evaluation vector and a determination value (second determination value) of the representative value are obtained from the difference in the feature amount of the shape information. The inspection unit F12 determines whether the condition of the surface of the object 100 is pass or fail based on the result of comparison between the object information and the reference information (S12d). In this embodiment, the inspection unit F12 determines that the state of the surface of the object 100 is acceptable when the first determination value E1 is less than the first specified value and the second determination value E2 is less than the second specified value. Suppose there is On the other hand, the inspection unit F12 determines that the state of the surface of the object 100 is unacceptable when the first determination value E1 is greater than or equal to the first specified value or the second determination value E2 is greater than or equal to the second specified value. and Then, in the inspection system 1, the presentation unit F13 outputs the inspection result of the inspection unit F12 to the external device through the input/output unit 11 (S13). Further, in the inspection system 1, the control unit F14 outputs control information based on the inspection result to the coating system 40 (S14).

(1.4) Summary The inspection system 1 described above includes an acquisition unit F11 and an inspection unit F12. An acquisition unit F11 and an inspection unit F12 are provided. The acquiring unit F11 is a step of acquiring a target image regarding the surface of the target object 100 . The inspection unit F12 inspects the state of the surface of the object 100 based on object information including shape information representing the shape of the frequency distribution of parameters based on pixel values obtained from the object image. According to this inspection system 1, the accuracy of inspection of the state of the surface of the object 100 can be improved.

In other words, it can be said that the inspection system 1 executes the following method (inspection method). The inspection method includes an acquisition step S11 and an inspection step S12. Acquisition step S<b>11 is a step of acquiring a target image of the surface of the target object 100 . The inspection step S12 inspects the state of the surface of the object 100 based on the object information including the shape information representing the shape of the frequency distribution of the parameters based on the pixel values obtained from the object image. According to this inspection method, as with the inspection system 1, the accuracy of inspection of the surface state of the object 100 can be improved.

The inspection method is implemented by one or more processors executing a program (computer program). This program is a program for causing one or more processors to execute an inspection method. According to such a program, the accuracy of inspection of the state of the surface of the object 100 can be improved in the same manner as the inspection method. And the program can be provided by a storage medium. This storage medium is a computer-readable non-temporary storage medium, and stores the above program. According to such a storage medium, the accuracy of inspection of the state of the surface of the object 100 can be improved as in the inspection method.

(2) Modifications Embodiments of the present disclosure are not limited to the above embodiments. The above-described embodiment can be modified in various ways according to design and the like, as long as the object of the present disclosure can be achieved. Modifications of the above embodiment are listed below.

In one variation, the target information includes shape information corresponding to each of multiple parameters, but the target information may include shape information corresponding to a single parameter instead of multiple parameters. Moreover, the feature amount included in the shape information may include at least one of a plurality of evaluation values (evaluation vectors) and a representative value.

In a modified example, the pixel values of the target image are not limited to the RGB color system. Examples of the color system include the CIE color system such as the XYZ color system, the xyY color system, the L ^* u ^* v ^* color system, the L ^* a ^* b ^* color system, and the Lch color system. be done. As an example, the imaging system 30 may generate an image in which pixel values are expressed in the XYZ color system instead of an image in which pixel values are expressed in the RGB color system. Alternatively, color system conversion may be performed by arithmetic processing. For example, an image whose pixel values are represented by the RGB color system may be converted into an image whose pixel values are represented by the XYZ color system. A model formula, a lookup table, or the like can be used for the arithmetic processing in this case. This makes it possible to convert the pixel values of the target image into parameters of a desired color system.

In a modified example, the inspection unit F12 does not need to use all of the shape information of the plurality of parameters obtained from the pixel values of the target image for inspection. For example, suppose the parameters are lightness, saturation, and hue. At this time, when the saturation of the surface of the object 100 is low and it is considered achromatic, a slight difference in color can greatly change the hue. In such cases, the use of hue in the inspection may reduce the accuracy of the inspection. Therefore, when the surface of the target object 100 is achromatic, it is better not to use the shape information regarding the hue. Thus, under certain conditions, certain parameters out of multiple parameters may not be valid for inspection. In short, when a plurality of parameters includes a first parameter and a second parameter, the inspection unit F12 does not use shape information related to the second parameter for inspection when the first parameter satisfies a predetermined condition. good too. For example, the first parameter is saturation and the second parameter is hue. The predetermined condition is that the saturation is equal to or less than the threshold. Here, the threshold can be determined based on whether it can be determined that the surface of the object 100 is achromatic if the saturation is equal to or less than the threshold. Note that the combination of the first parameter and the second parameter is not limited to saturation and hue, and may be other parameters.

In a modified example, weight values may be individually set for a plurality of parameters. That is, the inspection unit F12 may weight a plurality of parameters. For example, if the surface of the object 100 has low saturation and is considered achromatic, a slight difference in color can greatly change the hue. In such cases, the use of hue in the inspection may reduce the accuracy of the inspection. In this case, the closer the surface of the object 100 is to the achromatic color, the smaller the weighting value for the hue (lighter the weighting), so that the influence of the hue on the inspection result can be reduced. For example, as shown in FIG. 13, the inspection unit F12 sets a weighted value to the evaluation vector w11 to generate an evaluation vector w12 that suppresses the influence on the inspection, and conducts the inspection based on this evaluation vector w12. you can go In this manner, the inspection unit F12 may individually change weight values for a plurality of parameters according to the surface of the target object 100. FIG.

In a modified example, weight values may be individually set for a plurality of evaluation values. That is, the inspection unit F12 may weight each element of the evaluation vector corresponding to each parameter. For example, in the state of the surface of the object 100, there are cases where it is desired to perform an inspection focusing on the sense of brightness. At this time, if the area of the high-brightness region in the target image is small, the desired inspection accuracy may not be obtained for the sense of brightness. In such a case, the inspection unit F12 may set a larger weight value as the lightness increases. For example, as shown in FIG. 14, the inspection unit F12 sets a weight value to each element of the evaluation vector w11 to generate an evaluation vector w13 in which a region with high brightness is emphasized, and based on this evaluation vector w13 , may be inspected. In this manner, the inspection unit F12 may individually change the weight values for the plurality of evaluation values according to the surface of the target object 100. FIG. As for the feeling of brilliance, although the result of the inspection in the inspection unit F12 is acceptable, there are cases where the inspection result is not acceptable when visually observed by a person. In other words, the results of the inspection in the inspection unit F12 may not match the way people feel. In such a case, it is conceivable to use psychophysical quantities that reflect psychological quantities according to how people feel as weights. This makes it possible to reflect a person's psychological quantity in the examination in the examination section F12.

In one variant, the target information may be obtained from a deviation image obtained from the target image. A deviation image is an image having, as pixel values, values obtained by subtracting a reference value for the pixel values of the target image from the pixel values of the target image. The reference value can be selected from the average value, the mode value, the median value of the pixel values of the target image, and the representative value of the frequency distribution obtained from the target image. Such deviation images are effective in inspecting the texture of the surface of the object 100 . That is, when inspecting the texture of the surface of the object 100 as the state of the surface of the object 100, the inspection unit F12 generates a deviation image from the target image, and based on the deviation image, the object Generate information. In this case, a deviation image is also generated for a reference object to be compared with the object 100 by a similar method. By generating target information using the deviation image in this way, it is possible to improve the precision of inspection of the texture of the surface of the target object 100 .

In a modified example, the obtaining unit F11 may obtain a plurality of target images obtained by imaging the surface of the target object 100 under different imaging conditions. That is, the acquisition step S11 can be a step of acquiring a plurality of target images obtained by imaging the surface of the target object 100 under different imaging conditions. Examples of imaging conditions include imaging range, imaging direction, and resolution. For example, the imaging system 30 may use a plurality of cameras 31 and 32 with different imaging ranges and resolutions (see FIG. 15) in order to have different imaging conditions. The camera 31 can capture a wider range than the camera 32 , but the camera 32 has a higher resolution than the camera 31 . The camera 32 is arranged closer to the surface of the object 100 than the camera 31 is. As shown in FIG. 15, the camera 31 outputs the image of the first imaging range 111 of the surface 110 of the target object 100 as the target image P11 (see FIG. 16). The camera 32 outputs the image of the second imaging range 112 of the surface 110 of the target object 100 as the target image P12 (see FIG. 17). Here, the second imaging range 112 is narrower than the first imaging range 111 . Note that the second imaging range 112 does not have to be included in the first imaging range 111 . The camera 31 is a macro camera that is suitable for inspection of the surface 110 of the object 100 from a larger viewpoint than the camera 32 (for example, inspection regarding gradation). The camera 32 is a micro camera that is more suitable than the camera 31 for inspecting the surface 110 of the object 100 from a smaller viewpoint (for example, inspecting graininess). For example, FIG. 18 shows the histogram H11 corresponding to the target image P11, and FIG. 19 shows the histogram H12 corresponding to the target image P12. In this way, if the imaging conditions are different, naturally the obtained histograms are also different. The inspection unit F12 obtains target information for each of the target image P11 and the target image P12, and compares the target information with the corresponding reference information. As a result, the accuracy of inspection of the state of the surface 110 of the object 100 can be improved.

In a modified example, the inspection unit F12 may inspect the uneven state of the surface of the object 100 as the surface condition of the object 100 . Here, it is considered that the feature of the unevenness of the surface of the object 100 appears well in the specular reflection portion 130 of the surface of the object 100 in the object image P20 (see FIG. 20). Therefore, the acquisition unit F11 acquires a plurality of target images obtained by imaging the surface of the target object 100 under different imaging conditions. Examples of imaging conditions include imaging range, imaging direction, and resolution. For example, the imaging system 30 may use a plurality of cameras 33 and 34 with different imaging directions (see FIG. 21) in order to vary the imaging conditions. In FIG. 21 the surface 110 of the object 100 is illuminated with the illumination system 20 . A portion of the light 21 from the illumination system 20 becomes light 22 that is specularly reflected at the surface 110 and the rest of the light 21 from the illumination system 20 is diffusely reflected. Camera 33 is positioned to receive light diffusely reflected by surface 110 of object 100 . A diffuse reflection component on the surface 110 of the object 100 is reflected in the target image generated by the camera 33 . Camera 34 is positioned to receive light specularly reflected from surface 110 of object 100 . The object image generated by the camera 34 well reflects the specular reflection component on the surface 110 of the object 100 . The specular reflection component reflects the shape of the surface 110 of the object 100 better than the diffuse reflection component. well reflected. As a result, the state of the surface 110 of the object 100 can be inspected by considering not only the color of the surface of the object 100 itself but also the shape of the surface of the object 100 . As a result, according to the inspection system 1, the accuracy of inspection of the state of the surface of the object 100 can be improved. Note that when the imaging directions are varied, the surface 110 of the object 100 may be imaged from different positions with the same camera. Further, in order to better inspect the uneven state of the surface of the object 100, the specular reflection portion 130 may be imaged with a macro camera and a micro camera to obtain an object image. Target images obtained using high dynamic range imaging may also be used. The target image obtained by high dynamic range synthesis can handle a wider dynamic range than usual, so it is possible to change the amount of light from a relatively small amount of light such as a diffuse reflection component to a relatively large amount of light such as a specular reflection component. It is possible to capture up to the components in a single image.

In a modified example, the divisions Di for obtaining a plurality of evaluation values may not necessarily have the same width. For example, in the above embodiment, as shown in FIG. 3, the multiple segments D1-D7 have the same width. In contrast, the divisions Di may be of different widths. As the width of the division Di is narrowed, the resolution of the evaluation vector is improved, and the accuracy of inspection can be improved. In particular, it is preferable to reduce the width of the section Di closer to the peak of the histogram. For example, FIG. 22 shows a histogram H20 obtained from a target image containing many specular reflection components. The peak of histogram H20 is skewed to the bright side (right side). In FIG. 22, a plurality of divisions Di (D21 to D30) are set for the histogram H20. Here, the widths of the plurality of sections D21 to D30 are narrowed in this order. By setting the widths of the sections D21 to D30 in this manner, the shape of the histogram H20 is better reflected in the evaluation vectors obtained from these sections D21 to D30. Therefore, the accuracy of inspection of the surface state of the object 100 can be improved. In particular, in the range where the frequency changes greatly in the histogram H20, it is possible to better reflect the shape of the histogram H20 in the evaluation vector by reducing the width of the division. Conversely, in the range where the frequency change is small in the histogram H20, the width of the division may be increased, thereby reducing the data amount of the evaluation vector.

In a modified example, the inspection unit F12 may inspect the image clarity of the surface of the target object 100 as the state of the surface of the target object 100 . Image clarity is an index of sharpness of an object image (reflected image) reflected on the surface of the object 100 . When the image clarity is poor, the reflected image becomes unclear, and when the image clarity is improved, the reflected image becomes clear. Image clarity is affected by surface roughness and shape of object 100 . In the image clarity inspection, an inspection image 200 is used as shown in FIG. A chart with a predetermined test pattern is used as the inspection image 200 . The inspection image 200 is illuminated by light 23 from an illumination system 20 such that a reflected image 210 of the inspection image 200 appears on the surface 110 of the object 100 . Imaging system 30 is then positioned to receive light 24 from reflected image 210 . Thereby, the imaging system 30 captures the image of the reflected image 210 appearing on the surface 110 of the target object 100 and outputs it as the target image. That is, in the image clarity inspection, the target image is the image of the reflection image of the inspection image 200 from the surface 110 of the target object 100 . The inspection unit F12 may generate target information from such a target image, compare the target information with corresponding reference information, and inspect the state (image clarity) of the surface 110 of the target object 100 . A self-luminous display displaying a predetermined test pattern may be used instead of the chart. In this case, illumination system 20 may not be used to project the test pattern onto the surface of object 100 .

In a modified example, the presentation unit F13 may present difference information that visually indicates the difference between the target information and the reference information based on the inspection result of the inspection unit F12. FIG. 24 shows an example of difference information. FIG. 24 emphasizes the difference in the representative value of the frequency distribution as the difference information. More specifically, as shown in FIG. 24, the difference information is a graph showing the evaluation vector wT1 and representative value CT1 obtained from the target information and the evaluation vector wR1 and representative value CR1 obtained from the reference information. The difference information also includes an upper limit value Th1 and a lower limit value Th2 of the allowable range of the representative value CT1 of the target information. The allowable range is set around the representative value CR1 obtained from the reference information. According to the difference information in FIG. 24, the user can visually determine whether the representative value CT1 of the target information is within the allowable range with respect to the representative value CR1 of the reference information. FIG. 25 shows another example of difference information. FIG. 25 emphasizes the difference in evaluation vector between the target information and the reference information as the difference information. More specifically, in FIG. 25, the difference information is a graph showing the evaluation vector wT1 obtained from the target information and the evaluation vector wR1 obtained from the reference information. Furthermore, in FIG. 25, regions R11 and R12 that are the difference between the evaluation vector wT1 of the target information and the evaluation vector wR1 of the reference information are highlighted. Therefore, according to the difference information in FIG. 25, the user can visually grasp the difference between the evaluation vector wT1 of the target information and the evaluation vector wR1 of the reference information. Therefore, the user can determine whether the evaluation vector wT1 of the target information is within the allowable range with respect to the evaluation vector wR1 of the reference information. FIG. 26 shows another example of difference information. FIG. 26 shows, as the difference information, the vector of the difference between the evaluation vectors of the target information and the reference information. More specifically, the difference information in FIG. 26 displays areas R11 and R12 that are differences between the evaluation vector wT1 of the target information and the evaluation vector wR1 of the reference information in FIG. Therefore, according to the difference information in FIG. 26, the user can visually grasp the difference between the evaluation vector wT1 of the target information and the evaluation vector wR1 of the reference information. Therefore, the user can determine whether the evaluation vector wT1 of the target information is within the allowable range with respect to the evaluation vector wR1 of the reference information.

Further, the presentation unit F13 may present an image of the object 100 in which the difference between the object information and the reference information as a result of the inspection by the inspection unit F12 is reflected on the surface of the object 100 . Furthermore, the presentation unit F13 may present a presentation that clearly distinguishes between the portion determined to be acceptable by the inspection unit F12 and the portion determined to be unacceptable. For example, in an image P30 of the object 100 shown in FIG. 27, the surface of the object 100 is classified into four regions R31, R32, R33 and R34. Four regions R31, R32, R33, and R34 indicate that the difference in representative value between the target information and the reference information increases in this order. A region R34 is a portion of the surface of the object 100 that has been determined to be rejected by the inspection unit F12. By such presentation by the presentation unit F13, it is possible to present a portion of the target image that has been determined to be rejected (a portion with an abnormality) in an easy-to-understand manner. For example, in an image P40 of the object 100 shown in FIG. 28, the surface of the object 100 is classified into four regions R41, R42, R43 and R44. Four regions R41, R42, R43, and R44 indicate that the difference in evaluation vector between the target information and the reference information increases in this order. A region R44 is a portion of the surface of the object 100 that has been determined to be rejected by the inspection unit F12. By such presentation by the presentation unit F13, it is possible to present a portion of the target image that has been determined to be rejected (a portion with an abnormality) in an easy-to-understand manner.

In a modified example, the input/output unit 11 may include an image display device. In this case, the presentation unit F<b>13 may display the result of the inspection by the inspection unit F<b>12 on the image display device of the input/output unit 11 . Also, the input/output unit 11 may include an audio output device. In this case, the presentation unit F13 may output the result of the inspection by the inspection unit F12 from the voice output device of the input/output unit 11 .

In one variant, the wavelength of light emitted by illumination system 20 may be variable. This can be achieved using light sources with different emission colors, or color filters. Thus, in the inspection system 1, at least one of the wavelength of light emitted by the illumination system 20 and the wavelength of light detected by the imaging system 30 may be changeable.

In a modified example, the inspection system 1 (determination system 10) may be configured by a plurality of computers. For example, the functions of the inspection system 1 (determination system 10) (particularly, the acquisition unit F11, the inspection unit F12, the presentation unit F13, and the control unit F14) may be distributed among multiple devices. Furthermore, at least part of the functions of the inspection system 1 (determination system 10) may be realized by, for example, the cloud (cloud computing).

The executing subject of the inspection system 1 (determination system 10) described above includes a computer system. A computer system has a processor and memory as hardware. A processor executes a program recorded in the memory of the computer system, thereby realizing the function of the inspection system 1 (determination system 10) in the present disclosure as an execution entity. The program may be prerecorded in the memory of the computer system, or may be provided through an electric communication line. Also, the program may be provided by being recorded on a non-temporary recording medium such as a computer system-readable memory card, optical disk, hard disk drive, or the like. A processor in a computer system consists of one or more electronic circuits including semiconductor integrated circuits (ICs) or large scale integrated circuits (LSIs). A field programmable gate array (FGPA), an ASIC (application specific integrated circuit), or a reconfigurable field programmable gate array (FGPA) or ASIC (application specific integrated circuit), which can be programmed after the LSI is manufactured, or which can reconfigure junction relationships inside the LSI or set up circuit partitions inside the LSI. Logical devices can also be used for the same purpose. A plurality of electronic circuits may be integrated into one chip, or may be distributed over a plurality of chips. A plurality of chips may be integrated in one device, or may be distributed in a plurality of devices.

(3) Aspects As is clear from the above embodiments and modifications, the present disclosure includes the following aspects. In the following, reference numerals are attached with parentheses only for the purpose of clarifying correspondence with the embodiments.

A first aspect is an inspection method including an acquisition step (S11) and an inspection step (S12). The acquisition step (S11) is a step of acquiring a target image of the surface of the target object (100). The inspecting step (S12) inspects the state of the surface of the object (100) based on object information including shape information representing the shape of the frequency distribution of parameters based on pixel values obtained from the object image. It is a step to According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A second aspect is an inspection method based on the first aspect. In a second aspect, the shape information includes a feature quantity representing a shape feature of the frequency distribution. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A third aspect is an inspection method based on the second aspect. In a third aspect, the feature quantity includes a plurality of evaluation values based on areas (Si) of a plurality of divisions (Di) of the parameter of the shape of the frequency distribution. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A fourth aspect is an inspection method based on the third aspect. In a fourth aspect, in the inspection method, weight values are individually set for the plurality of evaluation values. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A fifth aspect is an inspection method based on the second aspect. In a fifth aspect, the feature quantity includes a representative value of the frequency distribution. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A sixth aspect is an inspection method based on the fifth aspect. In the sixth aspect, the representative value is the ratio (wi) of each of the plurality of divisions (Di) of the parameter of the shape of the frequency distribution to the entire shape of the frequency distribution, and the plurality of divisions (Di) It is a value obtained from the value (Li) of the corresponding parameter. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A seventh aspect is an inspection method based on any one of the first to sixth aspects. In a seventh aspect, the parameter is a color parameter. According to this aspect, it is possible to improve the accuracy of the color inspection as the state of the surface of the object (100).

An eighth aspect is an inspection method based on any one of the first to seventh aspects. In an eighth aspect, the target information includes shape information regarding each of the plurality of parameters. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A ninth aspect is an inspection method based on the eighth aspect. In the ninth aspect, weight values are individually set for the plurality of parameters. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A tenth aspect is an inspection method based on the eighth or ninth aspect. In a tenth aspect, the plurality of parameters includes a first parameter and a second parameter. The inspection step (S12) does not use the shape information related to the second parameter for inspection when the first parameter satisfies a predetermined condition. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

An eleventh aspect is an inspection method based on the tenth aspect. In an eleventh aspect, the first parameter is saturation. The second parameter is hue. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A twelfth aspect is an inspection method based on any one of the first to eleventh aspects. In the twelfth aspect, the inspection step (S12) compares the target information with reference information including the shape information obtained from a reference image regarding the surface of a reference object serving as a reference of the target object (100). do. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A thirteenth aspect is an inspection method based on any one of the first to twelfth aspects. In a thirteenth aspect, the target information is obtained from a deviation image obtained from the target image. The deviation image is an image having, as a pixel value, a value obtained by subtracting a reference value of the pixel value of the target image from the pixel value of the target image. According to this aspect, it is possible to inspect the texture as the condition of the surface of the object (100).

A fourteenth aspect is an inspection method based on any one of the first to thirteenth aspects. In the fourteenth aspect, the acquisition step (S11) acquires a plurality of target images obtained by imaging the surface of the target object (100) under different imaging conditions. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A fifteenth aspect is an inspection method based on any one of the first to fourteenth aspects. In a fifteenth aspect, the target image is an image of a reflection image (210) of an inspection image (200) from the surface of the target object (100). According to this aspect, it is possible to inspect the image clarity as the condition of the surface of the object (100).

A sixteenth aspect is an inspection method based on any one of the first to fifteenth aspects. In the sixteenth aspect, the inspection method further includes a presentation step (S13) of performing presentation based on the inspection result of the inspection step (S12). According to this aspect, the result of inspection of the object (100) can be presented.

A seventeenth aspect is an inspection method based on any one of the first to sixteenth aspects. In the seventeenth aspect, the inspection method includes a control step ( S14) is further included. According to this aspect, the quality of the coating on the surface of the object (100) can be improved.

An eighteenth aspect is an inspection method based on any one of the first to seventeenth aspects. In an eighteenth aspect, the shape of the frequency distribution is the shape of a histogram. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A nineteenth aspect is a program for causing one or more processors to execute the inspection method according to any one of the first to eighteenth aspects. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

A twentieth aspect is an inspection system (1) comprising an acquisition unit (F11) and an inspection unit (F12). The acquisition unit (F11) acquires a target image of the surface of the target object (100). The inspection unit (F12) inspects the state of the surface of the object (100) based on object information including shape information representing the shape of the frequency distribution of parameters based on pixel values obtained from the object image. do. According to this aspect, it is possible to improve the accuracy of inspection of the state of the surface of the object (100).

The second to eighteenth aspects can be applied to the nineteenth and twentieth aspects with appropriate modifications.

1 inspection system F11 acquisition unit F12 inspection unit 100 object 200 inspection image 210 reflection image S11 acquisition step S12 inspection step S13 presentation step S14 control step

Claims (20)

an acquisition step of acquiring an image of the object of the surface of the object;
an inspection step of inspecting the state of the surface of the object based on object information including shape information representing the shape of the frequency distribution of parameters based on pixel values obtained from the object image;
including,
Inspection method. The shape information includes a feature quantity representing the shape of the frequency distribution,
The inspection method according to claim 1. The feature amount includes a plurality of evaluation values based on the area of each of the plurality of sections of the parameter of the shape of the frequency distribution,
The inspection method according to claim 2. weighting values are individually set for the plurality of evaluation values;
The inspection method according to claim 3. The feature amount includes a representative value of the frequency distribution,
The inspection method according to claim 2. The representative value is a value obtained from the ratio of each of the plurality of sections of the parameter of the shape of the frequency distribution to the entire shape of the frequency distribution and the value of the parameter corresponding to the plurality of sections.
The inspection method according to claim 5. the parameter is a color parameter,
The inspection method according to any one of claims 1 to 6. The target information includes shape information about each of the plurality of parameters,
The inspection method according to any one of claims 1 to 7. Weight values are individually set for the plurality of parameters,
The inspection method according to claim 8. The plurality of parameters includes a first parameter and a second parameter,
In the inspection step, if the first parameter satisfies a predetermined condition, the shape information regarding the second parameter is not used for inspection.
The inspection method according to claim 8 or 9. the first parameter is saturation,
wherein the second parameter is hue;
The inspection method according to claim 10. The inspection step compares the target information with reference information including the shape information obtained from a reference image regarding the surface of a reference object serving as a reference of the target.
The inspection method according to any one of claims 1 to 11. The target information is obtained from a deviation image obtained from the target image,
The deviation image is an image having, as a pixel value, a value obtained by subtracting a reference value of the pixel value of the target image from the pixel value of the target image.
The inspection method according to any one of claims 1 to 12. The acquisition step acquires a plurality of target images obtained by imaging the surface of the target under different imaging conditions.
The inspection method according to any one of claims 1 to 13. The target image is an image of a reflected image of an inspection image from the surface of the target,
The inspection method according to any one of claims 1 to 14. further comprising a presentation step of presenting based on the results of the inspection in the inspection step;
The inspection method according to any one of claims 1 to 15. Further comprising a control step of outputting control information based on the result of the inspection in the inspection step to a painting system that paints the surface of the object,
The inspection method according to any one of claims 1 to 16. The shape of the frequency distribution is the shape of a histogram,
The inspection method according to any one of claims 1 to 17. for causing one or more processors to execute the inspection method according to any one of claims 1 to 18,
program. an acquisition unit that acquires a target image of a surface of an object;
an inspection unit that inspects the state of the surface of the object based on object information including shape information representing the shape of the frequency distribution of parameters based on pixel values obtained from the object image;
comprising
inspection system.

Images