JP2023127798A - Device for executing simulation of flaw detection of workpiece, method of executing the simulation and program of realizing the simulation - Google Patents

Abstract

To provide a simulation technique for constructing a system for detecting a flaw or chip of a workpiece.SOLUTION: A device 100 for simulation comprises: a camera setting unit which receives setting of a camera 102; a light source setting unit which receives setting of a light source 103; a workpiece setting unit which receives setting of a workpiece 150; a simulation execution unit which reproduces the operation of the workpiece 150; a flaw generation unit which generates a flaw 160 in the workpiece 150 when the workpiece 150 collides with another object; an imaging unit which images the workpiece 150; an image recognition unit which detects the flaw 160 from the captured image of the workpiece 150; a setting selection unit which selects setting of the camera 102 and setting of the light source 103 to be output on the basis of a correct answer rate of flaw detection when plural times of simulation are executed by using the combination of different setting of the camera 102 and setting of the light source 103; and an output unit which outputs the selected setting of the camera 102 and setting of the light source 103.SELECTED DRAWING: Figure 1

Description

The present disclosure relates to a simulation for detecting flaws on a workpiece, and more specifically, to a technique for outputting various settings for detecting flaws on a workpiece based on simulation results.

2. Description of the Related Art Systems or inspection devices using cameras and image processing devices are used in factory production lines and the like to detect scratches or chips on workpieces. In order to properly detect scratches on a workpiece, it is necessary to adjust camera placement, image processing settings, etc. in advance. With conventional technology, it was not possible to perform simulations that included detection of scratches on workpieces, so staff spent a long time on site adjusting camera placement, image processing settings, etc. Therefore, there is a need for a technology for constructing a system that more easily detects scratches or chips on a workpiece.

Regarding the technology for detecting flaws on a workpiece, for example, Japanese Patent Application Laid-Open No. 2019-158499 (Patent Document 1) discloses an appearance inspection system, which "includes an imaging condition determining section and a route determining section. The imaging condition determination section determines a plurality of imaging condition candidates including a relative position between the workpiece and the imaging device for at least one inspection target position among the plurality of inspection target positions. By selecting one of the plurality of imaging condition candidates so as to satisfy the specified requirements, a path for changing the imaging conditions is determined in order to sequentially image multiple inspection target positions. (See [Summary]).

Japanese Patent Application Publication No. 2019-158499

According to the technique disclosed in Patent Document 1, the system for detecting flaws on the workpiece cannot be verified until the line is constructed. Therefore, there is a need for simulation technology for constructing a system for detecting scratches or chips on a workpiece.

The present disclosure has been made in view of the above background, and an objective in one aspect is to provide a simulation technique for constructing a system for detecting scratches or chips on a workpiece.

According to an embodiment, an apparatus for simulating flaw detection on a workpiece is provided. The device includes a camera setting unit that accepts settings for a camera placed in 3D space, a light source setting unit that accepts settings for a light source in 3D space, a work setting unit that accepts settings for a workpiece, and a workpiece setting unit that accepts settings for a workpiece in 3D space. A simulation execution unit that reproduces the work, a scratch generation unit that generates scratches on the work when the work collides with another object based on the work settings, and an image of the work based on the camera settings and light source settings. an image recognition unit that uses an image recognition engine to detect flaws from images of the workpiece, and a simulation execution unit that executes the simulation multiple times using different combinations of camera settings and light source settings. a setting selection section that selects camera settings and light source settings to be output based on the correct answer rate of flaw detection by the image recognition section when Equipped with

According to this disclosure, the apparatus outputs camera and light source settings based on simulation results of detecting flaws on a workpiece in 3D space. By referring to the camera and light source settings output by the device, the user can consider or decide the configuration of the workpiece flaw detection system to be used on the production line, etc., before actually constructing the production line, etc. of the factory. obtain.

In the above disclosure, the device further includes an image recognition setting section that adjusts settings of the image recognition engine. The setting selection section selects an image recognition engine that outputs an image based on the correct answer rate of flaw detection by the image recognition section when the image recognition section executes flaw detection processing on the workpiece multiple times using different settings of the image recognition engine. Select settings. The output unit outputs the settings of the selected image recognition engine.

According to this disclosure, the apparatus outputs the settings of the image recognition engine based on the simulation result of detecting flaws on a workpiece in 3D space. By referring to the settings of the image recognition engine output by the device, the user can adjust the settings of the image recognition engine used in the production line or the like in the factory before actually constructing the production line or the like.

In the above disclosure, the camera setting unit is configured to be able to receive input of camera identification information and input of camera setting items associated with the identification information.

According to this disclosure, the device can receive in advance input of camera settings used in simulation.

In the above disclosure, the light source setting unit is configured to be able to accept settings for one or more lights and settings for ambient light.

According to this disclosure, the device can receive in advance input of settings for light sources used in simulation.

In the above disclosure, the workpiece setting unit is configured to be able to accept settings for the material of the workpiece, settings for the number of generated workpieces, settings for the environment in which the workpiece is placed, and some or all of the 3D model data for the workpiece. be done.

According to this disclosure, the apparatus can receive in advance the settings of a workpiece to be used in simulation and the input of 3D model data.

In the above disclosure, the flaw generation unit generates flaws on the workpiece by changing polygons or meshes that constitute the workpiece.

According to this disclosure, the device can generate scratches on the workpiece by changing the polygons or meshes that make up the workpiece.

According to another embodiment, a method is provided for a computer to perform a simulation of flaw detection on a workpiece. The method includes the following steps: accepting settings for a camera placed in 3D space, accepting settings for a light source in 3D space, accepting settings for a workpiece, and executing a simulation to reproduce the movement of the workpiece in 3D space. a step of generating a scratch on the workpiece when the workpiece collides with another object based on the settings of the workpiece; a step of imaging the workpiece based on the camera settings and the light source setting; The step of detecting flaws from images of the workpiece, and the camera settings to be output based on the correct answer rate for flaw detection when simulations using combinations of different camera settings and light source settings are executed multiple times. and selecting a light source setting; and outputting the selected camera setting and light source setting.

According to this disclosure, an apparatus that performs the above method outputs camera and light source settings based on a simulation result of detecting flaws on a workpiece in 3D space. By referring to the camera and light source settings output by the device, the user can consider or decide the configuration of the workpiece flaw detection system to be used on the production line, etc., before actually constructing the production line, etc. of the factory. obtain.

Furthermore, according to another embodiment, a program for causing a computer to execute a simulation of detecting flaws on a workpiece is provided. The program includes a step of receiving settings for a camera placed in 3D space, a step of receiving settings of a light source in 3D space, a step of receiving settings of a workpiece, and a simulation that reproduces the movement of the workpiece in 3D space. a step of generating a scratch on the workpiece when the workpiece collides with another object based on the settings of the workpiece; a step of imaging the workpiece based on the camera settings and the light source setting; The step of detecting flaws from images of the workpiece, and the camera settings to be output based on the correct answer rate for flaw detection when simulations using combinations of different camera settings and light source settings are executed multiple times. and causing the computer to execute the steps of selecting a light source setting and outputting the selected camera setting and light source setting.

According to this disclosure, a device that executes the above program outputs camera and light source settings based on a simulation result of detecting flaws on a workpiece in 3D space. By referring to the camera and light source settings output by the device, the user can consider or decide the configuration of the workpiece flaw detection system to be used on the production line, etc., before actually constructing the production line, etc. of the factory. obtain.

According to an embodiment, it is possible to provide a simulation technique for building a system for detecting flaws or chips in a workpiece.

These and other objects, features, aspects, and advantages of the disclosure will become apparent from the following detailed description of the disclosure, taken in conjunction with the accompanying drawings.

It is a figure showing an example of the outline of simulation by device 100 according to a certain embodiment. FIG. 1 is a diagram illustrating an example of a hardware configuration of a device 100 according to an embodiment. 1 is a diagram illustrating an example of functional blocks of a device 100 according to an embodiment. FIG. 4 is a diagram showing an example of a first work setting screen 400. FIG. 5 is a diagram showing an example of a second work setting screen 500. FIG. 6 is a diagram showing an example of simulation execution screens 600, 610, and 620. FIG. 5 is a diagram showing an example of a display of a flaw 160 on a workpiece 150. FIG. 8 is a diagram showing an example of a display of a camera setting screen 800. FIG. 9 is a diagram showing an example of a display of a light source setting screen 900. FIG. It is a figure which shows an example of a display of 900 A of light source setting screens when a light source is set to a 1st position. It is a figure which shows an example of a display of the light source setting screen 900B when a light source is set to a 2nd position. 3 is a diagram illustrating an example of the influence of the aperture (F number) of the camera 102. FIG. 13 is a diagram showing an example of an image recognition setting screen 1300. FIG. FIG. 2 is a diagram showing an example of a simulation execution procedure by the device 100 (simulator 110).

Hereinafter, embodiments of the technical idea according to the present disclosure will be described with reference to the drawings. In the following description, the same parts are given the same reference numerals. Their names and functions are also the same. Therefore, detailed descriptions thereof will not be repeated.

<A. Application example>
FIG. 1 is a diagram illustrating an example of an overview of a simulation performed by an apparatus 100 according to the present embodiment. The functions of the device 100 will be explained with reference to FIG.

The device 100 includes a simulator 110. In some aspects, simulator 110 may be implemented by software. In this case, the simulator 110 can be realized by executing a program on the hardware of the device 100 shown in FIG. 2. In other aspects, simulator 110 may be implemented partially or entirely in hardware.

Note that in this specification, the term "device" refers to a configuration consisting of one or more devices, a server, a virtual machine or container built in a cloud environment, or a system consisting of at least a portion of these. include. Further, the device may include an information processing device such as a personal computer, a workstation, a server device, a tablet, or a smartphone, and a PLC (Programmable Logic Controller), or may be a combination thereof. In one aspect, the device 100 may be used by a user by being connected to input/output devices such as a display and a keyboard. In other aspects, the device 100 may provide various functions to the user as a cloud service or a web application via a network. In this case, the user may use the functionality of the device 100 via a browser or client software installed on his or her terminal.

(a. Overview of functions of device 100)
The device 100 (simulator 110) provides a simulation function for detecting flaws 160 on the workpiece 150. The simulation function includes a first function (a function of simulating the generation of the flaw 160 on the workpiece 150) and a second function (a function of simulating the detection of the flaw 160 on the workpiece 150). In this specification, the term "work" refers to an object that is transported, inspected, processed, or subjected to any other treatment on a factory production line or the like. Further, the workpiece may be made of any material. As an example, the workpiece may be made of metal, resin, paper, food, any other material, or a combination thereof. Further, the workpiece may have any shape. As an example, the workpiece may be a hard object such as a box made of metal or resin, or may be a flexible object such as a film. Furthermore, the color of the workpiece may be any color, and may be translucent or transparent.

The first function (a function that simulates the generation of scratches 160 on the workpiece 150) is a function that generates virtual scratches 160 on the workpiece 150 when the workpiece 150 collides with another object in the 3D space 10. be.

Normally, when detecting the flaw 160 on the workpiece 150 using an image recognition engine, it is necessary to input a sample image to the image recognition engine and adjust the image recognition engine. However, with conventional simulation technology, it was not possible to reproduce the flaw 160 on the workpiece 150 in the 3D space 10, so the user had to adjust the image recognition engine before constructing a manufacturing line etc. (hereinafter referred to as "line"). It was not possible to sufficiently collect samples of various flaws 160 (cracks, scratches, chips, dents, etc.) that may occur on the required workpiece 150.

By using the first function (the function of simulating the generation of scratches 160 on the workpiece 150), the user can create workpieces 150A, 150B, 150C, and 150D (when collectively referred to simply as "workpiece 150") having different shapes and materials. It is possible to obtain image samples of respective scratches 160A, 160B, 160C, and 160D (collectively referred to simply as "wounds 160"). Note that, in reality, multiple types of scratches 160 may occur on each workpiece 150, so the user can obtain sample images of one or more scratches 160 on each workpiece. The apparatus 100 can generate samples of various scratches 160(1), 160(2) to 160(N) that may occur on each workpiece 150 through simulation.

The second function (a function that simulates detection of a flaw 160 on the workpiece 150) is to image the workpiece 150 using a virtual camera 102 and a light source 103 in the 3D space 10, and further use an image recognition engine to image the workpiece 150. This is a function to detect flaws 160 on the photographed workpiece 150. As used herein, the term "light source" includes any light, natural light, reflected light, other light, and combinations thereof. Furthermore, the device 100 (simulator 110) may have a function of arranging one or more (plural) light sources 103 in the 3D space 10. Further, the device 100 may provide the ability to select the type of light that is the light source. The type of light source may include any type of light such as a spotlight, a dome-shaped light, a combination of lights and their reflected light, and the like.

By using the second function (a function that simulates detection of flaws 160 on workpiece 150), the user can select camera 102, adjust camera 102 settings, and adjust camera 102 settings without actually constructing a line. The placement can be considered. Further, the user can select the light source (light) 103, adjust the settings of the light source 103, and consider the arrangement of the light source 103 without actually constructing the line. Additionally, the user may also configure the image recognition engine without actually constructing the line.

That is, by using the device 100, the user can determine the configuration of the system for detecting flaws 160 on the workpiece 150 before constructing the line. Here, the "configuration of a system for detecting scratches on a workpiece" includes, for example, selection of the camera 102, adjustment of the settings of the camera 102, arrangement of the camera 102, selection of the light source (light) 103, setting of the light source 103, This includes part or all of the arrangement of the light source 103 and the settings of the image recognition engine.

(b. Outline of operation procedure of device 100)
Next, an apparatus 100 for realizing the first function (function of simulating generation of scratches 160 on workpiece 150) and second function (function of simulating detection of scratches 160 on workpiece 150) described above. The operating procedure will be explained. Note that the operation procedure described below is an example, and the order of each process may be changed arbitrarily, or a plurality of processes may be executed simultaneously.

In the first step, the apparatus 100 receives input of settings for the workpiece 150 via a UI (User Interface) 300 (see FIG. 3). As an example, the settings for the work 150 may include the material of the work 150, the number of the work 150, the arrangement of the work 150, designation of 3D model data of the work 150 to be used, and the like. In one aspect, the settings for the work 150 may include environmental settings (temperature, humidity, etc.) for the location where the work 150 is placed. Note that, in addition to the settings of the workpiece 150, the apparatus 100 may also receive input of environmental settings (temperature, humidity, etc.) for the place where the workpiece 150 is placed, as environmental settings.

In the second step, the device 100 receives input of settings for the camera 102 via the UI 300. As an example, the settings of the camera 102 may include identification information of the camera 102, various settings of the camera 102, and arrangement of the camera 102. The identification information of the camera 102 is information that can specify the model number of the camera 102 and the like. Various camera settings may include arbitrary settings such as ISO sensitivity, aperture (F value), and shutter speed. The arrangement of the camera 102 is the arrangement of the camera 102 or the lens of the camera 102 within the 3D space 10. In one aspect, the device 100 may change various setting items of the camera 102 displayed on the UI 300 based on the identification information of the selected camera 102. In other aspects, the device 100 may accept input of settings for a plurality of cameras 102. In this case, the device 100 arranges two or more cameras 102 in the 3D space 10.

In the third step, the device 100 receives input of settings for the light source 103. As an example, the settings of the light source 103 may include identification information of the light source 103, various settings of the light source 103, arrangement of the light source 103, and settings of ambient light. The identification information of the light source 103 is information that can specify the model number and the like of the light source 103. Various settings of the light source 103 may include arbitrary settings such as brightness. The arrangement of the light sources 103 is the arrangement of the light sources 103 within the 3D space 10. Ambient light may be any light source other than light source 103, and may include light coming in through the windows of the factory where the line is installed, light from lights originally in the factory, and the like. In one aspect, the device 100 may change various setting items of the light source 103 displayed on the UI 300 based on the identification information of the selected light source 103. In other aspects, the device 100 may accept input of settings for a plurality of light sources 103. In this case, the device 100 places two or more light sources 103 in the 3D space 10.

In the fourth step, the apparatus 100 executes a collision simulation of the workpiece 150 based on the input settings of the workpiece 150. As an example, the apparatus 100 may perform a collision simulation by dropping one or more workpieces 150 to the ground in the 3D space 10. In other examples, the apparatus 100 may perform a collision simulation by shaking a container containing one or more workpieces 150 in the 3D space 10. Furthermore, in another example, the apparatus 100 may perform a collision simulation by reproducing a line to be constructed in the 3D space 10 and reproducing various collisions that may occur on the workpiece 150 on the line.

In the fifth step, the apparatus 100 generates scratches 160 on one or more workpieces 150 based on the collision simulation performed in the fourth step. The apparatus 100 generates a flaw 160 on the workpiece 150 when a force equal to or greater than a predetermined threshold is applied to the workpiece 150 in a collision simulation. In one aspect, the apparatus 100 may determine a predetermined threshold value based on the material of the workpiece 150. In other aspects, the apparatus 100 may accept input of a predetermined threshold value as part of setting the workpiece 150. The apparatus 100 generates scratches 160 on the workpiece 150 by changing the shape of 3D model data of the workpiece 150. For example, the apparatus 100 can generate (reproduce) the scratches 160 on the workpiece 150 by deforming or deleting part of the polygons or meshes that constitute the 3D model data of the workpiece 150. In some aspects, the apparatus 100 may repeatedly perform the fifth step and the sixth step until sufficient sample images of the flaw 160 on the workpiece 150 can be acquired. For example, the device 100 may repeatedly execute the fifth step and the sixth step for the number of collision simulation executions input via the UI 300. Alternatively, the fifth step and the sixth step may be repeatedly executed until sample images corresponding to the scheduled number of sample images input via the UI 300 are acquired.

In the sixth step, the apparatus 100 images one or more workpieces 150 in the 3D space 10 based on the input camera 102 settings and light source 103 settings. The device 100 associates a flag indicating whether or not the scratch 160 is shown with the image obtained by imaging, and stores the image in the primary storage device 2 (see FIG. 2) and/or the secondary storage device 3 (see FIG. 2). ). Furthermore, the apparatus 100 can repeatedly execute the simulation while changing at least some of the input camera 102 settings and light source 103 settings.

In the seventh step, the device 100 uses an image recognition engine to analyze the image obtained by imaging and detect the flaw 160. In this step, the device 100 analyzes the image and detects the flaw 160 based on the default image recognition engine settings or the image recognition engine settings input via the UI 300. The device 100 also calculates the correct answer rate of the image recognition engine when one or more images are analyzed. When the image recognition engine analyzes an image, the device 100 compares the flag associated with the image with the output result of the image recognition engine in the sixth step, and determines the analysis result of the image recognition engine. You can judge whether it is correct or not.

In one aspect, the apparatus 100 changes the settings of the camera 102 and/or the settings of the light source 103 until the correct answer rate of the image recognition engine reaches a predetermined first standard, and performs the fourth to seventh steps. You may perform the steps repeatedly. For example, the device 100 may repeatedly execute the fourth to seventh steps until the correct answer rate (first criterion) input via the UI 300 is reached. In another aspect, the apparatus 100 may repeatedly perform the fourth to seventh steps until finding a plurality of combinations of camera 102 settings and light source 103 settings that meet the first criterion. good. During the repetition, the apparatus 100 repeatedly executes the simulation while changing at least part of the input settings of the camera 102 and the settings of the light source 103. Thereby, the device 100 can discover settings for the camera 102 and settings for the light source 103 that result in a higher correct answer rate for the default image recognition engine. Furthermore, in another aspect, the device 100 may repeatedly execute the fourth to seventh steps until a predetermined number of executions input via the UI 300 is reached.

In an eighth step, the device 100 adjusts the settings of the image recognition engine. In one aspect, the device 100 may change the settings of the image recognition engine and repeatedly perform the eighth step until the correct answer rate of the image recognition engine reaches a predetermined second standard. Note that the device 100 may receive an input of the correct answer rate (second criterion) in advance via the UI 300. In other aspects, the apparatus 100 may repeatedly perform the eighth step until it finds a plurality of image recognition engine settings that meet the second criterion. As a result, the device 100 selects image recognition engine settings that have a high detection rate of flaws 160 for the captured image using the camera 102 settings and light source 103 settings obtained through the processing up to the seventh step. can be discovered. In other aspects, the device 100 may repeatedly execute the eighth step until reaching a predetermined number of executions input via the UI 300. Furthermore, in other aspects, the device 100 may repeatedly execute the eighth step while changing the settings of the image recognition engine input via the UI 300.

In the ninth step, the apparatus 100 selects a combination of settings (a combination of camera 102 settings, light source 103 settings, and image recognition engine settings) that increases the detection rate of the flaws 160 on the workpiece 150. In one aspect, the apparatus 100 may select only one combination of settings that increases the detection rate of the flaws 160 on the workpiece 150, or may select a plurality of combinations.

In the tenth step, the device 100 outputs the combination of settings selected in the ninth step. In some aspects, device 100 may display the combination of settings on a display connected to device 100. In other aspects, device 100 may transmit information regarding the selected configuration combination to other devices via the network. Furthermore, in another aspect, the device 100 outputs the output data without including the settings of the image recognition engine in the output data, but including only the combination of settings of the camera 102 and the light source 103 that have a high detection rate of the flaw 160. You may. For example, the device 100 may receive in advance, via the UI 300, an input from the user specifying the type of settings to be included in the output data. If the user already has image recognition engine settings with a certain level of precision, etc., by not including the image recognition engine settings in the output data, the user can omit the eighth step and obtain simulation execution results quickly. can do.

By performing the above series of processes, the device 100 can discover settings for the camera 102 and settings for the light source 103 that increase the correct answer rate of the default image recognition engine. That is, the device 100 can discover settings for the camera 102 and settings for the light source 103 that allow the image recognition engine to obtain an image that allows the image recognition engine to easily detect the flaw 160 on the workpiece 150. Furthermore, the apparatus 100 may discover image recognition engine settings that have a high detection rate of flaws 160 for images captured using the selected camera 102 settings and light source 103 settings. By constructing a line that reflects a combination of these settings, the user can realize a line or flaw detection system that can detect flaws 160 on workpiece 150 with high accuracy.

<B. Device configuration>
Next, the hardware configuration and functional configuration of the device 100 according to this embodiment will be described with reference to FIGS. 2 and 3.

FIG. 2 is a diagram showing an example of the hardware configuration of the device 100 according to this embodiment. The device 100 includes a CPU (Central Processing Unit) 1, a primary storage device 2, a secondary storage device 3, an external device interface 4, an input interface 5, an output interface 6, and a communication interface 7.

The CPU 1 can execute programs for realizing various functions of the device 100. The CPU 1 is configured by, for example, at least one integrated circuit. The integrated circuit may be configured by, for example, at least one CPU, at least one FPGA (Field Programmable Gate Array), or a combination thereof. The CPU 1 can implement the various functions described with reference to FIG. 1 by executing the simulator 110 loaded into the primary storage device 2 from the secondary storage device 3.

The primary storage device 2 stores programs executed by the CPU 1 and data referenced by the CPU 1. In one aspect, the primary storage device 2 may be implemented by DRAM (Dynamic Random Access Memory), SRAM (Static Random Access Memory), or the like.

The secondary storage device 3 is a nonvolatile memory, and may store programs executed by the CPU 1 including the simulator 110 and data referenced by the CPU 1. In that case, the CPU 1 executes the program read from the secondary storage device 3 to the primary storage device 2 and refers to the data read from the secondary storage device 3 to the primary storage device 2. In one aspect, the secondary storage device 3 is realized by a HDD (Hard Disk Drive), an SSD (Solid State Drive), an EPROM (Erasable Programmable Read Only Memory), an EEPROM (Electrically Erasable Programmable Read Only Memory), a flash memory, or the like. It's okay.

The external device interface 4 can be connected to any external device such as a printer, scanner, and external HDD. In one aspect, the external device interface 4 may be realized by a USB (Universal Serial Bus) terminal or the like.

Input interface 5 may be connected to any input device such as a keyboard, mouse, touch pad or game pad. In one aspect, the input interface 5 may be realized by a USB terminal, a PS/2 terminal, a Bluetooth (registered trademark) module, or the like.

The output interface 6 can be connected to any output device such as a cathode ray tube display, a liquid crystal display, or an organic EL (Electro-Luminescence) display. In one aspect, the output interface 6 may be realized by a USB terminal, a D-sub terminal, a DVI (Digital Visual Interface) terminal, an HDMI (registered trademark) (High-Definition Multimedia Interface) terminal, or the like.

The communication interface 7 is connected to wired or wireless network equipment. In one aspect, the communication interface 7 may be realized by a wired LAN (Local Area Network) port, a Wi-Fi (registered trademark) (Wireless Fidelity) module, or the like. In other aspects, the communication interface 7 may transmit and receive data using communication protocols such as TCP/IP (Transmission Control Protocol/Internet Protocol) and UDP (User Datagram Protocol).

In some aspects, the simulator 110 may be realized as independent software, may be realized in collaboration with other libraries or software, or may be realized as an IDE (Integrated Development Program) for developing programs for the control device 200. It may be realized as an add-in or a function of the ``Environment''. The control device 200 is, for example, a PLC (Programmable Logic Controller) or an integrated control device having both a robot control function and a PLC function.

In other aspects, some or all of the functions of the simulator 110 may be installed in the control device 200. In this case, the device 100 provides the UI 300 to the user and transmits the settings and simulation execution commands received from the user to the control device 200. The control device 200 executes a simulation based on the received settings and execution command, and selects a combination of settings (settings of the camera 102, settings of the light source 103, images combination of recognition engine settings) is sent to the device 100.

FIG. 3 is a diagram showing an example of functional blocks of the device 100 according to this embodiment. Each functional block shown in FIG. 3 can be realized as a program module, library, etc. that constitute the simulator 110. In some aspects, a portion of each functional block shown in FIG. 3 may be implemented as hardware. In this case, the device 100 further includes a hardware configuration for realizing a part of each functional block shown in FIG. 3.

The simulator 110 includes a UI 300, a work setting section 301, a camera setting section 302, a light source setting section 303, a 3D (dimensional) engine 304, a simulation execution section 305, and a collision determination section 306 as main functional blocks. , a flaw generation section 307, an imaging section 308, an image recognition section 309, an image recognition setting section 310, a setting selection section 311, and an output section 312. In some aspects, simulator 110 may use an external 3D engine instead of having 3D engine 304.

The UI 300 provides various functions of the simulator 110 to the user. As an example, the UI 300 includes a setting screen for the work 150, a setting screen for the camera 102, a setting screen for the light source 103, a simulation setting screen, a screen displaying the simulation execution status, a screen displaying the simulation results, and the simulator 110. may include any screens related to the functionality of. The simulator 110 displays various settings and simulation-related information to the user via the UI 300, and can also accept input of various settings from the user. The simulator 110 also provides various 3D model data (a model of the work 150, a model of the camera 102, a model of the light source (light) 103, a model of each part constituting the line to be constructed, It is possible to accept import operations for other arbitrary models, etc.).

The work setting unit 301 acquires the settings of the work 150 input via the UI 300 and outputs the settings of the work 150 to the simulation execution unit 305. Further, the work setting unit 301 stores the input setting information in the secondary storage device 3. The work setting unit 301 may read setting information of one or more previously registered works 150 from the secondary storage device 3 and display it on the UI 300 (work setting screen). The user can newly input setting information for the work 150 on the work setting screen. Further, the user may select any of the setting information of one or more previously registered works 150 on the work setting screen. Furthermore, the user can also select any of the setting information of one or more previously registered works 150 on the work setting screen and edit the setting information of the selected work 150.

The camera setting unit 302 acquires the settings of the camera 102 input via the UI 300 and outputs the settings of the camera 102 to the simulation execution unit 305. Further, the camera setting unit 302 stores the input setting information in the secondary storage device 3. The camera setting unit 302 may read setting information of one or more cameras 102 registered in the past from the secondary storage device 3 and display it on the UI 300 (camera setting screen). The user can newly input setting information for the camera 102 on the camera setting screen. Further, the user may select any of the setting information of one or more cameras 102 registered in the past on the camera setting screen. Furthermore, the user can also select any of the setting information of one or more cameras 102 registered in the past on the camera setting screen and edit the setting information of the selected camera 102.

The light source setting unit 303 acquires the settings of the light source 103 input via the UI 300 and outputs the settings of the light source 103 to the simulation execution unit 305. Further, the light source setting unit 303 stores the input setting information in the secondary storage device 3. The light source setting unit 303 may read setting information of one or more light sources 103 registered in the past from the secondary storage device 3 and display it on the UI 300 (light source setting screen). The user can newly input setting information for the light source 103 on the light source setting screen. Further, the user may select any of the setting information of one or more light sources 103 registered in the past on the light source setting screen. Furthermore, the user can also select any of the setting information of one or more light sources 103 registered in the past on the light source setting screen and edit the setting information of the selected light source 103.

3D engine 304 reproduces 3D space 10. The 3D engine 304 performs physical calculations such as display of the 3D space 10 and movement of objects within the 3D space 10. The 3D engine 304 may be referenced by the simulation execution unit 305 as a library, runtime, or the like.

The simulation execution unit 305 executes a simulation based on the setting information of the work 150 and the simulation setting information. The simulation execution unit 305 can execute any simulation using the 3D model data imported into the simulator 110. As an example, the simulation execution unit 305 may perform a collision simulation by dropping one or more works 150 onto the ground in the 3D space 10. In another example, the simulation execution unit 305 may perform a collision simulation by shaking a container containing one or more works 150 in the 3D space 10. In another example, the simulation execution unit 305 may perform a collision simulation by reproducing a line to be constructed in the 3D space 10 and reproducing various collisions that may occur on the workpiece 150 on the line. . Based on the simulation settings, the simulation execution unit 305 may repeatedly execute the simulation until the set number of times is reached, or the simulation execution unit 305 may repeatedly execute the simulation until the number of images output by the camera 102 reaches the set number. May be executed repeatedly. Furthermore, the simulation execution unit 305 can repeatedly execute the simulation while changing the settings acquired from the camera setting unit 302 and the settings acquired from the light source setting unit 303.

The collision determination unit 306 determines whether a collision has occurred for each workpiece 150 during the simulation. More specifically, the collision determination unit 306 calculates, in addition to the presence or absence of a collision, the location where the collision occurred for each workpiece 150 and the force applied to the workpiece 150 due to the collision (damage to the workpiece 150). The collision determination unit 306 outputs information regarding the presence or absence of a collision for each workpiece 150, the location where the collision occurred, and the force applied to the workpiece 150 due to the collision (damage to the workpiece 150) to the flaw generation unit 307. The collision determination unit 306 uses the setting information of the workpiece 150 to calculate the force applied to the workpiece 150 due to the collision.

The flaw generation unit 307 generates flaws 160 on each workpiece 150 based on the information acquired from the collision determination unit 306. For example, the flaw generation unit 307 generates flaws 160 on the workpiece 150 by deforming polygons or meshes of the 3D model data of the workpiece 150.

The imaging unit 308 images the workpiece 150 using the virtual camera 102 arranged in the 3D space 10. The imaging unit 308 stores the image obtained by imaging in the secondary storage device 3. The imaging unit 308 executes imaging processing based on the settings of the camera 102 and the settings of the light source 103. The imaging unit 308 associates the generated image with a flag indicating whether or not the workpiece 150 with the scratch 160 is included in the image. Note that the imaging unit 308 refers to the output result of the collision determination unit 306 (collision information for each workpiece 150) or the output result of the flaw generation unit 307 (information about the flaw 160 for each workpiece 150), thereby obtaining the generated image. It can be determined whether or not the workpiece 150 with the scratch 160 is included.

The image recognition unit 309 uses an image recognition engine to determine whether there is a workpiece 150 with a flaw 160 in the image captured by the imaging unit 308. Image recognition section 309 outputs the determination result to image recognition setting section 310.

If the correct answer rate of the determination result output by the image recognition unit 309 does not reach a predetermined standard, the image recognition setting unit 310 changes the settings of the image recognition engine and causes the image recognition unit 309 to perform image recognition. Submit the execution request again. Note that the image recognition setting unit 310 can determine whether the analysis result of the image recognition engine is correct by comparing the flag associated with the image and the output result of the image recognition engine.

The setting selection unit 311 selects a setting based on the correct answer rate of the determination result output by the image recognition unit 309 when the image recognition unit 309 executes the process of detecting 160 flaws on the workpiece 150 multiple times using different settings of the image recognition engine. Then, a combination of settings (settings of the camera 102, settings of the light source 103, settings of the image recognition engine) that gives the highest correct answer rate of the determination result output by the image recognition unit 309 is selected.

The output unit 312 outputs the combination of settings selected by the setting selection unit 311. In some aspects, the output unit 312 may display the combination of settings on a display connected to the device 100. In other aspects, the output unit 312 may transmit information regarding the selected setting combination to another device via the network. Furthermore, in another aspect, the output unit 312 does not include the settings of the image recognition engine in the output data, but only includes the combination of the settings of the camera 102 and the light source 103 that have a high detection rate of the flaw 160, and outputs the output data. You can also output it. For example, the device 100 may receive in advance, via the UI 300, an input from the user specifying the type of settings to be included in the output data. In cases where the user already has image recognition engine settings with a certain level of accuracy or higher, the user can quickly obtain simulation execution results by not including the image recognition engine settings in the output data.

<C. User interface>
Next, an example of the UI 300 provided by the simulator 110 according to this embodiment will be described with reference to FIGS. 4 to 13. In some aspects, simulator 110 may display each screen shown in FIGS. 4 to 13 on a display connected to device 100. In other aspects, the simulator 110 may transmit information on each screen shown in FIGS. 4 to 13 to other devices via a network.

FIG. 4 is a diagram showing an example of the first work setting screen 400. The first work setting screen 400 has, as main items, a display 401 of 3D model data of the work 150, an environment setting item 410, and 3D model data setting items 420A and 420B (collectively referred to as "3D model data "setting items 420"). Note that the screen shown in FIG. 4 is an example, and the first work setting screen 400 may include any number of 3D model data setting items 420.

The environment setting item 410 is configured to be able to set arbitrary environmental information around the workpiece 150 such as temperature or humidity. The 3D model data setting item 420 is configured to allow editing of arbitrary 3D model data settings used in the simulation (material, mass, friction coefficient, threshold of force for causing scratches, etc.). Note that the 3D model data setting items 420 are configured so that settings for any object other than the workpiece 150 (parts forming a line, etc.) can also be edited. The 3D model data display 401 displays the 3D model data displayed in the 3D model data setting item 420. In the example of FIG. 4, a work 150A corresponding to a 3D model data setting item 420A and a work 150B corresponding to a 3D model data setting item 420B are displayed in a 3D model data display 401.

FIG. 5 is a diagram showing an example of the second work setting screen 500. The simulator 110 may provide a second work setting screen 500 in an editor format in addition to the first work setting screen 400 in a dialog format. The second work setting screen 500 is configured such that a simulation program and/or script can be written. By using the second work setting screen 500 in the editor format, the user can create simulation settings for duplicating and using a large amount of works 150 with a small amount of description.

In one aspect, each workpiece setting screen may be configured to allow setting of the material, shape, color, transparency, etc. of the workpiece. As an example, the user can set the material of each work 150 (metal, resin, paper, food, any other material, or a combination thereof, etc.) and the shape of the work 150 (hard The shape of a flexible box, shape of a flexible film, or any other arbitrary shape) and the color of the workpiece (any color or combination of colors, transparency, etc.) may be settable.

FIG. 6 is a diagram showing an example of simulation execution screens 600, 610, and 620. The simulator 110 executes a collision simulation of the workpiece 150 based on the input settings of the workpiece 150. The simulator 110 also outputs a simulation execution screen. A simulation execution screen may be provided as part of the UI 300. For example, simulation execution screens 600, 610, and 620 show a simulation in which a plurality of workpieces 150 fall and collide with each other or the ground.

FIG. 7 is a diagram showing an example of a display of a flaw 160 on the workpiece 150. When the simulator 110 determines that a flaw 160 is formed on the workpiece 150, the simulator 110 generates (reproduces) the flaw 160 by deforming a polygon, mesh, or the like at the location of the flaw 160 on the surface of the workpiece 150. For example, the simulator 110 may create a depression representing the flaw 160 on the surface of the workpiece 150 by deleting the polygon where the flaw 160 is located. The simulator 110 generates a flaw 160 for each workpiece 150 on the simulation execution screen, and displays the generated flaw 160 on the simulation execution screen (for example, the simulation execution screen 600, 610, 620).

FIG. 8 is a diagram showing an example of a display of the camera setting screen 800. The camera setting screen 800 includes, as main items, a display 810 of the 3D space 10 seen from the lens of the camera 102, an overhead display 840 of the 3D space 10 including the camera 102, and a setting item 820 for the camera 102. A user may adjust the position of the camera 102 in the 3D space 10 while referring to a representation 810 of the 3D space 10 as seen through the lens of the camera 102 and/or an overhead representation 840 of the 3D space 10 including the camera 102. The setting items 820 include information that can identify the camera 102, arbitrary settings such as the resolution, brightness, ISO, focal length, brightness, pixel size, angle of view, field of view, drawing distance, etc. of the camera 102, and the arrangement of the camera 102. is configured to be configurable. In a certain aspect, the information that can identify the camera 102 may be an arbitrary ID, a model number of the camera 102, or the like. In other aspects, the simulator 110 may change the settings displayed in the setting item 820 based on information that allows the camera 102 to be identified. As the user edits settings 820 for camera 102 , simulator 110 may update display 810 of 3D space 10 as seen through the lens of camera 102 and/or overhead display 840 of 3D space 10 including camera 102 .

FIG. 9 is a diagram showing an example of a display of the light source setting screen 900. The light source setting screen 900 includes a display 910 of the 3D space 10 including the workpiece 150 and a setting item 920 for the light source 103 as main items. The user may adjust the position of light source 103 in 3D space 10 while referring to display 910.

The setting item 920 is configured to allow input of information that allows identification of the light source 103, arbitrary settings such as the position (or direction), hue, saturation, and brightness of the light source 103, and arbitrary settings for ambient light. . In a certain aspect, the information that can identify the light source 103 may be an arbitrary ID, a model number of the light source (light) 103, or the like. In other aspects, the simulator 110 may change the settings displayed in the setting item 920 based on information that allows the light source 103 to be identified. The simulator 110 may update the display 910 by the user editing the settings 920 of the light source 103.

In one aspect, the setting item 920 may be configured to allow selection of a plurality of (two or more) light sources 103. Further, the setting item 920 may be configured to allow selection of the type of light source 103 (spotlight, dome-shaped light, or other arbitrary light). Furthermore, the setting item 920 may be configured to allow selection of whether or not to include reflected light of the selected light in the light source 103. By including reflected light in the light source 103, the user can select the light source 103 that is most suitable for a type of workpiece 150 that is easily affected by reflection, such as a glossy workpiece 150 or a transparent workpiece 150.

In the example of FIG. 9, the user may adjust the position of light source 103 using light rotation settings 930. More specifically, the user may adjust the position of light source 103 in 3D space 10 by moving point 935 representing light source 103 on circle 936. Note that the center of the circle indicates the location where the work 150 is located. Changes in the display 910 when the light source 103 moves are as shown in FIGS. 10 and 11.

FIG. 10 is a diagram showing an example of the display of the light source setting screen 900A when the light source is set to the first position. FIG. 11 is a diagram showing an example of the display of the light source setting screen 900B when the light source is set to the second position. In the light source setting screen 900A, a point 935 exists at a first position 1010 on a circle 936. In the light source setting screen 900B, the point 935 exists at a second position 1110 on the circle 936. Comparing the displays 910A and 910B, it can be seen that the appearance of the workpiece 150 in the 3D space 10 changes significantly due to the change in the arrangement of the light source 103. As the position and brightness of the light source 103 change, the image obtained by the camera 102 also changes. As a result, the correct answer rate of the analysis results of the image recognition engine also changes. In one aspect, the light source setting screen 900 may include a setting item 1000 for a filter for the workpiece 150. The user can change the appearance of the workpiece 150 by selecting a filter from the filter setting items 1000.

FIG. 12 is a diagram showing an example of the influence of the aperture (F number) of the camera 102. Referring to FIG. 12, it can be seen that changing the aperture (F number) also affects the focus of the camera 102. By adjusting the aperture (F number) and the like on the camera setting screen 800, the camera 102 can output an image in which the image recognition engine can easily detect the flaw 160 on the workpiece 150.

FIG. 13 is a diagram showing an example of an image recognition setting screen 1300. The image recognition setting screen 1300 includes an image display 1310 and an image recognition engine setting item 1320 as main items. The user may adjust settings of the image recognition engine while viewing the display 1310 of the image and the detection results of the flaws 160 in the image.

In some aspects, simulator 110 may use settings entered by the user on image recognition settings screen 1300 as default settings for the image recognition engine. In another aspect, in the image recognition engine setting adjustment process (eighth step in FIG. 1), the simulator 110, instead of automatically adjusting the image recognition engine settings, via the image recognition setting screen 1300, An input to change the settings of the image recognition engine may be received from the user.

<D. Flowchart>
FIG. 14 is a diagram showing an example of a simulation execution procedure by the device 100 (simulator 110). In a certain aspect, the CPU 1 may read a program for performing the processing shown in FIG. 14 from the secondary storage device 3 into the primary storage device 2, and execute the program. In other aspects, part or all of the processing may be implemented as a combination of circuit elements configured to perform the processing.

In step S1405, the apparatus 100 displays a work setting screen. The work setting screen here corresponds to the first work setting screen 400 and/or the second work setting screen 500. Note that the work setting screen may be a part of the UI 300 or may be an independent screen.

In step S1410, the apparatus 100 receives settings for the work 150. As an example, the apparatus 100 can accept the material, mass, friction coefficient, threshold value of force for causing damage, number of pieces, etc. of the workpiece 150 as settings for the workpiece 150. Further, the apparatus 100 may receive input of environmental information (temperature, humidity, etc.) around the workpiece 150.

In step S1415, the device 100 displays a camera setting screen. The camera setting screen here corresponds to the camera setting screen 800. Note that the camera setting screen may be a part of the UI 300 or may be an independent screen.

In step S1420, the device 100 receives settings for the camera 102. As an example, the device 100 may accept information that can identify the camera 102, resolution, brightness, ISO, focal length, brightness, pixel size, angle of view, field of view, drawing distance, placement, etc. of the camera 102 as settings for the camera 102. .

In step S1425, the device 100 displays a light source setting screen. The light source setting screen here corresponds to the light source setting screen 900. Note that the light source setting screen may be a part of the UI 300 or may be an independent screen.

In step S1430, the device 100 receives settings for the light source 103. As an example, the device 100 can accept information that allows the light source 103 to be identified, as well as the position (or direction), hue, saturation, brightness, etc. of the light source 103 as settings for the light source 103.

In step S1435, the apparatus 100 simulates the movement of each workpiece 150 in the 3D space 10. More specifically, the apparatus 100 generates one or more workpieces 150 in the 3D space 10 based on the settings of the workpieces 150 input in step S1410, and executes a collision simulation of each workpiece 150. Furthermore, the device 100 outputs a simulation execution screen (corresponding to the simulation execution screens 600 to 620).

In step S1440, the apparatus 100 generates scratches on each workpiece 150 based on the hit determination. More specifically, when the apparatus 100 determines that a force of a certain value or more (a threshold value or more) is applied to the workpiece 150 based on the settings of the workpiece 150, the device 100 generates a scratch on the workpiece 150. The apparatus 100 expresses the flaws 160 on the workpiece 150 by changing the shape of the 3D model data of the workpiece 150 (by changing the polygons, mesh, etc.). The processing in this step corresponds to the processing in FIG.

In step S1445, the apparatus 100 images the workpiece 150 with the camera 102 placed in the 3D space 10. More specifically, the device 100 positions the camera 102 in the 3D space 10 based on the settings of the camera 102. The device 100 also adjusts the settings of the camera 102 based on the settings of the camera 102. The apparatus 100 captures an image of the workpiece 150 using the camera 102 whose arrangement and settings have been adjusted, and generates one or more images.

In step S1450, the device 100 uses the image recognition engine to detect flaws from the image. The image recognition engine analyzes one or more images obtained from the camera 102 placed in the 3D space 10 and detects a flaw 160 on the workpiece 150 that appears in the one or more images.

In step S1455, the device 100 determines whether imaging and image recognition have been repeated a specified number of times. In one aspect, the apparatus 100 may receive an input from the user via the UI 300 about the number of times imaging and image recognition are to be performed as part of simulation settings. If the device 100 determines that the imaging and image recognition have been repeated the specified number of times (YES in step S1455), the control moves to step S1465. Otherwise (NO in step S1455), the device 100 moves control to step S1460.

In step S1460, the device 100 changes the settings of the camera 102 and light source 103. The device 100 changes the settings of the camera 102 and light source 103 and repeatedly executes the processes from steps S1440 to S1450, thereby adjusting the settings of the camera 102 and light source 103 so that the default image recognition engine can easily detect the flaw 160. can be discovered.

In step S1465, the device 100 adjusts the settings of the image recognition engine. More specifically, the apparatus 100 repeatedly executes the process of detecting the flaw 160 on the workpiece 150 while adjusting the settings of the image recognition engine on the images obtained through the processes from steps S1440 to S1460. Thereby, the device 100 can discover image recognition engine settings that have a higher detection rate for the flaws 160. In one aspect, the device 100 may select settings for the camera 102 and light source 103 that allow the default image recognition engine to most easily detect the flaw 160 based on the execution result of the process in step S1450. The apparatus 100 may then perform the process of this step on one or more images captured using the selected settings of the camera 102 and light source 103. In another aspect, the device 100 executes the process of this step on one or more images captured using each of the settings of all the cameras 102 and light sources 103 used in the process of step S1450. It's okay. Through the processing in this step, the device 100 can discover image recognition engine settings that have a high detection rate for the flaws 160.

In step S1470, the device 100 outputs a combination of settings for the camera 102, light source 103, and image recognition engine. More specifically, the device 100 outputs the combination of settings of the camera 102, light source 103, and image recognition engine that has the highest detection rate of the flaw 160 based on the results of the processing from steps S1440 to S1465. In one aspect, the apparatus 100 may output a plurality of combinations of settings of the camera 102, light source 103, and image recognition engine that have a high detection rate of the flaw 160, based on the results of the processing from steps S1440 to S1465. Furthermore, in another aspect, the device 100 outputs the output data without including the settings of the image recognition engine in the output data, but including only the combination of settings of the camera 102 and the light source 103 that have a high detection rate of the flaw 160. You may. For example, the device 100 may receive in advance, via the UI 300, an input from the user specifying the type of settings to be included in the output data. If the user already has image recognition engine settings with a certain level of precision, etc., the user can omit step S1465 and obtain simulation execution results quickly by not including the image recognition engine settings in the output data. I can do it.

In one aspect, the device 100 may display a work setting screen, a camera setting screen, a light source setting screen, a simulation execution screen, and a simulation result screen on a display connected to the device 100. In another aspect, the apparatus 100 may transmit information on a work setting screen, a camera setting screen, a light source setting screen, a simulation execution screen, and a simulation result screen to the user's terminal. In this case, the user's terminal can display a work setting screen, a camera setting screen, a light source setting screen, a simulation execution screen, and a simulation result screen on the display using the browser or client software.

As described above, the apparatus 100 according to the present embodiment performs a collision simulation of the workpiece 150 in the 3D space 10 and generates a virtual scratch 160 on the workpiece 150. Furthermore, the apparatus 100 images the scratches 160 on the workpiece 150 in the 3D space 10 based on the input settings of the camera 102 and the settings of the light source 103. Additionally, the device 100 repeatedly performs the process of detecting flaws 160 in the image and adjusts the settings of the image recognition engine. Further, the apparatus 100 outputs a combination of settings for the camera 102, light source 103, and image recognition engine that has a high detection rate of the flaws 160 on the workpiece 150, based on the results of the simulation. By referring to the output combination of settings, the user can consider or adjust the settings of the camera, light source, and image recognition engine used in the line even before the line is constructed.

<E. Additional notes＞
As described above, this embodiment includes the following disclosures.
[Configuration 1]
A device (100) that performs a simulation of detecting a flaw (160) on a workpiece (150),
a camera setting section (302) that accepts settings for a camera placed in the 3D space (10);
a light source setting section (303) that accepts settings for the light source of the 3D space (10);
a work setting section (301) that accepts settings for the work (150);
a simulation execution unit (305) that reproduces the operation of the workpiece (150) in the 3D space (10);
a flaw generation unit (307) that generates a flaw (160) on the work (150) when the work (150) collides with another object based on settings of the work (150);
an imaging unit (308) that images the workpiece (150) based on the camera settings and the light source settings;
an image recognition unit (309) that detects the flaw (160) from an image of the workpiece (150) captured using an image recognition engine;
When the simulation execution unit (305) executes the simulation multiple times using different combinations of the camera settings and the light source settings, the correct answer rate for detecting the flaw (160) of the image recognition unit (309) a settings selection unit (311) that selects settings of the camera and settings of the light source to be output based on the information;
An apparatus (100) comprising an output section (312) for outputting the selected settings of the camera and the settings of the light source.
[Configuration 2]
further comprising an image recognition setting section (310) that adjusts settings of the image recognition engine,
The setting selection unit (311) selects the flaw (160) of the workpiece (150) when the image recognition unit (309) executes the process of detecting the flaw (160) on the workpiece (150) multiple times using different settings of the image recognition engine. Selecting the settings of the image recognition engine to be output based on the correct answer rate of the flaw (160) detection by the image recognition unit (309),
The device (100) according to configuration 1, wherein the output unit (312) outputs settings of the selected image recognition engine.
[Configuration 3]
The camera setting unit (302) is the device according to configuration 1 or 2 ( 100).
[Configuration 4]
The device (100) according to any one of configurations 1 to 3, wherein the light source setting unit (303) is configured to be able to accept settings for one or more lights and settings for ambient light.
[Configuration 5]
The workpiece setting section (301) sets the material of the workpiece (150), sets the number of the workpieces (150) to be generated, sets the environment in which the workpiece (150) is placed, and sets the workpiece (150). ) The device (100) according to any one of configurations 1 to 4, is configured to be able to accept some or all of the 3D model data.
[Configuration 6]
The flaw generating unit (307) generates the flaw (160) on the workpiece (150) by changing polygons or meshes constituting the workpiece (150), according to any one of configurations 1 to 5. (100).
[Configuration 7]
A method for a computer to perform a simulation of detecting a flaw (160) on a workpiece (150), the method comprising:
a step of accepting settings for a camera placed in the 3D space (10);
a step of receiving settings for a light source in the 3D space (10);
a step of accepting settings for the work (150);
executing a simulation that reproduces the movement of the workpiece (150) in the 3D space (10);
generating a scratch (160) on the workpiece (150) when the workpiece (150) collides with another object based on the settings of the workpiece (150);
capturing an image of the workpiece (150) based on the camera settings and the light source settings;
detecting the flaw (160) from the captured image of the workpiece (150);
When a simulation using a combination of different camera settings and light source settings is executed multiple times, the camera settings and the light source settings are output based on the correct answer rate for detecting the flaw (160). the step of selecting
outputting selected camera settings and light source settings.
[Configuration 8]
A program for causing a computer to execute a simulation of detecting a flaw (160) on a workpiece (150), the program comprising:
a step of accepting settings for a camera placed in the 3D space (10);
a step of receiving settings for a light source in the 3D space (10);
a step of accepting settings for the work (150);
executing a simulation that reproduces the movement of the workpiece (150) in the 3D space (10);
generating a scratch (160) on the workpiece (150) when the workpiece (150) collides with another object based on the settings of the workpiece (150);
capturing an image of the workpiece (150) based on the camera settings and the light source settings;
detecting the flaw (160) from the captured image of the workpiece (150);
When a simulation using a combination of different camera settings and light source settings is executed multiple times, the camera settings and the light source settings are output based on the correct answer rate for detecting the flaw (160). the step of selecting
A program that causes the computer to execute the step of outputting the selected camera settings and the light source settings.

The embodiments disclosed this time should be considered to be illustrative in all respects and not restrictive. The scope of the present disclosure is indicated by the claims rather than the above description, and it is intended that all changes are included within the meaning and scope equivalent to the claims. Furthermore, the disclosures described in the embodiments and each modification are intended to be implemented alone or in combination to the extent possible.

1 CPU, 2 Primary storage device, 3 Secondary storage device, 4 External device interface, 5 Input interface, 6 Output interface, 7 Communication interface, 10 3D space, 100 Device, 102 Camera, 103 Light source, 110 Simulator, 150 Work , 160 scratches, 200 control device, 301 work setting section, 302 camera setting section, 303 light source setting section, 304 3D engine, 305 simulation execution section, 306 collision determination section, 307 flaw generation section, 308 imaging section, 309 image recognition section , 310 image recognition setting section, 311 setting selection section, 312 output section, 400 first work setting screen, 401 3D model data display, 410 environment setting items, 420 3D model data setting items, 500 second work setting Screen, 600, 610, 620 Simulation execution screen, 800 Camera setting screen, 810 Display of 3D space 10 seen from the lens of camera 102, 820, 920, 1320 Setting items, 840 Overhead display of 3D space 10 including camera 102 , 900, 900A, 900B Light source setting screen, 910, 1310 Display, 930 Rotation setting item, 935 Point, 936 Yen, 1000 Filter setting item, 1010 First position, 1110 Second position, 1300 Image recognition setting screen.

Claims (8)

A device that performs a simulation of detecting scratches on a workpiece,
a camera setting section that accepts settings for cameras placed in 3D space;
a light source setting unit that accepts settings for a light source in the 3D space;
a work setting section that accepts settings for the work;
a simulation execution unit that reproduces the movement of the workpiece in the 3D space;
a flaw generation unit that generates flaws on the workpiece when the workpiece collides with another object based on settings of the workpiece;
an imaging unit that images the workpiece based on settings of the camera and settings of the light source;
an image recognition unit that detects the flaw from an image of the workpiece captured using an image recognition engine;
When the simulation execution unit executes the simulation multiple times using different combinations of the camera settings and the light source settings, the camera settings are output based on the correct answer rate of flaw detection of the image recognition unit. and a settings selection section that selects settings for the light source;
An apparatus comprising: an output unit that outputs selected settings of the camera and settings of the light source. further comprising an image recognition setting section that adjusts settings of the image recognition engine,
The setting selection unit is configured to select a method based on a correct answer rate of flaw detection by the image recognition unit when the image recognition unit executes the flaw detection process on the workpiece multiple times using different settings of the image recognition engine. , select the settings of the image recognition engine to output,
The apparatus according to claim 1, wherein the output unit outputs settings of the selected image recognition engine. The device according to claim 1 or 2, wherein the camera setting section is configured to be able to receive input of identification information of the camera and input of setting items of the camera associated with the identification information. The apparatus according to any one of claims 1 to 3, wherein the light source setting section is configured to be able to accept settings for one or more lights and settings for ambient light. The workpiece setting unit is configured to be able to accept settings for the material of the workpiece, settings for the number of the workpieces to be generated, settings for the environment in which the workpiece is placed, and part or all of 3D model data for the workpiece. The device according to any one of claims 1 to 4, wherein: The apparatus according to any one of claims 1 to 5, wherein the flaw generating section generates the flaws on the workpiece by changing polygons or meshes forming the workpiece. A method for a computer to perform a simulation of flaw detection on a workpiece, the method comprising:
a step of accepting settings for a camera placed in the 3D space;
accepting settings for a light source in the 3D space;
a step of accepting settings of the work;
executing a simulation that reproduces the movement of the workpiece in the 3D space;
generating scratches on the workpiece when the workpiece collides with another object based on settings of the workpiece;
imaging the workpiece based on the camera settings and the light source settings;
Detecting the flaw from the captured image of the workpiece;
When a simulation using a combination of different camera settings and light source settings is executed multiple times, the camera settings and the light source settings to be output are selected based on the correct answer rate of the flaw detection. step and
outputting selected camera settings and light source settings. A program for causing a computer to perform a simulation of detecting flaws on a workpiece,
a step of accepting settings for a camera placed in the 3D space;
accepting settings for a light source in the 3D space;
a step of accepting settings of the work;
executing a simulation that reproduces the movement of the workpiece in the 3D space;
generating scratches on the workpiece when the workpiece collides with another object based on settings of the workpiece;
imaging the workpiece based on the camera settings and the light source settings;
Detecting the flaw from the captured image of the workpiece;
When a simulation using a combination of different camera settings and light source settings is executed multiple times, the camera settings and the light source settings to be output are selected based on the correct answer rate of the flaw detection. step and
A program that causes the computer to execute the step of outputting the selected camera settings and the light source settings.

Images