Painting evaluation apparatus and painting evaluation method

Abstract

This painting evaluation device and this painting evaluation method involve: acquiring shape information representing the curved shape of a painted surface and surface roughness information representing the surface roughness of the painted surface; and estimating an evaluation value corresponding to a combination of the shape information and the surface roughness information, by using an evaluation model that outputs an evaluation value of the clarity of the painted surface in response to an input including the shape information and the surface roughness information.

Description

Technical Field

[0001] The present invention relates to a painting evaluation apparatus and a painting evaluation method.

Background Art

[0002] There is known an invention in which, among the surface undulations of a paint film, the amplitude of undulations with a wavelength of 1 mm to 10 mm is selectively measured, and the appearance of the paint film surface is evaluated based on the magnitude of this measurement result (Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] According to the invention described in Patent Document 1, when the painted surface is a curved surface, an error may occur in the amplitude of the undulations to be selectively measured, and there is a problem that the accuracy in evaluating the distinctness of image of the painted surface may be reduced.

[0005] The present invention has been made in view of the above problems, and an object thereof is to provide a painting evaluation apparatus and a painting evaluation method capable of accurately evaluating the distinctness of image of a painted surface even when the painted surface is a curved surface.

Means for Solving the Problems

[0006] A painting evaluation apparatus and a painting evaluation method according to an aspect of the present invention acquire shape information representing the curved shape of a painted surface and surface roughness information representing the surface roughness of the painted surface, and estimate an evaluation value corresponding to a combination of the shape information and the surface roughness information using an evaluation model that outputs an evaluation value of the distinctness of image of the painted surface for an input including the shape information and the surface roughness information.

Effects of the Invention

[0007] According to the present invention, even when the painted surface is a curved surface, the distinctness of image of the painted surface can be accurately evaluated.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Embodiments for Carrying Out the Invention

[0009] Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the description of the drawings, the same parts will be denoted by the same reference numerals and the description thereof will be omitted.

[0010] [Configuration of Painting Evaluation Apparatus] With reference to FIG. 1, a configuration example of the painting evaluation apparatus according to the present embodiment will be described. As shown in FIG. 1, the painting evaluation apparatus includes a shape acquisition unit 11, a surface roughness acquisition unit 17, and a controller 100. In addition, the painting evaluation apparatus may include a material acquisition unit 13, an image acquisition unit 21, and an output unit 400. The shape acquisition unit 11, the material acquisition unit 13, the surface roughness acquisition unit 17, the image acquisition unit 21, and the output unit 400 are connected to the controller 100.

[0011] The shape acquisition unit 11 acquires shape information representing the curved shape of the painted surface to be evaluated. More specifically, the shape acquisition unit 11 may acquire the design data of the painted surface as the shape information. For example, CAD (Computer-aided design) data can be cited as the design data. The design data is not limited to this as long as it represents the degree of curvature of the painted surface and the degree of inclination of the painted surface with respect to the surface roughness acquisition unit described later.

[0012] The shape acquisition unit 11 may acquire the stored design data of the painted surface from a database (not shown), or may acquire the design data of the painted surface from an external connection device (not shown) via a wired or wireless network. Alternatively, the shape acquisition unit 11 may acquire the design data based on user input.

[0013] Also, the shape acquisition unit 11 may acquire the measurement data obtained by measuring the painted surface as shape information. For example, a 3D scanner can be cited as the shape acquisition unit 11.

[0014] The material acquisition unit 13 acquires the material information of the painted surface. More specifically, the material acquisition unit 13 acquires information such as the type and color of the member as the material information of the painted surface. The material acquisition unit 13 may acquire the stored material information of the painted surface from a database (not shown), or may acquire the material information of the painted surface from an external connection device (not shown) via a wired or wireless network. Alternatively, the material acquisition unit 13 may acquire the material information based on user input.

[0015] The surface roughness acquisition unit 17 acquires surface roughness information representing the surface roughness of the painted surface. More specifically, the surface roughness acquisition unit 17 may acquire the surface roughness obtained by measuring the painted surface as surface roughness information. For example, a laser microscope can be cited as the surface roughness acquisition unit 17.

[0016] Alternatively, the surface roughness acquisition unit 17 may acquire the stored surface roughness information of the painted surface from a database (not shown), or may acquire the surface roughness information of the painted surface from an external connection device (not shown) via a wired or wireless network. Alternatively, the shape acquisition unit 11 may acquire the surface roughness information based on user input.

[0017] The image acquisition unit 21 acquires a captured image of the painted surface to be evaluated for painting. More specifically, the image acquisition unit 21 is a digital camera equipped with a solid-state imaging device such as a CCD or CMOS, and captures the painted surface to acquire a digital image.

[0018] The image acquisition unit 21 captures the painted surface to be evaluated for painting by setting parameters such as the focal length, the angle of view of the lens, and the vertical and horizontal angles of the camera.

[0019] The controller 100 is a general-purpose computer including a CPU (Central Processing Unit), a memory, a storage device, an input / output unit, and the like.

[0020] A computer program (painting evaluation program) for causing the controller 100 to function as a painting evaluation device is installed in the controller 100. By executing the computer program, the controller 100 functions as a plurality of information processing circuits included in the painting evaluation device.

[0021] Here, an example is shown in which a plurality of information processing circuits included in the painting evaluation device are realized by software. Of course, it is also possible to prepare dedicated hardware for executing each of the following information processes to configure the information processing circuit. Further, the plurality of information processing circuits may be configured by individual hardware.

[0022] The controller 100 includes an evaluation model setting unit 120, an evaluation value estimation unit 130, and a position identification unit 140.

[0023] The evaluation model setting unit 120 sets an evaluation model that outputs an evaluation value of the distinctness of image of the painted surface for an input including shape information and surface roughness information. Here, the evaluation model is a learning model generated by machine learning based on teacher data that is a set of shape information of the evaluated painted surface, surface roughness information of the evaluated painted surface, and an evaluation value of the distinctness of image of the evaluated painted surface.

[0024] Here, the evaluation value of distinctness of image is an index determined by at least one of, for example, the smoothness of the painted surface, the ratio of diffuse reflection in the reflected light on the painted surface, and the resolution of the image reflected on the painted surface. The evaluation value of the distinctness of image of the evaluated painted surface is a numerical value given in advance to the evaluated painted surface by another evaluation method for distinctness of image.

[0025] Note that the evaluation model setting unit 120 may set an evaluation model that outputs an evaluation value of the distinctness of image of the painted surface for an input including material information, shape information, and surface roughness information. In this case, the evaluation model is a learning model generated by machine learning based on teacher data that is a set of the material information of the evaluated painted surface, the shape information of the evaluated painted surface, the surface roughness information of the evaluated painted surface, and the evaluation value of the distinctness of image of the evaluated painted surface.

[0026] Examples of methods for generating a learning model by machine learning include methods using one or a combination of two or more of, for example, neural networks, support vector machines, Random Forest, XGBoost, LightGBM, PLS regression, Ridge regression, and Lasso regression. The method for generating a learning model by machine learning is not limited to the examples given here.

[0027] The evaluation model setting unit 120 may perform machine learning based on teacher data acquired from a database (not shown) to set an evaluation model. Further, an evaluation model may be stored in advance in a database (not shown), and the evaluation model setting unit 120 may set the evaluation model acquired from the database.

[0028] Note that the evaluation model may be configured by a neural network including an input layer and an output layer. More specifically, a neural network typically has an input layer, a plurality of hidden layers, and an output layer, and each layer (input layer, hidden layer, output layer) includes a plurality of neurons.

[0029] The input layer includes shape information representing the curved shape of the painted surface and surface roughness information representing the surface roughness of the painted surface as input data to be processed through each hidden layer. Also, the neurons in the output layer are assigned an evaluation value of the distinctness of image of the painted surface, which is assigned to the input data by the neural network. That is, the output data output from the output layer is the evaluation value of the distinctness of image of the painted surface.

[0030] The neural network constituting the evaluation model is trained to reproduce the teacher data that is a set of shape information representing the curved shape of the painted surface, surface roughness information representing the surface roughness of the painted surface, and the evaluation value of the distinctness of image of the painted surface. That is, when the input data that is a set of shape information and surface roughness information included in the teacher data is input, it is trained by machine learning to output the evaluation value of the distinctness of image of the painted surface as output data.

[0031] In this way, machine learning based on the teacher data is performed to generate an evaluation model that outputs the evaluation value of the distinctness of image of the painted surface for the input including the shape information and the surface roughness information.

[0032] The evaluation value estimation unit 130 estimates the evaluation value corresponding to the combination of the shape information and the surface roughness information using the set evaluation model. More specifically, when the set evaluation model is a learning model generated by machine learning based on the teacher data that is a set of the shape information of the evaluated painted surface, the surface roughness information of the evaluated painted surface, and the evaluation value of the distinctness of image of the evaluated painted surface, the evaluation value estimation unit 130 inputs the shape information and the surface roughness information to the evaluation model. Then, the evaluation value estimation unit 130 uses the value output from the evaluation model as the evaluation value corresponding to the combination of the shape information and the surface roughness information. The evaluation value corresponding to the combination of the shape information and the surface roughness information is the estimated evaluation value regarding the distinctness of image of the painted surface.

[0033] When the set evaluation model is a learning model generated by machine learning based on teacher data that includes material information of the evaluated painted surface, shape information of the evaluated painted surface, surface roughness information of the evaluated painted surface, and an evaluation value of the distinctness of image of the evaluated painted surface as a set, the evaluation value estimation unit 130 inputs the shape information, surface roughness information, and material information into the evaluation model. Then, the evaluation value estimation unit 130 uses the value output from the evaluation model as the evaluation value corresponding to the combination of the shape information, surface roughness information, and material information. The evaluation value corresponding to the combination of the shape information, surface roughness information, and material information is the estimated evaluation value regarding the distinctness of image of the painted surface.

[0034] The position specifying unit 140 associates the captured image and the shape information with the surface roughness information, and records the position on the painted surface where the surface roughness information was obtained. More specifically, the captured image and the shape information when the surface roughness information was obtained are associated with the surface roughness information and recorded in a database (not shown). Thereby, the position on the painted surface where the surface roughness information was obtained is specified.

[0035] The output unit 400 outputs the estimated evaluation value regarding the distinctness of image of the painted surface.

[0036] [Processing Procedure of Painting Evaluation Apparatus] Next, the processing procedure of the painting evaluation apparatus according to the present embodiment will be described with reference to the flowchart of FIG. 2. It is assumed that the evaluation model has already been set by the evaluation model setting unit 120 before the processing shown in the flowchart of FIG. 2 is started.

[0037] In step S102, the surface roughness acquisition unit 17 acquires surface roughness information representing the surface roughness of the painted surface to be evaluated for painting. Also, the shape acquisition unit 11 acquires shape information representing the curved shape of the painted surface. In addition, the material acquisition unit 13 acquires material information of the painted surface. The image acquisition unit 21 acquires a captured image of the painted surface.

[0038] In step S104, the position specifying unit 140 associates the captured image and the shape information with the surface roughness information, and records the position on the painted surface where the surface roughness information was obtained, thereby specifying the position on the painted surface when the surface roughness information was obtained.

[0039] In step S111, the evaluation value estimating unit 130 estimates an evaluation value corresponding to a combination of shape information and surface roughness information, using the set evaluation model.

[0040] In step S113, the output unit 400 outputs the evaluation value estimated by the evaluation value estimation unit .

[0041] [Effects of the embodiment] As described in detail above, the paint evaluation device, paint evaluation method, and paint evaluation program of this embodiment acquire shape information representing the curved shape of the painted surface and surface roughness information representing the surface roughness of the painted surface, and estimate an evaluation value corresponding to the combination of shape information and surface roughness information using an evaluation model that outputs an evaluation value of the sharpness of the painted surface in response to input including shape information and surface roughness information.

[0042] This makes it possible to accurately evaluate the sharpness of a coated surface even when the coated surface is curved. In addition, because the evaluation value is estimated using the evaluation model, no dedicated equipment or skilled personnel are required to evaluate the sharpness of a coated surface, and the cost of evaluating the sharpness of a coated surface can be reduced.

[0043] In the coating evaluation device, coating evaluation method, and coating evaluation program according to the present embodiment, the evaluation model may be a learning model generated by machine learning based on training data that is a set of shape information of the evaluated coated surface, surface roughness information of the evaluated coated surface, and evaluation value of the sharpness of the evaluated coated surface. This makes it possible to accurately evaluate the sharpness of the coated surface using the evaluation model generated from the training data on evaluated coated surfaces having various curved shapes.

[0044] Furthermore, in the coating evaluation device, coating evaluation method, and coating evaluation program according to the present embodiment, the evaluation value of the distinctness of the coated surface may be an index determined by at least one of the smoothness of the coated surface, the proportion of diffuse reflection in the reflected light on the coated surface, and the resolution of the image reflected on the coated surface. In this way, the standard for evaluating the distinctness of the coated surface is clearly indicated.

[0045] In addition, the painting evaluation apparatus, painting evaluation method, and painting evaluation program according to the present embodiment may acquire design data of a painted surface as shape information, or may acquire measurement data obtained by measuring the painted surface as shape information. Thereby, it is possible to accurately evaluate the distinctness of image of the painted surface in consideration of the curved shape of the painted surface.

[0046] Furthermore, the painting evaluation apparatus, painting evaluation method, and painting evaluation program according to the present embodiment may acquire a captured image of a painted surface, associate the captured image and the shape information with surface roughness information, and record the position on the painted surface where the surface roughness information is acquired. Thereby, the position on the painted surface when the surface roughness information is acquired is specified. In addition, the distinctness of image of the painted surface can be accurately evaluated.

[0047] In addition, the painting evaluation apparatus, painting evaluation method, and painting evaluation program according to the present embodiment may acquire material information of a painted surface, and use an evaluation model that outputs an evaluation value of the distinctness of image of the painted surface for an input including the material information, shape information, and surface roughness information to estimate an evaluation value corresponding to a combination of the shape information, surface roughness information, and material information. By using, in addition to the shape information and surface roughness information of the painted surface, the material information of the painted surface for the evaluation of the distinctness of image of the painted surface, the distinctness of image of the painted surface can be accurately evaluated.

[0048] Furthermore, in the painting evaluation apparatus, painting evaluation method, and painting evaluation program according to the present embodiment, the evaluation model may be a learning model generated by machine learning based on teacher data including the material information of the evaluated painted surface, the shape information of the evaluated painted surface, the surface roughness information of the evaluated painted surface, and the evaluation value of the distinctness of image of the evaluated painted surface as a set. Thereby, it is possible to accurately evaluate the distinctness of image of the painted surface by the evaluation model generated from the teacher data regarding the evaluated painted surfaces having various curved surface shapes and various materials.

[0049] Further, the evaluation model used in the painting evaluation apparatus, painting evaluation method, and painting evaluation program according to the present embodiment may be configured by a neural network including an input layer and an output layer. Here, it may be learned by associating input data including shape information representing the curved shape of the painted surface and surface roughness information representing the surface roughness of the painted surface with output data including an evaluation value of the distinctness of image of the painted surface, which is output from the output layer.

[0050] Thereby, the relationship established between the shape information and surface roughness information of the painted surface and the evaluation value of the distinctness of image of the painted surface can be expressed. As a result, it is not necessary to use dedicated equipment or skilled personnel for evaluating the distinctness of image of the painted surface, and the cost for evaluating the distinctness of image of the painted surface can be reduced.

[0051] Furthermore, the evaluation model used in the painting evaluation apparatus, painting evaluation method, and painting evaluation program according to the present embodiment may be generated by acquiring teacher data including shape information representing the curved shape of the painted surface, surface roughness information representing the surface roughness of the painted surface, and an evaluation value of the distinctness of image of the painted surface as a set, performing machine learning based on the teacher data, and configuring to output an evaluation value of the distinctness of image of the painted surface for an input including the shape information and surface roughness information. Thereby, an evaluation model that expresses the relationship established between the shape information and surface roughness information of the painted surface and the evaluation value of the distinctness of image of the painted surface can be obtained.

[0052] Each function shown in the above embodiment can be implemented by one or more processing circuits. The processing circuit includes a programmed processor, an electric circuit, etc., and further includes devices such as application-specific integrated circuits (ASICs) and circuit components arranged to execute the described functions.

[0053] As described above, the content of the present invention has been described in accordance with the embodiment. However, it is obvious to those skilled in the art that the present invention is not limited to these descriptions, and various modifications and improvements are possible. It should not be understood that the discussion and drawings forming part of this disclosure limit the present invention. Various alternative embodiments, examples, and operation techniques will be apparent to those skilled in the art from this disclosure.

[0054] The present invention naturally includes various embodiments and the like not described herein. Therefore, the technical scope of the present invention is defined only by the invention-specific matters according to the legitimate claims derived from the above description.

Explanation of Reference Numerals

[0055] 11 Shape acquisition unit 13 Material acquisition unit 17 Surface roughness acquisition unit 21 Image acquisition unit 100 Controller 120 Evaluation model setting unit 130 Evaluation value estimation unit 140 Position identification unit 400 Output unit

Claims

1. a shape acquisition unit that acquires shape information representing a curved shape of a coating surface; a surface roughness acquiring unit for acquiring surface roughness information representing the surface roughness of the coating surface; A coating evaluation device comprising: The controller: and estimating an evaluation value corresponding to a combination of the shape information and the surface roughness information using an evaluation model that outputs an evaluation value of the sharpness of the painted surface in response to an input including the shape information and the surface roughness information. A coating evaluation device characterized by:

2. The coating evaluation device according to claim 1, The evaluation model is A learning model generated by machine learning based on training data consisting of a set of shape information of an evaluated painted surface, surface roughness information of the evaluated painted surface, and an evaluation value of the sharpness of the evaluated painted surface. A coating evaluation device characterized by:

3. The coating evaluation device according to claim 1 or 2, The evaluation value of the distinctness of the coated surface is an index determined by at least one of the smoothness of the coated surface, the proportion of diffuse reflection in the reflected light on the coated surface, and the resolution of the image reflected on the coated surface. A coating evaluation device characterized by:

4. The coating evaluation device according to any one of claims 1 to 3, The shape acquisition unit acquires design data of the painted surface as the shape information. A coating evaluation device characterized by:

5. The coating evaluation device according to any one of claims 1 to 4, The shape acquisition unit acquires measurement data obtained by measuring the painted surface as the shape information. A coating evaluation device characterized by:

6. The coating evaluation device according to any one of claims 1 to 5, An image acquisition unit for acquiring an image of the painted surface, The controller: Linking the captured image and the shape information to the surface roughness information, and recording the position on the painted surface where the surface roughness information was acquired. A coating evaluation device characterized by:

7. The coating evaluation device according to any one of claims 1 to 6, A material acquisition unit that acquires material information of the painted surface, The controller: Using an evaluation model that outputs an evaluation value of the sharpness of the painted surface in response to an input including the material information, the shape information, and the surface roughness information, an evaluation value corresponding to a combination of the shape information, the surface roughness information, and the material information is estimated. A coating evaluation device characterized by:

8. The coating evaluation device according to claim 7, The evaluation model is a learning model generated by machine learning based on training data consisting of a set of material information of the evaluated painted surface, shape information of the evaluated painted surface, surface roughness information of the evaluated painted surface, and evaluation value of the sharpness of the evaluated painted surface; A coating evaluation device characterized by:

9. Acquire shape information representing the curved shape of the painted surface; Obtaining surface roughness information representing the surface roughness of the painted surface; and estimating an evaluation value corresponding to a combination of the shape information and the surface roughness information using an evaluation model that outputs an evaluation value of the sharpness of the painted surface in response to an input including the shape information and the surface roughness information. A coating evaluation method comprising the steps of:

10. a shape acquisition unit that acquires shape information representing a curved shape of a coating surface; a surface roughness acquiring unit for acquiring surface roughness information representing the surface roughness of the coating surface; The computer that controls acquiring the shape information using the shape acquisition unit; acquiring the surface roughness information using the surface roughness acquisition unit; a step of estimating an evaluation value corresponding to a combination of the shape information and the surface roughness information using an evaluation model that outputs an evaluation value of the sharpness of the painted surface in response to an input including the shape information and the surface roughness information; A coating evaluation program to carry out the above.

11. An evaluation model configured by a neural network including an input layer and an output layer, input data input to the input layer, the input data including shape information representing a curved shape of a coating surface and surface roughness information representing a surface roughness of the coating surface; output data output from the output layer, the output data including an evaluation value of the sharpness of the coating surface; The learning process involved associating the An evaluation model characterized by:

12. acquiring training data including a set of shape information representing a curved shape of a coating surface, surface roughness information representing a surface roughness of the coating surface, and an evaluation value of a sharpness of the coating surface; performing machine learning based on the training data to generate an evaluation model that outputs an evaluation value of the sharpness of the painted surface in response to an input including the shape information and the surface roughness information; The evaluation model generating method is characterized by the above.

Images