Image testing device

Abstract

Provided is an image testing device capable of easily performing an image test with high accuracy. An image testing device 100 comprises: an input means for inputting test data Dt for displaying a test image Pt on a vehicle display device 200; a capturing means for capturing the test image Pt displayed on the vehicle display device 200 in response to the input of the test data Dt, and storing the captured test image Pt as a captured image Pc; an expected value image storing means for storing a predetermined expected value image Pe; and a determining means for determining a test result R for the captured image Pc on the basis of a difference between the expected value image Pe and the captured image Pc. The test result R is expressed in a plurality of levels. The determining means includes a trained model that has been subjected to machine learning using teacher data so as to output the test result R when the difference is input.

Description

Image Test Equipment

[0001] The present disclosure relates to an imaging test device.

[0002] For example, Patent Document 1 describes a vehicle display device that displays images on a display configured with a TFT (Thin Film Transistor) liquid crystal display, an organic EL (Electro-Luminescence) display, or the like.

[0003] JP 2023-104941 A

[0004] In this type of vehicle display device, when testing whether the displayed image is the desired image before shipping, it is common to rely on manual work. For example, a procedure is performed in which an operator inputs test data to display a test image on the vehicle display device, and then visually checks whether the test image displayed in accordance with the input is in the desired state. This procedure is not only cumbersome, but also likely to result in inconsistent test results.

[0005] The present disclosure has been made in consideration of the above-described circumstances, and aims to provide an image testing device that can easily perform image testing with high accuracy.

[0006] In order to achieve the above object, the image test device of the present disclosure comprises: an input means for inputting test data for displaying a test image on a vehicle display device; a capture means for capturing the test image displayed on the vehicle display device in response to the input of the test data and storing it as a captured image; an expected image storage means for storing a predetermined expected image; and a judgment means for judging a test result of the captured image based on the difference between the expected image stored in the expected image storage means and the captured image stored in the capture means, wherein the test result is expressed in multiple stages, and the judgment means has a trained model that has been subjected to machine learning using training data so as to output the test result when the difference is input.

[0007] According to the present disclosure, it is possible to easily perform high-precision image testing.

[0008] 1 is a block diagram of an image test system according to an embodiment of the present invention;

[0009] An embodiment of the present disclosure will be described with reference to the drawings.

[0010] As shown in FIG. 1, the image test system according to this embodiment includes an image test device 100 and a vehicle display device 200.

[0011] The vehicle display device 200 is mounted on a vehicle and is, for example, a combination meter that displays various meters using images. The vehicle display device 200 includes a display unit 210, a display control unit 220, and a function control unit 230.

[0012] (Vehicle Display Device 200) The display unit 210 is configured with a TFT liquid crystal display, an organic EL display, or the like, and has an area capable of displaying an image (hereinafter, referred to as the display area). The display unit 210 displays an image related to the vehicle across the entire display area. The display control unit 220 controls the display of the display unit 210, and the function control unit 230 controls the overall operation of the vehicle display device 200. For example, the function control unit 230 is configured with a microcontroller, communicates with various electronic components such as the vehicle's ECU (Electronic Control Unit) via a CAN (Controller Area Network), and monitors various vehicle conditions. For example, the display control unit 220 is configured with a GPU (Graphics Processing Unit), and generates an image to be displayed on the display unit 210 based on a control signal from the function control unit 230 and image data for drawing stored in a memory (not shown), and outputs a drive signal to the display unit 210. The display control unit 220 and the function control unit 230 may be realized as a function of a control device configured from one or more IC (Integrated Circuit) chips, and their configurations are arbitrary.

[0013] (Image test device 100) The image test device 100 includes a test data input unit 110, a capture unit 120, an expected image storage unit 130, a judgment unit 140, and a judgment result notification unit 150. Of these, the test data input unit 110, the capture unit 120, the expected image storage unit 130, and the judgment unit 140 are each realized by a function of a controller configured with one or more IC chips. In the following description, "storing or saving an image" refers to storing or saving data representing the image. This saving or storage function is realized as one function of a well-known storage device.

[0014] The test data input unit 110 inputs test data Dt for displaying a test image on the vehicle display device 200. The test data input unit 110 automatically inputs the test data Dt to the display control unit 220 of the vehicle display device 200 by interprocess communication (IPC) in accordance with a test case 1 that has been created and prepared in advance. The test case 1 is composed of a combination of input data (test data Dt) for switching the image to be displayed on the display unit 210.

[0015] The capture unit 120 includes an image acquisition unit 121 , an image processing unit 122 , and a captured image storage unit 123 .

[0016] The image acquisition unit 121 receives a display signal St indicating the content of an image (test image Pt) displayed on the display unit 210, and acquires the test image Pt (saves the data of the test image Pt). As shown in FIG. 2, the test image Pt is a test object screen displayed across the entire display area 210a of the display unit 210, and displays information related to a vehicle (hereinafter, vehicle information). Depending on the various test objects, the test image Pt may include one or more pieces of vehicle information. For example, in FIG. 2, the test image Pt displays vehicle speed information, but other vehicle information (not shown) may be displayed in areas of the test image Pt other than the area displaying the vehicle speed information.

[0017] The image processing unit 122 generates a capture image Pc by clipping a set area C included in the test image Pt acquired by the image processing unit 121. The set area C is a partial area of ​​the test image Pt displayed across the display area 210a, and is set in advance as an area for narrowing down the display parts necessary for the test.

[0018] The capture image storage unit 123 stores the capture image Pc generated by the image processing unit 122. When testing the entire area of ​​the test image Pt, clipping does not need to be performed by the image processing unit 122. In this case, the capture image storage unit 123 stores the image that has not been clipped by the image processing unit 122, that is, the test image Pt acquired by the image processing unit 121, as the capture image Pc as is.

[0019] The expected image storage unit 130 stores the expected image Pe. The expected image Pe is stored in advance in the expected image storage unit 130 as an image to be displayed on the display unit 210 of the vehicle display device 200. The expected image Pe is an image generated in advance using computer graphics, i.e., CGI (Computer Generated Imagery). The expected image Pe shown in FIG. 2 is an example of a setting area image corresponding to the setting area C as a display part required for the test. As described above, when testing the entire test image Pt, the expected image Pe may be prepared in advance as an image to be displayed across the entire display area 210a of the display unit 210. In other words, the expected image Pe may or may not include a setting area image corresponding to the setting area C, depending on the test target area.

[0020] The determination unit 140 includes a comparison unit 141 and a machine learning model 142 .

[0021] The comparison unit 141 compares the captured image Pc stored in the captured image storage unit 123 with the expected image Pe stored in the expected image storage unit 130 to generate a difference image ΔP that indicates the difference between the two images. The example shown in FIG. 2 shows a case where the vehicle speed display in the captured image Pc is underlined, while the vehicle speed display in the desired expected image Pe is not underlined. In other words, this example shows a case where a design change has occurred but the change is not reflected in the test image Pt. In this case, the difference image ΔP shows the difference between the two images, i.e., the underline as the design change.

[0022] The machine learning model 142 is a trained neural network that has undergone machine learning. Machine learning using training data is applied to the machine learning model 142 so that the model outputs a test result R for the captured image Pc when the difference is input from the comparison unit 141 (i.e., when the difference image ΔP is input). The training data includes the captured image Pc, the expected image Pe, and the test result R. The test result R is expressed in multiple levels, for example, two levels, "OK" (correct) and "NG" (incorrect), in this embodiment. Note that, in the annotation stage where training data is created in advance, the test result R is set to "NG" if (i) a visually recognizable specification change (such as a change in font, icon, color, or coordinates) occurs between the expected image Pe and the captured image Pc, and the test result R is set to "OK" if (ii) a visually unrecognizable or difficult-to-recognize change occurs between the expected image Pe and the captured image Pc. That is, the difference image ΔP shown in FIG. 2 illustrates an example in which the test result R is "NG."

[0023] In the above (ii), the test result R is judged to be "OK" if, for example, the filtering effect differs due to differences in the drawing processing methods between the expected image Pe created as CGI and the captured image Pc created based on the display signal St (such as when the degree of blurring of the edges of the characters or designs included in the two images differs). In other words, the test result R is judged to be "OK" not only when there is no difference between the expected image Pe and the captured image Pc, but also when a slight difference occurs due to differences in the drawing processing methods.

[0024] Any machine learning algorithm may be used to create the machine learning model 142. The test result R is not limited to two levels of correct or incorrect, but may be three or more levels including a "needs confirmation" level.

[0025] The determination result notification unit 150 notifies the worker of the test result R output by the machine learning model 142. The determination result notification unit 150 is composed of a display, a speaker, etc., and may notify the test result R by at least one of an image and a sound.

[0026] (1) The image test device 100 described above includes: a test data input unit 110 (an example of an input means) that inputs test data Dt for displaying a test image Pt on the vehicle display device 200; a capture unit 120 (an example of a capture means) that captures the test image Pt displayed on the vehicle display device 200 in response to the input of the test data Dt and stores it as a captured image Pc; an expected image storage unit 130 (an example of an expected image storage means) that stores a predetermined expected image Pe; and a judgment unit 140 (an example of a judgment means) that judges a test result R of the captured image Pc based on the difference (difference image ΔP) between the expected image Pe stored in the expected image storage unit 130 and the captured image Pc stored in the capture unit 120, where the test result R is expressed in multiple stages, and the judgment unit 140 has a machine learning model 142 (a trained model) that has been subjected to machine learning using teacher data so as to output the test result R when the difference is input.

[0027] According to the configuration (1) above, the test data Dt can be automatically input to the vehicle display device 200, and once teacher data is created, the test result R is output accordingly, making it possible to easily test images. Furthermore, according to the configuration (1) above, the accuracy of the test result R can be improved by machine learning, making it possible to test images with high accuracy.

[0028] (2) In the image test device 100, the capture unit 120 may clip a set area C included in the test image Pt and store it as a captured image Pc, and the expected image Pe may include a set area image corresponding to the set area.

[0029] According to the above configuration (2), the display objects can be narrowed down to only the display parts necessary for the test when performing image comparison, so that it is possible to prevent unnecessary differences from occurring during comparison.

[0030] The present invention is not limited to the above-described embodiments and drawings, and modifications (including the deletion of components) can be made as appropriate within the scope of the present invention.

[0031] In the above description, in order to facilitate understanding of the present disclosure, descriptions of well-known technical matters have been omitted as appropriate.

[0032] This invention allows various embodiments and modifications without departing from the broad spirit and scope of this invention. Furthermore, the above-described embodiments are intended to explain this invention and do not limit the scope of this invention. That is, the scope of this invention is defined by the claims, not the embodiments. Various modifications made within the scope of the claims and the meaning of the invention equivalent thereto are considered to be within the scope of this invention.

[0033] DESCRIPTION OF SYMBOLS 100...Image test device 110...Test data input unit (an example of an input means) 120...Capture unit (an example of a capture means) 121...Image acquisition unit 122...Image processing unit 123...Captured image storage unit 130...Expected value image storage unit (an example of an expected value image acquisition means) 140...Determination unit (an example of a determination means) 141...Comparator 142...Machine learning model 150...Determination result notification unit 200...Vehicle display device 210...Display unit, 210a...Display area 220...Display control unit 230...Function control unit 1...Test case Dt...Test data, St...Display signal, Pt...Test image Pc...Captured image, C...Setting area Pe...Expected value image ΔP...Difference image R...Test result

Claims

1. An image test device comprising: an input means for inputting test data for displaying a test image on a vehicle display device; a capture means for capturing the test image displayed on the vehicle display device in response to the input of the test data and storing it as a captured image; an expected image storage means for storing a predetermined expected image; and a judgment means for judging a test result of the captured image based on a difference between the expected image stored in the expected image storage means and the captured image stored in the capture means, wherein the test result is expressed in a plurality of stages, and the judgment means has a trained model that has been subjected to machine learning using training data so as to output the test result when the difference is input.

2. The image test device according to claim 1, wherein said capture means clips a set area included in said test image and stores the clipped image as said captured image, and said expected image includes a set area image corresponding to said set area.

Images