Test specimen display recognition device, test specimen testing system, test specimen display setting device, test specimen display recognition method, and test specimen display recognition program

Abstract

The present invention provides a test specimen display recognition device for identifying a test specimen display for indicating a state of a test specimen, from an image obtained by imaging a meter panel of the test specimen by means of a camera, the test specimen display recognition device comprising: a setting information accepting unit for accepting a selection signal indicating the type of an image recognition algorithm for identifying the test specimen display from the image; a setting information storage unit for storing the selection signal accepted by the setting information accepting unit as setting information; and a test sample display recognition unit for acquiring the setting information from the setting information storage unit and identifying an instructed value indicated by test specimen display on the basis of the setting information.

Description

Specimen marking recognition device, specimen testing system, specimen marking setting device, specimen marking recognition method, and specimen marking recognition program

[0001] The present invention relates to a specimen marking recognition device, a specimen testing system, a specimen marking setting device, a specimen marking recognition method, and a specimen marking recognition program.

[0002] An example of this type of specimen marking recognition device is one that identifies specimen marks that indicate the state of a specimen based on an image of a vehicle or a part thereof, that is, a meter panel of a specimen, as disclosed in Patent Document 1. This specimen marking recognition device identifies specimen marks such as meters, warning lights, or indicator lights on the meter panel based on an image of the specimen's meter panel and known data that has been stored in advance in a data storage unit.

[0003] WO2020 / 111192

[0004] The above-mentioned specimen display recognition device takes advantage of the fact that all specimen displays on the meter panel are lit when the specimen is started, etc., and identifies the position or type of the specimen display using an image of the meter panel with all specimen displays lit (hereinafter referred to as the fully lit image).

[0005] However, the above-mentioned test specimen marking recognition device identifies all test specimen marks at once using a fully lit image, and requires the preparation of a template image of a fully lit image according to the type of test specimen. However, if a template image does not exist, the test specimen marking cannot be identified. Furthermore, simply comparing the template image with the fully lit image makes it difficult to identify test specimen marks that change depending on the state of the test specimen, such as pointers, gauges, or letters. Furthermore, if an image in which some test specimen marks are not lit is used, it is difficult to identify the position or type of the unlit test specimen marking.

[0006] Therefore, the present invention has been made in consideration of the above-mentioned problems, and its main objective is to easily set an image recognition algorithm according to the specimen display to be recognized, and to accurately identify the indicated value shown by the specimen display on the meter panel.

[0007] In other words, the specimen marking recognition device of the present invention is a specimen marking recognition device that identifies a specimen marking that indicates the state of a specimen from an image of the meter panel of a vehicle or a part thereof captured by a camera, and is characterized by comprising: a setting information receiving unit that receives a selection signal that indicates the type of image recognition algorithm for identifying the type of specimen marking from the image; a setting information storage unit that stores the selection signal received by the setting information receiving unit as setting information; and a specimen marking recognition unit that identifies the indication value indicated by the specimen marking based on the setting information stored in the setting information storage unit.

[0008] With this type of test specimen marking recognition device, the user can select an appropriate image recognition algorithm that matches each test specimen marking on the meter panel. The test specimen marking recognition unit then uses the selected image recognition algorithm to identify the test specimen marking, allowing each test specimen marking on the meter panel to be easily and accurately identified. Furthermore, since the present invention individually identifies each test specimen marking on the meter panel, it is not necessary to use an image of the meter panel with all test specimen marks lit, as in the past, and this prevents erroneous recognition or missed identification of test specimen marks.

[0009] It is desirable that the setting information receiving unit further receives change information, which is information for identifying a change in the state of the specimen display, and stores it in the setting information storage unit, and the specimen display recognition unit acquires the change information from the setting information storage unit and identifies the indication value of the specimen display based on the image recognition algorithm and the change information. With this configuration, the indication value indicated by the specimen display is identified using the change information indicating the change in the state of the specimen display, making it possible to identify the indication value more easily and accurately.

[0010] It is desirable that the setting information receiving unit receives, as the change information, information indicating a change in the state of a bar meter, a circle meter, character information, and / or graphic information, and the specimen display recognition unit acquires the setting information from the setting information storage unit and identifies the indicated value indicated by the specimen display based on the image recognition algorithm and the change information. With this configuration, the specimen display recognition unit can more easily and accurately identify the indicated value of the bar meter, the indicated value of the circle meter, the character information, and / or the graphic information.

[0011] It is desirable that the setting information storage unit receives and stores a range designation signal indicating a range of the image designated by a user, and the specimen marking recognition unit uses the image recognition algorithm to identify the specimen marking included in the range. With this configuration, the image recognition algorithm is used within the designated image range, making it possible to more accurately identify the specimen marking than when the image recognition algorithm is used within the entire image range.

[0012] Specific examples of the display on the specimen include at least one of a circle meter, a bar meter, letters, and figures.

[0013] The setting information receiving unit receives angle-of-view correction information for correcting the angle of view of the camera, and may further include an angle-of-view correction storage unit that stores the angle-of-view correction information received by the setting information receiving unit, and the image correction unit that corrects the angle of view based on the angle-of-view correction information stored in the angle-of-view correction storage unit. With this configuration, when changing cameras or testing multiple specimens of the same type, setting for identifying the indicated value indicated by the specimen display is performed once, and thereafter, the indicated value indicated by the specimen display can be identified simply by adjusting the angle of view of the camera or the distortion, size, position, or angle of the image, so that setting for identifying the indicated value indicated by the specimen display can be omitted.

[0014] When adjusting the angle of view, there are cases where the angle of view changes significantly, for example, when a camera is replaced. On the other hand, there are cases where the angle of view changes intermittently and slightly, for example, when the specimen vibrates while traveling. In order to accurately correct the angle of view, it is desirable to correct the angle of view in response to each of these cases. Therefore, it is desirable that the angle of view correction information includes at least first angle of view correction information, which is one of the angle of view correction information, and second angle of view correction information, which is angle of view correction information with a smaller correction amount than the first angle of view correction information. It is desirable that the image correction unit corrects the angle of view based on the first angle of view correction information immediately after the specimen marking recognition unit starts to identify the specimen marking, and corrects the angle of view based on the second angle of view correction information while the specimen marking recognition unit is identifying the specimen marking.

[0015] It is preferable that the device further comprises a mask processing unit that performs a masking process to mask a predetermined area selected by a user from the captured image of the meter panel.

[0016] In this way, when the specimen display recognition unit identifies the indication value of the specimen display, the unit identifies the indication value indicated by the specimen display in an area other than the masked specified area, thereby enabling more accurate identification of the indication value indicated by the specimen display.

[0017] Furthermore, a specimen testing system according to the present invention is a specimen testing system for testing a specimen that is a vehicle or a part thereof, and is characterized by comprising: test equipment for testing the specimen, a camera for capturing an image of the meter panel of the specimen, the specimen marking recognition device, and a control unit for controlling the test equipment based on the specimen marking identified by the specimen marking recognition device. Such a specimen testing system can achieve the same effects as the above-mentioned specimen marking recognition device.

[0018] In addition, a specimen marking recognition method according to the present invention is a method for identifying a specimen marking indicating the state of a specimen from an image of a vehicle or a part thereof, the method comprising the steps of: receiving a selection signal indicating a type of image recognition algorithm for identifying the type of specimen marking from the image; storing setting information including the received selection signal; and identifying an indication value indicated by the specimen marking based on the setting information and the image recognition algorithm. Such a specimen marking recognition method can achieve the same effects as the above-described specimen marking recognition device.

[0019] Furthermore, a specimen marking recognition program according to the present invention is a specimen marking recognition program used in a specimen marking recognition device that identifies a specimen marking indicating the state of a specimen from an image of a vehicle or a part thereof, a meter panel of the specimen, and is characterized in that it causes a computer to function as a setting information receiving unit that receives a selection signal indicating the type of image recognition algorithm for identifying the type of specimen marking from the image, a setting information storage unit that stores the setting information received by the setting information receiving unit, and a specimen marking recognition unit that identifies the indicated value indicated by the specimen marking based on the setting information stored in the setting information storage unit. Such a specimen marking recognition program can achieve the same effects as the above-mentioned specimen marking recognition device.

[0020] According to the present invention configured in this manner, it is possible to easily set an image recognition algorithm according to the specimen display to be recognized, and to accurately identify the indicated value shown by the specimen display on the meter panel.

[0021] FIG. 1 is a schematic diagram showing a specimen testing system according to one embodiment of the present invention; FIG. 2 is a functional block diagram of a specimen marking recognition device according to the same embodiment; FIG. 3 is a schematic diagram of an image output to a display by the specimen marking recognition device according to the same embodiment; (a) a flowchart showing a method for setting setting information, and (b) a flowchart showing a method for recognizing a specimen mark according to the same embodiment; FIG. 4 is a functional block diagram of a specimen marking recognition device according to another embodiment; (b) a schematic diagram of an image output to a display by the specimen marking recognition device according to another embodiment when setting pattern matching according to another embodiment; (c) a schematic diagram of an image output to a display by the specimen marking recognition device according to another embodiment when setting meter identification according to another embodiment; (d) a schematic diagram of an image output to a display by the specimen marking recognition device according to another embodiment.

[0022] An embodiment of a specimen testing system using a specimen marking recognition device according to the present invention will be described below with reference to the drawings. Note that in all of the drawings shown below, some parts may be omitted or exaggerated as needed for clarity. Identical components will be assigned the same reference numerals and their descriptions will be omitted as needed.

[0023] <Device Configuration> The specimen testing system 100 according to this embodiment tests a specimen W, which is a vehicle or a part thereof.

[0024] Here, the specimen W may be a vehicle or a part thereof, such as a vehicle equipped with an internal combustion engine, a hybrid vehicle, a plug-in hybrid vehicle, an electric vehicle, or a fuel cell vehicle. The specimen W may also be an ADAS vehicle or a part thereof having an advanced driver assistance system (ADAS) such as adaptive cruise control (ACC), or a vehicle (AD vehicle) or a part thereof having an automated driving system (AD). The vehicle is not limited to a four-wheeled vehicle, but may also be a two-wheeled vehicle or other vehicle. The specimen W may also be a ship, an aircraft, or the like.

[0025] Specifically, as shown in Figure 1, the specimen testing system 100 comprises a test facility 10 for testing a specimen W, a control device 20 for controlling the test facility 10, a specimen display recognition device 30 for identifying a specimen display H indicating the state of the specimen W, and a camera 4.

[0026] The following describes each of the devices 10 to 30. The test facility 10 includes a dynamometer 11 for simulating driving the test specimen W, and an autonomous driving robot 12 for operating the brakes, accelerator, steering, buttons, or shift of the test specimen W.

[0027] The dynamometer 11 is a so-called chassis dynamometer, and includes rollers 111 on which the wheels of the test specimen W are placed, and a load device 112 connected to the rollers 111. Note that the dynamometer 11 is not limited to a chassis dynamometer, and may be a hub dynamometer connected to the rotating shaft of the wheels of the test specimen W. The dynamometer 11 may also be attached to either the driving wheel or the driven wheel of the test specimen. Furthermore, the dynamometer 11 may be an engine dynamometer connected to the rotating shaft of the engine of the test specimen W, a motor dynamometer connected to the rotating shaft of the motor of the test specimen W, or a powertrain dynamometer for evaluating the powertrain.

[0028] The autonomous driving robot 12 is an operating actuator that operates the brake pedal, accelerator pedal, steering wheel, buttons, or shift of the test specimen W. For example, if the test specimen W is equipped with an advanced driver assistance system (AD / ADAS), the autonomous driving robot 12 may link an active driving state in which the test specimen W actively drives using the autonomous driving system or the advanced driver assistance system (AD / ADAS) with a passive driving state in which the test specimen W passively drives using the brake, accelerator, or steering operation.

[0029] The control device 20 controls the test equipment 10 so that the test specimen W can travel in a predetermined travel pattern. Specifically, the control device 20 is a computer equipped with a CPU, internal memory, an input / output interface, an A / D converter, etc. The control device 20 includes a load control unit 21 that controls the dynamometer 11, a robot control unit 22 that controls the autonomous driving robot 12, and a host control unit 23 that controls these respective control units in an integrated manner. These control units may be included in a common computer, or may be included in multiple computers each having their own function.

[0030] The specimen display recognition device 30 identifies one or more specimen displays H indicating the state of the specimen W from an image captured of the meter panel M of the specimen W (including other displays such as a display provided on the specimen W).

[0031] In this embodiment, the specimen display recognition device 30 is a computer equipped with a CPU, internal memory, an input / output interface, an A / D converter, etc. Based on a program stored in the internal memory, the specimen display recognition device 30 performs the functions of a setting information receiving unit R, an image acquisition unit 31, a setting information storage unit 32, a specimen display recognition unit 33, a primary storage unit 34, a data transmission unit 35, and a result storage unit 36, as shown in FIG. 2 . Note that the setting information receiving unit R and the setting information storage unit 32 may be provided separately from the specimen display recognition device 30. In this case, a device equipped with the setting information receiving unit R and the setting information storage unit 32 is referred to as a specimen display setting device.

[0032] Here, the specimen display H is an indicator displayed on the meter panel M, which indicates the state of the specimen W. Specifically, the specimen display H is at least one of the following: - A circle meter indicating a scale arranged circumferentially (for example, a speedometer or engine tachometer, etc.); - A bar meter indicating a scale arranged in a straight or curved line (for example, a battery level gauge or a fuel level gauge, etc.); - Characters representing the state of the specimen W (for example, characters indicating the driving mode of the specimen W, the total distance traveled, the set temperature of the air conditioner, or the outside temperature, etc.); - A graphic representing the state of the specimen W (for example, the distance between the specimen W and the vehicle ahead, whether the AD / ADAS function is enabled or disabled, whether it is in standby mode, setting a target distance between vehicles, setting a target vehicle speed, a brake press recommendation, or a steering wheel operation recommendation, etc.). - Warning lights indicating abnormalities in the test specimen W (for example, water temperature abnormality warning light, gasoline remaining warning light, battery warning light, AT fluid warning light, oil pressure warning light, fuel filter warning light, ABS warning light, hybrid system warning light, engine abnormality warning light, exhaust temperature abnormality warning light, brake warning light, airbag warning light, seat belt warning light, anti-skid device operation indicator light, door ajar warning light, or washer fluid level warning light, etc.) - Indicator lights indicating the status of the test specimen W (for example, turn signal indicator light, headlamp upward indicator light, tail lamp indicator light, fog lamp indicator light, high beam assist indicator light, VDC OFF indicator light, security indicator, low water temperature indicator light, go ahead indicator light, EV indicator light, or vehicle approach warning sound pause indicator light, etc.).

[0033] The image acquisition unit 31 acquires an image captured or recorded by the camera 4. In this embodiment, the image acquisition unit 31 acquires an image of the meter panel M captured by the camera 4. The image acquired by the image acquisition unit 31 can be displayed on a display unit such as a display D.

[0034] 3, the display D is provided with an image selection area D1 for selecting an image and / or a display area D2 in which the image selected in the image selection area D1 is displayed. The image selection area D1 is provided with an image source selection area D1a for selecting an image source indicating the image capture medium and / or an image file selection area D1b indicating the storage location of the image file. In this embodiment, when the user selects the camera 4 as the image source from the image source selection area D1a, an image of the meter panel captured by the camera 4 is displayed in the display area D2. The user may also select a recorded image from the image source selection area D1a and select the storage location of the recorded image in the image file selection area D1b.

[0035] The setting information receiving unit R receives information (hereinafter, "setting information") that is set to identify the specimen indication H. Here, the setting information includes, but is not limited to, for example, a parameter name, a type of image recognition algorithm, an area to be analyzed, a threshold value, and / or a transmission order. The setting information received by the setting information receiving unit R is then stored in the setting information storage unit 32.

[0036] The setting information receiving unit R functions as an image recognition algorithm receiving unit R1 that receives a selection signal, which is a signal indicating an image recognition algorithm selected by a user from multiple image recognition algorithms. In this embodiment, an algorithm selection area D3 is provided on the display D, where an arbitrary image recognition algorithm can be selected from multiple image recognition algorithms. When the user selects an image recognition algorithm from the algorithm selection area D3 via input means such as a touch panel, a mouse, or a keyboard, a selection signal is output based on the selected image recognition algorithm. The image recognition algorithm receiving unit R1 receives the output selection signal and stores the selection signal in the setting information storage unit 32.

[0037] In this embodiment, the image recognition algorithm refers to an algorithm for identifying the specimen marking H from the image of the meter panel M. More specifically, the image recognition algorithm identifies the type of specimen marking H from the image of the meter panel M. Furthermore, the multiple image recognition algorithms can include, for example, at least one of an algorithm for identifying a circle meter, an algorithm for identifying a bar meter, an algorithm for identifying characters, or pattern matching.

[0038] Specific examples of image recognition algorithms are shown below. (1) Circle meter: An algorithm for identifying a circle meter indicating a scale arranged in a circular shape (in FIG. 3, this algorithm "Circle meter" is selected). (2) Bar meter: An algorithm for identifying a bar meter indicating a scale arranged in a straight or curved shape. (3) Character recognition: An algorithm such as optical character recognition (OCR) for recognizing characters from an image of the meter panel M. (4) Pattern matching: An algorithm such as pattern matching for identifying a graphic, warning light, or indicator light from an image of the meter panel M.

[0039] Furthermore, the setting information receiving unit R functions as a change information receiving unit R2 that receives change information, which is information for identifying a change in the state of the specimen display H. The change information receiving unit R2 receives the change information and stores the change information in the setting information storage unit 32.

[0040] Specifically, the change information is input by the user via an input device into the editing area D4, and is accepted by the change information accepting unit R2. Examples of the change information include the following: (1) When the specimen display H is a circle meter, the change information is information for identifying a change in the state of the circle meter. Examples of the change information include at least one of the center of the circle meter, a scale such as a maximum or minimum value, the position of the scale such as a maximum or minimum value, a value for scaling, and a threshold value for edge or line segment extraction. (2) When the specimen display H is a bar meter, the change information is information for identifying a change in the state of the bar meter. Examples of the change information include at least one of the shape of the bar meter (e.g., linear or curved), a scale such as a maximum or minimum value, the position of the scale such as a maximum or minimum value, a value for scaling, and a threshold value. (3) When the specimen display H is a character, the change information can include at least one of character information such as alphabets, numbers, regions, or maximum or minimum values ​​for determining abnormal values, or information for identifying a change in the state of such character information (including at least a change in state when characters are displayed and when characters are not displayed). (4) When the specimen display H is a graphic, the change information can include at least one of graphic information such as an area, registration of a template image, a threshold for determining similarity, an output value based on similarity, or a selection between grayscale and color scale, or information for identifying a change in the state of the graphic information (including at least a change in state when a graphic is displayed and when it is not displayed).

[0041] Furthermore, the setting information receiving unit R functions as a range designation signal receiving unit R3 that receives a range designation signal indicating a range of an image designated by the user. The range designation signal receiving unit R3 receives the range designation signal and stores the range designation signal in the setting information storage unit 32. In this embodiment, the user can designate a range designation area D5 on the display D via an input device, which designates a range of an image to which a selected image recognition algorithm is to be applied. When the user designates the range designation area D5, the range designation signal receiving unit R3 receives a range designation signal indicating that the range designation area D5 has been designated. Note that the user can modify the range designation area D5 even after the range designation signal receiving unit R3 receives the range designation signal. In this case, the setting information storage unit 32 stores the range designation signal based on the modified range designation area D5.

[0042] Furthermore, the setting information receiving unit R may function as a parameter name receiving unit R4 that receives the name of a parameter to be analyzed. In this embodiment, a parameter designation area D6 in which the user designates a parameter to be analyzed via an input means is displayed on the display D. When the user inputs a parameter to be analyzed in the parameter designation area D6, the parameter name receiving unit R4 accepts the parameter to be analyzed and stores it in the setting information storage unit 32. At this time, the parameter to be analyzed is associated with an image recognition algorithm, a range designation signal, and / or change information and stored in the setting information storage unit 32 as setting information.

[0043] The specimen display recognition unit 33 identifies the indicated value indicated by the specimen display H based on the setting information stored in the setting information storage unit 32. In this embodiment, the specimen display recognition unit 33 acquires the setting information stored in the setting information storage unit 32, and identifies the indicated value indicated by the specimen display H on the meter panel M based on the setting information and an image recognition algorithm corresponding to the setting information.

[0044] Specifically, the specimen marking recognition unit 33 identifies the specimen marking H using one or more image recognition algorithms, as described below.

[0045] (1) The DUT display recognition unit 33 uses the image recognition algorithm of the circle meter to detect and line-divide edges passing near the center of the range specification area D5 and calculate their angles. The unit then converts the detected values ​​into values ​​based on the preset circle meter scale. In this embodiment, the DUT display recognition unit 33 identifies the positions of the maximum and minimum values ​​of the circle meter and the position or angle of the circle meter needle between the maximum and minimum values, thereby identifying the value indicated by the circle meter needle (indicated value) from the preset circle meter scale and scaling value. Note that the DUT display recognition unit 33 may set additional scales or values ​​between the maximum and minimum values ​​to improve accuracy.

[0046] (2) The specimen display recognition unit 33 uses an image recognition algorithm for the bar meter to identify the position or indicated scale or value of the bar meter in the range specification region D5. In this embodiment, the specimen display recognition unit 33 identifies the positions of the maximum and minimum values ​​of the bar meter and identifies the position of the bar of the bar meter between the maximum and minimum values, thereby identifying the scale or value indicated by the bar of the bar meter. The specimen display recognition unit 33 may also identify additional scales or values ​​located between the maximum and minimum values ​​to improve accuracy.

[0047] (3) When the image recognition algorithm is character reading, the specimen display recognition unit 33 identifies the characters (indication values) included in the range specification area D5 using, for example, OCR.

[0048] (4) When the image recognition algorithm is graphic recognition, the specimen display recognition unit 33 uses, for example, pattern matching to identify the graphic, warning light, or indicator light (indication value) that is lit in the range specification area D5.

[0049] The primary storage unit 34 temporarily stores the identification result of the specimen marking H identified by the specimen marking recognition unit 33. Here, "temporarily storing" means that the identification result of the specimen marking H is stored in a predetermined memory of the specimen marking recognition device 30 from the time the specimen marking recognition unit 33 identifies the specimen marking H until the time the result is output to the control device 20 and / or display D.

[0050] The data transmission unit 35 transmits the identification results stored in the primary storage unit 34 to the upper control unit 23 .

[0051] The result storage unit 36 ​​stores the identification results in a predetermined memory of the specimen display recognition device 30 when the data transmission unit 35 transmits the identification results to the upper control unit 23. The primary storage unit 34 and the result storage unit 36 ​​may be the same storage unit. The identification results stored in the result storage unit 36 ​​may be displayed on the display D.

[0052] <Specimen Mark Recognition Method> Next, a setting method for making settings to identify the indicated value indicated by the specimen mark H using the specimen mark recognition device 30 of this embodiment will be described with reference to FIG. 4(a).

[0053] (1) Setting the setting information

[0054] First, the image acquisition unit 31 acquires an image of the meter panel via the camera 4 and outputs the image data to the display D (S1). Note that the image acquisition unit 31 may acquire images or videos that have been stored in advance.

[0055] When the user inputs a parameter to be analyzed in the parameter designation area D6, the parameter name receiving unit R4 receives the parameter to be analyzed and stores it in the setting information storage unit 32 (S2).

[0056] When the user designates the range designation area D5, the range designation signal receiving unit R3 receives the range designation signal and stores it in the setting information storage unit 32 (S3).

[0057] The user selects an image recognition algorithm in the algorithm selection area D3, and the image recognition algorithm receiving unit R1 receives the selection signal and stores it in the setting information storage unit 32 (S4).

[0058] A setting screen corresponding to the selected image recognition algorithm is displayed on the display D (S5). This allows the user to input change information into the editing area D4, and the change information receiving unit R2 receives the change information and stores it in the setting information storage unit 32 (S6). Note that the order of S1 to S6 may be reversed.

[0059] Next, a method for recognizing a mark on a specimen using the specimen mark recognition device 30 of this embodiment will be described with reference to FIG.

[0060] (2) Identification of Specimen Mark H The specimen mark recognition unit 33 acquires the setting information stored in the setting information storage unit 32 (S7).

[0061] The image acquisition unit 31 acquires an image of the meter panel M captured by the camera 4 (S8). Note that the image acquisition unit 31 may acquire image data captured by the camera 4 and stored in advance.

[0062] The specimen display recognition unit 33 acquires the range designation signal stored in the setting information storage unit 32, and cuts out the range designation region D5 from the image of the meter panel M (S9).

[0063] The specimen display recognition unit 33 acquires the selection signal and the change information from the setting information storage unit 32. Then, the specimen display recognition unit 33 analyzes the image in the range specification region D5 based on the image recognition algorithm corresponding to the selection signal and the change information, and identifies the indication value indicated by the specimen display H (S10).

[0064] The specimen display recognition unit 33 collects the indication values ​​output for each range specification region D5 (S11).

[0065] The specimen display recognition unit 33 stores the collected indication values ​​in the primary storage unit 34 (S12).

[0066] <Specimen Testing Method> Next, a description will be given of a specimen testing method using the specimen testing system 100 of this embodiment. In this specimen testing method, an example of a specimen W is shown as an AD / ADAS vehicle, but the specimen is not limited to an AD / ADAS vehicle.

[0067] First, the test specimen W is placed on the dynamometer 11. A predetermined driving pattern is set using the control device 20. This predetermined driving pattern includes setting the vehicle speed pattern of the test specimen W, setting road surface information including road surface gradient, setting the position of the vehicle in front, setting the vehicle speed pattern of the vehicle in front, and / or setting various parameters of the AD / ADAS vehicle in the test specimen W.

[0068] Once the predetermined driving pattern has been set, a test is performed on the specimen W. When the test is performed, the host control unit 23 sends a signal to the camera 4 to start capturing an image of the meter panel M, and starts controlling the dynamometer 11 or the autonomous driving robot 12 via the load control unit 21 or the robot control unit 22, respectively. Specifically, the host control unit 23 instructs the robot control unit 22 to drive the specimen W, and the robot control unit 22 controls the autonomous driving robot 12 to drive the specimen W. A load may also be applied to the specimen W by the load device 112.

[0069] The camera 4 then sends in real time an image of the meter panel M to the specimen display recognition device 30. The specimen display recognition device 30 identifies the specimen display H from the image captured by the camera 4 and identifies the state of the specimen W. Note that the host control unit 23 may control the robot control unit 22 based on the meter display content, such as a brake press instruction or a restart instruction from the vehicle, while the vehicle is traveling under adaptive cruise control, so that the autonomous driving robot 12 performs various operations.

[0070] <Effects of this embodiment> With the specimen marking recognition device 30 of this embodiment configured as described above, the user can select an image recognition algorithm suited to each specimen marking H on the meter panel M. Then, the specimen marking recognition unit 33 identifies the specimen marking H using the selected image recognition algorithm, making it possible to easily identify each specimen marking H on the meter panel M. Furthermore, in this embodiment, each specimen marking H on the meter panel M is individually identified, and there is no need to use an image of the meter panel M in which all specimen marks H are lit, as in the conventional case, making it possible to prevent erroneous recognition or omission of identification of a specimen marking H.

[0071] According to this embodiment, the specimen indicator H is identified using the setting information, so that the specimen indicator H can be more easily identified.

[0072] According to this embodiment, the specimen display recognition device 30 uses an image recognition algorithm in the range specification area D5, and therefore can more accurately identify the indication value of the specimen display H than when an image recognition algorithm is used in the entire range of the image.

[0073] Other Embodiments The present invention is not limited to the above-described embodiment.

[0074] 5, the setting information receiving unit R may be configured to function as a camera setting information receiving unit R5 that further receives camera setting information, which is information related to the settings of the camera 4. In this case, the specimen display recognition unit 33 may include a camera setting storage unit 37 that stores the camera setting information received by the camera setting information receiving unit R5, and an image correction unit 39 that corrects an image captured by the camera 4 that has been set based on the camera setting information. Here, the camera setting information refers to parameters that are manually or automatically set when capturing an image with the camera 4, such as aperture value, shutter speed, exposure time, white balance, and / or ISO sensitivity.

[0075] With this configuration, the specimen display recognition unit 33 identifies the specimen display H based on the image corrected by the image correction unit 39, so the specimen display recognition unit 33 can accurately identify the state of the specimen display H regardless of the environment around the meter panel.

[0076] 5, the setting information receiving unit R may be configured to perform the function of a field of view correction receiving unit R6 that receives field of view correction information, which is information regarding the amount of correction for correcting the field of view of the camera 4. In this case, the specimen display recognition device 30 may further include a field of view correction storage unit 38 that stores the field of view correction information received by the field of view correction receiving unit R6, and an image correction unit 39 that corrects the field of view of the camera 4 based on the field of view correction information stored in the field of view correction storage unit 38. Note that the image correction unit 39 that corrects the image captured by the camera 4 set based on the camera setting information and the image correction unit 39 that corrects the field of view of the camera 4 based on the field of view correction information stored in the field of view correction storage unit 38 may be the same correction unit or different correction units.

[0077] 5, when setting the camera 4, the angle-of-view correction receiving unit R6 receives the angle-of-view correction information and stores it in the angle-of-view correction storage unit 38. Then, the image correcting unit 39 corrects the angle of view of the image captured by the camera 4 based on the angle-of-view correction information stored in the angle-of-view correction storage unit 38. For the angle-of-view correction information, a reference image (for example, the reference image that is the image of the specimen display H acquired when the setting information was set) can be used.

[0078] With this configuration, when changing the camera 4 or testing multiple specimens of the same type, the settings for identifying the specimen markings are made once, and then the settings for identifying the indicated value indicated by the specimen markings H can be omitted by simply adjusting the camera's angle of view or image distortion, size, position and / or angle.

[0079] 5 , the angle-of-view correction storage unit 38 may store, for example, at least first angle-of-view correction information and second angle-of-view correction information, which is angle-of-view correction information with a smaller correction amount than the first angle-of-view correction information. In this case, immediately after the specimen display recognition unit 33 starts identifying the specimen display H, the image correction unit 39 obtains the first angle-of-view correction information from the angle-of-view correction storage unit 38 and corrects the angle of view based on the first angle-of-view correction information.

[0080] Meanwhile, while the specimen display recognition unit 33 is identifying the specimen display H, the image correction unit 39 acquires second angle of view correction information from the angle of view correction storage unit 38 and corrects the angle of view based on the second angle of view correction information.

[0081] With this configuration, when the angle of view changes significantly, for example, when the camera is replaced, the angle of view is corrected based on the first angle of view correction information. When small intermittent changes in the angle of view occur, for example, when the test specimen vibrates while traveling, the angle of view is corrected based on the second angle of view correction information. This changes the angle of view correction information in accordance with the magnitude of the correction amount, allowing the angle of view to be accurately corrected regardless of the magnitude of the correction amount.

[0082] Furthermore, the specimen display recognition device 30 may further include a processing process output unit (not shown) that outputs the processing process of the analysis target parameters to the display unit.

[0083] 6 shows an example of a display D for setting up the area specification area D5 to identify a graphic displayed in the area specification area D5 when pattern matching is selected as the image recognition algorithm. In this case, the change information includes a threshold value (e.g., a threshold value for the degree of similarity with a template image) and a template image selected by the user, and this change information is input into the editing area D4. A processing process confirmation window D7 for checking the processing process is also displayed on the left side of the display D.

[0084] The processing process output unit displays the similarity in the image of the meter panel M, the set threshold, and the output value in the range specification area D5. If the similarity exceeds the threshold, the processing process output unit outputs an output value (e.g., a number such as 1.00) indicating that the similarity exceeds the threshold. On the other hand, if the similarity is equal to or less than the threshold, the processing process output unit outputs an output value (e.g., a number such as 0) indicating that the similarity does not exceed the threshold.

[0085] With this configuration, the user can confirm whether the specimen mark recognition unit 33 has been able to identify the specimen mark in accordance with a predetermined process.

[0086] Furthermore, the specimen display recognition device 30 may further include a mask processing unit (not shown) that performs a masking process to mask a predetermined area G selected by the user from the captured image of the meter panel M.

[0087] The example shown in Figure 7 shows a display D that performs mask processing on the original image of a captured circle meter when, for example, setting is made to identify the indicated value shown by the circle meter. As shown in Figure 7, the original image may display information that may be mistakenly recognized as an indicated value by the specimen display recognition unit 33 when identifying the indicated value of the circle meter, such as the needle, scale, and center of the circle meter, as well as line segments displayed on the circle meter.

[0088] Therefore, the user selects a predetermined area G in which information is displayed that will be mistakenly recognized as an indicated value when the specimen display recognition unit 33 identifies the indicated value of the circle meter. When the user selects the predetermined area G, a mask image in which the predetermined area G is masked is displayed on the display D, as shown in Fig. 7. When identifying the indicated value shown by the circle meter, the specimen display recognition unit 33 identifies the area of ​​the mask image other than the masked predetermined area G.

[0089] With this configuration, when the specimen indication recognition unit 33 identifies the indicated value of the specimen indication H, it identifies the specimen indication H in an area other than the masked predetermined area G, so that information that is erroneously recognized when the specimen indication recognition unit 33 identifies the indicated value of the specimen indication H can be excluded from the analysis target. As a result, the specimen indication recognition unit 33 can more accurately identify the indicated value of the specimen indication H.

[0090] Furthermore, the specimen display recognition device 30 may further include a recording / playback unit (not shown) that records images captured by the camera 4 and plays them back on the display unit when the specimen display recognition unit 33 does not identify the specimen display H.

[0091] In the example shown in Fig. 8, a recording / playback instruction area D8 is displayed on the display D, in which the user can instruct playback or stop of the recorded video via the input means. As shown in Fig. 8, when a recording / playback instruction is given, the recording / playback unit outputs the recorded video to the display area D2, thereby playing back the video. At this time, the recording / playback unit may display an indication value or the like that identifies the specimen indication H included in the played back video.

[0092] The specimen display recognition device 30 may further include an overlay unit (not shown) that overlays a reference image of the meter panel M and a captured image of the meter panel M and displays them on the display D. Based on the deviation between the overlaid reference image and the captured image, the user can determine whether or not a field of view correction process related to the first field of view correction information is required.

[0093] 8, the angle of view correction process is performed using an angle of view adjustment menu D9 on the display D. The angle of view adjustment menu D9 displays automatic video angle adjustment D9a for automatically correcting the angle of view, and manual video angle adjustment D9b for manually correcting the angle of view, which can be selected by the user. The manual video angle adjustment D9b allows adjustment of, for example, the rotation angle, scale value, horizontal movement distance, and / or vertical movement distance of the camera 4.

[0094] With this configuration, the reference image and the captured image are superimposed and displayed on the display D, making it easier for the user to identify the misalignment between the captured image and the reference image, and allowing the angle of view of the captured image to be corrected automatically or manually.

[0095] In this embodiment, the image acquisition unit 31 acquires images captured by the camera 4, but may also acquire images captured or recorded by other media. Furthermore, the image acquisition unit 31 may acquire still images or moving images.

[0096] In this embodiment, the specimen display recognition device 30 is provided separately from the control device 20, but the specimen display recognition device 30 may be provided on the same computer as the control device 20, or some of the functions of the specimen display recognition device 30 may be incorporated into the control device 20.

[0097] In this embodiment, the specimen testing system 100 is configured to include an autonomous driving robot 12, but it does not have to be configured to include an autonomous driving robot 12, and the user may drive the vehicle.

[0098] In this embodiment, the specimen testing system 100 is configured to include a dynamometer 11, but it does not have to include a dynamometer 11 and may be used for a specimen W that runs on a road or a test course.

[0099] In this embodiment, the specimen marking recognition device 30 functions as an image acquisition unit 31, a setting information storage unit 32, a specimen marking recognition unit 33, a primary storage unit 34, a data transmission unit 35, and a result storage unit 36, but it is sufficient for the specimen marking recognition device 30 to function as at least the setting information storage unit 32 and the specimen marking recognition unit 33. Furthermore, in order to easily set an image recognition algorithm according to the specimen marking H and accurately identify the specimen marking on the meter panel, it is sufficient for the setting information storage unit 32 to at least receive and store a selection signal.

[0100] In this embodiment, the specimen display recognition device 30 includes the specimen display recognition unit 33. However, in other embodiments, the specimen display setting device may not include the specimen display recognition unit 33, and the specimen display recognition unit 33 performs settings for reading the indication value of the specimen display H. The specimen display setting device may also include a setting information receiving unit R that receives a selection signal indicating the type of image recognition algorithm for identifying the type of specimen display H from an image, and a setting information storage unit 32 that stores the setting information received by the setting information receiving unit R. With this configuration, the setting information receiving unit R receives the setting information and stores it in the setting information storage unit 32, so the user only needs to perform settings for identifying the indication value of the specimen display H once, thereby reducing the user's effort during setting.

[0101] In addition, various modifications and combinations of the embodiments may be made as long as they do not go against the spirit of the present invention.

[0102] According to the present invention, an image recognition algorithm can be easily set in accordance with the specimen marking to be recognized, and the specimen marking on the meter panel can be accurately identified.

[0103] REFERENCE SIGNS LIST 100... Specimen testing system 10... Test equipment 20... Control device R... Setting information receiving section R1... Image recognition algorithm receiving section R2... Change information receiving section R3... Range specification signal receiving section R4... Parameter receiving section R5... Camera setting information receiving section R6... View angle correction information receiving section 30... Specimen display recognition device 31... Image acquisition section 32... Setting information storage section 33... Specimen display recognition section 34... Primary storage section 35... Data transmission section 36... Result storage section 37... Camera setting storage section 38... View angle correction storage section 39... Image correction section 4... Camera W... Specimen H... Specimen display M... Meter panel

Claims

1. A test specimen marking recognition device that identifies a test specimen marking for indicating the state of a test specimen from an image of a vehicle or a part thereof taken by a camera, a setting information receiving unit that receives a selection signal indicating a type of image recognition algorithm for identifying the specimen marking from the image; a setting information storage unit that stores the selection signal received by the setting information receiving unit as setting information; a specimen display recognition unit that acquires the setting information from the setting information storage unit and identifies an indication value indicated by the specimen display based on the setting information.

2. the setting information receiving unit receives change information that is information for identifying a state change of the specimen display, and stores the change information as the setting information in the setting information storage unit; The specimen display recognition device of claim 1, wherein the specimen display recognition unit acquires the setting information from the setting information storage unit and identifies the indication value indicated by the specimen display based on the image recognition algorithm and the change information.

3. the setting information receiving unit receives, as the change information, information for identifying a state change of a bar meter, a circle meter, character information, and / or graphic information; The specimen display recognition device of claim 2, wherein the specimen display recognition unit acquires the setting information from the setting information storage unit and identifies the indication value indicated by the specimen display based on the image recognition algorithm and the change information.

4. the setting information receiving unit receives a range designation signal indicating a range of the image designated by a user, and stores the range designation signal in the setting information storage unit; The specimen marking recognition device according to claim 1 , wherein the specimen marking recognition unit uses the image recognition algorithm to identify an indication value indicated by the specimen marking that falls within the range.

5. 4. The device for recognizing a specimen marking according to claim 1, wherein the specimen marking is at least one of a circle meter, a bar meter, characters, and figures.

6. the setting information receiving unit receives angle-of-view correction information, which is information for correcting an angle of view of the camera; The specimen display recognition device a view angle correction storage unit that stores the view angle correction information received by the setting information receiving unit; The specimen display recognition device according to claim 1 , further comprising an image correction unit that corrects the angle of view based on the angle of view correction information stored in the angle of view correction storage unit.

7. the view angle correction information includes at least first view angle correction information, which is one of the view angle correction information, and second view angle correction information, which is view angle correction information having a smaller correction amount than the first view angle correction information, The specimen display recognition device of claim 6, wherein the image correction unit corrects the angle of view based on the first angle of view correction information immediately after the specimen display recognition unit starts identifying the specimen display, and corrects the angle of view based on the second angle of view correction information while the specimen display recognition unit is identifying the specimen display.

8. 4. The apparatus for recognizing an indication of a specimen according to claim 1, further comprising a mask processing unit that performs a masking process to mask a predetermined area selected by a user from the image of the meter panel.

9. A test piece testing system for testing a test piece that is a vehicle or a part thereof, Test equipment for testing the specimen; a camera for capturing an image of the meter panel of the test specimen; A specimen display recognition device according to any one of claims 1 to 3; a control unit that controls the test equipment based on the specimen marking identified by the specimen marking recognition device.

10. A specimen display setting device that identifies a specimen display for indicating the state of a specimen from an image of a meter panel of a vehicle or a part thereof captured by a camera, a setting information receiving unit that receives a selection signal indicating a type of image recognition algorithm for identifying the specimen marking from the image; a setting information storage unit that stores the selection signal received by the setting information receiving unit as setting information.

11. A test specimen marking recognition method for identifying a test specimen marking indicating a state of a test specimen from an image of a vehicle or a part thereof taken by a camera, receiving a selection signal indicating a type of image recognition algorithm for identifying the specimen indicia from the image; storing setting information including the received selection signal; A specimen marking recognition method for identifying an indication value indicated by the specimen marking based on the setting information.

12. A test piece marking recognition program used in a test piece marking recognition device that identifies a test piece marking indicating a state of a test piece from an image of a vehicle or a part thereof, the image being captured by a camera, a function as a setting information receiving unit that receives a selection signal indicating a type of image recognition algorithm for identifying the specimen marking from the image; a function as a setting information storage unit that stores setting information including the selection signal received by the setting information receiving unit; A specimen display recognition program that causes a computer to function as a specimen display recognition unit that acquires the setting information from the setting information storage unit and identifies the indication value indicated by the specimen display based on the setting information.