Automatic driving system test method and device, electronic equipment and storage medium

Abstract

The invention provides an automatic driving system test method and device, electronic equipment and a storage medium, and the method comprises the steps: determining a test target object according to a test demand of a to-be-tested automatic driving system; outputting multiple frames of test images and / or test cloud pictures in different test scenes according to the test requirements of the to-be-tested automatic driving system and the test target object; obtaining an identification result of the to-be-tested automatic driving system on the test image and / or the test cloud picture; and determining a score of the to-be-tested automatic driving system according to the identification result and a preset scoring rule. According to the automatic driving system test method and device, the electronic equipment and the storage medium provided by the invention, the problem that the identification capability of an automatic driving system of a vehicle is inconvenient to evaluate or test in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to an autonomous driving system testing method, device, electronic equipment and storage medium. Background Art

[0002] With the development of autonomous driving technology, the requirements for vehicle perception capabilities are getting higher and higher. Cameras and lidar, as the main perception devices of autonomous vehicles, require precise testing and calibration to ensure their reliability and accuracy under different environmental conditions. However, traditional technologies lack testing methods for the perception capabilities of autonomous driving systems, which makes it inconvenient to test and evaluate the perception capabilities of vehicle autonomous driving systems, further affecting technicians' maintenance and development of autonomous driving systems. Summary of the invention

[0003] The present invention provides an automatic driving system testing method, device, electronic device and storage medium, aiming to solve the problem of inconvenience in evaluating or testing the recognition capability of a vehicle's automatic driving system in traditional technologies.

[0004] In view of the problems existing in the prior art, an embodiment of the present invention provides a method for testing fusion perception of an autonomous driving system, comprising: Determine the test target according to the test requirements of the autonomous driving system to be tested; Outputting multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the test target object; Obtaining a recognition result of the automatic driving system to be tested on the test image and / or the test cloud map; A score of the autonomous driving system to be tested is determined according to the recognition result and preset scoring rules.

[0005] According to a fusion perception test method for an autonomous driving system provided by the present invention, the test targets include non-semantic obstacles and semantic targets, the non-semantic obstacles include cones and temporary roadblocks, the semantic targets include signs and traffic lights, and the test scenes include lighting test scenes, weather test scenes, viewing angle test scenes and distance test scenes.

[0006] According to a fusion perception test method for an autonomous driving system provided by the present invention, the brightness values ​​of the lighting test scenes are different, the weather test scenes include but are not limited to sunny test scenes, rainy test scenes, foggy test scenes and snowy test scenes, the perspective test scenes have different perspectives relative to the test target, and the distance test scenes have different distances compared to the test target.

[0007] According to a fusion perception testing method for an autonomous driving system provided by the present invention, each of the test images or the test cloud images corresponds to one or more of the test scenes and one or more of the test targets.

[0008] According to a method for testing fusion perception of an autonomous driving system provided by the present invention, outputting multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the test target object includes: Determining that the autonomous driving system to be tested is used to test image recognition capability, and outputting a plurality of frames of the test images under different test scenarios; Determine that the autonomous driving system to be tested is used to test the laser radar cloud image recognition capability, and output multiple frames of the test cloud images under different test scenarios; Determine the perception capability of the autonomous driving system to be tested for the fusion of test images and lidar cloud maps, output multiple frames of the test images and the test cloud maps under different test scenarios, and output the test images and the test cloud maps alternately.

[0009] According to a method for testing fusion perception of an autonomous driving system provided by the present invention, determining the score of the autonomous driving system to be tested according to the recognition result and a preset scoring rule includes: Determine that the automatic driving system to be tested recognizes the i The test target in the test image or the test cloud image is scored n i =1; Determine that the autonomous driving system under test does not recognize the i The test target in the test image or the test cloud image is scored n i =0; After the test is completed, the score of the autonomous driving system to be tested is calculated according to the following formula; in, P For rating, w A weighting coefficient of the test scene and the test target object corresponding to each of the test images or the test cloud maps.

[0010] According to a fusion perception test method for an autonomous driving system provided by the present invention, the calculation formula of the weighting coefficient is as follows: w = S + V + C + R + E

[0011] in,S Indicates the degree of shape distinctness of the test object, S The value range is 0-a. V Indicates the volume of the target to be measured, V The value range is bc, C Indicates the color brightness of the target object to be measured, C The value range of is de, R represents the reflectance or retroreflectance of the target object to be measured, R The value range is 0-f, E Indicates the difficulty of the test scenario, E The value range is gh.

[0012] The present invention also provides an autonomous driving system fusion perception test device, which uses the autonomous driving system fusion perception test method as described in any one of the above items, and the device comprises: A test target object determination module, used to determine the test target object according to the test requirements of the automatic driving system to be tested; An image output module, configured to output multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the determined test target; A scoring module is used to determine the score of the autonomous driving system to be tested based on the recognition result of the autonomous driving system to be tested on the test image and / or the test cloud map and preset scoring rules.

[0013] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the autonomous driving system fusion perception test method as described in any one of the above items are implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the autonomous driving system fusion perception testing method as described in any one of the above items are implemented.

[0015] The autonomous driving system fusion perception testing method provided by the present invention provides test images or test cloud maps under different test scenarios to the autonomous driving system to be tested according to the test requirements of the autonomous driving system to be tested. According to the recognition result of the autonomous driving system to be tested, the score of the autonomous driving system to be tested can be obtained, so as to evaluate the perception ability of the autonomous driving system to be tested, and further evaluate and calibrate the autonomous driving system's recognition ability of targets in real road scenes. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is a flowchart of the fusion perception testing method of the autonomous driving system provided by the present invention; Figure 2 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] In the description of the embodiments of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the embodiments of the present invention can be understood according to specific circumstances.

[0020] In the description of the embodiments of the present invention, it should be noted that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the embodiments of the present invention.

[0021] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0022] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0023] The disclosure below provides many different embodiments or examples to implement different structures of the present invention. In order to simplify the disclosure of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the present invention. In addition, the present invention can repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplicity and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed. In addition, the present invention provides various specific processes and examples of materials, but those of ordinary skill in the art can be aware of the applicability of other processes and / or the use of other materials.

[0024] Combine the following Figure 1-Figure 2 The present invention describes an automatic driving system testing method, device, electronic device and storage medium.

[0025] See also Figure 1 , an embodiment of the present invention provides a method for testing fusion perception of an autonomous driving system, comprising: S100, determining a test target object according to a test requirement of the autonomous driving system to be tested; S200, outputting multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the test target; S300, obtaining a recognition result of the test image and / or the test cloud map by the automatic driving system to be tested; S400: Determine a score of the autonomous driving system to be tested according to the recognition result and preset scoring rules.

[0026] Specifically, the test targets include non-semantic obstacles and semantic targets. Non-semantic obstacles include cones and temporary roadblocks. Semantic targets include signs and traffic lights. The test scenes include lighting test scenes, weather test scenes, viewing angle test scenes and distance test scenes.

[0027] The test target needs to be determined for the test requirements of the autonomous driving system to be tested. The test requirements of some autonomous driving systems may be for non-semantic obstacles or semantic targets alone. In this case, the test image or test cloud map only contains non-semantic obstacles or semantic targets, and the autonomous driving system's ability to recognize roadblocks or traffic lights can be tested separately. The test requirements of some autonomous driving systems are for non-semantic obstacles and semantic targets. In this case, the test image or test cloud map includes both non-semantic obstacles and semantic targets.

[0028] Among the non-semantic obstacles, such as common cones or temporary roadblocks, cones are usually used to mark construction areas, dangerous areas on the road or temporarily changed road layouts to alert drivers. Temporary roadblocks are used to close lanes, set up temporary traffic control or protect specific areas. Cones or temporary roadblocks are defined as non-semantic obstacles to evaluate the recognition ability of the autonomous driving system for temporarily set obstacles. Non-semantic obstacles are not limited to the above two. Among the semantic targets, such as common road signs or traffic lights, signs are common traffic signs on the road, which are used to provide information such as direction, distance, road conditions, etc. to help drivers make correct driving decisions. Traffic lights are traffic lights used to control traffic flow and ensure road safety. Their color changes (red, yellow, green) represent different traffic instructions. In autonomous driving tests, semantic targets are used to evaluate the system's ability to recognize and understand traffic signs and signals. The present invention can evaluate the recognition and understanding ability of the autonomous driving system by setting test targets.

[0029] Furthermore, after determining the test target, it is still necessary to determine the test scene according to the test requirements of the autonomous driving system. Even if the test target is the same, the parameters of the test image or test cloud map formed by the environmental influence of different test scenes will be quite different, affecting the recognition of the autonomous driving system. Therefore, in the technical solution provided by the present invention, it is necessary to assist the transformation of the test scene to comprehensively evaluate the recognition ability of the autonomous driving system for the test target. It should be noted that the test images or test cloud maps under different test scenes can be obtained from real road scenes, and the road scenes can be real scenes or construction scenes in real life. In comparison, the construction scene has higher controllability and is easy to control manually. For example, the lighting test scene can be completed in the construction scene by adjusting the brightness of the light source. Further, for the test image, in some test scenes, different test images can also be generated by image processing technology. For example, different lighting test scenes can be simulated by adjusting the brightness or contrast of the test image. For the test cloud map, in order to ensure the accuracy of the test, it is recommended to obtain a real test cloud map by changing the environmental factors in the real scene or the construction scene.

[0030] It should be noted that the test images or test cloud maps have been stored and classified before the test. When the autonomous driving system is tested, they will be allocated and output according to the test requirements of the autonomous driving system. After receiving the test image or test cloud map, the autonomous driving system will output the recognition result. By acquiring and calculating the recognition result of the autonomous driving system, the score of the autonomous driving system to be tested can be obtained. Under the positive scoring mechanism, the higher the score, the higher the recognition ability of the autonomous driving system. The autonomous driving system fusion perception test method provided by the present invention can evaluate the perception ability of the autonomous driving system to be tested, and further can evaluate and calibrate the autonomous driving system's recognition ability of targets in real road scenes. On the one hand, it is conducive to establishing a mechanism for scoring and detecting autonomous driving systems, and on the other hand, it is also conducive to technicians to maintain and redevelop autonomous driving systems based on recognition results and scoring results.

[0031] Specifically, in order to improve the comprehensive coverage of the test scene, each test scene has more embodiments. For example, for the illumination test scene, the brightness value of each illumination test scene is different. Taking the construction scene as an example, it can be obtained by adjusting the brightness of the light source in the construction scene. Each weather test scene includes but is not limited to a sunny test scene, a rainy test scene, a foggy test scene, and a snowy test scene. In the construction scene, the brightness of the light can be used to simulate a sunny day, water can be sprayed to simulate a rainy day, dry ice can be sprayed to simulate a foggy day, and foam can be sprayed to simulate a snowy day. Each viewing angle test scene has a different viewing angle relative to the test target. In the construction scene, it can be obtained by adjusting the shooting angle of the test target. Conventional ones include front view, side view, or oblique view. Each distance test scene has a different distance from the test target. In the construction scene, it can be obtained by adjusting the distance between the test target and the test target.

[0032] Furthermore, in order to improve the integration between the environment and the target object to be tested, each test image or test cloud map corresponds to one or more test scenes, and one or more test targets. It is understandable that each test image or test cloud map may include one or more test scenes, such as the integration of a brightness test scene and a weather test scene, the integration of a weather test scene and a viewing angle test scene, or the integration of three or more test scenes. In general, it is necessary to provide at least 100 separate or integrated test scenes to the autonomous driving system to be tested. Furthermore, there may be one or more test targets in the test scene. Correspondingly, when the more types of test scene integration and the more test targets are set, the higher the difficulty of identifying the test image or test cloud map will be, which will test the recognition ability of the autonomous driving system.

[0033] Specifically, S200, outputting multiple test images and / or test cloud images under different test scenarios according to the test requirements of the automatic driving system to be tested and the test target object, includes: S210, determining that the autonomous driving system to be tested is used to test image recognition capability, and outputting a plurality of test images under different test scenarios; S220, determining that the autonomous driving system to be tested is used to test the laser radar cloud image recognition capability, and outputting multiple frames of the test cloud images under different test scenarios; S230: Determine the perception capability of the autonomous driving system to be tested for the fusion of test images and lidar cloud maps, output multiple frames of the test images and the test cloud maps under different test scenarios, and output the test images and the test cloud maps alternately.

[0034] As mentioned above, the test requirements of the autonomous driving systems to be tested are different. Some may test the image recognition ability alone, and then output the test images in different test scenarios separately; some may test the cloud map recognition ability alone, and then output the test cloud maps in different test scenarios separately; in more cases, it is necessary to test the fusion perception ability of images and lidar cloud maps, so the test images and test cloud maps are output alternately. It should be noted that in order to give the autonomous driving system sufficient recognition time, the output of two adjacent frames of images generally has a time interval, which can be set to 3-5s. If the target object to be tested is not recognized within the time limit, the autonomous driving system is judged to have failed in recognition. It should also be noted that, in general, a fixed test scene can generate test images and test cloud maps respectively. When conducting a fusion perception ability test, it is preferred that the test image and test cloud map of a fixed test scene are output at adjacent intervals.

[0035] Further, S400, determining the score of the automatic driving system to be tested according to the recognition result and a preset scoring rule includes: S410: Determine whether the automatic driving system to be tested recognizes the i The test target in the test image or the test cloud image is scored n i =1; S420: Determine that the automatic driving system to be tested does not recognize i The test target in the test image or the test cloud image is scored n i =0; S430: After the test is completed, the score of the autonomous driving system to be tested is calculated according to the following formula; in, P For rating, w A weighting coefficient of the test scene and the test target object corresponding to each of the test images or the test cloud maps.

[0036] Specifically, the calculation formula of the weighting coefficient is as follows: w = S + V + C + R + E

[0037] in, S Indicates the degree of shape distinctness of the test object, S The value range is 0-a. V Indicates the volume of the target to be measured, V The value range is bc, C Indicates the color brightness of the target object to be measured, C The value range of is de, R represents the reflectance or retroreflectance of the target object to be measured, R The value range is 0-f, E Indicates the difficulty of the test scenario, E The value range is gh.

[0038] As mentioned above, the recognition difficulty of each test image or test cloud map will be different. The more test scenes are integrated in the test image and the more test targets are included, the greater the recognition difficulty of the test image or test cloud map. Furthermore, the recognition difficulty is also different for different test scenes and test targets of different forms. In order to improve the representation effect of the score, in the technical solution provided by the present invention, different weighting coefficients are set for different test scenes and test targets. By combining the weighting coefficients and the number of recognitions of the test images or test cloud maps, the comprehensive score of the autonomous driving system can be obtained. It should be noted that when all autonomous driving systems are tested using the same test method, different autonomous driving systems will obtain different scores. Under the positive scoring mechanism, the higher the score, the higher the fusion perception ability of the autonomous driving system. The mainstream autonomous driving systems on the market can be selected to form samples for scoring. According to the distribution of the scores, several evaluation conditions such as excellent, good, qualified, and unqualified can be determined, thereby assisting in establishing the scoring criteria for the autonomous driving system, which is conducive to forming a standard autonomous driving system scoring test method and standard.

[0039] Specifically, the present invention provides several methods for setting weighting coefficients. In the calculation formula of the weighting coefficients, S Indicates the degree of shape prominence of the test target. SThe value range is 0-a. Specifically, a regular circle or square takes a value close to 1, and a complex irregular shape takes a value close to 0. In actual application, a specific value can be set for each different shape, and a specific value can be uniformly determined for irregular shapes that cannot be confirmed. V Indicates the volume of the target to be measured. V The value range of V is bc. The smaller the volume of the target to be measured, the larger the value of V is. Specific values ​​can be set for targets to be measured in different volume ranges. C Indicates the color vividness of the target object to be measured. C The value range of is de, and the chroma C value in the CIELAB color system can be used for assignment. The higher the chroma, the larger C is. R Indicates the reflectance or retroreflectance of the target object to be measured. R The value range is 0-f. The R value of a completely non-reflective target is 0, and the R value of a reflective mark is usually above 1. Different R values ​​can be set according to the type of target. E Indicates the difficulty of the test scenario, E In an optional embodiment, the E value of a daytime scene is 1, the E value of a night scene is 2, and the E value of a rainy scene is 3. When multiple different scenes are combined, the sum of the E values ​​of the multiple different scenes can be used for calculation.

[0040] On the basis of the above test method, the present invention further provides an autonomous driving system fusion perception test device, which applies the above autonomous driving system fusion perception test method, and the device comprises: A test target object determination module, used to determine the test target object according to the test requirements of the automatic driving system to be tested; An image output module, configured to output multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the determined test target; A scoring module is used to determine the score of the autonomous driving system to be tested based on the recognition result of the autonomous driving system to be tested on the test image and / or the test cloud map and preset scoring rules.

[0041] The present invention also provides an electronic device, Figure 2 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 2As shown, the electronic device may include: a processor 110, a communications interface 120, a memory 130 and a communication bus 140, wherein the processor 110, the communications interface 120 and the memory 130 communicate with each other through the communication bus 140. The processor 110 may call the logic instructions in the memory 130 to execute the autonomous driving system fusion perception test method. The method includes: S100, determining a test target object according to a test requirement of the autonomous driving system to be tested; S200, outputting multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the test target; S300, obtaining a recognition result of the test image and / or the test cloud map by the automatic driving system to be tested; S400: Determine a score of the autonomous driving system to be tested according to the recognition result and preset scoring rules.

[0042] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0043] On the other hand, the present invention also provides a computer program product, the computer program product includes a computer program, the computer program can be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer can execute the autonomous driving system fusion perception test method provided by the above methods. The method includes: S100, determining a test target object according to a test requirement of the autonomous driving system to be tested; S200, outputting multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the test target; S300, obtaining a recognition result of the test image and / or the test cloud map by the automatic driving system to be tested; S400: Determine a score of the autonomous driving system to be tested according to the recognition result and preset scoring rules.

[0044] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to perform the autonomous driving system fusion perception test method provided by the above methods. The method includes: S100, determining a test target object according to a test requirement of the autonomous driving system to be tested; S200, outputting multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the test target; S300, obtaining a recognition result of the test image and / or the test cloud map by the automatic driving system to be tested; S400: Determine a score of the autonomous driving system to be tested according to the recognition result and preset scoring rules.

[0045] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0046] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0047] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for testing fusion perception of an autonomous driving system, characterized in that: include: Determine the test target according to the test requirements of the autonomous driving system to be tested; Outputting multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the test target object; Obtaining a recognition result of the automatic driving system to be tested on the test image and / or the test cloud map; A score of the autonomous driving system to be tested is determined according to the recognition result and preset scoring rules.

2. The method for testing the fusion perception of an autonomous driving system according to claim 1, characterized in that: The test targets include non-semantic obstacles and semantic targets. The non-semantic obstacles include cones and temporary roadblocks. The semantic targets include signboards and traffic lights. The test scenes include lighting test scenes, weather test scenes, viewing angle test scenes and distance test scenes.

3. The method for testing the fusion perception of an autonomous driving system according to claim 2, characterized in that: The brightness values ​​of the lighting test scenes are different, the weather test scenes include but are not limited to sunny test scenes, rainy test scenes, foggy test scenes and snowy test scenes, the viewing angles of the viewing angles of the test targets are different, and the distances of the distances of the distance test scenes are different compared to the test targets.

4. The method for testing the fusion perception of an autonomous driving system according to claim 3, characterized in that: Each of the test images or the test cloud images corresponds to one or more of the test scenes and one or more of the test objects.

5. The autonomous driving system fusion perception testing method according to claim 1, characterized in that: According to the test requirements of the autonomous driving system to be tested and the test target, outputting multiple test images and / or test cloud images under different test scenarios includes: Determining that the autonomous driving system to be tested is used to test image recognition capability, and outputting a plurality of frames of the test images under different test scenarios; Determine that the autonomous driving system to be tested is used to test the laser radar cloud image recognition capability, and output multiple frames of the test cloud images under different test scenarios; Determine the perception capability of the autonomous driving system to be tested for the fusion of test images and lidar cloud maps, output multiple frames of the test images and the test cloud maps under different test scenarios, and output the test images and the test cloud maps alternately.

6. The method for testing the fusion perception of an autonomous driving system according to claim 1, characterized in that: Determining the score of the autonomous driving system to be tested according to the recognition result and the preset scoring rules includes: Determine that the automatic driving system to be tested recognizes the i The test target in the test image or the test cloud image is scored n i =1; Determine that the autonomous driving system under test does not recognize the i The test target in the test image or the test cloud image is scored n i =0; After the test is completed, the score of the autonomous driving system to be tested is calculated according to the following formula; in, P For rating, w A weighting coefficient of the test scene and the test target object corresponding to each of the test images or the test cloud maps.

7. The method for testing the fusion perception of an autonomous driving system according to claim 6, characterized in that: The calculation formula of the weighting coefficient is as follows: w = S + V + C + R + E in, S Indicates the degree of shape distinctness of the test object, S The value range is 0-a. V Indicates the volume of the target to be measured, V The value range is bc, C Indicates the color brightness of the target object to be measured, C The value range of is de, R represents the reflectance or retroreflectance of the target object to be measured, R The value range is 0-f, E Indicates the difficulty of the test scenario, E The value range is gh.

8. An autonomous driving system fusion perception test device, characterized in that: The method for testing the fusion perception of an autonomous driving system according to any one of claims 1 to 7 is applied, wherein the device comprises: A test target object determination module, used to determine the test target object according to the test requirements of the automatic driving system to be tested; An image output module, configured to output multiple test images and / or test cloud images under different test scenarios according to the test requirements of the autonomous driving system to be tested and the determined test target; A scoring module is used to determine the score of the autonomous driving system to be tested based on the recognition result of the autonomous driving system to be tested on the test image and / or the test cloud map and preset scoring rules.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the autonomous driving system fusion perception testing method as described in any one of claims 1 to 7 are implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the autonomous driving system fusion perception testing method as described in any one of claims 1 to 7 are implemented.

Images