Scene detection method and device, equipment, storage medium and product

Abstract

The embodiment of the invention provides a scene detection method, device and equipment, a storage medium and a product, and the method comprises the steps: obtaining real-time measurement data detected by a plurality of vehicle sensors in a vehicle driving process; determining an incidence matrix by using the real-time measurement data and historical measurement data of the plurality of vehicle sensors, the incidence matrix being used for representing the similarity between the real-time measurement data and the historical measurement data; target matching is conducted on the real-time measurement data through the incidence matrix and the Hungary algorithm, a matched target object is obtained, and the target object is an object detected in the multiple vehicle sensors; determining an unmatching rate of the plurality of vehicle sensors by using the target object; and under the condition that the non-matching rate is greater than a preset threshold value, determining a vehicle running scene corresponding to the non-matching rate. Through the steps, the safety and reliability of the automatic driving system can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle networking, and in particular relates to a scene detection method, device, equipment, storage medium and product. Background Art

[0002] With the rapid development of smart cars, autonomous driving technology has become the core driving force for the transformation of the automotive industry. However, with the in-depth application of technology, an increasingly prominent technical problem is that when smart cars face unknown and unsafe scenarios, data discovery is difficult and inefficient. These unknown and unsafe scenarios often fall into the category of long-tail effects, that is, rare or extreme situations that are difficult to cover or simulate in conventional testing or training data, and they pose a serious threat to the safety and reliability of autonomous driving systems. Summary of the invention

[0003] The embodiments of the present application provide a scene detection method, apparatus, device, storage medium and product to solve the problem that rare or extreme situations that are difficult to cover or simulate in existing conventional test or training data pose a serious threat to the safety and reliability of the autonomous driving system.

[0004] In a first aspect, an embodiment of the present application provides a scene detection method, the method comprising:

[0005] Acquire real-time measurement data detected by multiple vehicle sensors during vehicle driving;

[0006] Determine a correlation matrix using the real-time measurement data and the historical measurement data of the plurality of vehicle sensors, wherein the correlation matrix is ​​used to represent the similarity between the real-time measurement data and the historical measurement data;

[0007] Performing target matching on the real-time measurement data using the correlation matrix and the Hungarian algorithm to obtain a matched target object, wherein the target object is an object detected by all of the multiple vehicle sensors;

[0008] determining a mismatch rate of the plurality of vehicle sensors using the target object;

[0009] When the non-matching rate is greater than a preset threshold, a vehicle operation scenario corresponding to the non-matching rate is determined.

[0010] In a second aspect, an embodiment of the present application provides a scene detection device, the device comprising:

[0011] An acquisition module, used to acquire real-time measurement data detected by multiple vehicle sensors during vehicle driving;

[0012] A first determination module, configured to determine a correlation matrix using the real-time measurement data and the historical measurement data of the plurality of vehicle sensors, wherein the correlation matrix is ​​used to represent the similarity between the real-time measurement data and the historical measurement data;

[0013] A matching module, used for performing target matching on the real-time measurement data using the association matrix and the Hungarian algorithm to obtain a matched target object, wherein the target object is an object detected by all the vehicle sensors;

[0014] A second determination module, configured to determine a mismatch rate of the plurality of vehicle sensors using the target object;

[0015] The third determination module is used to determine the vehicle operation scenario corresponding to the non-matching rate when the non-matching rate is greater than a preset threshold.

[0016] In a third aspect, an embodiment of the present application provides a terminal device, the device comprising: a processor and a memory storing computer program instructions;

[0017] When the processor executes the computer program instructions, the scene detection method of the first aspect is implemented.

[0018] In a fourth aspect, an embodiment of the present application provides a computer storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the scene detection method of the first aspect is implemented.

[0019] In a fifth aspect, an embodiment of the present application provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the scene detection method of the first aspect.

[0020] The scene detection method, device, equipment, storage medium and product of the embodiment of the present invention can calculate the target mismatch rate in real time in this embodiment, so that the autonomous driving system can instantly evaluate the matching quality of sensor data, thereby timely detecting and responding to potential problems in data matching, greatly enhancing the reliability and stability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0022] Figure 1 is a flow chart of a scene detection method provided in an embodiment of the present application;

[0023] Figure 2is a structural schematic diagram of a scene detection device provided in an embodiment of the present application;

[0024] Figure 3 It is a schematic diagram of the structure of the terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0025] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0026] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0027] In order to solve the problems of the prior art, the embodiments of the present application provide a scene detection method, device, equipment, storage medium and product. The following first describes an example of the scene detection method provided by the embodiments of the present application that can be applied.

[0028] Figure 1 FIG. 1 is a flow chart of a scene detection method provided by an embodiment of the present application. Figure 1 As shown, the method includes:

[0029] Step 101, obtaining real-time measurement data detected by multiple vehicle sensors during vehicle driving;

[0030] Step 102, using the real-time measurement data and the historical measurement data of the plurality of vehicle sensors, determining an association matrix, wherein the association matrix is ​​used to represent the similarity between the real-time measurement data and the historical measurement data;

[0031] Step 103, using the correlation matrix and the Hungarian algorithm to perform target matching on the real-time measurement data to obtain a matched target object, wherein the target object is an object detected by all the vehicle sensors;

[0032] Step 104, using the target object to determine the mismatch rate of the plurality of vehicle sensors;

[0033] Step 105 : when the non-matching rate is greater than a preset threshold, determining a vehicle operation scenario corresponding to the non-matching rate.

[0034] In this embodiment, the scene detection method is applied to a scene detection system, which is collectively referred to as a system in the following embodiments. The vehicle sensors include cameras, laser radars, and millimeter-wave radars. During the driving process of the vehicle, the system continuously obtains real-time measurement data from multiple sensors installed on the vehicle. These data may include various information such as the distance, speed, direction, shape, and color of the object.

[0035] The system then uses these real-time measurements and the sensor's past (i.e., historical) measurements to calculate a correlation matrix. The main purpose of this correlation matrix is ​​to quantify the similarity or correlation between the real-time measurements and the historical data. The higher the similarity, the more similar the currently measured object or environment is to a certain moment in the past, which helps to identify persistent objects or environmental features.

[0036] The Hungarian algorithm is a combinatorial optimization algorithm that solves the task allocation problem in polynomial time. In this scenario, it is used to match targets (such as pedestrians, vehicles, obstacles, etc.) measured in real time to find common targets detected by multiple sensors. Through matching, the system can more accurately identify and track objects in the environment and reduce false positives and false negatives.

[0037] After completing the target matching, the system will calculate the mismatch rate of each sensor or the entire system. The mismatch rate refers to the proportion of targets detected in one or some sensors but failed to match (i.e. failed to be jointly confirmed) in other sensors. This indicator reflects the consistency of data fusion between sensors and the reliability of the system.

[0038] Finally, if the mismatch rate exceeds the preset threshold, the system will consider that the current environment or operating scenario may have an impact on the performance of the sensor and the reliability of the data. At this point, the system will further analyze the correlation between the mismatch rate and different vehicle operating scenarios (such as complex traffic environments, bad weather conditions, night driving, etc.) to determine the specific scenario the vehicle is currently in. This scenario recognition helps the system take corresponding strategies or measures to improve the accuracy and safety of environmental perception.

[0039] In this embodiment, the target mismatch rate can be calculated in real time, so that the autonomous driving system can instantly evaluate the matching quality of sensor data, thereby promptly detecting and responding to potential problems in data matching, greatly enhancing the reliability and stability of the system.

[0040] In one embodiment of the present application, the Hungarian algorithm is used to perform target matching on the image data, the millimeter wave radar data, and the point cloud data to obtain a matched target object, including:

[0041] Using the image data, the millimeter wave radar data and the point cloud data, create a tracking list and a detection list

[0042] determining an association matrix between the tracking list and the detection list;

[0043] The target object is determined using the association matrix.

[0044] In this embodiment, the real-time measurement data includes image data, millimeter-wave radar data, and point cloud data; wherein the image data comes from the vehicle-mounted camera, providing two-dimensional or three-dimensional visual information of the vehicle's surroundings, including rich color, texture, and shape details. Millimeter-wave radar data, generated by millimeter-wave radar sensors, provides distance, speed, and angle information of target objects, and has good penetration in adverse weather conditions (such as rain and fog). Point cloud data, usually generated by lidar, is a collection of points in three-dimensional space, representing the shape and position of an object.

[0045] The tracking list is a continuously updated list of objects that the system has identified and is tracking. Each object has a unique identifier and state information associated with it (such as position, speed, size, etc.), which is estimated and predicted based on previous measurement data (including historical image, radar and point cloud data). The detection list is a list generated based on the current real-time measurement data (image, radar and point cloud) and contains all potential objects detected in the current time step.

[0046] The association matrix is ​​a two-dimensional array that represents the possible correspondence between the targets in the tracking list and the potential targets in the detection list. Each element in the matrix represents the similarity or association between a pair of targets.

[0047] In an embodiment of the present application, determining the association matrix between the tracking list and the detection list includes:

[0048] obtaining a distance and a Welsh variance between objects in any two vehicle sensors of the plurality of vehicle sensors;

[0049] Determine the existence time and area of ​​the object;

[0050] The association matrix is ​​constructed using the distance, the Welsh variance, the existence time and the area.

[0051] In this embodiment, distance refers to the relative distance between objects detected by different sensors in a multidimensional space. This distance can be Euclidean distance, Manhattan distance, or any other measurement method suitable for a specific scenario. It reflects the proximity of objects in physical space.

[0052] Welsh variance can reflect the similarity or difference between two objects in data characteristics.

[0053] Existence time refers to the length of time that an object is continuously detected.

[0054] Area,The area (or volume, in 3D) of an object in an image or point cloud data is an important feature that provides information about the size of the object. Objects of different sizes may have different importance in the scene, and changes in area may also indicate movement or deformation of the object.

[0055] Using the above four features (distance, Welsh variance, existence time, area), we can construct an association matrix, where each row of the matrix represents a target in the tracking list and each column represents a potential target in the detection list. The elements in the matrix (i.e. the value of each cell) are the similarity or association between the two objects calculated based on these four features.

[0056] In this embodiment, a matrix reflecting the object similarity or association between the tracking list and the detection list can be constructed. This matrix can then be used in subsequent target matching algorithms (such as the Hungarian algorithm) to achieve more accurate and reliable target tracking and recognition.

[0057] In an embodiment of the present application, obtaining the distance between objects in any two vehicle sensors among the plurality of vehicle sensors includes:

[0058] Determining the distance using a first formula;

[0059] The first formula is:

[0060]

[0061] Among them, Distance is the distance, is the horizontal coordinate of the object detected by one of any two vehicle sensors, is the horizontal coordinate of the object detected by the other of any two vehicle sensors, is the vertical coordinate of the object detected by one of any two vehicle sensors, and is the vertical coordinate of the object detected by the other of any two vehicle sensors.

[0062] In an embodiment of the present application, obtaining the Welsh variance between objects in any two vehicle sensors among the plurality of vehicle sensors includes:

[0063] Determine the mean and variance corresponding to the data detected by any two vehicle sensors;

[0064] determining a Welsh statistic and degrees of freedom using the mean and the variance;

[0065] The Welsh variance is determined using the Welsh statistic and the degrees of freedom.

[0066] In this embodiment, for any two data sets about the same object detected by vehicle sensors, the mean (ie, the average value of the data) and variance (ie, the average value of the square of the difference between the data and its mean) of each data set are first calculated.

[0067] Degrees of Freedom (df) is another important concept related to statistical tests. It represents the number of values ​​that can vary independently without affecting the distribution of the sample statistic. When comparing two data sets, the calculation of degrees of freedom depends on the sample size of the data set and the statistical test used.

[0068] In one embodiment, the time and area in the above embodiment can be calculated by the following process:

[0069] The target's trajectory information can be used to consider historical data. If the target is stably matched by a tracker in the past few frames, the matching cost with the tracker in the current frame can be reduced. That is, if the target exists within the threshold in the historical trajectory, its matching cost will decrease with time until it exceeds the threshold and becomes 0.

[0070] Calculation formula:

[0071] Y=AX+B

[0072] Area calculation: Use the shoelace formula to calculate the area of ​​the target. Specific calculation formula:

[0073]

[0074] Among them, (xk ,y k ) represents the coordinates of the kth target, (x k+1 ,y k+1 ) represents the coordinates of the k+1th target.

[0075] Finally, according to the method of DS evidence theory, the above four values ​​are comprehensively processed to obtain weights, and the weight cost matrix between each target is constructed.

[0076] In an embodiment of the present application, when the non-matching rate is greater than a preset threshold, after determining the vehicle operation scenario corresponding to the non-matching rate, the method further includes:

[0077] storing the operation data segment of the current frame corresponding to the vehicle operation scenario into a first data packet;

[0078] In the first data packet, a first data segment of a preset duration corresponding to before the current frame and a second data segment of a preset duration corresponding to after the current frame are concatenated to obtain a target data packet.

[0079] In this embodiment, when the mismatch rate exceeds the preset threshold, the system will first analyze the current environmental conditions, vehicle status, and sensor data characteristics to determine one or more vehicle operation scenarios associated with a high mismatch rate. These scenarios may include, but are not limited to: bad weather (such as rain, fog, snow), complex traffic environments (such as congestion, intersections), large vehicle dynamic changes (such as sudden acceleration, sudden braking, frequent lane changes), etc.

[0080] Once the vehicle operation scenario corresponding to the mismatch rate is determined, the system will store the operation data fragment of the current frame (i.e., the frame that triggers the mismatch rate to exceed the threshold) into a specific data packet, which we call the "first data packet." This data packet contains the raw data or processed data of all relevant sensors (such as images, millimeter-wave radars, point clouds, etc.) in the current frame for subsequent detailed analysis.

[0081] In order to obtain more comprehensive context information, the system not only saves the data of the current frame, but also takes a data segment of a preset length from the current frame forward and backward. We call these segments "first data segment" and "second data segment" respectively. They represent the vehicle state before the current frame and the possible changes after that.

[0082] The system then concatenates these two data fragments with the data of the current frame to form a data packet with a longer time span and richer information, which is called the "target data packet."

[0083] In this embodiment, by storing and splicing the vehicle operation scene and its related data (including the current frame and the data segments of the preset time before and after) when the mismatch rate exceeds the preset threshold into a target data packet, a rich and continuous data source is provided for subsequent data analysis. This enables researchers or engineers to have a deeper understanding of the characteristics of sensor data in a specific operation scenario, thereby optimizing the target tracking and matching algorithm and improving the accuracy and robustness of the system.

[0084] like Figure 2 As shown, the embodiment of the present application further provides a scene detection device 200, and the scene detection device 200 includes:

[0085] The acquisition module 201 is used to acquire real-time measurement data detected by multiple vehicle sensors during the vehicle's driving process;

[0086] A first determination module 202, configured to determine a correlation matrix using the real-time measurement data and the historical measurement data of the plurality of vehicle sensors, wherein the correlation matrix is ​​used to represent the similarity between the real-time measurement data and the historical measurement data;

[0087] A matching module 203 is used to perform target matching on the real-time measurement data using the correlation matrix and the Hungarian algorithm to obtain a matched target object, where the target object is an object detected by all of the multiple vehicle sensors;

[0088] A second determination module 204 is used to determine the mismatch rate of the plurality of vehicle sensors using the target object;

[0089] The third determination module 205 is configured to determine a vehicle operation scenario corresponding to the non-matching rate when the non-matching rate is greater than a preset threshold.

[0090] Optionally, the first determining module 202 includes:

[0091] Create a submodule for creating a tracking list and a detection list using the image data, the millimeter wave radar data and the point cloud data

[0092] The first determination submodule is used to determine the association matrix between the tracking list and the detection list.

[0093] Optionally, the first determining submodule includes:

[0094] A first acquisition unit, configured to acquire a distance and a Welsh variance between objects in any two vehicle sensors among the plurality of vehicle sensors;

[0095] Determine a unit for the existence time and area of ​​the object;

[0096] A construction unit is used to construct the association matrix using the distance, the Welsh variance, the existence time and the area.

[0097] Optionally, the first acquiring unit is specifically configured to:

[0098] Determining the distance using a first formula;

[0099] The first formula is:

[0100]

[0101] Where Distance is the distance, x measurement is the horizontal coordinate of the object detected by one of the two vehicle sensors, x prediction is the horizontal coordinate of the object detected by the other sensor of any two vehicle sensors, y measurement is the ordinate of the object detected by one of the two vehicle sensors, y prediction is the ordinate of the object detected by the other sensor of any two vehicle sensors.

[0102] Optionally, the first acquisition unit is specifically used to:

[0103] Determine the mean and variance corresponding to the data detected by any two vehicle sensors;

[0104] determining a Welsh statistic and degrees of freedom using the mean and the variance;

[0105] The Welsh variance is determined using the Welsh statistic and the degree of freedom. Optionally, the scene detection device 200 further includes:

[0106] A storage module, used for storing the operation data segment of the current frame corresponding to the vehicle operation scene into a first data packet;

[0107] The splicing module is used to splice, in the first data packet, a first data segment of a preset time length corresponding to before the current frame and a second data segment of a preset time length corresponding to after the current frame to obtain a target data packet.

[0108] Figure 3 A schematic diagram of the hardware structure of a terminal device provided in an embodiment of the present application is shown.

[0109] The terminal device may include a processor 301 and a memory 302 storing computer program instructions.

[0110] Specifically, the processor 301 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0111] The memory 302 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 302 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 302 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 302 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 302 is a non-volatile solid-state memory.

[0112] In certain embodiments, the memory 302 may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present disclosure.

[0113] The processor 301 implements any one of the scene detection methods in the above embodiments by reading and executing computer program instructions stored in the memory 302 .

[0114] In one example, the terminal device may further include a communication interface 303 and a bus 310. Figure 3 As shown, the processor 301, the memory 302, and the communication interface 303 are connected via a bus 310 and communicate with each other.

[0115] The communication interface 303 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0116] Bus 310 includes hardware, software or both, and the components of online data flow billing equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front-end bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, first video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 310 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the present application considers any suitable bus or interconnection.

[0117] In addition, in combination with the scene detection method in the above embodiment, the embodiment of the present application can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any one of the scene detection methods in the above embodiment is implemented.

[0118] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0119] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0120] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0121] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and computer program product according to the embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable scene detection device to produce a machine so that these instructions executed by the processor of the computer or other programmable scene detection device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0122] The above are only specific implementation methods of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A scene detection method, characterized in that: The method comprises: Acquire real-time measurement data detected by multiple vehicle sensors during vehicle driving; Determine a correlation matrix using the real-time measurement data and the historical measurement data of the plurality of vehicle sensors, wherein the correlation matrix is ​​used to represent the similarity between the real-time measurement data and the historical measurement data; Performing target matching on the real-time measurement data using the correlation matrix and the Hungarian algorithm to obtain a matched target object, wherein the target object is an object detected by all of the multiple vehicle sensors; determining a mismatch rate of the plurality of vehicle sensors using the target object; When the non-matching rate is greater than a preset threshold, a vehicle operation scenario corresponding to the non-matching rate is determined.

2. The scene detection method according to claim 1, characterized in that: The real-time measurement data includes image data, millimeter wave radar data and point cloud data; The determining of a correlation matrix using the real-time measurement data and the historical measurement data of the plurality of vehicle sensors comprises: Using the image data, the millimeter wave radar data and the point cloud data, create a tracking list and a detection list An association matrix between the tracking list and the detection list is determined.

3. The scene detection method according to claim 2, characterized in that: The determining of the association matrix between the tracking list and the detection list comprises: obtaining a distance and a Welsh variance between objects in any two vehicle sensors of the plurality of vehicle sensors; Determine the existence time and area of ​​the object; The association matrix is ​​constructed using the distance, the Welsh variance, the existence time and the area.

4. The scene detection method according to claim 3, characterized in that: Obtaining a distance between objects in any two vehicle sensors of the plurality of vehicle sensors includes: Determining the distance using a first formula; The first formula is: Where Distance is the distance, x measurement is the horizontal coordinate of the object detected by one of the two vehicle sensors, x prediction is the horizontal coordinate of the object detected by the other sensor of any two vehicle sensors, y measurement is the ordinate of the object detected by one of the two vehicle sensors, y prediction is the ordinate of the object detected by the other sensor of any two vehicle sensors.

5. The scene detection method according to claim 3, characterized in that: Obtaining the Welsh variance between objects in any two vehicle sensors in the plurality of vehicle sensors, comprising: Determine the mean and variance corresponding to the data detected by any two vehicle sensors; determining a Welsh statistic and degrees of freedom using the mean and the variance; The Welsh variance is determined using the Welsh statistic and the degrees of freedom.

6. The scene detection method according to claim 1, characterized in that: When the non-matching rate is greater than a preset threshold, after determining a vehicle operation scenario corresponding to the non-matching rate, the method further includes: storing the operation data segment of the current frame corresponding to the vehicle operation scenario into a first data packet; In the first data packet, a first data segment of a preset duration corresponding to before the current frame and a second data segment of a preset duration corresponding to after the current frame are concatenated to obtain a target data packet.

7. A scene detection device, characterized in that: The device comprises: An acquisition module, used to acquire real-time measurement data detected by multiple vehicle sensors during vehicle driving; A first determination module, configured to determine a correlation matrix using the real-time measurement data and the historical measurement data of the plurality of vehicle sensors, wherein the correlation matrix is ​​used to represent the similarity between the real-time measurement data and the historical measurement data; A matching module, used for performing target matching on the real-time measurement data using the association matrix and the Hungarian algorithm to obtain a matched target object, wherein the target object is an object detected by all the vehicle sensors; A second determination module, configured to determine a mismatch rate of the plurality of vehicle sensors using the target object; The third determination module is used to determine the vehicle operation scenario corresponding to the non-matching rate when the non-matching rate is greater than a preset threshold.

8. A terminal device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the scene detection method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the scene detection method according to any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The computer product comprises a computer program, and when the computer program is executed by a processor, the scene detection method according to any one of claims 1 to 6 is implemented.

Images