Radar installation angle detection method and system based on image recognition

Abstract

The invention discloses a radar installation angle detection method and system based on image recognition, and relates to the technical field of intelligent detection, and the method comprises the steps: carrying out the multi-angle image collection of a target vehicle and surrounding objects, and generating a plurality of image data; predicting the relative distance between the target vehicle and the surrounding object; building a radar detection twinborn model in the visual simulation platform; and taking an installation angle threshold value of the vehicle-mounted radar equipment as an optimization space, carrying out simulation optimization on the installation angle based on the radar detection twin model and the plurality of relative distance prediction results, generating an optimal radar installation angle, carrying out angle detection on the vehicle-mounted radar equipment of the target vehicle, and carrying out optimization adjustment according to an angle detection result. According to the invention, the technical problem that the automatic driving function is influenced by the installation angle error of the vehicle-mounted radar and the accident probability is improved in the prior art is solved, and the technical effects of improving the precision and accuracy of the installation angle detection of the vehicle-mounted radar and further improving the installation quality of the radar are achieved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent detection technology, and in particular to a radar installation angle detection method and system based on image recognition. Background Art

[0002] As the core sensor of autonomous driving, the on-board radar sensor relies on the precise installation angle to accurately locate obstacles in the detection area. If the installation angle is offset, the trajectory information of the obstacle it outputs will also be offset. Inaccurate positioning information can lead to the false triggering of autonomous driving related functions, thereby increasing the probability of safety accidents. Summary of the invention

[0003] The present application provides a radar installation angle detection method and system based on image recognition, which is used to solve the technical problem in the prior art that the installation angle error of the vehicle-mounted radar will affect the automatic driving function and increase the probability of accidents.

[0004] The first aspect of the present application provides a radar installation angle detection method based on image recognition, the method comprising: performing multi-angle image acquisition of a target vehicle and surrounding objects through a CCD image sensor to generate multiple image data; predicting the relative distance between the target vehicle and surrounding objects based on the multiple image data to generate multiple relative distance prediction results, wherein the relative distance prediction results have a mapping relationship with a single object in the surrounding objects; in a visual simulation platform, building a 3D twin building model of the target vehicle and surrounding objects and a radar twin model of a vehicle-mounted radar device of the target vehicle, and embedding the radar twin model into the 3D twin building model to generate a radar detection twin model; taking the installation angle threshold of the vehicle-mounted radar device as the optimization space, simulating and optimizing the installation angle of the vehicle-mounted radar device based on the radar detection twin model and the multiple relative distance prediction results to generate an optimal radar installation angle; performing angle detection on the vehicle-mounted radar device of the target vehicle based on the optimal radar installation angle, and optimizing and adjusting the installation angle of the vehicle-mounted radar device according to the angle detection result.

[0005] The second aspect of the present application provides a radar installation angle detection system based on image recognition, the system comprising: a multi-angle image acquisition module, the multi-angle image acquisition module is used to perform multi-angle image acquisition of the target vehicle and surrounding objects through a CCD image sensor to generate multiple image data; a relative distance prediction result generation module, the relative distance prediction result generation module is used to predict the relative distance between the target vehicle and the surrounding objects according to the multiple image data, and generate multiple relative distance prediction results, wherein the relative distance prediction results have a mapping relationship with a single object in the surrounding objects; a radar detection twin model generation module, the radar detection twin model generation module is used to build a target vehicle and surrounding objects in a visual simulation platform. A 3D twin building model and a radar twin model of the on-board radar equipment of the target vehicle, and embedding the radar twin model into the 3D twin building model to generate a radar detection twin model; an optimal radar installation angle generation module, the optimal radar installation angle generation module is used to use the installation angle threshold of the on-board radar equipment as the optimization space, simulate and optimize the installation angle of the on-board radar equipment based on the radar detection twin model and the multiple relative distance prediction results, and generate an optimal radar installation angle; an installation angle optimization module, the installation angle optimization module is used to perform angle detection on the on-board radar equipment of the target vehicle based on the optimal radar installation angle, and optimize and adjust the installation angle of the on-board radar equipment according to the angle detection result.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The present application provides a radar installation angle detection method based on image recognition, which relates to the field of intelligent detection technology. By collecting multi-angle images of a target vehicle and surrounding objects, the relative distance between the target vehicle and surrounding objects is predicted, and a radar detection twin model is built in a visual simulation platform. The installation angle threshold of the vehicle-mounted radar equipment is used as the optimization space, and the installation angle is simulated and optimized to generate the optimal radar installation angle. The vehicle-mounted radar equipment of the target vehicle is angle detected, and optimized and adjusted according to the angle detection result. The technical problem that the installation angle error of the vehicle-mounted radar in the prior art will affect the automatic driving function and increase the probability of accidents is solved, and the precision and accuracy of the vehicle-mounted radar installation angle detection are improved, thereby improving the radar installation quality and distance measurement accuracy, and reducing the probability of accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1 A schematic diagram of a radar installation angle detection method based on image recognition provided in an embodiment of the present application;

[0010] Figure 2 A schematic diagram of a process for generating a radar detection twin model in a radar installation angle detection method based on image recognition provided in an embodiment of the present application;

[0011] Figure 3 A schematic diagram of a process for generating an optimal radar installation angle in a radar installation angle detection method based on image recognition provided in an embodiment of the present application;

[0012] Figure 4 A schematic diagram of the structure of a radar installation angle detection system based on image recognition provided in an embodiment of the present application.

[0013] Explanation of the accompanying drawings: multi-angle image acquisition module 11, relative distance prediction result generation module 12, radar detection twin model generation module 13, optimal radar installation angle generation module 14, installation angle optimization module 15. DETAILED DESCRIPTION

[0014] The present application provides a radar installation angle detection method based on image recognition, which is used to solve the technical problem in the prior art that the installation angle error of the vehicle-mounted radar will affect the automatic driving function and increase the probability of accidents.

[0015] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0016] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices.

[0017] Embodiment 1

[0018] like Figure 1 As shown, the present application provides a radar installation angle detection method based on image recognition, the method comprising:

[0019] P10: Use CCD image sensor to collect multi-angle images of the target vehicle and surrounding objects to generate multiple image data;

[0020] It should be understood that basic image acquisition is performed through a CCD image sensor, which is also called a charge coupled device. It is a special semiconductor material that can directly convert optical signals into analog current signals. The current signals are amplified and converted from analog to digital to achieve image acquisition, storage, transmission, processing and reproduction. It has the advantages of small size, low energy consumption, stable performance, high sensitivity, etc. The CCD image sensor is used to capture images of the target vehicle itself and objects around the target vehicle according to multiple preset acquisition angles, and multiple image data are generated through signal conversion, which are used as basic data for predicting the distance between the target vehicle and surrounding objects.

[0021] P20: Predicting the relative distance between the target vehicle and surrounding objects according to the plurality of image data, generating a plurality of relative distance prediction results, wherein the relative distance prediction results have a mapping relationship with a single object among the surrounding objects;

[0022] Furthermore, step P20 of the embodiment of the present application further includes:

[0023] P21: Perform grayscale image enhancement on the multiple image data to generate multiple grayscale image data;

[0024] P22: Construct a distance prediction model based on the BP neural network, and use the distance prediction model to perform relative distance prediction on the multiple grayscale image data in turn to generate multiple relative distance prediction results.

[0025] Specifically, the multiple image data are used to predict the relative distance between the target vehicle and the surrounding objects. The relative distance refers to the relative distance between the target vehicle and the nearby obstacles, which can be used as a reference indicator for adjusting the installation angle of the vehicle-mounted radar. The specific prediction process can be to first perform grayscale image enhancement on the multiple image data, first perform grayscale processing on the multiple images, that is, convert the color pictures into black and white grayscale images, and then perform binarization processing to convert the pixels in the image into only two colors, white and gray, so as to simplify the image and highlight the characteristics of the target object, and generate multiple grayscale image data.

[0026] Furthermore, based on big data, multiple groups of sample image distance prediction data are collected, including multiple sample grayscale images and corresponding multiple sample vehicle-object distances, and a distance prediction model is constructed in combination with a BP neural network. The BP neural network is a multi-layer feedforward neural network trained according to the error back propagation algorithm. It learns certain rules through its own training and obtains the result closest to the expected output value when the input value is given. The multiple groups of sample image distance prediction data are used as training data, and supervised training is performed on the distance prediction model until the output of the model converges and meets the preset accuracy requirements, and the training of the distance prediction model is completed. Furthermore, the multiple grayscale image data are input into the distance prediction model in turn to predict the vehicle-object distance, and multiple relative distance prediction results are obtained, which can be used as a reference indicator for adjusting the installation angle of the vehicle-mounted radar of the target vehicle.

[0027] P30: In the visual simulation platform, build a 3D twin building model of the target vehicle and surrounding objects and a radar twin model of the on-board radar equipment of the target vehicle, and embed the radar twin model into the 3D twin building model to generate a radar detection twin model;

[0028] Further, such as Figure 2 As shown, step P30 of the embodiment of the present application also includes:

[0029] P31: Obtaining specification parameters and operating parameters of the vehicle-mounted radar device;

[0030] P32: In the visual simulation platform, simulation modeling is performed on the vehicle-mounted radar device based on the specification parameters and the operating parameters to generate a radar twin model;

[0031] P33: Obtain the radar installation position coordinates of the target vehicle, and embed the radar twin model into the 3D twin building model based on the radar installation position coordinates to obtain the radar detection twin model.

[0032] Optionally, through a visual simulation platform, first use the coordinate parameters of the objects around the target vehicle and the structural parameters of the target vehicle to build a 3D twin building model of the target vehicle and the surrounding objects. Further, collect the specification parameters and operating parameters of the vehicle-mounted radar equipment to be installed, including equipment structure, size, transmission power, etc., and use the specification parameters and operating parameters to simulate and model the vehicle-mounted radar equipment to generate a radar twin model. Further, obtain the radar installation position coordinates of the target vehicle, and embed the radar twin model into the 3D twin building model with reference to the position coordinates, together forming a radar detection twin model, which can be used to optimize the installation angle of the vehicle-mounted radar.

[0033] P40: Taking the installation angle threshold of the vehicle-mounted radar device as the optimization space, simulating and optimizing the installation angle of the vehicle-mounted radar device based on the radar detection twin model and the multiple relative distance prediction results to generate an optimal radar installation angle;

[0034] It should be understood that the installation angle threshold of the vehicle-mounted radar equipment, that is, the recommended value range of the installation angle, is obtained, and this is used as the optimization space for the installation angle of the vehicle-mounted radar equipment, and the radar detection twin model is used to perform radar detection simulations at multiple different installation angles, and the multiple relative distance prediction results are used as evaluation criteria for the detection simulation results, so as to simulate and optimize the installation angle of the vehicle-mounted radar equipment, and select the test installation angle that is closest to the radar detection simulation result and the relative distance prediction result as the optimal radar installation angle, which can be used as a reference for adjusting the installation angle of the vehicle-mounted radar equipment of the target vehicle.

[0035] Further, such as Figure 3 As shown, step P40 of the embodiment of the present application also includes:

[0036] P41: randomly selecting an installation angle within the installation angle threshold without replacement as the first installation angle;

[0037] P42: In the radar detection twin model, based on the first installation angle, controlling the radar twin model to perform distance detection on surrounding objects, and generating a plurality of first simulated distance detection results;

[0038] P43: Evaluate the plurality of first simulated distance detection results by using a detection quality evaluation function to generate a first detection quality;

[0039] The expression of the detection quality evaluation function is:

[0040]

[0041] Among them, F jis the detection quality of the jth installation angle, M is the number of relative distance prediction results, G i is the i-th relative distance prediction result among multiple relative distance prediction results, J i It is the i-th simulated distance detection result among multiple simulated distance detection results of the j-th installation angle.

[0042] P44: randomly selecting an installation angle within the installation angle threshold as a second installation angle without replacement, and calculating a second detection quality of the second installation angle;

[0043] P45: When the first detection mass is less than or equal to the second detection mass, the second installation angle is used as the current optimal installation angle; when the first detection mass is greater than the second detection mass, the second installation angle is used as the current optimal installation angle according to probability, wherein the probability is the ratio of the difference between the total optimization number and the current optimization number to the total optimization number;

[0044] P46: Continuously perform iterative optimization until the current number of optimization times meets the total number of optimization times, then stop the optimization and output the current optimal installation angle as the optimal radar installation angle.

[0045] In a possible embodiment of the present application, an installation angle is randomly selected from the installation angle threshold without replacement as the first installation angle, and the radar detection twin model is used to adjust the installation angle of the radar twin model according to the first installation angle, and the distance detection simulation of the surrounding objects is performed to obtain multiple first simulated distance detection results based on different objects. Furthermore, the multiple first simulated distance detection results are evaluated using a detection quality evaluation function to generate a first detection quality, and the detection quality evaluation function is: Among them, F j is the detection quality of the jth installation angle, M is the number of relative distance prediction results, G i is the i-th relative distance prediction result among multiple relative distance prediction results, J i It is the i-th simulated distance detection result among multiple simulated distance detection results of the j-th installation angle.

[0046] By analogy, an installation angle is randomly selected from the installation angle threshold without replacement as the second installation angle, and the second detection quality of the second installation angle is obtained through distance detection simulation and detection quality calculation, and the first detection quality is compared with the second detection quality. When the first detection quality is less than or equal to the second detection quality, the second installation angle is used as the current optimal installation angle. When the first detection quality is greater than the second detection quality, the second installation angle is used as the current optimal installation angle according to probability to avoid falling into a local optimal solution, wherein the probability is the ratio of the difference between the total optimization number and the current optimization number to the total optimization number, that is, the probability decreases with the increase of the optimization number, and iterative optimization is continuously performed until the current optimization number meets the total optimization number, then the optimization is stopped, and the optimal installation angle of the last iteration is used as the optimal radar installation angle.

[0047] P50: Performing angle detection on the on-board radar device of the target vehicle based on the optimal radar installation angle, and optimizing and adjusting the installation angle of the on-board radar device according to the angle detection result.

[0048] Furthermore, step P50 of the embodiment of the present application also includes:

[0049] P51: Get the real-time installation angle of the target vehicle’s onboard radar device;

[0050] P52: Performing angle deviation analysis on the real-time installation angle according to the optimal radar installation angle to generate an angle deviation value;

[0051] P53: Optimize and correct the installation angle of the vehicle-mounted radar device based on the angle deviation value.

[0052] It should be understood that, based on the optimal radar installation angle, the angle detection of the on-board radar equipment of the target vehicle is performed, that is, the real-time installation angle of the on-board radar equipment of the target vehicle is obtained, and the angle deviation of the real-time installation angle is calculated according to the optimal radar installation angle to generate an angle deviation value, and the installation angle of the on-board radar equipment is adjusted with reference to the angle deviation value to improve the distance measurement accuracy of the on-board radar equipment and reduce the probability of accidents.

[0053] In summary, the embodiments of the present application have at least the following technical effects:

[0054] This application collects multi-angle images of the target vehicle and surrounding objects, predicts the relative distance between the target vehicle and surrounding objects, builds a radar detection twin model in a visual simulation platform, uses the installation angle threshold of the on-board radar equipment as the optimization space, simulates and optimizes the installation angle, generates the optimal radar installation angle, performs angle detection on the on-board radar equipment of the target vehicle, and optimizes and adjusts according to the angle detection results.

[0055] The technical effect of improving the precision and accuracy of vehicle-mounted radar installation angle detection, thereby improving radar installation quality and distance measurement accuracy and reducing accident probability is achieved.

[0056] Embodiment 2

[0057] Based on the same inventive concept as the radar installation angle detection method based on image recognition in the aforementioned embodiment, Figure 4 As shown, the present application provides a radar installation angle detection system based on image recognition, and the system and method embodiments in the present application are based on the same inventive concept. The system includes:

[0058] A multi-angle image acquisition module 11, which is used to acquire multi-angle images of the target vehicle and surrounding objects through a CCD image sensor to generate multiple image data;

[0059] A relative distance prediction result generating module 12, wherein the relative distance prediction result generating module 12 is used to predict the relative distance between the target vehicle and the surrounding objects according to the plurality of image data, and generate a plurality of relative distance prediction results, wherein the relative distance prediction results have a mapping relationship with a single object among the surrounding objects;

[0060] A radar detection twin model generation module 13, wherein the radar detection twin model generation module 13 is used to build a 3D twin building model of the target vehicle and surrounding objects and a radar twin model of the on-board radar device of the target vehicle in a visual simulation platform, and embed the radar twin model into the 3D twin building model to generate a radar detection twin model;

[0061] An optimal radar installation angle generation module 14 is used to use the installation angle threshold of the vehicle-mounted radar device as the optimization space, simulate and optimize the installation angle of the vehicle-mounted radar device based on the radar detection twin model and the multiple relative distance prediction results, and generate an optimal radar installation angle;

[0062] The installation angle optimization module 15 is used to perform angle detection on the vehicle-mounted radar device of the target vehicle based on the optimal radar installation angle, and optimize and adjust the installation angle of the vehicle-mounted radar device according to the angle detection result.

[0063] Furthermore, the relative distance prediction result generating module 12 is further used to perform the following steps:

[0064] Performing grayscale image enhancement on the multiple image data to generate multiple grayscale image data;

[0065] A distance prediction model is constructed based on the BP neural network, and relative distance prediction is performed on the plurality of grayscale image data in sequence through the distance prediction model to generate a plurality of relative distance prediction results.

[0066] Furthermore, the radar detection twin model generation module 13 is also used to perform the following steps:

[0067] Obtaining specification parameters and operating parameters of the vehicle-mounted radar device;

[0068] In a visual simulation platform, simulation modeling is performed on the vehicle-mounted radar device based on the specification parameters and the operating parameters to generate a radar twin model;

[0069] The radar installation position coordinates of the target vehicle are obtained, and based on the radar installation position coordinates, the radar twin model is embedded into the 3D twin building model to obtain a radar detection twin model.

[0070] Furthermore, the optimal radar installation angle generation module 14 is further configured to perform the following steps:

[0071] Randomly selecting an installation angle within the installation angle threshold without replacement as a first installation angle;

[0072] In the radar detection twin model, based on the first installation angle, the radar twin model is controlled to perform distance detection on surrounding objects to generate a plurality of first simulated distance detection results;

[0073] Evaluate the plurality of first simulated distance detection results by using a detection quality evaluation function to generate a first detection quality;

[0074] randomly selecting an installation angle within the installation angle threshold as a second installation angle without replacement, and calculating and obtaining a second detection mass of the second installation angle;

[0075] When the first detection mass is less than or equal to the second detection mass, the second installation angle is used as the current optimal installation angle; when the first detection mass is greater than the second detection mass, the second installation angle is used as the current optimal installation angle according to probability, wherein the probability is the ratio of the difference between the total optimization number and the current optimization number to the total optimization number;

[0076] The iterative optimization is continuously performed until the current optimization times meet the total optimization times, then the optimization is stopped and the current optimal installation angle is output as the optimal radar installation angle.

[0077] Furthermore, the optimal radar installation angle generation module 14 is further configured to perform the following steps:

[0078] The expression of the detection quality evaluation function is:

[0079]

[0080] Among them, F j is the detection quality of the jth installation angle, M is the number of relative distance prediction results, G i is the i-th relative distance prediction result among multiple relative distance prediction results, J i It is the i-th simulated distance detection result among multiple simulated distance detection results of the j-th installation angle.

[0081] Furthermore, the installation angle optimization module 15 is also used to perform the following steps:

[0082] Obtain the real-time installation angle of the on-board radar device of the target vehicle;

[0083] Performing angle deviation analysis on the real-time installation angle according to the optimal radar installation angle to generate an angle deviation value;

[0084] The installation angle of the vehicle-mounted radar device is optimized and corrected based on the angle deviation value.

[0085] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. Other embodiments are within the scope of the attached claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0086] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0087] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A radar installation angle detection method based on image recognition, characterized in that: The method comprises: Capture multi-angle images of the target vehicle and surrounding objects through the CCD image sensor to generate multiple image data; Predicting the relative distance between the target vehicle and surrounding objects based on the multiple image data to generate multiple relative distance prediction results, wherein the relative distance prediction results have a mapping relationship with a single object among the surrounding objects; In the visual simulation platform, a 3D twin building model of the target vehicle and surrounding objects and a radar twin model of the on-board radar equipment of the target vehicle are constructed, and the radar twin model is embedded in the 3D twin building model to generate a radar detection twin model; Taking the installation angle threshold of the vehicle-mounted radar device as the optimization space, simulating and optimizing the installation angle of the vehicle-mounted radar device based on the radar detection twin model and the multiple relative distance prediction results, and generating an optimal radar installation angle; An angle detection is performed on the on-board radar device of the target vehicle based on the optimal radar installation angle, and the installation angle of the on-board radar device is optimized and adjusted according to the angle detection result.

2. The method according to claim 1, characterized in that The step of predicting the relative distance between the target vehicle and surrounding objects based on the plurality of image data to generate a plurality of relative distance prediction results further includes: Performing grayscale image enhancement on the multiple image data to generate multiple grayscale image data; A distance prediction model is constructed based on the BP neural network, and relative distance prediction is performed on the plurality of grayscale image data in sequence through the distance prediction model to generate a plurality of relative distance prediction results.

3. The method according to claim 1, characterized in that The generating of the radar detection twin model further includes: Obtaining specification parameters and operating parameters of the vehicle-mounted radar device; In a visual simulation platform, simulation modeling is performed on the vehicle-mounted radar device based on the specification parameters and the operating parameters to generate a radar twin model; The radar installation position coordinates of the target vehicle are obtained, and based on the radar installation position coordinates, the radar twin model is embedded into the 3D twin building model to obtain a radar detection twin model.

4. The method according to claim 3, characterized in that The method of taking the installation angle threshold of the vehicle-mounted radar device as the optimization space, simulating and optimizing the installation angle of the vehicle-mounted radar device based on the radar detection twin model and the multiple relative distance prediction results to generate the optimal radar installation angle also includes: Randomly selecting an installation angle within the installation angle threshold without replacement as a first installation angle; In the radar detection twin model, based on the first installation angle, the radar twin model is controlled to perform distance detection on surrounding objects to generate a plurality of first simulated distance detection results; Evaluate the plurality of first simulated distance detection results by using a detection quality evaluation function to generate a first detection quality; randomly selecting an installation angle within the installation angle threshold as a second installation angle without replacement, and calculating and obtaining a second detection mass of the second installation angle; When the first detection mass is less than or equal to the second detection mass, the second installation angle is used as the current optimal installation angle; when the first detection mass is greater than the second detection mass, the second installation angle is used as the current optimal installation angle according to probability, wherein the probability is the ratio of the difference between the total optimization number and the current optimization number to the total optimization number; The iterative optimization is continuously performed until the current optimization times meet the total optimization times, then the optimization is stopped and the current optimal installation angle is output as the optimal radar installation angle.

5. The method according to claim 4, characterized in that The method further comprises: The expression of the detection quality evaluation function is: Among them, F j is the detection quality of the jth installation angle, M is the number of relative distance prediction results, G i is the i-th relative distance prediction result among multiple relative distance prediction results, J i It is the i-th simulated distance detection result among multiple simulated distance detection results of the j-th installation angle.

6. The method according to claim 1, characterized in that The method further comprises: Obtain the real-time installation angle of the on-board radar device of the target vehicle; Performing angle deviation analysis on the real-time installation angle according to the optimal radar installation angle to generate an angle deviation value; The installation angle of the vehicle-mounted radar device is optimized and corrected based on the angle deviation value.

7. A radar installation angle detection system based on image recognition, characterized in that: The system comprises: A multi-angle image acquisition module, which is used to acquire multi-angle images of the target vehicle and surrounding objects through a CCD image sensor to generate multiple image data; A relative distance prediction result generation module, the relative distance prediction result generation module is used to predict the relative distance between the target vehicle and the surrounding objects according to the multiple image data, and generate multiple relative distance prediction results, wherein the relative distance prediction results have a mapping relationship with a single object among the surrounding objects; A radar detection twin model generation module, wherein the radar detection twin model generation module is used to build a 3D twin building model of the target vehicle and surrounding objects and a radar twin model of the on-board radar device of the target vehicle in a visual simulation platform, and embed the radar twin model into the 3D twin building model to generate a radar detection twin model; An optimal radar installation angle generation module, wherein the optimal radar installation angle generation module is used to simulate and optimize the installation angle of the vehicle-mounted radar device based on the radar detection twin model and the multiple relative distance prediction results, using the installation angle threshold of the vehicle-mounted radar device as the optimization space, to generate an optimal radar installation angle; The installation angle optimization module is used to perform angle detection on the on-board radar device of the target vehicle based on the optimal radar installation angle, and optimize and adjust the installation angle of the on-board radar device according to the angle detection result.

Images