A storage and reading method of an automatic driving simulation 3D model radar reflectivity

Abstract

The application relates to the technical field of automatic driving, and discloses a storage and reading method for radar reflectivity of an automatic driving simulation 3D model, which comprises the following steps: collecting and calculating real reflectivity of objects in an automatic driving related scene; embedding the real reflectivity of the objects in the automatic driving related scene into a texture map on the surface of a 3D model; and calculating and reading radar reflection intensity feedback data. According to the application, a radar reflectivity field is embedded in the texture of the 3D model, the radar reflection characteristics of different objects are accurately simulated, the influence of environmental factors on radar reflection is considered, and efficient storage and reading of data are realized, so that the support capability of an automatic driving simulation system for end-to-end testing is improved, and the authenticity and effectiveness of simulation are enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, in particular to a storage and reading method of automatic driving simulation 3D model radar reflectivity. BACKGROUND

[0002] With the rapid development of automatic driving technology, simulation testing has become an important means for the development and verification of automatic driving systems. Simulation testing can provide a flexible and safe environment to verify the performance of automatic driving systems in complex scenarios. However, current simulation technology still faces many challenges, especially in the simulation of real perception data, the simulation of radar (including lidar and millimeter wave radar) data is particularly prominent.

[0003] Currently, simulation systems usually rely on three main testing methods:

[0004] 1. Road testing and structured testing in real scenarios:

[0005] This method verifies the performance of automatic driving systems through actual road testing, which can provide the most realistic environment data. However, it is costly and difficult to test in some dangerous or difficult-to-reproduce scenarios. In addition, real scenario testing consumes a lot of time, and environmental changes (such as weather, road conditions, traffic conditions) are difficult to control, and there are certain safety hazards.

[0006] 2. Testing based on real data playback:

[0007] This method simulates scenarios by playing back collected road data. It ensures the authenticity of the test environment, but lacks interactivity and flexibility. For example, the vehicle may face environmental changes that do not match the preset during the playback data process, and the test cannot reflect the environmental response after the autonomous decision of the autonomous vehicle. This method also cannot simulate complex physical effects (such as reflection, occlusion, etc.).

[0008] 3. Testing based on simulator simulation:

[0009] Simulators can flexibly construct various traffic scenarios, reduce testing costs, and simulate extreme or rare scenarios. However, existing simulators usually rely on simplified physical models, such as car dynamics simulation and game engine technology, which are difficult to accurately simulate the reflection characteristics of radar sensors, resulting in a large gap between simulation results and the real world. The simulation of radar data is insufficient, especially in terms of reflectivity, propagation characteristics, and multipath effects, which often affects the reliability of simulation results and cannot meet the end-to-end verification needs of automatic driving systems.

[0010] Therefore, improving the simulation accuracy of radar sensor data, especially the simulation of radar reflection characteristics, has become a key technical problem in current autonomous driving simulation research. Radar sensors (laser radar and millimeter wave radar) are important components of the autonomous driving perception system, and accurate radar data is crucial for target detection, environment modeling and path planning. How to accurately and realistically simulate radar data in the simulation environment, especially to simulate the interaction between radar reflectivity and actual objects, has become a key to improving simulation accuracy and testing efficiency. SUMMARY

[0011] To solve the above problems, the present application provides a storage and reading method for 3D model radar reflectivity in autonomous driving simulation, to solve the problem of insufficient radar reflectivity simulation in current autonomous driving simulation testing. Specifically, the present application can solve the following core problems:

[0012] 1. How to accurately simulate the radar reflection characteristics of different objects in the simulator.

[0013] The reflection characteristics of objects in current simulation systems usually use uniform reflectivity parameters, ignoring the influence of different materials, surface roughness and geometric shapes on radar reflection characteristics.

[0014] The present application accurately embeds different radar reflectivity fields in each object's 3D model, dynamically adjusts the reflectivity according to the material type, surface structure and geometric shape of the object, so that each object can exhibit consistent radar reflection characteristics with the real scene in the simulation environment.

[0015] 2. How to consider the influence of environmental factors on radar reflection.

[0016] Radar reflection intensity is not only related to the reflectivity of the object itself, but also affected by distance, environmental attenuation, weather conditions (such as rain, fog, snow, etc.) and multipath effect. Existing simulation methods are difficult to fully simulate these influences.

[0017] The present application proposes a comprehensive calculation model combining distance attenuation, object reflectivity, environmental attenuation and multipath effect, which dynamically generates radar feedback data consistent with real conditions during simulation.

[0018] 3. How to achieve efficient storage, reading and portability of radar reflectivity data.

[0019] Traditional simulation methods usually store radar reflectivity separately from 3D models, causing difficulties in data synchronization and management.

[0020] The application innovatively embeds radar reflectivity information directly into the texture map (such as a JPEG file) of a 3D model, storing the reflectivity data by embedding a "reflection" field in the JPEG metadata area. This method not only improves storage efficiency, but also enables fast reading and parsing of reflectivity information during simulation, ensuring efficient use of data and real-time performance.

[0021] 4. How to improve the support capability of simulation system for end-to-end automatic driving test.

[0022] The current simulation system lacks accurate simulation of radar reflection characteristics when conducting complex end-to-end testing of automatic driving systems, often leading to inconsistent test results.

[0023] By accurately simulating the changes in radar reflectivity and combining dynamic changes in the scene and real-time feedback, the simulation system can better support end-to-end testing of automatic driving systems, especially comprehensive verification of perception modules such as object detection, path planning, and environment modeling.

[0024] By solving the above core problems, the application can significantly improve the authenticity of radar data in automatic driving simulation systems and enhance the simulation environment's ability to simulate complex scenarios and dynamic changes, providing more efficient and accurate support for the research and verification of automatic driving technology.

[0025] To achieve the above purpose, the technical solution adopted by the application is as follows:

[0026] In a first aspect, the application provides a method for storing and reading radar reflectivity of automatic driving simulation 3D models, comprising:

[0027] Collecting and calculating the real reflectivity of automatic driving related scene objects;

[0028] Embedding the real reflectivity of automatic driving related scene objects into the texture map of the surface of the 3D model;

[0029] Calculating and reading radar reflectivity feedback data.

[0030] Further, collecting and calculating the real reflectivity of automatic driving related scene objects includes:

[0031] According to the object in the automatic driving scene, the object is classified, and the physical characteristics of the object surface are analyzed; wherein the object in the automatic driving scene includes a vehicle, a windshield, a building, a road surface, a street lamp column, the physical characteristics of the object surface include material properties, surface roughness and geometric characteristics, the material properties include metal, plastic, glass, concrete and asphalt, and the geometric characteristics include the shape of the object; based on the object classification and physical characteristic analysis, a basic reflectivity parameter range is set for each type of object to provide a reference for subsequent collection and modeling;

[0032] The correction method combining theoretical method and actual collected data is used to calibrate the data;

[0033] Real scene data is acquired and denoising and normalization processing is performed; wherein the normalization formula is:

[0034]

[0035] In the formula, R 归一化 is the normalized reflectivity value; R max and R min are the minimum and maximum reflectivity values in different objects respectively; R is the reflectivity value of the object;

[0036] A reflection intensity calculation formula is constructed and is expressed as:

[0037] P = f (material, roughness, angle, distance, wavelength)

[0038] In the formula, P is the reflection intensity; in the function f, the material determines the absorption and reflection proportion of the object to the electromagnetic wave; the roughness affects the distribution characteristics of the reflection direction; the angle and the distance change the intensity distribution of the received signal; the wavelength determines the penetration and scattering characteristics of the wave;

[0039] The reflection intensity characteristics of different object categories are classified, distributed, fitted and standardized stored.

[0040] Further, the correction method combining theoretical method and actual collected data is used to calibrate the data, including:

[0041] If it is a laser radar, the reflectivity and the laser echo energy are positively correlated, and the formula for modeling the laser radar is:

[0042]

[0043] In the formula, P r is the laser receiving power; P s is the laser transmitting power; D r is the size of the radar receiving aperture; η atm is an atmospheric effect constant; η sysis the optical transmission coefficient; R is the distance; β is the laser beam width; and σ is the effective backscattering cross-sectional area.

[0044] Further, the reflection intensity characteristics of different object categories are classified, distribution fitted, and standardized stored, including:

[0045] Based on the characteristic distribution curve, the collected reflectivity data is classified by material, surface roughness, and geometric shape for each category of object;

[0046] A distribution fitting algorithm is used to generate a reflectivity characteristic curve for the corresponding category, representing the law of reflectivity of the object with respect to the change of incident angle and wavelength;

[0047] A standardized radar reflectivity database is established, and the reflectivity characteristics of each object category are stored as a parameterized model, including numerical values or functions of reflectivity with respect to variables such as distance, angle, and wavelength;

[0048] The reflectivity characteristics of each material are stored in JSON format.

[0049] Further, the real reflectivity of objects in an autonomous driving related scene is embedded into the texture map of the 3D model surface, including:

[0050] A 3D model of the target object in the simulation scene is established, with the requirements of matching the surface material of the target object with the actual scene, and matching the scale, texture, and details of the target object with the real scene.

[0051] The 3D model surface is mapped to a two-dimensional image space to generate a texture map, and a JPEG format texture map is exported, which includes an APP1 segment for storing reflectivity data, and the APP1 segment is used for EXIF metadata.

[0052] The reflectivity is stored using the EXIF UserComment field to embed the reflectivity of each object.

[0053] The reflectivity value is represented in color on the texture map, the reflectivity value is mapped to a heat map, and the reflectivity heat map is superimposed on the original texture map.

[0054] Further, the radar reflection intensity feedback data is calculated and read, including:

[0055] In the simulation system, the direction and angle range of the current radar scan line beam are calculated according to the installation position and design parameters of the radar; the direction and angle range of the current radar scan line beam determine the object area that each radar beam line may touch in the simulation scene, and the simulator calculates the emission path of the radar point at the current angle step by step to form complete radar point cloud data through step-by-step scanning.

[0056] Based on the radar point cloud data, it is determined by a ray tracing algorithm whether the propagation path of each radar point collides with a certain object surface. When the collision occurs, the simulator records the 3D object in the radar click, the collision position and the corresponding surface normal information; in combination with the scene tree or acceleration structure of the simulation environment, the target object in the complex scene is located;

[0057] The 3D object surface based on the radar point collision contains a pre-embedded radar reflectivity field. The simulation system obtains the radar reflectivity parameter of the current object by reading the set field in the texture map. If the reflectivity field is empty or undefined, a default parameter library is called to give a reasonable reflectivity value to the unknown material. According to the read reflectivity value, the direction and intensity of the radar echo are calculated in combination with the surface normal and the incident angle;

[0058] A physical simulation model is established according to the radar reflectivity of the object surface, the distance attenuation and the environmental conditions, the radar reflection intensity is comprehensively calculated, and radar reflection intensity feedback data is generated; wherein the radar reflection intensity feedback data includes the position coordinates and reflection intensity values of each radar point;

[0059] Based on the position coordinates and reflection intensity values of each radar point, the point cloud data is output.

[0060] Further, after outputting the point cloud data based on the position coordinates and reflection intensity values of each radar point, the method further comprises:

[0061] If the current radar scan beam has not yet completed the preset angle range, the calculation and reflectivity calculation of the next radar beam line are continued until all scans are completed. After the scan is completed, the complete point cloud data is stored and output as a simulation result.

[0062] In a second aspect, the application provides a storage and reading device for automatic driving simulation 3D model radar reflectivity, the device comprising:

[0063] The reflectivity calculation module is configured to collect and calculate the real reflectivity of the automatic driving related scene object;

[0064] The reflectivity embedding module embeds the real reflectivity of the automatic driving related scene object into the texture map of the surface of the 3D model;

[0065] The feedback data calculation module is configured to calculate and read the radar reflection intensity feedback data.

[0066] In a third aspect, the application provides a storage and reading system for automatic driving simulation 3D model radar reflectivity, the system comprising: a memory for storing a computer program; a processor for executing the computer program to realize the method as described above.

[0067] In a fourth aspect, the application provides a non-transitory computer-readable storage medium storing instructions that, when executed by a processor, perform the method described above.

[0068] The application has at least the following beneficial effects:

[0069] The application embeds the storage and reading of radar reflectivity in the simulation environment, achieving high-precision simulation of real radar signals and providing strong support for the testing of autonomous driving systems. The beneficial effects are reflected in the following two points:

[0070] 1. The application can improve the accuracy of radar reflection characteristic simulation.

[0071] Fine reflectivity modeling: By embedding radar reflectivity fields in the texture map of 3D models, the application can dynamically reflect the influence of object material, surface roughness, and geometric shape on radar reflection.

[0072] Multi-material support: Different materials (such as metal, glass, plastic, concrete, etc.) have different reflection characteristics in the simulation environment. The application can define exclusive reflectivity parameters for each material to ensure that the simulation data is highly consistent with the actual scene.

[0073] Realism enhancement: By accurately depicting the reflection characteristics of each object, the application breaks through the limitations of traditional simulation methods using uniform reflectivity parameters, providing simulation results that are closer to the real world and improving the overall realism and credibility of the simulation environment.

[0074] 2. Provide an efficient storage and reading scheme.

[0075] Embedded storage design: The application innovatively embeds radar reflectivity data directly into the metadata area of the texture map (such as a JPEG file) of a 3D model by adding a "reflection" field in the JPEG file to store reflectivity information.

[0076] Data synchronization and management simplification: Reflectivity information is tightly coupled with 3D model data, avoiding the data synchronization and management difficulties brought by traditional separate storage, improving the convenience and consistency of data management.

[0077] Fast reading and real-time analysis: The simulator automatically parses the reflectivity field in the JPEG file when loading the 3D model through a plug-in module, achieving efficient reading and real-time application of reflectivity data, significantly improving the running efficiency and response speed of the simulation system.

[0078] Extensibility and compatibility: The present application is compatible with existing 3D modeling and image storage formats, can be seamlessly integrated into mainstream simulation engines (such as Unity, Unreal Engine), and supports the extension storage of future other physical properties (such as absorption rate, refractive index), providing good extensibility and adaptability. BRIEF DESCRIPTION OF DRAWINGS

[0079] Figure 1 A flowchart of a method for storing and reading radar reflectivity of an autonomous driving simulation 3D model according to an embodiment of the present application is shown.

[0080] Figure 2 A comparison chart of a simulation traffic scene and a real traffic scene according to an embodiment of the present application is shown.

[0081] Figure 3 A test equipment schematic diagram according to an embodiment of the present application is shown.

[0082] Figure 4 A laser radar point cloud diagram according to an embodiment of the present application is shown.

[0083] Figure 5 A flowchart of a radar reflection intensity acquisition process in a simulation environment according to an embodiment of the present application is shown.

[0084] Figure 6 A structural diagram of a device for storing and reading radar reflectivity of an autonomous driving simulation 3D model according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0085] The embodiments of the present application will be described in detail with specific reference to particular examples. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied in other different specific embodiments, and the details in this specification can be modified or changed based on different views and applications without departing from the spirit of the present application. It should be noted that the following examples and features in the examples can be combined with each other without conflict.

[0086] The specific embodiments of the present application will be described in further detail below with reference to the accompanying drawings and examples.

[0087] The embodiments of the present application provide a method for storing and reading radar reflectivity of an autonomous driving simulation 3D model, which accurately simulates the radar reflection characteristics of different objects by embedding a radar reflectivity field in the 3D model texture, considers the influence of environmental factors on radar reflection, and realizes efficient storage and reading of data, to improve the support capability of the autonomous driving simulation system for end-to-end testing, and enhance the reality and effectiveness of the simulation.

[0088] Specifically, please refer toFigure 1 A flowchart of a storage and reading method for an autonomous driving simulation 3D model radar reflectivity, the storage and reading method for the autonomous driving simulation 3D model radar reflectivity comprising steps S1 to S3, which are described in detail as follows.

[0089] S1: Collection and calculation of real reflectivity of objects in an autonomous driving related scene.

[0090] In order to realize accurate simulation of radar reflectivity in a real scene, the first step of the method is to collect and calculate real reflectivity data for various typical objects. The following are the specific implementation details and technical explanations of step S1:

[0091] S11: Object classification and reflectivity characteristic analysis.

[0092] According to the common objects in the autonomous driving scene (such as vehicles, windshields, buildings, road surfaces, street lamp posts, etc.), these objects are classified and their surface physical characteristics are analyzed: material properties: including metal, plastic, glass, concrete, asphalt, etc., different materials have significant differences in electromagnetic wave reflection characteristics. Surface roughness: the smoothness or roughness of the surface directly affects the reflection direction and intensity of electromagnetic waves. Geometric characteristics: the shape of the object (such as flat, curved, angular) will cause different multi-path effects and scattering characteristics.

[0093] Through the analysis of these properties, a basic reflectivity parameter range can be defined for each type of object, providing a reference for subsequent data collection and modeling.

[0094] S12: Data calibration.

[0095] In early studies, intensity information was generally used directly without any processing, which had a great impact on the reflected intensity received by the radar later, therefore, the laser radar used in the embodiments of the present application is calibrated in multiple steps to ensure the accuracy and consistency of the reflectivity data. Overall, intensity correction can be mainly divided into two categories: theoretical correction and empirical correction. Theoretical correction focuses on understanding the factors that affect intensity changes under ideal conditions, such as atmospheric attenuation, scanner characteristics, target surface characteristics, and the geometric relationship of data collection. In contrast, empirical correction relies on establishing an actual functional relationship between intensity and factors such as distance, incidence angle, surface roughness, and color in field data. In terms of accuracy, the theoretical correction method is higher than the empirical correction method, and the embodiments of the present application use a correction method that combines theoretical methods with actual collected data to further improve the accuracy of radar reflection intensity.

[0096] Taking a laser radar as an example, the reflectivity and laser echo energy are positively correlated, and the formula for modeling them is:

[0097]

[0098] where P r is the laser receiving power; P s is the laser transmitting power; D r is the size of the radar receiving aperture; η atm is the atmospheric effect constant; η sys is the optical transmission coefficient; R is the distance; β is the laser beam width; and σ is the effective backscattering cross-sectional area.

[0099] In the actual intensity correction process, due to the deviation between the theoretical value and the actual situation, in order to improve the accuracy of correction, several commonly used parameter estimation techniques can be used: first, set some combinations of parameters to remain constant; second, specifications provided by the equipment manufacturer can be used; finally, the specific values of the parameters can be determined by analyzing statistical data. These methods help to reduce the error in the model, thereby improving the accuracy of the correction.

[0100] S13: Real scene data acquisition and preprocessing.

[0101] As shown in Figure 2 , it is a comparison chart of the simulation traffic scene and the real traffic scene. The "real scene data" described in this embodiment is collected by a test device in a real traffic scene. As shown in Figure 3 , it is a schematic diagram of the test device. In this embodiment, a car is used as the test device, which carries a sensor device to collect real scene data, and a vehicle-mounted chip carried by the car can be used to realize the preprocessing of the collected data. The specific data acquisition and preprocessing method will be described in detail later.

[0102] Using sensor devices such as millimeter wave radars and laser radars, the reflectivity data of typical objects are collected in experimental environments and actual roads:

[0103] Device selection:

[0104] The millimeter wave radar selected is ARS 408-21 produced by Germany's Continental Company. ARS 408-21 millimeter wave radar has the following technical parameters: working frequency band 77GHz, detection distance 0.20m to 250m, azimuth angle field of view ±60° to ±9.0°, pitch angle field of view 14° to 20°, distance resolution 1.79m to 0.39m, speed range -400km / h to +200km / h, speed accuracy ±0.1km / h, and support CAN bus interface. It is mainly used to measure the large-area reflection intensity of objects, and is sensitive to material and distance.

[0105] The laser radar selected is RS-Ruby-80 produced by Speedtech, which is an 80-line laser radar with a 0.1° vertical angle resolution and a ranging capability of 180 meters @10% NIST. It outputs approximately 1.44 million points per second (single echo) and 2.88 million points per second (double echo) in balanced mode, and approximately 2.88 million points per second (single echo) and 5.76 million points per second (double echo) in high-performance mode. The operating voltage is 9-32V, the operating temperature range is -40°C to +60°C, the weight is approximately 1.85kg, and the protection level is IP67 and IP6K9K. The 80-line laser radar has higher angle resolution, suitable for precise measurement of small objects and edge reflection characteristics.

[0106] Collection process:

[0107] The radar sensor is installed on a fixed platform to scan the surface of the object at different angles and distances, and record the intensity of the radar return signal. In the experiment, environmental conditions such as light, humidity, rain and fog are controlled to ensure the comprehensiveness and diversity of data collection. The schematic diagram of the collected laser radar point cloud is shown in Figure 4 .

[0108] The reflectivity values under different materials and angles are obtained by measurement as shown in Table 1:

[0109] Table 1: Reflectivity value table under different materials and angles

[0110]

[0111] Denoising: Apply filtering algorithm (median filter) to remove random noise and improve data smoothness. For abnormal points (such as sudden high or low reflection values), use anomaly detection technology for rejection or correction.

[0112] Normalization and standardization: According to the calibration results, normalize the data collected by different devices to a unified scale range (such as 0 to 1) for subsequent processing and storage.

[0113] Normalization formula:

[0114]

[0115] In the formula, R 归一化 is the normalized reflectivity value; R max and R min are the minimum and maximum reflectivity values in different objects, respectively; R is the reflectivity value of the object.

[0116] Group the objects by material and scene, and record the normalized values as shown in the table:

[0117] Table 2 Normalized reflectivity of different materials

[0118]

[0119] S14: Construction of reflection intensity calculation formula

[0120] The reflection intensity not only depends on the properties of the object, but also closely related to the measurement conditions (such as distance, angle of incidence, wavelength, etc.). Therefore, it is necessary to construct a mathematical model to summarize and calculate the collected data:

[0121] Basic reflection intensity formula:

[0122] ρ = f(material, roughness, angle, distance, wavelength)

[0123] Where: ρ is the reflection intensity; in the function f, the material determines the absorption and reflection ratio of the object to electromagnetic waves; roughness affects the distribution characteristics of the reflection direction; angle and distance change the intensity distribution of the received signal; wavelength (millimeter wave and laser wavelength) determines the penetration and scattering characteristics of the wave.

[0124] S15: Generation and classified storage of reflectivity data distribution.

[0125] After completing the collection and calibration of the basic reflection intensity parameters, the reflection intensity characteristics of different object categories need to be classified, distribution fitted and standardized stored for subsequent simulation modeling. The specific implementation steps of step S15 are as follows:

[0126] S151: Generate characteristic distribution curve, for each type of object (such as metal vehicles, glass windows, asphalt pavement, etc.), classify the collected reflectivity data by material, surface roughness and geometric shape.

[0127] S152: Use distribution fitting algorithm (Gaussian distribution) to generate the reflectivity characteristic curve of this category, which represents the law of reflectivity change with incident angle and wavelength.

[0128] Example: The surface of a vehicle may have a high reflectivity peak point, while a concrete pavement may exhibit a more uniform low reflectivity distribution.

[0129] S153: Classified storage: Establish a standardized radar reflectivity database, store the reflectivity characteristics of each object as a parameterized model, including the numerical or functional form of reflectivity change with distance, angle, wavelength, etc.

[0130] S154: The data format needs to be compatible with the simulator requirements, and the reflectivity characteristics of each material are stored in JSON format for subsequent parsing.

[0131] S2: Generation of 3D model surface image and reflectivity embedding.

[0132] The measured reflectance data is embedded into the texture map (JPEG file) of the 3D model surface, providing a convenient format for the simulation engine to load and read the data.

[0133] In some embodiments, the implementation steps of step S2 are as follows:

[0134] S21: Generate surface texture image of 3D model.

[0135] The implementation steps of step S21 are as follows:

[0136] S211: Create 3D model of target object.

[0137] Use professional modeling software (such as Blender, 3ds Max, Maya) to create the target object in the simulation scene.

[0138] Modeling requirements: The surface material of the object should match the actual scene (such as smooth metal, rough concrete). Ensure that the scale, texture, and details of the object are as realistic as possible.

[0139] S212: Generate surface texture map.

[0140] Use UV unwrapping technology to map the 3D model surface to a two-dimensional image space to generate a texture map. Export the texture map in JPEG format.

[0141] S22: Add reflectance field to JPEG metadata.

[0142] It should be noted that a JPEG file is composed of multiple segments, common segments include: SOI (Start of Image): File start flag. APPn (Application-specific segments): Used to store user-defined data. DQT / DHT / DNL: Quantization table, Huffman table, etc. Image data. The APP1 segment of JPEG (usually used for EXIF metadata) supports extended fields, which is very suitable for storing reflectance data.

[0143] Therefore, in certain embodiments, the reflectance field is embedded by the following steps:

[0144] S221: Store reflectance using EXIF metadata.

[0145] S222: Embed the reflectance of each object using the UserComment field of EXIF.

[0146] Exemplarily:

[0147] mmwave_reflectivity: 0.85;

[0148] lidar_reflectivity: 0.9.

[0149] Example Chinese pseudocode:

[0150] from PIL import Image

[0151] from PIL.ExifTags import TAGS,GPSTAGS

[0152] # Open JPEG image

[0153] img = Image.open("texture.jpg")

[0154] # Add reflectivity metadata

[0155] exif_data = img.getexif()

[0156] exif_data

[37510] = "mmwave_reflectivity: 0.85; lidar_reflectivity: 0.9"

[0157] UserComment field

[0158] # Save image

[0159] img.save("texture_with_reflection.jpg", exif=exif_data)

[0160] S23: Reflectivity heat map generation and manual identification.

[0161] Represent reflectivity values in color on the texture map: Specifically, low reflectivity: dark blue. High reflectivity: red. Use Python or OpenCV to map reflectivity values to a heat map. Overlay the reflectivity heat map on the original texture map and manually confirm whether the distribution is reasonable.

[0162] S3: Calculation and reading of radar reflectivity feedback data.

[0163] This step S3 will describe in detail how the simulation system combines the radar reflectivity field and environmental parameters to calculate the radar reflectivity feedback data and read relevant information, generating high-fidelity simulation radar data.

[0164] The flow chart of radar reflectivity intensity acquisition in simulation environment is as followsFigure 5 Step S3 can be implemented by the following steps:

[0165] S31: Calculate the angle of the radar scan beam:

[0166] In the simulation system, the direction and angle range of the current radar scan beam are calculated based on the installation position and design parameters of the radar. The scan angle determines the area of the object that each radar beam line may touch in the simulation scene. The simulator needs to calculate the emission path of the radar point at the current angle step by step. Through step-by-step scanning, complete radar point cloud data is formed, laying the foundation for subsequent reflectivity calculation.

[0167] S32: Determine the 3D object surface corresponding to the radar point:

[0168] During the propagation of the radar point, interaction with the 3D model in the scene occurs. Through the RayTracing algorithm, it is determined whether the propagation path of each radar point collides with a certain object surface. When a collision occurs, the simulator records the 3D object, collision position, and corresponding surface normal information in the radar click.

[0169] In addition, combined with the scene tree or acceleration structure (such as BVH tree) of the simulation environment, the target object in the complex scene is quickly located, ensuring the real-time and efficiency of radar calculation.

[0170] S33: Read the radar reflectivity field of the 3D object surface map:

[0171] The 3D object surface collided by the radar point contains a pre-embedded radar reflectivity field (stored in the texture map). The simulation system obtains the radar reflectivity parameters of the current object by reading the "reflection" field in the texture map. If the reflectivity field is empty or undefined, the system will call the default parameter library to assign a reasonable reflectivity value to unknown materials, ensuring the completeness of the calculation. According to the read reflectivity value, combined with the surface normal and incident angle, the direction and intensity of the radar echo are calculated.

[0172] S34: Comprehensive calculation of radar reflection intensity feedback data:

[0173] Factors affecting reflection intensity: Radar reflection intensity not only depends on the radar reflectivity of the object surface, but also is affected by the following factors: Distance attenuation: As the propagation distance of the radar point increases, the echo intensity decreases exponentially. Environmental conditions: Weather factors such as rain, fog, and snow further reduce echo intensity.

[0174] Calculation method:

[0175] The system establishes a physical simulation model based on the above factors to comprehensively calculate the radar reflection intensity and generate feedback data highly consistent with the real scene.

[0176] S35: Output radar point position and reflection intensity:

[0177] After the calculation of each radar point is completed, the simulator writes its position coordinates (XYZ) and reflection intensity value (Intensity) into the output point cloud data format (such as.pcd,.las). The generated point cloud data can be directly used for testing the perception module of the autonomous driving system, verifying the reliability of core algorithms such as target detection and path planning.

[0178] S36: Determine whether the radar scanning is completed:

[0179] If the current radar scan beam has not yet completed the preset angle range, continue to execute the calculation and reflection calculation of the next radar beam line, until all scans are completed. After the scanning is completed, the complete point cloud data is stored and output as the simulation result.

[0180] The embodiments of the present application also provide a storage and reading device for simulating 3D model radar reflectivity of autonomous driving, as shown in FIG. 6, the device comprises: Figure 6

[0181] The reflectivity calculation module 601 is configured to collect and calculate the real reflectivity of the objects in the autonomous driving related scene;

[0182] The reflectivity embedding module 602 embeds the real reflectivity of the objects in the autonomous driving related scene into the texture map of the surface of the 3D model;

[0183] The feedback data calculation module 603 is configured to calculate and read the radar reflection intensity feedback data.

[0184] In some embodiments, the reflectivity calculation module is further configured to:

[0185] According to the objects in the autonomous driving scene, the objects are classified, and the physical properties of the object surface are analyzed; wherein the objects in the autonomous driving scene include vehicles, windshields, buildings, road surfaces, street lamp posts, the physical properties of the object surface include material properties, surface roughness and geometric characteristics, the material properties include metal, plastic, glass, concrete and asphalt, and the geometric characteristics include the shape of the object; based on the object classification and physical property analysis, a basic reflectivity parameter range is set for each type of object to provide a reference for subsequent collection and modeling;

[0186] A correction method combining theoretical methods and actual collected data is used to calibrate the data;

[0187] Real scene data is obtained and denoising and normalization processing is performed; wherein the normalization formula is:

[0188] ​

[0189] where R 归一化 is the normalized reflectivity value; R max and R min are the minimum and maximum reflectivity values in different objects, respectively; and R is the reflectivity value of the object.

[0190] The reflection intensity calculation formula is constructed and expressed as:

[0191] p = f(material, roughness, angle, distance, wavelength)

[0192] where p is the reflection intensity; in the function f, the material determines the absorption and reflection proportion of the object to the electromagnetic wave; the roughness affects the distribution characteristics of the reflection direction; the angle and the distance change the intensity distribution of the received signal; and the wavelength determines the penetration and scattering characteristics of the wave.

[0193] The reflection intensity characteristics of different object categories are classified, distribution fitted, and standardized stored.

[0194] In some embodiments, the reflectivity calculation module is further configured to:

[0195] If it is a laser radar, the reflectivity and the laser echo energy are positively correlated, and the formula for modeling the laser radar is:

[0196]

[0197] where P r is the laser receiving power; P s is the laser transmitting power; D r is the size of the radar receiving aperture; η atm is the atmospheric effect constant; η sys is the optical transmission coefficient; R is the distance; β is the laser beam width; and σ is the effective backscattering cross-sectional area.

[0198] In some embodiments, the reflectivity calculation module is further configured to:

[0199] Based on the characteristic distribution curve, the collected reflectivity data is classified according to the material, surface roughness, and geometric shape for each category of object;

[0200] The distribution fitting algorithm is used to generate the reflectivity characteristic curve of the corresponding category, representing the law of the reflectivity of the object changing with the incident angle and the wavelength;

[0201] A standardized radar reflectivity database is established, and the reflectivity characteristics of each category of object are stored as a parameterized model, including the numerical value or function form of the reflectivity changing with variables such as distance, angle, and wavelength;

[0202] Reflectance properties of each material are stored in JSON format.

[0203] In some embodiments, the reflectance embedding module is further configured to:

[0204] A 3D model of the target object in the simulation scene is established, with the modeling requirements of matching the surface material of the target object with the actual scene and matching the scale, texture, and details of the target object with the real scene.

[0205] The surface of the 3D model is mapped to a two-dimensional image space to generate a texture map, and a JPEG format texture map is exported, which includes an APP1 segment for storing reflectance data, and the APP1 segment is used for EXIF metadata.

[0206] Reflectance is stored using EXIF metadata, and the reflectance of each object is embedded using the UserComment field of EXIF.

[0207] Reflectance values are represented in color on the texture map, mapped as a heat map, and the reflectance heat map is superimposed on the original texture map.

[0208] In some embodiments, the feedback data calculation module is further configured to:

[0209] In the simulation system, the direction and angle range of the current radar scan beam are calculated according to the installation position and design parameters of the radar; the direction and angle range of the current radar scan beam determine the area of the object that each radar beam line may touch in the simulation scene, and the simulator calculates the emission path of the radar point at the current angle step by step to form complete radar point cloud data through step-by-step scanning;

[0210] Based on the radar point cloud data, the propagation path of each radar point is determined whether it collides with a certain object surface through a ray tracing algorithm, and when a collision occurs, the simulator records the 3D object, collision position, and corresponding surface normal information in the radar click; the target object in a complex scene is located in combination with the scene tree or acceleration structure of the simulation environment.

[0211] The 3D object surface based on the radar point collision contains a pre-embedded radar reflectance field, and the simulation system obtains the radar reflectance parameters of the current object by reading the specified field in the texture map, and if the reflectance field is empty or undefined, it calls the default parameter library to assign a reasonable reflectance value to the unknown material, and according to the read reflectance value, the direction and intensity of the radar echo are calculated in combination with the surface normal and the incident angle.

[0212] The physical simulation model is established according to radar reflectivity of the object surface, distance attenuation and environmental conditions, radar reflection intensity is comprehensively calculated, and radar reflection intensity feedback data is generated; wherein the radar reflection intensity feedback data includes position coordinates and reflection intensity values of each radar point.

[0213] Based on the position coordinates and reflection intensity values of each radar point, point cloud data is output.

[0214] In some embodiments, the feedback data calculation module is further configured to:

[0215] If the current radar scan beam has not yet completed the preset angle range, the calculation and reflectivity calculation of the next radar beam line are continued to be performed until all scans are completed, and after the scan is completed, the complete point cloud data is stored and output as a simulation result.

[0216] It should be noted that the device described in the embodiment belongs to the same technical concept as the method described above, has the same technical principle, and can achieve the same beneficial effects, so no further description is given here.

[0217] The embodiments of the present application also provide a storage and reading system for automatically driving a simulation 3D model radar reflectivity, the system comprising: a memory for storing a computer program; a processor for executing the computer program to realize the method according to any one of the above embodiments.

[0218] The embodiments of the present application also provide a non-transitory computer readable storage medium storing instructions, when the instructions are executed by a processor, the method according to any one of the above embodiments is executed.

[0219] The above embodiments are only used to illustrate the present application, and are not intended to limit the present application. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application. Therefore, all equivalent technical solutions also belong to the scope of the present application, and the patent protection scope of the present application should be defined by the claims.

Claims

1. A method for storing and retrieving radar reflectivity from a 3D simulation model of autonomous driving, characterized in that, The method includes; Collect and calculate the true reflectivity of objects in autonomous driving-related scenes; Embed the real reflectivity of objects in autonomous driving-related scenes into the texture map of the 3D model surface; Calculate and read the radar reflection intensity feedback data; Collect and calculate the true reflectivity of objects in autonomous driving-related scenes, including: Based on objects in autonomous driving scenarios, objects are classified and their surface physical properties are analyzed. These objects include vehicles, windshields, buildings, roads, and lampposts. The surface physical properties include material properties, surface roughness, and geometric features. Material properties include metals, plastics, glass, concrete, and asphalt. Geometric features include the object's shape. Based on object classification and physical property analysis, a basic reflectivity parameter range is defined for each object category to provide a reference for subsequent data collection and modeling. The data is calibrated using a correction method that combines theoretical approaches with actual data collection. Acquire real-world scene data and perform denoising and normalization processing; the normalization formula is as follows: where R 归一化 is the normalized reflectance value; R max and R min are the minimum and maximum reflectance values in different objects, respectively; and R is the reflectance value of the object. The formula for calculating reflection intensity is as follows: ρ = f(material, roughness, angle, distance, wavelength) In the formula, ρ is the reflection intensity; in the function f, the material determines the ratio of electromagnetic wave absorption and reflection by the object; roughness affects the distribution characteristics of the reflection direction; angle and distance change the intensity distribution of the received signal; wavelength determines the wave's penetration and scattering characteristics. The reflection intensity characteristics of different object categories are classified, fitted with distributions, and standardized for storage. Embedding the true reflectivity of objects in autonomous driving-related scenes into texture maps on the surface of 3D models includes: The modeling requirements are to match the surface material of the target object with the actual scene, and to match the proportion, texture, and details of the target object with the real scene. A 3D model of the target object in the simulation scene is then established. The 3D model surface is mapped to a two-dimensional image space to generate a texture map, and the texture map is exported in JPEG format. The exported JPEG texture map includes an APP1 segment for storing reflectivity data, and the APP1 segment is used for EXIF ​​metadata. Store reflectance using EXIF ​​metadata, and embed the reflectance of each object using the EXIF ​​UserComment field; Reflectance values ​​are represented by color on the texture map, the reflectance values ​​are mapped to a heatmap, and the reflectance heatmap is overlaid on the original texture map; Calculate and read radar reflection intensity feedback data, including: In the simulation system, the direction and angle range of the current radar scanning beam are calculated based on the radar's installation location and design parameters. The direction and angle range of the current radar scanning beam determine the object area that each radar beam may touch in the simulation scene. The simulator calculates the transmission path of the radar point at the current angle line by line, and forms complete radar point cloud data through step-by-step scanning. Based on the radar point cloud data, a ray tracing algorithm is used to determine whether the propagation path of each radar point collides with the surface of an object. When a collision occurs, the simulator records the 3D object in the radar point, the collision position, and the corresponding surface normal information. Combined with the scene tree or acceleration structure of the simulation environment, the target object in the complex scene is located. The 3D object surface based on radar point collision contains a pre-embedded radar reflectivity field. The simulation system obtains the radar reflectivity parameters of the current object by reading the set fields in the texture map. If the reflectivity field is empty or undefined, the default parameter library is called to assign a reasonable reflectivity value to the unknown material. Based on the read reflectivity value, the direction and intensity of the radar echo are calculated in combination with the surface normal and the incident angle. A physical simulation model is established based on the radar reflectivity of the object's surface, distance attenuation, and environmental conditions. The radar reflection intensity is then comprehensively calculated to generate radar reflection intensity feedback data. The radar reflection intensity feedback data includes the position coordinates and reflection intensity values ​​of each radar point. Point cloud data is output based on the location coordinates and reflection intensity values ​​of each radar point.

2. The method for storing and retrieving radar reflectivity of an autonomous driving simulation 3D model as described in claim 1, characterized in that, The data is calibrated using a correction method that combines theoretical approaches with actual data collection, including: If it is a lidar, then reflectivity and laser echo energy are positively correlated. The formula for modeling lidar is: where P r is the laser received power; P s is the laser transmitted power; D r is the radar receiving aperture size; η atm is the atmospheric effect constant; η sys is the optical transmission coefficient; R is the range; β is the laser beam width; and σ is the effective backscatter cross-sectional area.

3. The method for storing and retrieving radar reflectivity of an autonomous driving simulation 3D model as described in claim 1, characterized in that, The reflection intensity characteristics of different object categories are classified, fitted to distributions, and standardized for storage, including: Based on the characteristic distribution curve, the collected reflectivity data are classified according to material, surface roughness and geometry for each type of object. The distribution fitting algorithm is used to generate reflectance characteristic curves for the corresponding categories, which show the variation of reflectance of the object with incident angle and wavelength. Establish a standardized radar reflectivity database and store the reflectivity characteristics of each type of object as a parameterized model, including numerical or functional forms of reflectivity as a function of distance, angle, and wavelength. The reflectivity characteristics of each material are stored in JSON format.

4. The method for storing and retrieving radar reflectivity of an autonomous driving simulation 3D model as described in claim 1, characterized in that, After outputting point cloud data based on the location coordinates and reflection intensity values ​​of each radar point, the method further includes: If the current radar scan line has not yet completed the preset angle range, the calculation of the next radar line and the reflectivity calculation will continue until all scans are completed. After the scan is completed, the complete point cloud data will be stored and output as the simulation result.

5. A device for storing and retrieving radar reflectivity of a 3D model for autonomous driving simulation, characterized in that, The apparatus for implementing the method as described in any one of claims 1 to 4 comprises: The reflectivity calculation module is configured to collect and calculate the true reflectivity of objects in autonomous driving-related scenes; The reflectivity embedding module embeds the real reflectivity of objects in autonomous driving-related scenes into the texture map of the 3D model surface; The feedback data calculation module is configured to calculate and read radar reflection intensity feedback data.

6. A system for storing and retrieving radar reflectivity of a 3D simulation model for autonomous driving, characterized in that: The system includes: Memory, used to store computer programs; A processor for executing the computer program to implement the method as described in any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, perform the method according to any one of claims 1 to 4.

Images