Digital optical simulation test method and device, storage medium and product
By establishing a target scene model and virtual camera sensors in a purely digital environment to generate simulated images, the problem of incomplete testing in existing testing methods is solved, and efficient and comprehensive testing of driver assistance systems is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-12
- Publication Date
- 2026-04-03
AI Technical Summary
Existing testing methods for driver assistance systems are difficult to conduct comprehensive testing efficiently. Real-vehicle testing is time-consuming and costly, and the scenario simulation is incomplete. Hardware-in-the-loop testing is costly and has low accuracy, and vehicle-in-the-loop testing scenarios are limited.
By establishing a target scene model, introducing virtual cameras and virtual sensors, generating simulation images, and conducting simulation tests in a purely digital environment based on these simulation images, we can simulate different environments and the optical materials of objects for comprehensive testing.
It enables efficient and comprehensive testing of driver assistance systems in a purely digital environment, reducing hardware and time costs while improving the comprehensiveness and accuracy of testing.
Smart Images
Figure CN121787076A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of simulation testing technology, and in particular to digital optical simulation testing methods, equipment, storage media and products. Background Technology
[0002] Vehicle driver assistance systems (ADAS) frequently require comprehensive testing during development to ensure their safe and reliable operation in real-world use. Therefore, optimizing and upgrading testing methods for ADAS is crucial.
[0003] Current testing methods primarily include real-vehicle testing and in-the-loop testing. Real-vehicle testing requires the actual vehicle to travel a considerable distance on the road, which is time-consuming and makes it difficult to simulate certain scenarios. In-the-loop testing places the hardware or vehicle in a virtual environment, but it cannot simulate certain scenarios and still suffers from incomplete testing scenarios. Therefore, current testing methods for driver assistance systems are insufficient for efficiently and comprehensively testing driver assistance systems. Summary of the Invention
[0004] The main objective of this application is to provide a digital optical simulation testing method, equipment, storage medium, and product, aiming to solve the technical problem that current methods are difficult to efficiently and comprehensively test driver assistance systems.
[0005] To achieve the above objectives, this application proposes a digital optical simulation testing method, the method comprising: Obtain modeling parameters and modeling objects, and establish a target scene model based on the modeling parameters and modeling objects. The modeling objects include the target vehicle to be tested, and the modeling parameters include environmental parameters, physical and optical parameters of the modeling objects, and motion parameters of the modeling objects. A virtual camera and virtual sensors are created in the target scene model; Based on the environmental parameters, the physical optical parameters, the motion parameters, the virtual camera, and the virtual sensor, a simulated image is generated; Based on the simulated images, the driver assistance system of the target vehicle is simulated and tested in a purely digital environment.
[0006] In one embodiment, the modeling object includes an ambient light source, and the step of generating a simulation image based on the environmental parameters, the physical optical parameters, the motion parameters, the virtual camera, and the virtual sensor includes: Based on the environmental parameters, the physical optical parameters, and the motion parameters, in the target scene model, the light rays from the ambient light source to the virtual camera are traced using a preset ray tracing method to obtain a linear exposure map; Based on the virtual sensor, the photons in the linear exposure image are converted into digital values to obtain a RAW image; The simulated image is obtained by correcting and enhancing the single-channel data in the RAW image.
[0007] In one embodiment, the virtual camera has corresponding camera parameters, including physical defect parameters, and the environmental parameters include the light source parameters of the ambient light source. The step of tracing the light from the ambient light source to the virtual camera in the target scene model based on the environmental parameters, the physical optical parameters, and the motion parameters to obtain a linear exposure map includes: Based on the light source parameters, the physical optical parameters, and the motion parameters, the energy distribution of photons emitted by the ambient light source when they reach the virtual camera after being reflected and / or refracted by the modeling object is determined in the target scene model. The modeling object also includes scene objects in the target scene model. Based on the energy distribution and the physical defect parameters, the linear exposure map is generated by the virtual camera, wherein the physical defect parameters include the aberrations of the real camera corresponding to the virtual camera.
[0008] In one embodiment, the virtual sensor has corresponding sensor parameters, including sensor imaging parameters, sensor noise parameters, and sensor defect parameters. The step of converting photons in the linear exposure image into digital values based on the virtual sensor to obtain a RAW image includes: Based on the sensor defect parameters, sensor noise parameters, and sensor imaging parameters, the virtual sensor converts the photons into digital values in the order of photon, electron, voltage, and digital value conversion to obtain the RAW image.
[0009] In one embodiment, the target scene model is equipped with multiple virtual cameras, and the step of simulating and testing the driver assistance system of the target vehicle in a purely digital environment based on the simulated images includes: The simulated images corresponding to each of the virtual cameras are fused to obtain a fused perception map; Based on the fused perception map, a simulation test is conducted on the driver assistance system of the target vehicle.
[0010] In one embodiment, the step of fusing the simulated images corresponding to each of the virtual cameras to obtain a fused perception map includes: The simulated images are fused to obtain pre-fused data; Feature information is extracted from each of the simulated images, and the feature information is fused to obtain feature fusion data; Target tracking and recognition are performed based on the simulated images, and the results of target tracking and recognition are fused to obtain post-fused data. The fused perception map is determined based on the pre-fusion data, the feature fusion data, and the post-fusion data.
[0011] In one embodiment, the step of simulating and testing the driver assistance system of the target vehicle based on the fused perception map includes: Obtain high-precision maps and global path planning information; The fused perception map, the high-precision map, and the global path planning information are input into the assisted driving system to obtain a local motion trajectory. Based on the local motion trajectory, the vehicle control command for the target vehicle is determined; The vehicle control commands are fed back to the target vehicle to complete the simulation test.
[0012] In addition, to achieve the above objectives, this application also proposes a digital optical simulation testing device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the digital optical simulation testing method as described above.
[0013] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the digital optical simulation testing method described above.
[0014] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the digital optical simulation testing method described above.
[0015] One or more technical solutions proposed in this application have at least the following technical effects: This application obtains modeling parameters and modeling objects, and establishes a target scene model based on the modeling parameters and modeling objects. The modeling objects include the target vehicle to be tested, and the modeling parameters include environmental parameters, physical and optical parameters of the modeling objects, and motion parameters of the modeling objects. A virtual camera and virtual sensors are established in the target scene model. Based on the environmental parameters, physical and optical parameters, motion parameters, virtual camera, and virtual sensors, a simulation image is generated. Based on the simulation image, the driver assistance system of the target vehicle is simulated and tested in a purely digital environment.
[0016] To address the problem that current testing methods for advanced driver assistance systems (ADAS) struggle to efficiently and comprehensively test such systems, this application addresses this issue by establishing scene models, virtual cameras, and virtual sensors to generate simulated images. Simulation tests are then performed based on these images, thus enabling efficient and comprehensive testing of ADAS systems. Specifically, this application utilizes model-based simulation testing, eliminating the need for actual vehicles to travel long distances on public roads, thereby improving testing efficiency. Furthermore, this application sets environmental, physical-optical, and motion parameters when building the scene model, conducting simulations in a purely digital environment. This allows for the simulation of different environments, object optical materials, and vehicle driving scenarios, enhancing the comprehensiveness of the test scenarios. Therefore, overall, this application can efficiently and comprehensively test ADAS systems. Attached Figure Description
[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.
[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 This is a flowchart illustrating an embodiment of the digital optical simulation testing method of this application. Figure 2 This is a flowchart illustrating Embodiment 2 of the digital optical simulation testing method of this application. Figure 3 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the digital optical simulation testing method in this application embodiment; Figure 4 This is a schematic diagram illustrating the data acquisition consent process involved in the digital optical simulation testing method described in this application.
[0020] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0021] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.
[0022] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.
[0023] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device or digital optical simulation testing device capable of performing the above functions. The following description uses a digital optical simulation testing device as an example to illustrate this embodiment and the subsequent embodiments.
[0024] Based on this, the embodiments of this application provide a digital optical simulation testing method, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the digital optical simulation testing method of this application.
[0025] In this embodiment, the specific application scenario is the testing of driver assistance systems. The development process of driver assistance systems involves long-term testing, making the upgrading and iteration of testing methods particularly important. Good testing methods can shorten the development cycle, reduce capital investment, and improve development efficiency. The current mainstream testing methods include the following: 1. Real-vehicle testing: Verifying safety, reliability, and generalization ability through tens of thousands of kilometers of testing on real vehicles. Real-vehicle testing is extremely costly in terms of time and money, and cannot cover long-tail scenarios such as extreme weather, rare traffic participant behavior, and dangerous accident scenarios, which are difficult to reproduce in reality and have extremely high testing risks, resulting in low test comprehensiveness.
[0026] 2. Hardware-in-the-Loop Testing: This method integrates real hardware such as domain controllers and sensors into a simulation environment, forming a semi-physical simulation closed loop. The simulator runs vehicle dynamics and sensor models, sending simulated CAN (CAN / LIN / Ethernet, Controller Area Network) / LIN (Local Interconnect Network) / Ethernet signals to the controller. The controller calculates the signals, issues control commands, and sends them back to the simulator to complete the loop. Hardware-in-the-loop testing is safe and efficient, capable of reproducing complex and dangerous scenarios at a lower cost. However, the initial setup of the test platform is expensive, and the coupling accuracy between the actual hardware and the virtual scenario is relatively low.
[0027] 3. Vehicle-in-the-loop (V2L) testing: By placing a real vehicle in a laboratory environment and injecting a virtual environment into the vehicle's real sensors, the vehicle feels as if it is driving in a virtual scenario. V2L testing is closer to the realism of actual road testing, but building a real V2L laboratory is more expensive. Furthermore, it cannot simulate scenarios such as vehicles rapidly approaching from behind, resulting in limited simulation scenarios and insufficient comprehensiveness.
[0028] This embodiment aims to: efficiently conduct comprehensive testing of driver assistance systems.
[0029] Specifically, this embodiment constructs a complete physical model encompassing light source emission, scene lighting interaction, sensor photoelectric conversion, and ISP (Image Signal Processor) image processing. In a purely digital environment, this embodiment can simulate arbitrary weather conditions, lighting conditions, and extreme scenarios, enhancing the scenario freedom of assisted driving system testing and thus improving the comprehensiveness of the testing.
[0030] Specifically, compared to traditional simulation methods that only provide the algorithm with computer-rendered images, this embodiment generates image data based on the optical properties of materials in the real world. By generating image data based on the optical properties of materials in the real world, this embodiment can make the noise patterns, optical artifacts, color deviations, etc. of the perception algorithm highly consistent with the real world, greatly improving the reliability of simulation tests, narrowing the gap between simulation and reality, and thus improving the accuracy of software and hardware coupling.
[0031] Specifically, this embodiment uses a purely digital environment for simulation testing, eliminating the need for developers to purchase numerous real cameras and processors. Testing can be conducted simply by adjusting parameters within the simulation platform, significantly reducing hardware procurement and experimental setup costs. Furthermore, during the hardware selection phase, this embodiment allows inputting candidate camera sensor parameters into the platform to quickly test the performance of different sensors in the same scenario, providing intuitive and quantitative data support for hardware selection and substantially reducing selection time and financial costs.
[0032] In this embodiment, the digital optical simulation testing method includes steps S10 to S40: Step S10: Obtain modeling parameters and modeling objects; establish a target scene model based on the modeling parameters and modeling objects, wherein the modeling objects include the target vehicle to be tested, and the modeling parameters include environmental parameters, physical and optical parameters of the modeling objects, and motion parameters of the modeling objects; It should be noted that modeling parameters refer to various numerical or attribute data used to construct the simulation scene model, including environmental parameters, physical optical parameters, and motion parameters. The modeling object refers to the entity that needs to be modeled in the simulation. In this embodiment, the modeling object is the target vehicle and other related objects. Modeling mainly includes a scene model, a test vehicle model, and a virtual camera sensor model. The target vehicle refers to the test vehicle model that serves as the test subject in the intelligent driving optical link simulation. The test vehicle model is placed within the target scene model and includes a geometric model, optical characteristics, and motion parameters. The optical characteristics include headlights and cabin light sources, etc.
[0033] The target scene model is a 3D virtual environment constructed using modeling parameters and modeling objects for simulation, including the target vehicle and its physical and optical environment. Environmental parameters describe data about the external conditions of the simulation scene, including, in this embodiment, parameters such as light intensity, weather conditions, light source type, color temperature, and spectral distribution. Physical and optical parameters are physical quantities describing the interaction characteristics between the surface of the modeling object and light, including diffuse reflectance, specular reflectance, refractive index, transmittance, absorption coefficient, and surface roughness. Motion parameters describe the dynamic behavior of the modeling object during the simulation, including position, velocity, acceleration, and trajectory.
[0034] It should also be noted that, based on the modeling parameters and the modeling object, the target scene model is established in the following order: 3D scene construction, material property assignment, light source and environment definition, and dynamic scene configuration.
[0035] The construction of 3D scenes specifically includes: using high-precision 3D modeling software to build or import complex static and dynamic scene models containing elements such as vehicles, pedestrians, bicycles, roads, traffic signs, trees, and buildings. Among them, vehicles have internal details such as dashboards and steering wheels when modeling.
[0036] The assignment of material properties specifically includes assigning physical material parameters to the surface of each object in the scene, describing its interaction with light. These are physical optical parameters such as diffuse reflectance, specular reflectance, transmittance, refractive index, and absorption coefficient in this embodiment. Physical optical parameters are the foundation for achieving physically accurate ray tracing.
[0037] The definition of light source and environment specifically includes: precisely defining the light source parameters of natural light (sun) and artificial light sources (vehicle lights, street lights).
[0038] Dynamic scene configuration specifically includes setting the motion trajectory, speed, and acceleration for dynamic objects such as the main vehicle, target vehicle, and pedestrians in the scene to simulate complex traffic interaction scenarios such as cut-in and pedestrian crossing.
[0039] It is understood that in establishing the target scene model, this embodiment simultaneously introduces environmental parameters, physical and optical parameters of the modeling object, and motion parameters, and uses the target vehicle as the core modeling object for complete modeling, so that the constructed target scene model can realistically reflect the combined effects of lighting, material reflection / refraction behavior, and dynamic traffic participants in the actual road environment.
[0040] Furthermore, since this embodiment replaces traditional visual textures with physical optical parameters, and can support the setting of any extreme or rare conditions through the setting of environmental parameters, it ensures the high fidelity of the simulation scene at the optical level and the wide coverage of the scene.
[0041] Step S20: Establish a virtual camera and virtual sensors in the target scene model; It should be noted that a virtual camera refers to an optical imaging device that simulates a real vehicle-mounted camera in a target scene model. It includes physical characteristics such as lens distortion, vignetting, and aberrations, as well as geometric parameters such as installation position and orientation. In this embodiment, a virtual sensor is a component that digitally models an image sensor in a simulation environment. The virtual camera and virtual sensor in this embodiment are modeled based on real modules, and they are positioned both inside and outside the test vehicle's cabin.
[0042] Step S30: Generate a simulated image based on the environmental parameters, the physical optical parameters, the motion parameters, the virtual camera, and the virtual sensor; It should be noted that simulated images refer to images generated after a complete image signal processing flow. They can be visual images such as RGB images, or infrared images. RGB images are color images that conform to human visual perception or display device standards, where each pixel contains three channels of values: red (R), green (G), and blue (B). Infrared images are images formed by capturing infrared electromagnetic waves radiated or reflected by an object using an infrared sensor.
[0043] It is understood that this embodiment uses a virtual camera to capture images based on a constructed target scene model and corresponding parameters, and then uses a virtual sensor to convert the generated images to obtain a simulated image. The entire imaging process described above in this embodiment is based entirely on physical laws and does not rely on real vehicles or physical sensors. Therefore, any imaging conditions, including extreme weather, rare lighting, or dangerous traffic scenarios, can be reproduced in a purely digital environment. This significantly improves the comprehensiveness of the assisted driving system algorithm testing and reduces the time and economic costs of testing.
[0044] In one feasible implementation, the specific method for generating simulated images based on the environmental parameters, the physical optical parameters, the motion parameters, the virtual camera, and the virtual sensor can also be: Based on the environmental parameters, the physical optical parameters, and the motion parameters, in the target scene model, the light rays from the ambient light source to the virtual camera are traced using a preset ray tracing method to obtain a linear exposure map. Based on the virtual sensor, the photons in the linear exposure map are converted into digital values to obtain a RAW image. The single-channel data in the RAW image are corrected and enhanced to obtain the simulation image.
[0045] It should be noted that ray tracing is an image generation technique based on the principles of physical optics. It simulates the process of light tracing originating from a light source, undergoing reflection, refraction, and transmission within a scene before reaching the imaging device. In this embodiment, it is used to calculate the light energy received by each pixel. A linear exposure map refers to the raw image data generated by ray tracing, representing the number of photons received by each pixel on the sensor surface, before any image signal processing. Single-channel data refers to the raw sample value in a RAW image where each pixel contains only one color component.
[0046] Understandably, in this embodiment, based on pre-defined environmental, physical optical, and motion parameters, a preset ray tracing method is used in the target scene model to physically simulate the path of light rays originating from the ambient light source and ultimately reaching the virtual camera. This accurately calculates the interaction process between the light rays and the surfaces of various modeled objects in the scene, thereby generating a linear exposure map. By inputting this linear exposure map into the virtual sensor model, the number of photons is converted into corresponding digital values based on its photoelectric conversion characteristics, generating a RAW image with realistic noise characteristics. By sequentially performing image data processing operations on the single-channel data in the generated RAW image, it is converted into a simulated image where each pixel contains information from the red, green, and blue channels, thus realizing a complete end-to-end simulation imaging process from physical optical propagation to a visually perceptible image.
[0047] This embodiment strictly simulates light propagation based on environmental parameters, physical optical parameters, and motion parameters. It also combines a virtual sensor to perform photoelectric conversion on the linear exposure image to generate a RAW image. The single-channel data of the RAW image is then systematically corrected and enhanced, thereby ensuring that the final simulated image closely approximates the output of a real vehicle camera in terms of optical response, noise distribution, and color reproduction.
[0048] In one feasible implementation, the virtual camera has corresponding camera parameters, including physical defect parameters, and the environmental parameters include the light source parameters of the ambient light source. The specific implementation of tracing the light rays from the ambient light source to the virtual camera in the target scene model using a preset ray tracing method based on the environmental parameters, the physical optical parameters, and the motion parameters to obtain a linear exposure map can also be: Based on the light source parameters, the physical optical parameters, and the motion parameters, in the target scene model, the energy distribution of photons emitted by the ambient light source after reflection and / or refraction by the modeling object and reaching the virtual camera is determined. The modeling object also includes scene objects in the target scene model. Based on the energy distribution and the physical defect parameters, the linear exposure map is generated by the virtual camera. The physical defect parameters include the aberrations of the real camera corresponding to the virtual camera.
[0049] It should be noted that camera parameters refer to the set of parameters used to describe the imaging characteristics of the virtual camera, including geometric mounting parameters and optical characteristic parameters. Physical defect parameters refer to the parameters in the virtual camera used to simulate the non-ideal optical behavior of a real lens, including radial distortion, tangential distortion, chromatic aberration, vignetting, astigmatism, and field curvature. Ambient light sources refer to the light source entities in the target scene model used to provide illumination, which in this embodiment can be the sun, streetlights, vehicle lights, etc.
[0050] Light source parameters are data describing the characteristics of ambient light sources, including spectral power distribution (based on blackbody radiation temperature or a custom spectrum), light intensity distribution (Lambert or Gaussian models), beam angle, and luminous flux. This embodiment uses light source parameters to define the global illumination effect of ambient light (such as cloudy days or dusk). Energy distribution refers to the spatial distribution of light energy density carried by light rays at various locations on the entrance pupil plane of the virtual camera after reflection and / or refraction from the surface of the modeled object.
[0051] It is understood that this embodiment simulates the optical interaction process between photons emitted by ambient light sources and the modeled object in the target scene model based on light source parameters, physical optical parameters, and motion parameters. Specifically, it includes the reflection and / or refraction behavior of photons on the surface of the modeled object, and calculates the energy distribution formed on the entrance pupil plane when these photons reach the virtual camera. By combining this energy distribution with the physical defect parameters of the virtual camera, such as distortion, vignetting, and chromatic aberration, the energy is spatially remapped and attenuated, ultimately generating a linear exposure image that considers the non-ideal effects of real lenses, thereby achieving a high degree of reproduction of the actual optical imaging process of the vehicle-mounted camera front end.
[0052] In generating a linear exposure map, this embodiment not only considers the photon propagation path and energy distribution determined by the light source parameters, physical optical parameters, and motion parameters, but also introduces the physical defect parameters of the virtual camera to correct the energy distribution. This allows the generated linear exposure map to accurately reflect the optical distortion and light intensity attenuation introduced by the real lens due to manufacturing process or material limitations.
[0053] Furthermore, since this embodiment integrates the ambient light source, the optical material of the modeled object, the dynamic behavior, and the physical defects of the camera into the same scene for joint modeling, it avoids the imaging distortion caused by ignoring lens defects or using an ideal pinhole model in traditional graphics rendering. This significantly improves the closeness between the subsequent RAW images and simulation images and the actual images. Based on the final generated simulation images, simulation testing can effectively improve the accuracy of simulation testing.
[0054] In one feasible implementation, the virtual sensor has corresponding sensor parameters, including sensor imaging parameters, sensor noise parameters, and sensor defect parameters. A further implementation of converting photons in the linear exposure image into digital values based on the virtual sensor to obtain the RAW image can be: Based on the sensor defect parameters, sensor noise parameters, and sensor imaging parameters, the virtual sensor converts the photons into digital values in the order of photon, electron, voltage, and digital value conversion to obtain the RAW image.
[0055] It should be noted that sensor parameters refer to the set of parameters used to describe the photoelectric conversion characteristics of a virtual sensor, including parameters related to imaging capabilities, noise behavior, and manufacturing defects. Sensor imaging parameters refer to parameters that affect the basic imaging performance of a virtual sensor, including pixel size, resolution, quantum efficiency, full-well capacity, dynamic range, and response linearity. Sensor noise parameters refer to parameters characterizing various types of noise introduced by the virtual sensor during the photoelectric conversion process, including readout noise, dark current noise, shot noise, and fixed-pattern noise. Sensor defect parameters refer to parameters used to simulate the non-ideal characteristics of real image sensors, including dead pixels, hot spots, pixel response non-uniformity, column / row noise, and microlens crosstalk.
[0056] It should also be noted that the virtual sensor model constructed in this embodiment conforms to the EMVA 1288 (European Machine Vision Association Standard 1288) standard, and the specific input parameters of the model include: Color filter arrangement, such as Bayer Filter Array, RCCC (a type of filter arrangement), RGB-IR (red-green-blue-infrared filter array), etc.; quantum efficiency (QE) curve, used to define the efficiency of photons at different wavelengths generating electrons; system gain, used to define the ratio of the number of electrons to digital values; dark current, used for the signal generated under no-light conditions; full-well capacity, used to define the maximum number of electrons that a single pixel can hold, determining the dynamic range; photoelectric conversion nonlinearity, used to define the nonlinear relationship between sensor response and light intensity; spatial non-uniformity, specifically including fixed-mode noise (FPN) and pixel response non-uniformity (PRNU).
[0057] Understandably, after obtaining the linear exposure image, this embodiment first converts the number of incident photons into the number of generated electrons based on quantum efficiency; then, it maps the number of electrons into an analog voltage signal by combining the full-well capacity and response linearity; then, it superimposes various types of noise defined by sensor noise parameters into this signal and introduces pixel-level non-ideal effects described by sensor defect parameters; finally, it quantizes the analog voltage into an integer digital value through an analog-to-digital converter (ADC) model to form a RAW image conforming to Bayer arrangement or other color filter array formats, thereby achieving a high degree of simulation of the generated image.
[0058] This embodiment converts photons into digital values in the physical order of photons, electrons, voltage, and digital values. By introducing sensor imaging parameters, sensor noise parameters, and sensor defect parameters, the generated RAW image not only contains real illumination information but also incorporates the inherent noise characteristics and manufacturing defects of the actual sensor.
[0059] Furthermore, since the conversion process in this embodiment is entirely based on physical mechanism modeling and all sensor parameters can be flexibly configured, it is possible to quantitatively compare and verify the imaging quality of different candidate sensor schemes in an early stage without the need for physical hardware, which significantly improves the efficiency of intelligent driving system testing.
[0060] In one embodiment, the specific implementation of correcting and enhancing single-channel data in the RAW image to obtain the simulated image may also be: The generated RAW image is input into a programmable ISP simulation software or pipeline to gradually simulate the processing flow of a real ISP chip and obtain a simulation image.
[0061] It should be noted that the simulation of the processing flow of a real ISP chip in this embodiment specifically includes the following steps in sequence: Black level correction is used to subtract the sensor's base signal value; lens shading correction is used to compensate for the decrease in brightness around the edges of the image caused by lens vignetting; dead pixel correction is used to repair or compensate for faulty pixels; automatic white balance is used to adjust the gain of the R, G, and B channels according to the ambient light color temperature to correct the color; demosaic is used to convert Bayer format RAW data into a full-color image where each pixel contains R, G, and B channel information through interpolation algorithms (such as bilinear and edge-adaptive); color correction matrix is used to convert the image to a standard color space (such as sRGB); gamma correction is used to perform non-linear transformations on the image to adapt to human vision or display devices; image enhancement is used to further optimize image quality by applying noise reduction and sharpening algorithms; output is used to finally generate a simulated image that conforms to human visual perception.
[0062] It is understood that this embodiment simulates the processing flow of a real ISP chip, making the final generated simulation image closer to the image obtained during actual vehicle driving, thereby improving the credibility of the simulation test of the assisted driving system.
[0063] In one embodiment, the process of simulating a tunnel driving scenario includes: After obtaining the tunnel dimensions, a tunnel model is built in Speos (a professional-grade optical system simulation software) according to the actual dimensions; the vehicle's digital model file is imported into Speos; the optical properties of materials such as glass, paint, concrete, and asphalt are obtained by consulting literature or actual measurement; the light brightness is set, the light type is set to Lambertian, the integral angle is 180°, and the color temperature or spectral composition is set; the camera sensor is set in Speos, correctly positioned at the camera installation location, and the lens's downscaled file and motion file are imported; reverse simulation is set up to perform ray tracing simulation and obtain the exposure image; the EMVA1288 data from Sonser is imported, and a RAW image is obtained based on this data; the image is then imported into the ISP simulation software to generate the simulation image.
[0064] This embodiment simulates a vehicle driving in a tunnel using the steps described above, thereby obtaining a simulated image that closely resembles the actual vehicle driving.
[0065] Step S40: Based on the simulation image, perform a simulation test on the driver assistance system of the target vehicle in a purely digital environment.
[0066] It is understood that this embodiment directly uses the simulated image generated by physical precision modeling as the input of the driver assistance system. This image has fully preserved key features such as optical effects, sensor noise and lens defects in the real world, thereby ensuring that the data processed by the perception module is highly consistent with the images collected from the actual vehicle, and significantly improving the credibility of the simulation test results.
[0067] Furthermore, since the simulation test of the target vehicle's driver assistance system in this embodiment is conducted in a purely digital environment, while ensuring a high degree of consistency between the data in this embodiment and the images acquired from the actual vehicle, this embodiment allows for direct simulation testing of corresponding functions during product development in a purely digital environment, thereby identifying product risks and related defects in advance. Moreover, because the entire testing process is conducted in a purely digital environment, this embodiment can accurately reproduce problems that arise during testing, thereby assisting developers in quickly locating issues and improving development efficiency.
[0068] In summary, this embodiment obtains modeling parameters and modeling objects, establishes a target scene model based on the modeling parameters and modeling objects, wherein the modeling objects include the target vehicle to be tested, the modeling parameters include environmental parameters, physical and optical parameters of the modeling objects, and motion parameters of the modeling objects, establishes virtual cameras and virtual sensors in the target scene model, generates simulation images based on the environmental parameters, physical and optical parameters, motion parameters, virtual cameras, and virtual sensors, and performs simulation tests on the driver assistance system of the target vehicle in a purely digital environment based on the simulation images.
[0069] To address the problem that current testing methods for advanced driver assistance systems (ADAS) struggle to efficiently and comprehensively test such systems, this embodiment addresses this issue by establishing scene models, virtual cameras, and virtual sensors to generate simulated images. Simulation tests are then performed based on these images, thus enabling efficient and comprehensive testing of ADAS systems. Specifically, this embodiment uses model-based simulation testing, eliminating the need for actual vehicles to travel long distances on the road, thereby improving testing efficiency. Furthermore, this embodiment sets environmental, physical-optical, and motion parameters when building the scene model, allowing for simulation testing of different environments, object optical materials, and vehicle driving scenarios, enhancing the comprehensiveness of the test scenarios. Therefore, overall, this embodiment can efficiently and comprehensively test ADAS systems.
[0070] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 The digital optical simulation testing method further includes steps S41-S42 in step S40: Step S41: The simulated images corresponding to each of the virtual cameras are fused to obtain a fused perception map; It should be noted that the fused perception map refers to the unified environmental perception result generated by spatiotemporally aligning and integrating simulated images from multiple virtual cameras through a multi-sensor fusion algorithm. In this embodiment, it includes enhanced information such as target detection boxes, semantic labels, depth estimation, or 3D scene structure.
[0071] It is understandable that this embodiment fuses the simulated images corresponding to multiple virtual cameras, making full use of the redundant and complementary information from different perspectives, and achieving an architecture consistent with the multi-view vision of the real vehicle at the simulation level, thereby improving the credibility of the simulation test of the assisted driving system.
[0072] In one feasible implementation, the specific implementation of fusing the simulated images corresponding to each of the virtual cameras to obtain the fused perception map can also be: The simulated images are fused to obtain pre-fusion data. Feature information is extracted from each simulated image and fused to obtain feature fusion data. Target tracking and recognition are performed based on each simulated image. The results of the target tracking and recognition are fused to obtain post-fusion data. The fused perception map is determined based on the pre-fusion data, the feature fusion data, and the post-fusion data.
[0073] It should be noted that pre-fusion data refers to the fusion result obtained by aligning, stitching, or projecting multiple simulated images at the original data level; it is low-level fusion data without high-level semantic processing. Feature information refers to intermediate representations extracted from simulated images through algorithms, including structured information that can be used for perception tasks, such as edges, textures, semantic regions, keypoints, or target candidate boxes. Feature fusion data refers to the fusion result generated after aligning and integrating the feature information from each simulated image in the feature space. Post-fusion data refers to the high-level fusion result obtained by associating, deduplicating, and weighting the recognition results of each simulated image after they have independently completed target tracking and recognition.
[0074] It is understood that this embodiment simultaneously employs three fusion strategies—pre-fusion, feature-level fusion, and post-fusion—when generating the fused perception map, and generates pre-fusion data, feature-fusion data, and post-fusion data respectively, thereby making full use of all information from the original pixels to high-level semantics and avoiding the limitations of a single fusion level under occlusion, blind spots, or sensor noise.
[0075] Furthermore, this embodiment combines low-level geometric consistency, mid-level semantic complementarity, and high-level decision redundancy. Moreover, the simulation image input in this embodiment is obtained based on the simulation of the constructed target scene. Therefore, the generated fused perception map not only has the advantages of multi-view collaboration, but also inherits the optical and sensor characteristics in real imaging. This provides a perception input that is closer to the actual vehicle operating state for the testing of the assisted driving system, significantly enhancing the credibility of the simulation test.
[0076] Step S42: Based on the fused perception map, perform a simulation test on the driver assistance system of the target vehicle.
[0077] Understandably, this step uses the fused perception map as the perception input of the assisted driving system. This fused perception map integrates multi-level information from front-fused data, feature-fused data, and back-fused data, and has stronger target integrity, positioning accuracy, and scene robustness. This makes the perception conditions faced by the assisted driving system in simulation testing closer to the actual output of a real multi-camera system, thereby improving the credibility of the simulation test.
[0078] In one feasible implementation, the specific implementation of simulating and testing the driver assistance system of the target vehicle based on the fused perception map can also be: A high-precision map and global path planning information are acquired, and the fused perception map, the high-precision map, and the global path planning information are input into the assisted driving system to obtain a local motion trajectory. Based on the local motion trajectory, the vehicle control command for the target vehicle is determined, and the vehicle control command is fed back to the target vehicle to complete the simulation test.
[0079] It should be noted that high-precision maps refer to digital maps containing centimeter-level geographic information such as lane lines, traffic signs, curbs, slopes, curvature, and static obstacles, used to assist autonomous vehicles in localization and path planning. Global path planning information refers to the macroscopic driving route from start to finish, typically generated by a navigation system, including high-level path guidance information such as roads encountered and turning points. Local motion trajectory refers to the executable vehicle trajectory generated within a short time window based on current environmental perception and the global path, including dynamic information such as position, speed, acceleration, and heading. Vehicle control commands refer to signals output by the driver assistance system to control the target vehicle to perform specific driving actions, including throttle opening, braking pressure, and steering angle.
[0080] Understandably, this embodiment acquires high-precision maps and global path planning information, and inputs the fused perception map, high-precision map, and global path planning information into the target vehicle's driver assistance system. Based on these three types of input information, the driver assistance system comprehensively judges the current traffic environment and driving intention, generating a safe, smooth, and traffic-compliant local motion trajectory. In this embodiment, the local motion trajectory calculates the corresponding vehicle control commands, and feeds these commands back to the target vehicle dynamics model in the simulation environment, driving the virtual vehicle to perform corresponding actions, thereby completing a closed-loop end-to-end simulation test.
[0081] Because this embodiment introduces fused perception map, high-precision map and global path planning information as input to the assisted driving system in the simulation test, it constructs a test environment that is completely consistent with the actual intelligent driving system operation logic, enabling the decision planning module to generate reasonable local motion trajectories with the support of highly realistic multi-source information.
[0082] Furthermore, since the fused perception map itself originates from physically accurate optical link simulation, and the high-precision map and global path planning information provide accurate static environment and task objectives, the generated vehicle control commands can comprehensively reflect the overall performance of the assisted driving system in complex and dynamic scenarios. This allows simulation testing to not only test various aspects of the assisted driving system, such as simulation and path planning, but also significantly improve the comprehensiveness and depth of the testing.
[0083] In one embodiment, the specific implementation method after inputting the fused perception map, the high-precision map, and the global path planning information into the assisted driving system to obtain the local motion trajectory can also be: Based on the local motion trajectory, the pose and state of the target vehicle in the target scene model are updated; and the optical simulation process from ray tracing to simulation image generation is re-executed to obtain a new simulation image corresponding to the local motion trajectory. That is, after the pose and state of the target vehicle are updated, simulation images are generated based on the updated environmental parameters, physical optical parameters, motion parameters, virtual camera, and virtual sensor corresponding to the target vehicle, ultimately obtaining a complete sequence of simulation images corresponding to the local motion trajectory.
[0084] It is understood that after obtaining the local motion trajectory, this embodiment feeds it back into the target scene model, dynamically updates the position, orientation and motion state of the target vehicle, and re-runs the complete optical link simulation process, including environmental parameters, physical optical parameters, virtual cameras and virtual sensors, based on the updated scene state, thereby generating a new simulation image that is strictly spatiotemporally aligned with the current local motion trajectory.
[0085] Since the new simulated images are generated by scene changes caused by the system's own behavior under closed-loop control, their imaging content realistically reflects the dynamic effects such as perspective switching, lighting changes, or occlusion evolution caused by the decisions of the assisted driving system. This provides effective visual input close to the real scene for perception, planning, and control in the next cycle, enhancing the dynamic realism and continuity of end-to-end simulation testing.
[0086] In summary, this embodiment performs pre-fusion, feature-level fusion, and post-fusion on each simulated image to determine a highly robust fused perception map. This fused perception map, along with a high-precision map and global path planning information, is input into the target vehicle's driver assistance system. Its decision planning module generates a local motion trajectory and calculates vehicle control commands based on this trajectory. These commands are then fed back to the target vehicle in the simulation environment to complete the closed-loop test.
[0087] This embodiment introduces a three-level fusion mechanism—pre-fusion, feature-level fusion, and post-fusion—to generate a fused perception map during simulation testing. This fused perception map, along with high-precision maps and global path planning information, is then collaboratively input into the assisted driving system. This ensures that the perception input not only possesses multi-view geometric consistency and semantic complementarity but also inherits the realistic noise, lens defects, and illumination effects inherent in physically accurate optical simulations, significantly improving the realism of the assisted driving system's perception. Furthermore, since each stage of the testing process is based on the same set of physically consistent modeling parameters, the temporal and spatial consistency of the simulation system is ensured. Ultimately, this enables the testing of the assisted driving system in complex and dynamic scenarios, greatly enhancing the comprehensiveness and reliability of the assisted driving system simulation testing.
[0088] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the digital optical simulation testing method of this application. Any simple modifications based on this technical concept are within the protection scope of this application.
[0089] This application provides a digital optical simulation testing device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the digital optical simulation testing method in the above embodiment 1.
[0090] The following is for reference. Figure 3 The diagram illustrates a structural schematic of a digital optical simulation testing device suitable for implementing embodiments of this application. The digital optical simulation testing device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, tablets, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 3 The digital optical simulation testing equipment shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.
[0091] like Figure 3As shown, the digital optical simulation testing equipment may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the digital optical simulation testing equipment. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. The communication device 1009 allows the digital optical simulation test equipment to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a digital optical simulation test equipment with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.
[0092] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.
[0093] The digital optical simulation testing equipment provided in this application, employing the digital optical simulation testing method described in the above embodiments, can solve the technical problem that current methods are unable to efficiently and comprehensively test assisted driving systems. Compared with the prior art, the beneficial effects of the digital optical simulation testing equipment provided in this application are the same as those of the digital optical simulation testing method provided in the above embodiments, and other technical features of this digital optical simulation testing equipment are the same as those disclosed in the previous embodiment method, and will not be repeated here.
[0094] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.
[0095] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0096] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, which are used to execute the digital optical simulation test method in the above embodiments.
[0097] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.
[0098] The aforementioned computer-readable storage medium may be included in the digital optical simulation and testing equipment; or it may exist independently and not be assembled into the digital optical simulation and testing equipment.
[0099] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the digital optical simulation testing equipment, cause the digital optical simulation testing equipment to perform the aforementioned digital optical simulation testing method.
[0100] Computer program code for performing the operations of this application can be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, and C++, and conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0101] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.
[0102] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.
[0103] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described digital optical simulation testing method. This solves the technical problem that current methods struggle to efficiently and comprehensively test assisted driving systems. Compared to the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the digital optical simulation testing method provided in the above embodiments, and will not be elaborated upon here.
[0104] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the digital optical simulation testing method described above.
[0105] The computer program product provided in this application can solve the technical problem that current methods are unable to efficiently and comprehensively test driver assistance systems. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the digital optical simulation testing method provided in the above embodiments, and will not be repeated here.
[0106] All user-related data involved in this application was obtained with the user's permission or consent, as per [reference]. Figure 4 In other words, when this application is applied to a specific product or technology, user permission is required to acquire and process the relevant data, and the processing of the relevant data must comply with the relevant laws, regulations and regulatory standards of the relevant countries and regions.
[0107] The above description is only a part of the embodiments of this application and does not limit the scope of protection of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.
Claims
1. A digital optical simulation testing method, characterized in that, The method includes: Obtain modeling parameters and modeling objects, and establish a target scene model based on the modeling parameters and modeling objects. The modeling objects include the target vehicle to be tested, and the modeling parameters include environmental parameters, physical and optical parameters of the modeling objects, and motion parameters of the modeling objects. A virtual camera and virtual sensors are created in the target scene model; Based on the environmental parameters, the physical optical parameters, the motion parameters, the virtual camera, and the virtual sensor, a simulated image is generated; Based on the simulated images, the driver assistance system of the target vehicle is simulated and tested in a purely digital environment.
2. The method as described in claim 1, characterized in that, The modeling object includes an ambient light source, and the step of generating a simulation image based on the environmental parameters, the physical optical parameters, the motion parameters, the virtual camera, and the virtual sensor includes: Based on the environmental parameters, the physical optical parameters, and the motion parameters, in the target scene model, the light rays from the ambient light source to the virtual camera are traced using a preset ray tracing method to obtain a linear exposure map; Based on the virtual sensor, the photons in the linear exposure image are converted into digital values to obtain a RAW image; The simulated image is obtained by correcting and enhancing the single-channel data in the RAW image.
3. The method as described in claim 2, characterized in that, The virtual camera has corresponding camera parameters, including physical defect parameters. The environmental parameters include the light source parameters of the ambient light source. The step of tracing the light rays from the ambient light source to the virtual camera in the target scene model based on the environmental parameters, the physical optical parameters, and the motion parameters to obtain a linear exposure map includes: Based on the light source parameters, the physical optical parameters, and the motion parameters, the energy distribution of photons emitted by the ambient light source when they reach the virtual camera after being reflected and / or refracted by the modeling object is determined in the target scene model. The modeling object also includes scene objects in the target scene model. Based on the energy distribution and the physical defect parameters, the linear exposure map is generated by the virtual camera, wherein the physical defect parameters include the aberrations of the real camera corresponding to the virtual camera.
4. The method as described in claim 2, characterized in that, The virtual sensor has corresponding sensor parameters, including sensor imaging parameters, sensor noise parameters, and sensor defect parameters. The step of converting photons in the linear exposure image into digital values based on the virtual sensor to obtain a RAW image includes: Based on the sensor defect parameters, sensor noise parameters, and sensor imaging parameters, the virtual sensor converts the photons into digital values in the order of photon, electron, voltage, and digital value conversion to obtain the RAW image.
5. The method as described in claim 1, characterized in that, The target scene model is equipped with multiple virtual cameras. The step of simulating and testing the driver assistance system of the target vehicle in a purely digital environment based on the simulated images includes: The simulated images corresponding to each of the virtual cameras are fused to obtain a fused perception map; Based on the fused perception map, a simulation test is conducted on the driver assistance system of the target vehicle.
6. The method as described in claim 5, characterized in that, The step of fusing the simulated images corresponding to each of the virtual cameras to obtain a fused perception map includes: The simulated images are fused to obtain pre-fused data; Feature information is extracted from each of the simulated images, and the feature information is fused to obtain feature fusion data; Target tracking and recognition are performed based on the simulated images, and the results of target tracking and recognition are fused to obtain post-fused data. The fused perception map is determined based on the pre-fusion data, the feature fusion data, and the post-fusion data.
7. The method as described in claim 5, characterized in that, The step of simulating and testing the driver assistance system of the target vehicle based on the fused perception map includes: Obtain high-precision maps and global path planning information; The fused perception map, the high-precision map, and the global path planning information are input into the assisted driving system to obtain a local motion trajectory. Based on the local motion trajectory, the vehicle control command for the target vehicle is determined; The vehicle control commands are fed back to the target vehicle to complete the simulation test.
8. A digital optical simulation testing device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the digital optical simulation testing method as described in any one of claims 1 to 7.
9. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the digital optical simulation test method as described in any one of claims 1 to 7.
10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the digital optical simulation testing method as described in any one of claims 1 to 7.