Research On Vehicle Environment Perception Technology Based On Multi-sensor

Traffic safety is the focus of the society.As an important factor in the field of automatic driving,weather conditions are important for real-time feedback of road weather conditions and reduction of traffic accidents.However,a single sensor can not guarantee the accurate detection results,so it is very important to study the vehicle environment sensing technology based on multi-sensor.In this paper,based on the understanding of the research status of vehicle environmental awareness technology at home and abroad,the software and hardware design and target detection algorithm research of vehicle environmental awareness technology based on multi-sensor are carried out.The overall scheme of the system is designed,and the camera,millimeter wave radar and lidar sensors are taken as the research objects.The principle of target detection is analyzed and explained respectively.On the basis of multi-sensor target detection,using data fusion technology can effectively improve the detection accuracy.In this paper,based on the multi-sensor data fusion technology,the three sensor elements are respectively.The layout and detection area are designed to ensure the 360 ° blind area free detection of vehicles.In view of the complexity of the system working environment,the target detection algorithm of the vehicle environment awareness system is studied.This paper analyzes the defect of camera detection accuracy reduction caused by image atomization and other factors in rainy and snowy weather conditions,and proposes a detection algorithm combining MSRCR image enhancement preprocessing algorithm with improved YOLOv3 target detection algorithm;proposes a multi-target speed and distance detection method based on variable period symmetric triangular wave,which effectively improves the multi-target detection of millimeter wave radar.In order to improve the detection accuracy and output point cloud data and motion information data at the same time,a lidar target detection algorithm based on DBSCAN clustering algorithm and a target detection algorithm based on camera,lidar and millimeter wave radar data fusion are studied.The design and research of vehicle environment perception control system are carried out.In the hardware system design,according to the accuracy requirements of the system,the hardware of each sensor is selected,and then the pin part of the sensor is described to determine the power supply voltage that meets the normal operation of the system.In the design of the software system,according to the working principle of the system,the system software is designed in a modular way,and the working process of the software is determined.The general program and subprogram of some modules of the system are designed by using Matlab / Simulink software.Finally,the data processing and real-time display of each sensor are designed to meet the requirements of the system for the data receiving,transmitting and real-time display of the target motion state.Through the simulation experiment,the target detection algorithm is analyzed to verify the feasibility of the proposed algorithm,and effectively improve the target detection accuracy of camera image under the conditions of rain and snow complex weather;the fusion of camera and millimeter wave radar can obtain a higher precision relative target distance;the fusion of millimeter wave radar and laser radar can achieve the relative between the environmental point cloud information and the target.The real-time observation of distance,relative velocity and azimuth information provides convenience for the observer to observe the dynamic information of the target.