Design Of Vehicle Panoramic Vision System With Adaptive Viewing Angle

Advanced driving assistance system improves the safety of vehicles,and the panoramic vision system is one of the important ones,which can reduce the safety accidents caused by the visual blind area of vehicles.Existing solutions mainly project the images around the vehicle to a two-dimensional plane to form a overlooking image,which will lose a lot of image information,and the driver can only observe the scene around the vehicle in a close range.The observation perspective is single,and the tension distortion of the three-dimensional objects in the observable area is serious,and the adaptability is poor.In order to solve the above defects,this thesis designed an intelligent vehicle-mounted 3D panoramic vision system with self-adaptive perspective.The specific research contents are as follows:1.In this thesis,the method of front and back frame splicing based on feature detection is studied.For the front and back cameras,in the process of the vehicle moving forward or in reverse,the current frame image is fused with the previous frames to reduce the blind area in the front and rear direction of the vehicle,so as to expand the observable area at the current moment.2.This thesis studies the target detection method and anti-collision warning of monocular camera.The monocular camera is used to collect the images in real time,and the target detection results in the image plane are obtained through the target detection network.Combined with the vehicle-road constraints,the actual distance of the target in the vehicle coordinate system is calculated by the camera calibration parameters.Then,a collision warning model considering both the collision time threshold and the ratio of the vehicle speed to distance is constructed to evaluate the target danger level and realize collision prevention warning.3.In this thesis,an adaptive texture mapping model and an adaptive perspective transformation method are proposed to realize the adaptive transformation of panoramic vision with the speed of vehicle.Firstly,the function of the model datum curve changing with the speed is constructed.Then,the 3D model is generated by the model construction method,and the image is mapped to the model by the texture mapping method.In order to provide the scene around the vehicle adaptively,this thesis constructs the function of the angle of view changing with the speed,and provides the required observation area for the driver in real time.According to the above research content,the viewpoint adaptive vehicle panoramic vision system is designed and developed,and experiments are carried out through the simulation platform and the actual scene.The experimental results show that the proposed system improves the problems of large image distortion and small observable area of the two-dimensional panoramic vision system,and the problems of poor adaptability of the existing three-dimensional panoramic vision system model and observation angle,the driving observation sense of the driver can be improved,and the driving safety of the vehicle can be enhanced correspondingly.