Research On Slope Obstacle Ranging Method Based On Monocular Vision

This paper focuses on the universal application of computer vision technology on smart cars.Though image processing technology to detect obstacle vehicles on the pictures which captured by monocular camera that installed in cars.Then calculate distances of obstacle vehicles and provide complete position information for the active braking and active avoidance of the test vehicle.The algorithm mainly base on obstacle-vehicle have been detected,by establishing the geometric model which consider the road slope to realize position calculation of obstacle vehicle.Firstly,the internal and external parameters are calibrated by camera calibration.Using Zhang Zhengyou's camera calibration method to calculate internal parameters of the camera accurately and quickly.Considering the vanish point of the lane line and the disappearance line of the road are affected by the camera installation,the external parameters including pitch angle,yaw angle and rotation angle of the camera are calculated by using the image coordinates of the vanishing point and the gradient of the vanishing line.Then,the image preprocessing operation is performed to extract the edge features of the road lane line.Meanwhile adding constraint and using Hough transform algorithm to detect road lines.And then,analyze the gradient of the road in front of camera according to the slope change of the lane line,calculate the image coordinates of vanish point;On account of many traffic elements involve transportation,take vehicle as obstacles and use adaboost algorithm to detect the vehicles in front and extract the coordinates of the measurement points.Then,according to the image-forming principle of camera,take road gradient into account and build a measurement model to show the distance between the vehicle and obstacle.According to the position where the self-vehicle is and the positional relationship between the vehicle and the obstacle vehicle,derive the formulas to calculate the angle of road gradient and vehicle distance,after that,analyze the measurement calculate methods while drive a car following the vehicles Finally,the experimental vehicle platform built by this subject is introduced,and the visual ranging algorithm is tested and verified on the platform.