Study On Real-time Distance Detection Based On Monocular Vision Technique

Faced with serious traffic safety problem increasingly, the Intelligent Vehicle as an important part of Intelligent Transportation System more and more has to be taken seriously. With the progress of science and technology and the further study of scientific research personnel, many new methods of theory and practical technology applied in the Intelligent Vehicle to improve the intelligent level and ensure the driving safety of the vehicle. In this paper, a vehicle traveling on the highway as the research target, we propose a study on real-time distance detection based on monocular vision technique on the basis of the existing methods for improvement.Firstly, this paper adopts monocular vision measuring system which has obvious advantages by comparing the advantages and disadvantages of various ranging technology. Secondly, according to the lane line of commonly used model assumption, combined the experimental environment and the practical application demand, this paper has made the specific assumptions on the lane line model in the lane line detection and extraction process, through the analysis and comparison of various lane line detection algorithm, it sums up the essential condition of the lane line detection algorithm, on the basis of the above work, it uses a Canny edge detection and Hough transform linear fitting method to complete the lane line detection and extraction. Then according to the characteristics of freeway lane line, used the properties of parallel lines and vanishing point, found out the intrinsic relationship between the camera parameters and the vanishing point and slope, this paper completes the camera parameter calibration using the two parallel lanes on the road. Finally, established appropriate mathematical ranging model and deduced the distance between the vehicle and the front of the target, this paper achieves to measure the distance of the target objects ahead by the internal parameters and the calibration of the external parameters of camera. Through the experiment, the method can more accurately measure the distance of the target ahead in line with the requirements of high speed vehicles for real-time.