Research On Forward Vehicle Ranging Based On Monocular Vision

With the number of civil vehicles increasing year by year,the number of traffic rear-end accidents is also increasing sharply,which makes people pay more attention to road traffic safety.As a kind of driving assistant system,the forward anti-collision alarm system has been paid more and more attention by automobile manufacturers.In recent years,due to the development of computer vision technology and hardware support,the front-car detection and ranging system based on vision has the advantages of low cost,comprehensive detection information and so on.Based on the object detection algorithm of vision,the forward distance detection method of monocular vision is studied in this paper,and the following work is completed :(1)Lane detection is used to extract the vehicle region in front of the vehicle,and optimize the extraction of unstructured road region to obtain the region of interest.Image enhancement is used to detect the vehicle.(2)In order to improve the detection’s efficiency and accuracy of vehicle and plate detection,the data set of real vehicle driving camera is constructed,and optimizes the clustering algorithm of YOLO.(3)The distance between the vehicle and the vehicle is obtained by perspective geometry operation of the vehicle license plate feature points,which increases the validation condition of the vehicle license plate area.(4)Use PyQt5 to design the graphical interface,realize the function encapsulation,and verify the feasibility and stability of the experiment.Experimental results show that the system has high accuracy and low false alarm rate,and has a good performance in the township road and urban road environment.It can be effectively applied to vehicle detection,ranging and other assistant driving systems.