Research On The Method Of Distance And Speed Measurement In Front Of Vehicle Based On Vehicle Camera

A large number of researches on the road traffic accidents show that the rear-end collision is one of the most common phenomenon.A very important cause results in the rear-end collision is due to the safe distance between the vehicles is too small.According to incomplete statistics,more than 80%of the accident is caused by the corresponding driver failed to keep the safe distance,which is mainly caused by failing to control the driving speed.So keeping the safe driving speed and the safe inter-vehicle distance is the key to reduce road accidents and ensure the driving safety.On the basis of the analysis to the researches and applications of video-based intelligent vehicle speed and distance measurement technology at home and abroad,to solve the problems in the existing intelligent transportation system with fixed bayonet cameras to capture the overspeed behavior,such as the poor mobility,hard to achieving the whole road network coverage,easy avoidance by the malicious drivers,and etc.,and the high-cost problem for many in-vehicle intelligent systems,for the dual purpose of improving the surveillance ability to the overspeed behavior in the whole road network and ensuring the driving safety of the self-vehicle,research on the speed and distance measurement to the front vehicle base on vehicle-mounted camera is carried out in this paper.The main research contents and achievements include:1.Front-vehicle detection and feature recognition technology.Proposed a target recognition method which combined the multi-features with improved Hough-forest algorithm.At the classifier-training phase,multiple features fusion technology is used to solve the problem in which single feature cannot fully describe the vehicle characteristics;at the vehicle detection stage,by adopting the improved Hough-forest voting scheme to further improve the accuracy and achieving the real-time vehicle detection.2.The technology of front-vehicle distance measurement.A monocular vision measurement method based on the pitch angle dynamic compensation is proposed.By combining the geometric-ranging model with camera calibration technology,and by adopting dynamic compensation based on the change of the pitch angle of the vehicle-mounted camera,the distance measurement error caused by the change of the pitch angle in the dynamic environment is reduced.Experimental verification shows that the proposed method has good real-time performance and high accuracy,which can meet the requirement of distance measurement and get the accurate distance information of the front vehicle.3.Front-vehicle speed measurement technology.By utilizing the vehicle-mounted camera,and the distance measurement change of the front vehicle at the same lane,a method based on the variations of soft measurement of the vechicle speed is proposed.The experiments result show that this method can meet the real-time performance and accuracy requirements for the surveillance on vehicle overspeed behavior.