Monocular Vision-based Vehicles, Active Safety Technology Research,

In recent years, with the rapid increase of vehicles, traffic accidents also increased rapidly caused by vehicle collision and lane departure, the number of casualties is rising year by year, and become a public hazard in modern society. Recently, vehicle active safe driving has attracted more and more attention of some research institutions and vehicle manufacturers.Computer Vision based vehicle detection, vehicle collision warning, lane detection and lane departure warning is one important research area of vehicle active safety. Due to the low price of Vision sensors, flexible algorithm, and collecting information-rich, is widely used in vehicle active safety.In this paper, we mainly study on the technology of active safety based on monocular vision, collection of environmental information by the camera installed in front of the vehicles and processing meanwhile, realization vehicle detection, vehicle distance measuring, lane detection, and Lane Departure Warning. The main works and research results are as follows:1.In order to build up the mapping relationship between image and spatial object and vision measurement, the camera need to calibration. At first we introduced the main camera calibration Method currently, then according to the research subject using monocular camera, draw up a simple, practical and high-accuracy calibration method.2.Study a vehicle detection method using based on gradient orientation histogram, using linear support vector machine training the HOG characteristics extracted from samples then received a bin classifier, apply the classifier to vehicle detection. According image processing and geometric transform theory, combined with vehicle detection result information to realize distance measuring of the front vehicles.3.Using canny operator extraction image's edge and edge direction's information to improve the traditional Hough transform, apply improved Hough transform realize lane detection. According lane Detection result and the current driving situation determines the lane parameters.4.First give a brief introduction of decision-make method for lane departure warning. Present a calculation method for vehicle lateral shift distance and vehicle yaw angle, according to these information establishments a warning method based on time to lane cross estimation...