The Visual Supervisory System Design Of Intelligent Vehicle

Intelligent vehicle is an integration system of optical system and mechatronics system combined with environmental perception, path planning and decision making and automatic driving functions using computer, sensing, information, communications, navigation, artificial intelligence and automatic control technology.A experimental prototype of intelligent vehicle is discussed in this paper. At first, the key technology of intelligent vehicle is summarized. And then the design of MVC1000SA-GE30 camera, with low energy consumption, high processing speed and gigabit ethernet interface for the visual supervisory system of intelligent vehicle is presented. It can afford an extensible platform for the intelligent vehicle.The flat surface template-based method that named zhang's method is proposed for the calibration of the intrinsic and extrinsic parameters of camera, which adopts an ideal pin-hole model. The intrinsic and extrinsic parameters of camera are getting by using Transform Matrix. The HIS color model is selected and the method of using the threshold of H value to detect the stumbling block rapidly.The sense of vision supervision software system is described in the paper. The software system has 6 function parts: pictures displaying part, deciding part, alarming and communication part, status displaying part, files saving part and pictures management part. These 6 parts cooperate in harmony to fulfill the processing together.The test shows that the design for the visual supervisory system of intelligent vehicle is the effectiveness and correctness.