Research On Vision Navigation And Path Planning Of Intelligent Vehicle

Vision navigation and path planning are the key technique and resourceful hotspot of intelligent vehicle. This paper contains the two parts.The author reasearchs the vision navigation of intelligent vehicle based on the algorithm of road detection and recognition, And also studies the path planning of intelligent vehicle based on the algorithm of improved bionic algorithm—Ant Colony Algorithm(ACA).In the vision navigation, this paper primarily studies the problem about detecting and recognizing the edges of the road for navigation and adopts following operations sequence: Image obtained, image preprocessing, road edge detection, morphology processing. For image preprocessing, the OTSU method is used to extract the road region;in the edge detection, the author analyze the different algorithm and then choose sobel to detect the edge of the road, then use morphology Open-Closing processing to filter the edge. In the lane recgnition, Hough transform is used for fitting line after extracting the lane feature, and the mathod of image-sequence information comparison is used for recognizing the lane based on the linear road Mode and prior knowledge constrains. The lane image is restituted and it testifies algorithm is right and effective.This paper uses the ACA solving path planning of the intelligent vehicle problem based on the advantage and disadvantage of ACA according with the theory and model analysis of original ACA, the author proposes an improve method of adjust global pheromone updated strategy and dynamic updating random variables q0 to overcome these potential shortcomin- gs of partial optimal solution and long searching time. And the algoritm Complex analysis, convergence Analysis is also designed. Then, by solving the analogue problem of path planning for the intelligent vehicle which contains 32 nodes in a logistics distribution system in a city, the improved ant colony algorithm which gets good results is proved to be more effective than the original ant colony algorithm.Many experiments results show that the algorithm proposed by the paper are efficient, stable and accurate.