| Summary: | 碩士 === 義守大學 === 電機工程學系碩士班 === 94 === In this paper, we propose a new method for land detection and tracking. We use the Particle Swarm Optimization (PSO) which is high adaptive to detect and track the lane position of the source image which is gotten in real time. Firstly, this method convert the color image to the gray image which has rich gradation. Secondly, we aim on the region of interest to use a spatial filter for extract the lane feature, and then the vanishing point is used to build a fitness function. Finally, the fitness function is used by the Particle Swarm Optimization for searching correct the lane position.
We built an image database which is actual and outdoor lane condition that include with lane mark, without lane mark, shadowy, cloudy, rainy, curvy, rough, etc. In experiments, 8,000 images were processed to evaluate the system performance. The experiment results show that the performance and the efficiency of the proposed method achieve the objection for real time lane detection and the tracing.
|
|---|
