摘要
针对智能车辆导航常运用图像处理技术检测前方道路障碍物的现状,指出传统的道路检测方法一般运用空间域或频率域增强去除噪声,再使用图像分割或模板匹配技术检测道路线,认为这些方法基于线性系统,缺乏针对图像形态、结构等非线性因素的处理。介绍了3种有效、可行性强的算法,并根据实验结论对这3种算法进行对比和分析,得出了每种算法的适用范围。
The intelligence vehicle navigation is a technique commonly used to detect the obstacle in front of the road by image processing. The traditional methods of the lane detection used space area or frequency area to strengthen or weaken the voice, and then detected the road line with the image partition or template match technique. However; the methods based on line system lacked the process to the non - linear factors aiming at image appearance, structure, etc. Three effective and feasible methods were introduced. By comparing and analyzing the methods according to their their test results, the suitable algorithm range of each method was obtained.
出处
《武汉理工大学学报（信息与管理工程版）》
CAS
2008年第2期185-188,共4页
Journal of Wuhan University of Technology：Information & Management Engineering
基金
宁波市软件产业基金项目(R200505)
