基于小波分解的弯道识别技术  

Technology of Winding Roads Based on Wavelet Decomposition

作　　者：杜毓皓[1] 张亚岐[1] 李健[1] 

机构地区：[1]长安大学汽车学院,陕西西安710064

出　　处：《电子科技》2012年第8期93-95,共3页Electronic Science and Technology

基　　金：国家道路交通安全科技行动计划基金资助项目(2009BAG13A05)

摘　　要：为了保障车辆安全有效的通过弯道,准确识别弯道曲率尤为重要。因此,文中提出一种基于小波分解的弯道识别技术,该算法通过车载CCD实时采集自车前方的车道线的图像信息,对采集的图像进行小波去噪以及小波边缘检测,最终准确辨识车道线,通过曲线拟合确定道路的曲率半径。经试验验证,该算法能准确辨识出弯道的曲率半径。In order to ensure that vehicles go through winding roads safely and effectively, accurate identifica- tion of the bend curvature is particularly important. Therefore, this article proposes a technology of identifying wind- ing roads based on wavelet decomposition. This algorithm collects the real-time road images with vehicle-mounted CCD. The wavelet of the collected images is then de-noised and its edge detected. Finally the lane＇s line is accu- rately identified. The curvature radius is determined through curve fitting. Experiments show that this algorithm can accurately identify the curvature radius of the winding roads.

关 键 词：弯道识别 小波分解 曲线拟合 

分 类 号：TP206[自动化与计算机技术—检测技术与自动化装置]