基于梯形模型和支撑向量机的非结构化道路检测  

Unstructured-lane detection based on trapezoidal model and SVM

作　　者：张玉颖[1] 顾晓东[1] 汪源源[1] 

机构地区：[1]复旦大学电子工程系,上海200433

出　　处：《计算机工程与应用》2010年第15期138-141,175,共5页Computer Engineering and Applications

基　　金：国家自然科学基金No.60671062;国家重点基础研究发展规划(973)No.2005CB724303~~

摘　　要：在H.Jeong的梯形模型的基础上,提出了基于梯形模型和支撑向量机——SVM(Support Vector Machine)的道路检测算法。算法先对视频中提取的图像帧进行预处理,然后采用Kalman滤波及EM算法进行处理,接着用SVM得到道路检测结果,并进行滤波处理得到最终的检测结果。由于算法采用了比BP(Back Propagation)网络具有更好的分类识别效果的SVM,所以比采用BP网络的H.Jeong等人提出的模型具有更好的检测效果。该算法在预处理部分采用脉冲耦合神经网络即(PCNN-Pulse Coupled Neural Network)消除道路上的阴影,减少了光照变化对最终检测结果的不利影响。实验表明,与H.Jeong的梯形及BP算法相比,道路的检测效果更好。This paper presents a method of unstructured lane detection based on trapezoidal model proposed by H.Jeong,et al and SVM（Support Vector Machine）.The frames extracted from the video are pretreated by PCNN（Pulse Coupled Neural Network）, and then processed by Kalman filter and EM（Expectation Maximization） algorithm.Using SVM get the result of the lane detection and using morphological filter get the final detecting result.Because this method uses SVM,which has a better classified performance than BP（Back Propagation）neural network,it obtains a better detection result than that using BP neural network（H.Jeong, et al）.Furthermore,this method uses PCNN to process the frames to remove the shadow in the road,reducing the effect of illumination variations.Experimental results show that this method can receive better lane detecting results than the trapezoidal model and BP neural network proposed by H.Jeong,et al.

关 键 词：非结构化道路 道路检测 梯形模型 KALMAN滤波 支撑向量机 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]