基于椭圆拟合的车辆遮挡处理算法  

Vehicle occlusion processing algorithm based on ellipse fitting

作　　者：朱周[1] 路小波[2] 

机构地区：[1]东南大学交通学院,南京210096 [2]东南大学自动化学院,南京210096

出　　处：《仪器仪表学报》2015年第1期209-214,共6页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(60972001;61374194)资助项目

摘　　要：为了对遮挡车辆进行分割,提出了一种基于椭圆拟合的车辆遮挡处理算法。在遮挡检测方面,利用最小二乘椭圆拟合法检测独立前景区域中的车尾阴影。如果检测到2个车尾阴影,则判断前景区域中存在着遮挡现象。在遮挡处理方面,利用与车道线方向平行的直线将前景区域轮廓分割为2个子轮廓。之后,提出了长轴方向固定的最小二乘椭圆拟合法对2个子轮廓分别进行拟合,并定义2个子轮廓到各自椭圆的距离之和作为分割评价函数,最后选择具有最小评价值的直线作为车辆遮挡的分割直线。实验结果表明,该算法能够准确地检测和分割车辆间的遮挡现象。To segment vehicles in occlusion,a vehicle occlusion processing algorithm based on ellipse fitting is proposed. For the purpose of detecting occlusion,the rear shadows in independent foreground region which are parallel to the vehicle＇s width direction are detected with the least square ellipse fitting. If two rear shadows are detected,it＇s considered that there is an occlusion in the foreground region. For occlusion processing,a line parallel to the lane is used to divide the foreground boundary into two sub-boundaries. Then the least square ellipse fitting based on a fixed direction of major axis is used to fit the sub-boundaries,and the sum of distances between the sub-boundaries and their ellipses is defined as the evaluation function of segmentation. Finally,the line with the minimum evaluation value is selected as the segmentation line. Experimental results indicate that the proposed method can detect and segment the occlusion between vehicles accurately.

关 键 词：遮挡检测 遮挡分割 椭圆拟合 最小二乘 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TH741[自动化与计算机技术—计算机科学与技术]