基于蒙特卡罗算法的车辆跟踪方法.pdfVIP

40 2 Vol.40, No.2 2011 4 Acta Geodaetica et Cartographica Sinica Apr., 2011 2011) 1, 2 1, 2 王文龙 ,李清泉 1. , 430079; 2. , 430079 A Vehicle Tracking Algorithm with MonteCarlo Method 1,2 1, 2 WANG Wenlong , LI Qingquan 1. State Key Laboratory for Information ngineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan430079, China;2.Transportation ResearchCenter, Wuhan University, Wuhan 430079, China AbstractAiming to the highdefinitiontraffic video collected by airship flying along road,a vehicle tracking a pproach based on MonteCarlo a lgorithm is proposed. The particles state characteristics such as position a nd scale ca n be obta ined by MonteCarlo a lgorithm to forecast. Since the similarity between eachforecasted state a nd the state of reference target can be derived, the state with maximum similarity can be identified, which is treated as the track ing result. The xperiment research shows that this approach cantrack the irregular a nd nonlinea r moving vehicles real time a nd accurately, deal withthe situation of short time occlusion as well. Key words : vehicle tracking; MonteCa rlo algorithm; lowa ltitude data collection platform; traffic video :针对飞艇沿公路飞行获取的高清晰交通视频, 研究基于蒙特卡罗 法的车辆跟踪方法该方法用蒙特卡罗 法 预测目标下 一时刻的一组状态(位置和尺度) ;然后通过计 得到各预测状态中与参考目标的状态相似度最大的一个, 即 为车辆目标跟踪的结果研究结果表明:该方法能实时准确地跟踪道路上无规律非线性运动的车辆,并能够应对跟踪 过程中的短时遮挡情况 :车辆跟踪; 蒙特卡罗 法;低空数据采集平台; 交通视频 :U491. 254 : A : 国家科技支撑计划( 2008BAK49B02) ; 国家863 计划( 2009AA11Z213,2009AA11Z218) ;测绘遥感信息工程国家 重点实验室专项研究基金 , , , , ; [ 9] , , , ,