摘要
针对传统的mean shift算法在复杂场景下跟踪目标易丢失的问题,提出一种mean shift算法与卡尔曼预测相结合的算法.根据mean shift跟踪位置与卡尔曼预测位置的比较,确定最终的跟踪位置,并将其作为下一次跟踪的起点,跟踪结束后显示车辆运动的轨迹.实验结果表明:当目标与背景的颜色相近时,该文算法更具鲁棒性,且仍可满足实时性的要求.
In complex cases, the traditional mean shift algorithm easily loses the target. An algo- rithm which combines the mean shift and Kalman prediction algorithm is proposed. By comparing the location of mean shift tracking with Kalman prediction, the algorithm determines the final tracking position which is regarded as the starting point of the next tracking, and the trajectory of vehicle motion is shown after tracking is finished. The experimental results show that the algorithm improves the problems of easy losing tracking target under complex scenarios, and the proposed algorithm can still meet the requirement of real-time.
出处
《扬州大学学报(自然科学版)》
CAS
北大核心
2014年第2期57-59,64,共4页
Journal of Yangzhou University:Natural Science Edition
基金
江苏省省级现代服务业(软件产业)发展专项引导资金项目(苏财建[2010]401号)
扬州市-扬州大学科技合作资金计划项目(YZ2011149)