Moving Object Detecting and Tracking Based on Predicti on Model Mechanism

摘要：

Aiming at detecting and tracking of moving object successfully, the paper sets up one time sequence autoregression(AR) model of the movement object and realizes the prediction of the movement parameters, such as displacement and speed of the object ,etc. Adopts the normalization to eliminate the uneven among different images caused by the change of illumination in the vision field. And utilizing the difference information among different frame of images, add multi-images and then averages it to realize dividing of object image in the statistics region from background. Uses one-order moment to obtain its gray centroid and takes it to represent the moving object itself and fits its moving orbit.

关键词： moving object detecting and tracking image sequence AR model transfer function

作者: Li Rui-hong Han Yue-ping

作者单位: Key Laboratory of Instrumentation Science & Dynamic Measurement North China University of Science and Technology,Ministry of Education. North University of China. Taiyuan 030051, China

会议类型: 国际会议

会议名称: 第七届国际测试技术研讨会

会议地点: 北京

会议语种:英文

在线出版日期: 2007-08-05（万方平台首次上网日期，不代表论文的发表时间）