摘要
采用一种自适应背景差分法中背景更新的改进方案对车辆进行检测,并结合迭代式阈值法和人工阈值法完成目标分割。检测出车辆后,运用改进的卡尔曼滤波器运动跟踪算法跟踪车辆以缩小特征搜索范围。
During the implementation of Adaptive Background Difference method,the paper adopts an improved scheme on updating background for specific applications to detect the vehicles.The iterative threshold method and threshold-choosing method by manual are combined to segment the objects.Based on these objects segmented,the paper discusses an improved tracking algorithm based on Kalman Filter（KF） used to accurately estimate the feature＇s position so as to decrease the size of the region matching feature.
出处
《煤炭技术》
CAS
北大核心
2011年第3期195-197,共3页
Coal Technology
关键词
背景提取
背景更新
目标检测
卡尔曼滤波
目标跟踪
Background extraction
Background update
Object detection
Kalman filter
Object tracking
