摘要
提出了一种改进的基于核相关滤波(KCF)的目标跟踪算法,该算法在传统的KCF算法的基础上,根据响应峰值来检测目标是否异常,当目标连续异常达到一定阈值时,通过引入线性预测和模板匹配方法对于丢失目标进行重新定位。实验结果表明,该算法相比传统的KCF算法,能够显著提高算法对目标遮挡后处理能力,具有更高的鲁棒性。
This paper proposes an improved target tracking algorithm based on kernel correlation filter(KCF).The algorithm detects whether the target is abnormal according to the response peak based on the traditional KCF algorithm.When the continuous anomaly of the target reaches a certain threshold,the lost target is relocated by introducing linear prediction and template matching.The experimental results show that this algorithm can significantly improve the post-processing ability of target occlusion compared with the traditional KCF algorithm,and has higher robustness.
作者
李兴阳
赵天亮
李雅欣
桑飞飞
王晨宇
肖春宝
Li Xingyang;Zhao Tianliang;Li Yaxin;Sang Feifei;Wang Chenyu;Xiao Chunbao(College of Information Engineering,Henan University of Science and Technology,Luoyang Henan 471023,China)
出处
《山西电子技术》
2023年第6期23-25,共3页
Shanxi Electronic Technology
关键词
核相关滤波
目标遮挡
丢失检测
模板匹配
kernel correlation filter
target occlusion
loss detection
template matching
