摘要
针对采用单一特征跟踪鲁棒性不高的问题,该文提出一种自适应多特征融合均值迁移红外目标跟踪算法。为了增强对目标的表征能力,对局部均值对比度算法进行改进,利用局部均值对比度和灰度特征表征目标。在特征融合中引入特征不确定度量方法,自适应调整不同特征对跟踪结果的贡献,有效地提高均值迁移算法的鲁棒性。为了进一步提高对尺度变化目标的跟踪性能,采取尺度算子更新跟踪窗的大小。实验结果表明,该算法提取的目标特征具有较强的鲁棒性,能实现复杂场景下的目标跟踪。
For target tracking by using single feature results in a poor performance in robustness,an infrared object tracking method based on adaptive multi-features fusion and Mean Shift（MS） is presented.In order to enhance the important features,the proposed method advances local contrast mean difference characteristic and uses advanced local contrast mean difference characteristic and grey features to present the interested target.Uncertainty measurement method is introduced in features fusion to adjust the relative contributions of different features adaptively,and the robustness of MS algorithm is significantly enhanced.Furthermore,scale operator is introduced to update tracking window in order to improve the tracking performance in size-changing target.Experimental results indicate the proposed method is more robust to present object and has good performance in complex scene.
出处
《电子与信息学报》
EI
CSCD
北大核心
2012年第5期1137-1141,共5页
Journal of Electronics & Information Technology
基金
国家863计划项目(2008AA8012320B)资助课题
关键词
目标跟踪
多特征融合
均值迁移
局部均值对比度
Target tracking
Multiple features fusion
Mean Shift（MS）
Local contrast mean difference characteristic
