摘要
本文针对受光线干扰影响的复杂背景下目标检测问题,实现运动靶标的实时跟踪。研究采用方向滤波器克服场景光照以及大气湍流对目标特征造成模糊的影响。采用归一化积相关法计算目标模板与实时图像的相关性定位靶标位置,并研究分层金字塔模板搜索方法减少计算量实现了靶标的快速跟踪。为了提高系统的鲁棒性,消除场景变化对目标跟踪位置产生的偏差及错判,本文采用了目标轨迹运动预测技术。通过实验验证,结果表明本系统目标跟踪稳定可靠,计算效率高,满足工程中的实际应用需要。
This paper research target detection on the light interference effect under complicated background. We implement the real-time track of motion target. We research direction filter the fuzzy target characteristics which caused by thescene illumination and atmospheric turbulence. We locate target position by the NProd method. This Methodcan obtain the correlation of target template and real-time image.We study the layered pyramid template search method to reduce computation. Then we achieve the fast targettracking. In order to improve the robustness of the systemand eliminate the scene changes on target tracking position deviation and wrongly, this paper adopted the target trajectory prediction technology. Through the experiment, the results show that the target tracking system is stable, reliable and high computational efficiency. The methods of this paper can meet the needs of the practical application in engineering.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2014年第S1期34-38,共5页
Chinese Journal of Scientific Instrument
基金
吉林省科技发展计划项目(20130102017JC)
国家高技术研究发展计划(863)(006AA703405F)资助