摘要
提出了一种新的低信噪比红外序列图像运动点目标检测与跟踪算法,该算法有机地结合了TBD检测算法与粒子滤波跟踪算法。首先通过多帧TBD处理后,检测出运动目标的初始位置和运动速度,然后在跟踪阶段采用粒子滤波算法估计目标运动状态,可实现信噪比为2的点目标的检测与跟踪。对真实红外图像序列进行实验仿真,仿真结果证明,该算法具有良好的实时性与很高的精确性。
This paper presents a new low SNR infrared image sequences point target detection and tracking algorithm, the algorithm organically combines TBD detection algorithm with particle filter tracking algorithm. Through the first multi-frame TBD processed, the detection of moving objects initial position, velocity, and then tracking phase is estimated using particle filter target state, achievable signal to noise ratio is 2 points target detection and tracking. The infrared image sequences are actualized in experimental. The simulation results show that the algorithm has a good real-time with high accuracy.
出处
《电子技术应用》
北大核心
2008年第4期81-83,共3页
Application of Electronic Technique
基金
国家自然科学基金资助项目(No.60507005)
新疆维吾尔自治区教育厅高校科研计划科学研究重点资助项目(No.XJEDU2005I04)
关键词
点目标
粒子滤波器
序列图像
跟踪
point target
particle filter
image sequences
tracking
