复杂红外背景中运动小目标快速跟踪技术  

Fast tracking technology of moving dim and small targets under complex infrared background

作　　者：汪嘉鑫 徐贵川 于婷洋 刘正君[1] WANG Jiaxin;XU Guichuan;YU Tingyang;LIU Zhengjun(Center of Ultra-Precision Optoelectronic Instrument Engineering,College of Instrument Science and Engineering,Harbin Institute of Technology,Harbin 150001,China)

机构地区：[1]哈尔滨工业大学仪器科学与工程学院超精密光电仪器工程研究所,黑龙江哈尔滨150001

出　　处：《应用光学》2021年第3期443-453,共11页Journal of Applied Optics

基　　金：国家自然科学基金(61975044)。

摘　　要：在地面、海面、天空复杂背景下对红外小目标稳定跟踪是亟需解决的难题。为兼顾鲁棒性和实时性,以判别尺度空间跟踪算法为基础,应用可有效表征目标区域灰度突变特性和目标形状信息的广义结构张量算法作为特征提取方法。改进后的算法更适用于红外图像快速处理,能提高跟踪鲁棒性,且计算量小、效率高,目标特征维度低。为提高跟踪稳定性,依据置信度决定跟踪模型更新,避免模型被错误样本干扰。与判别尺度空间跟踪算法相比,所研究算法在准确性、实时性、鲁棒性方面均具有显著优势,且实现300 fps的跟踪速度@256×256像素图像。The stable tracking of small infrared targets in the complex backgrounds such as ground,sea surface and sky is an urgent problem to be solved.Taking into account both the robustness and real-time performances,based on the discriminative scale space tracking algorithm,the generalized structure tensor algorithm was applied that could effectively characterize the gray-scale mutation characteristics of the target area and the target shape information as a feature extraction method.The improved algorithm was more suitable for the rapid infrared image processing,which could improve the tracking robustness,and had a small amount of calculation,high efficiency,and low target feature dimension.In order to improve the stability of the model,it was decided to track the model update according to the confidence level,so as to avoid the model being disturbed by the wrong samples.Compared with the discriminative scale space tracking algorithm,the proposed algorithm has significant advantages in accuracy,real time and robustness,and achieves a tracking speed of 300 fps@256×256 pixels of image.

关 键 词：相关运算 目标跟踪 广义结构张量 

分 类 号：TN215[电子电信—物理电子学] TP391[自动化与计算机技术—计算机应用技术]