融合MEMS-IMU信息的亚像素相关跟踪方法  

Subpixel Correlation Tracking Method Based on MEMS-IMU Information Fusion

作　　者：唐湘成[1] 田金文 黄正中[1] 刘东升[1] TANG Xiang-cheng;TIAN Jin-wen;HUANG Zheng-zhong;LIU Dong-sheng(Southwest Institute of Technical Physics,Chengdu 610041,China;School of Automation,Huazhong University of Science&Technology,Wuhan 430074,China)

机构地区：[1]西南技术物理研究所,四川成都610041 [2]华中科技大学自动化学院,湖北武汉430074

出　　处：《红外》2018年第3期29-35,43,共8页Infrared

摘　　要：结合图像制导武器的工程应用,提出了一种融合微机电系统(Micro-Electro-Mechanical System,MEMS)-惯性测量单元(Inertial Measurement Unit,IMU)信息的亚像素相关跟踪方法。该方法采用基于速率测定的误差补偿方法对MEMS--IMU的测量误差进行补偿,以实现MEMS--IMU信息与图像数据的异质数据融合,同时采用基于模板插值的亚像素相关跟踪算法来提高跟踪精度。结果表明,该方法不仅可以较好地解决目前单靠图像信息提升图像跟踪精度时所面临的算法实时性与复杂度之间的冲突问题,而且还可以有效提升动基座下的图像跟踪精度。In combination with the engineering applications of image guided weapons,a subpixel correlation tracking method with MEMS-IMU information fusion is proposed.In this method,an error compensation method based on rate measurement is used to compensate the measurement error of MEMS-IMU,so as to achieve the fusion of MEMS-IMU data and image heterogeneous data.Meanwhile,a subpixel correlation tracking algorithm based on template interpolation is employed to improve tracking precision.The result shows that this method not only can well solve the conflict problem between the real-time capacity and the complexity of the algorithm used when the image tracking precision is improved solely by image information,but also can improve the image tracking precision under moving bases effectively.

关 键 词：图像制导 异质数据融合 亚像素相关跟踪 模板插值 

分 类 号：TP765.3[自动化与计算机技术—检测技术与自动化装置]