一种改进相位相关算法的亚像元像移检测方法  

A Subpixel Image Shift Detection Method Based on Improved Phase Correlation Algorithm

作　　者：李锋[1,2,3] 董峰 冯旗[1,3] 王文秀[1,2,3] LI Feng;DONG Feng;FENG Qi;WANG Wenxiu(Shanghai Institute of Technical Physics,Chinese Academy of Science,Shanghai 200083,China;University of Chinese Academy of Sciences,Beijing 100049,China;CAS Key Laboratory of Infrared System Detection and Imaging Technology,Shanghai Institute of Technical Physics,Shanghai 200083,China)

机构地区：[1]中国科学院上海技术物理研究所,上海200083 [2]中国科学院大学,北京100049 [3]中国科学院红外探测与成像技术重点实验室,上海200083

出　　处：《红外技术》2018年第8期805-811,共7页Infrared Technology

基　　金：国家863计划探测技术(2015AAXXX5097)

摘　　要：稳像系统能够保证对地观测视频小卫星"凝视"观测时获得稳定的视频序列。不稳定视频序列的像移精确检测是稳像系统能够正常工作的前提。本文研究了一种改进的相位相关算法提取亚像元像移的方法用于视频卫星稳像系统中的像移检测。传统的相位相关算法只能得到垂直和水平方向上的整像元像移,通常视频序列图像之间的变换关系存在一定的旋转角度。本文在传统的相位相关算法中加入旋转矢量,为方便计算将空间域的旋转映射成了频域的平移,同时在求得的极值点附件利用曲面拟合得到亚像元的精度,从而保证稳像系统能获得亚像元的稳像精度。实验时,当图像尺寸为200×200、固定像移为8个像元时,检测误差均值Mean=0.0191 pixel,均方根RMS=0.01067。Image stabilization systems can ensure a stable video sequence for small satellites"staring"at the earth.Accurate detection of image movement of unstable video sequence is the precondition for normal operation of the image stabilization system.In this study,an improved phase correlation algorithm is developed to extract sub-pixel image shift for image shift detection in video satellite image stabilization systems.The traditional phase correlation algorithm can only obtain the whole pixel shift in the vertical and horizontal directions.Generally,there is a certain rotation angle between video sequence images.In this study,a rotation vector is added to the traditional phase correlation algorithm for the convenience of calculation,mapping the rotation of the spatial domain into a frequency domain translation,while simultaneously using curved surface fitting near the extreme value point to obtain subpixel precision.Thus,the image stabilization system can obtain subpixel precision.Experiments show that for an image size of 200×200 and fixed image motion of eight pixel,the test error mean is 0.0191 pixels and RMS is 0.01067.

关 键 词：稳像系统 相位相关 像移检测 亚像元 

分 类 号：TN215[电子电信—物理电子学]