低曝光条件下遥感相机微振动量检测  

Micro-vibration detection of remote sensing camera under low exposure condition

作　　者：乐国庆 郭永飞[1] 刘春香[1] 马天波[1] 石俊霞[1] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所,吉林长春130033 [2]中国科学院大学,北京100049

出　　处：《中国光学》2014年第6期917-924,共8页Chinese Optics

基　　金：国家自然科学基金资助项目(No.61036015)

摘　　要：针对搭载平台颤振对遥感相机成像的影响,利用辅助高速面阵CCD相机拍摄高帧频图像序列,通过像元合并来补偿曝光时间不足,增加图像的亮度、对比度和信噪比,同时结合一种区域选择算法来选择用于计算的图像区域,最后使用灰度投影算法对振动位移量进行估计。实验数据表明,提出的改进算法误差为一个像元,在准确性和稳定性方面均明显优于原始算法。To deal with the effect of flutter of carrying platform on remote sensing camera, motion estimation algorithm can be used to the image sequences taken by auxiliary high-speed array CCD camera during exposure to estimate the vibration. Pixel binnng is proposed to compensate the time of exposure and to improve the brightness and contrast as well as SNR of the images. And a region selection algorithm is used to select the im- age region by which we can estimate the displacement using gray projection algorithm. The experiment results show that the error is 1 pixel by utilizing the improved algorithm, which greatly enhance the accuracy and sta- bility.

关 键 词：遥感相机 微振动检测 运动估计 灰度投影算法 

分 类 号：V447.33[航空宇航科学与技术—飞行器设计] TP391[自动化与计算机技术—计算机应用技术]