摘要
针对地球静止轨道卫星平台引起其面阵凝视成像系统的图像模糊,提出基于分时积分亚像元融合的方法来削弱平台颤振的影响,提高景物辨识度。设计并搭建了一维振动平台;利用高速图像采集系统采集不同曝光时间下不同频率振动的序列图像;使用基于能量区域质心法的相位相关法进行亚像元图像配准,计算每帧图像相对偏移量;与一维振动平台中的位移传感器数据进行对比,得到配准算法的精度优于0.1pixel。最后,选择短曝光序列图像中清晰度较高的图像完成亚像元融合,与长曝光图像相比,其融合图像清晰度更高,信噪比与长曝光图像相当。得到的结果验证了分时积分亚像元融合法对颤振模糊的抑制作用。
According to the image blurs caused by platform vibration in a staring imaging system in the geostationary orbit, a vibration suppression method based on multiple integration was proposed to e- liminate the motion blur degradation . A 1D vibration platform with a high-speed CMOS image captu- ring system was built to collect image sequences with different vibration frequencies under different exposure conditions. Next a phase correlation method based on the energy area center was proposed to calculate the relative drift of each image in sub-pixels. Then, The sub-pixel displacements between the different frames estimated by image registration algorithm were compared with the motion data meas- ured by the displacement sensor ,which proved that the algorithm precision was better than 0.1 pixels. Furthermore, several sharpest short-exposure frames were chosen for sub-pixel image fusion. Compared with long-exposure images, the motion blur in the fusion image has been weakened. It has proved that the method of multiple integration and sub-pixel image fusion is suitable for the vibration suppression of staring imaging systems in the geostationary orbit.
出处
《光学精密工程》
EI
CAS
CSCD
北大核心
2013年第8期2169-2179,共11页
Optics and Precision Engineering
基金
国家自然科学基金重点项目(No.61036015)
关键词
地球静止轨道
遥感图像
振动抑制
图像配准
图像融合
geostationary orbit
remote sensing image
vibration suppression
image registration
image fusion
