基于旋转阵列的多帧合成孔径图像复原  

Multi-frame synthetic aperture image restoration based on rotated array

作　　者：魏小峰[1] 耿则勋[1] 娄博[1] 王洛飞[1] 

机构地区：[1]解放军信息工程大学地理空间信息学院,河南郑州450052

出　　处：《光电子．激光》2013年第4期820-825,共6页Journal of Optoelectronics·Laser

基　　金：国家自然科学基金(60778051)资助项目

摘　　要：针对光学合成孔径成像系统的调制传递函数(MTF)在中-低频段的衰减以及造成的合成图像降质问题,以6孔径Y型阵列为例,提出子孔径阵列旋转成像以获取包含互补信息的多帧图像;通过截止频率与中频能量比确定旋转角度,并利用多帧迭代盲解卷积算法进行图像复原。实验表明,与现有的代表性工作相比,本文方法对于Y型阵列的合成孔径图像有较好的复原效果和重建的准确性。Optical synthetic aperture imaging technique,which substitutes the single aperture with multiple sub-apertures of a certain arrangement, can meet the high resolution imaging demand while reducing the weight of the system. However ,the modulation transfer function （MTF） of optical synthetic aperture system is attenuated in the low and middle frequency band,which causes the degradation of image. Aiming at this problem, the characteristics and image restoration algorithm of rotated array based on Ytype array with six apertures are studied. Firstly,some of the Y-type array sub-apertures are rotated to respectively acquire the multi-frame images containing integrated information. Secondly,the rotation an- gles are determined via the cut-off frequency and mid-band energy ratio which correspond to the resolution and spectrum energy, respectively. Finally, image restoration is performed by multi-frame iterative blind deconvolution algorithm. The experimental results show that the proposed method works well on the synthetic aperture image restoration of Y-type array, and the reconstruction accuracy has also been validated. Moreover, the design of relative sensor can be innovated by this method.

关 键 词：图像复原 光学合成孔径 旋转阵列 调制传递函数(MTF) 多帧迭代盲解卷积 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]