预估计混叠度的MAP超分辨率处理算法-软件学报.PDFVIP

1000-9825/2004/15(02)0207 ©2004 Journal of Software 软 件 学 报 Vol.15, No.2 ∗ 预估计混叠度的 MAP 超分辨率处理算法 孟庆武+ (哈尔滨工业大学 航天学院,黑龙江 哈尔滨 150001) An Aliasing Degree Pre-Estimated MAP Algorithm of Super-Resolution Processing MENG Qing-Wu+ (School of Astronautics, Harbin Institute of Technology, Harbin 150001, China) + Corresponding author: Phn: +86-451 E-mail: mengqw2001@, Received 2003-01-21; Accepted 2003-06-04 Meng QW. An aliasing degree pre-estimated MAP algorithm of super-resolution processing. Journal of Software, 2004,15(2):207~214. /1000-9825/15/207.htm Abstract: In this paper, a sort of aliasing degree pre-estimated MAP (maximum a posteriori) algorithm, PEMAP, is proposed. It is used in super-resolution processing of satellite images on the ground. The aliasing degree of satellite images is firstly determined in frequency domain by spectrum analysis, and then used in a joint estimation of the registration parameters and high-resolution image in space domain as a priori information. This algorithm overcomes the blindness and instability of the MAP algorithm and improve its flexibility. Good results have been shown in a practical satellite image processing. Key words: satellite image; super-resolution processing; aliasing degree; registration parameter; MAP (maximum a posteriori) estimation 摘 要: 提出一种预估计混叠度的PEMAP(pre-estimated MAP (maximum a posteriori))算法,用于卫星图像的地面 超分辨率处理.它通过频域分析确定卫星图像的混叠度,将其作为先验信息在空域控制 MAP 估计的循环迭代,联合 估计帧间位移和高分辨率图像.该算法克服了最大后验概率MAP 算法的盲目性和不稳定性,使其适应性更好.实际 的卫星图像处理显示了较好的处理效果. 关键词: 卫星图像;超分辨率处理;混叠度;位移参数;MAP(maximum a posteriori)估计 中图法分类号: TP391 文献标识码: A 图像复原及超分辨率处理是图像处理的一项重要内容.它以改善图像质量为目标, 以求获得更清晰的高分 辨率图像.在过去的 30 年中,对于退化图像的复原以及高分辨率图像的获得,人们已经形成了较为成型的系统 理论. 用线性的退化过程模拟图像的退化, 由品质已经下降的图像恢复原始