一种改进的隐含相似性光学和SAR图像配准算法

Optical and SAR Image Registration Via Improving Implicit Similarity

光学和SAR(synthetic aperture radar)图像信息的互补性和特征表现的差异性使得两者的配准成为目前多源遥感图像处理的研究重点.隐含相似性配准从图像间存在结构上的相似性出发,将传统复杂的特征匹配过程简化为特征点集的迁移和仅需在单幅图像上对配准参数进行迭代搜索的过程,为光学和SAR图像配准提供新的思路.基于上述配准思想,研究用Canny算子改进特征点集提取过程,引入联合马尔可夫模型提高SAR图像去噪质量,以改进后的量子粒子群算法优化配准参数搜索过程,最终实现光学和SAR图像的配准.经实验证明,改进后的隐含相似性光学和SAR图像配准算法能达到像素级甚至亚像素级的配准精度.

Optical and synthetic aperture radar （SAR） image registration has become a research focus in the area of multisensory image processing for their information complementarity and feature difference. Based on the structural similarity between images, registration via implicit similarity simplifies the traditional feature matching process as a migration of the feature points and the iterative search of registration parameters on a single image. This method provides a new idea for optical and SAR image registration. As a result, the Canny operator is adopted to modify extraction process of feature points. The joint Markov model （JMM） is employed to improve denoising quality of SAR image. The search process of registration parameters is optimized with the modified quantum particle swarm optimization （QPSO） algorithm, and the optical and SAR image registration is finally realized. The experiment proves that the improved implicit similarity algorithm on optical and SAR image registration can reach a high accuracy of pixel level or even sub-pixel level.