SAR Images Co-registration Based on Phase Congruency Algorithm

Abstract

As the first step in image processing operations, corner detection is a vital procedure where it can be applied to many applications such as matching features, image registration, image mosaicking, change detection…. Registration of images can be described as the process of getting the pixel location misaligned between two or more images. A modified corner detector is proposed in this paper based on a combination of both phase congruence, later named PC, and Harris corner detector where PC image can provide fundamental and meaningful features despite complex changes in intensity. The performance was similar to detectors for the Shi-Tomasi, FAST, and Harris corner. Experiments are carried out using simulated images. As metric metrics, MSE (mean square error) and PSNR (peak signal-to-noise ratio) are used. The experimental results verify the effectiveness where the advantages of image constitutional representation are utilized, allowing the extraction of the powerful key points since it preserves the robustness of the coregistration process using image frequency properties that are not variant to illumination. It also has the ability to produce reasonable results as opposed to state-of-the-art such as Shi-Tomasi, FAST, and Harris algorithms.