1：Experimental Design and Method Selection:

The study proposes an improved SIFT registration algorithm based on phase congruency (PC) to implement the registration of SAR and optical images. It avoids gradient direction deflection due to gray difference by removing the principal direction of SIFT and importing a symmetric gradient. The descriptor has high stability due to the combination of the PC and gradient information of the feature points.

2：Sample Selection and Data Sources:

Two groups of images are chosen in this experiment. The first one is an Airport, where the SAR image with a size of 281×329 and a 2m resolution; the optical image is from Google Earth with a size of 268×328 and a 2m resolution. The second one is from Kunming Lake, where the SAR image is from the Cosmo, with a size of 1200×1200 and a 2m resolution, and the optical image is from Google Earth, with a size of 1200×1200 and a resolution of

3：2 m. List of Experimental Equipment and Materials:

Not explicitly mentioned.

4：Experimental Procedures and Operational Workflow:

The improved descriptor based on a PC flow diagram is shown where the descriptor is constructed using gradient information and the PC-SIFT method. According to the traditional SIFT method, the scale-space pyramid is first established, the feature points are extracted based on the DOG, and the gradient of the neighborhood of the feature points is calculated. The maximum PC moment M(x,y) of the original image is then extracted.

5：Data Analysis Methods:

The distance between the homologous points, and the number and proportion of correct matching points are chosen to verify the validity of the method.