Skip to Main Content
The registration of high resolution panchromatic image and the low resolution multi-spectrum image plays a growing vital role in the remote sensing applications. Aiming at this problem, during this paper, a locally adaptive registration algorithm of the remote sensing image based on SIFT is proposed. The algorithm can keep matching points uniformly distributed. First, obtain SIFT points feature descriptors from base image and warp image by cutting up them, which realizes the uniform distribution of feature points and also speeds up the extraction rate of SIFT feature points. Then, select the best block matching points adaptively using geometric constraint to improve the image matching accuracy. Finally, some mismatched points are eliminated by using consistency checking, which further improved matching accuracy. Numerous experiments have been conducted for different type images. The experimental results show that the improved algorithm can realize registration of great solution difference images effectively and accurately.
Date of Conference: 25-30 July 2010