Underwater Image Stitching Algorithm Based on RANSAC+SIFT
Kangyou Su · 2023
Underwater resources are abundant; however, obtaining clear global views of underwater images is often challenging due to factors such as light and other obstacles. In this paper, we propose an improved method for underwater image stitching based on the RANSAC+SIFT algorithm. Firstly, two underwater images are optimized by applying an equalization algorithm to improve their balance, thereby enhancing the distribution of feature points. Secondly, the RANSAC algorithm is used for random sampling in order to reduce the number of feature points. Finally, the SIFT algorithm is used to complete the stitching of underwater images. The experimental results demonstrate that this method can effectively enhance the quality and accuracy of underwater image stitching, thereby providing robust technical support for underwater resource detection.