An Improved Algorithm for Image Mosaic Based on Speeded-Up Robust Features

Due to the shortcomings of the large data and time consuming in the conventional image mosaic algorithms, an improved algorithm based on Speeded-up Robust Features(SURF) is presented in this paper. The integral image is introduced for image convolutions to reduce the computation and Hessian matrix is adopted for the detector to extract the interest points. For each feature point, the dominant orientation is assigned by Haar-wavelet response, and then the descriptor is determined. Finally, the Homography matrix between two images is obtained by RANSAC. Experimental result shows that this algorithm with strong robustness performs rapidly and effectively.