Image Mosaic Algorithm Based on the Combination of Harris and Edge Extraction

A fast image mosaic algorithm based on improved speeded up Harris is proposed to overcome the threshold selection problem of the original Harris algorithms. The improved algorithm detects image edge by Canny and extracted the corner by Harris. The algorithm is used to find the point satisfying these two conditions simultaneously and determine the precise location of corners by local optimization. NCC and RANSAC algorithm is used to realize image registration. The weighted average method for image fusion is applied at last and realize seamless splice. Test results show that this algorithm increases the precision and efficiency of mosaic and reduces the time consumed by image stitching and improves the accuracy and efficiency of the image mosaic. The algorithm can efficiently achieve a goal of seamless mosaic and satisfy the actual demand.