Mosaic method based on feature points detection and tracking for unmanned aerial vehicle videos

This paper explains and implements a fast image mosaic algorithm for unmanned aerial vehicle (UAV) sequence images. In this algorithm, feature points between images are matched using a modified Harris corner detection algorithm, and then an improved pyramid Lucas-Kanade optical flow algorithm and corner points matching authentication algorithm is used to achieve an effective match of the feature points. Next, the transformation parameters between two images are obtained for image transformation. Finally, synthetic algorithm of gradually progressive is used for image fusion. Experimental results show that the fast image mosaic method can get good results for the proposed video images which existing rotation, translation and noise distraction.