Optimization method based on unmanned aerial vehicle camera image stitching algorithm

An optimization method based on an unmanned aerial vehicle photographed image stitching algorithm comprises the following steps: step 1, finding an intersection area of a plurality of photos, and taking the intersection area as a target area for feature point detection; 2, feature points are extracted from the intersection area of the two images through an SIFT algorithm; step 3, adding an epipolar constraint method on the basis of original BF violent matching to carry out rough matching of feature point pairs; 4, performing fine purification on the feature point matching pairs by combining a grid-based motion statistics GMS algorithm and an RANSAC algorithm so as to obtain a more accurate spliced image; 5, aligning the two images by using the homography transformation matrix; and step 6, smoothing the spliced image by adopting a weighted average fusion algorithm to obtain a complete panorama. According to the method, the problems of low registration precision, low recovery rate and the like can be effectivel