Aiming at the low accuracy of UAV image registration algorithm, and the complex calculation of SIFT algorithm feature descriptor and the low time efficiency, an image matching algorithm based on improved SIFT is proposed. Firstly, the image of UAV is preprocessed, and the feature point operator is detected by Shi-Tomasi corner, then is extracted by SIFT feature vector, and the dimension of feature vector is reduced by PCA. Then the nearest neighbor method is used for feature matching, and the BBF algorithm is used to search the feature values to improve the matching speed. Finally, the RANSAC algorithm is used to purify the feature points for accurate matching. Experiments show that compared with the traditional image matching method, the proposed algorithm improves the efficiency and accuracy of the UAV image registration technology.
展开▼