UAV Image Feature Extraction Based on Local Invariant Features

According to the problems of different scales image feature extraction, this paper proposed an algorithm based on local invariant features. Referring to the classic SIFT algorithm, the paper analyzed the most popular SURF feature detection algorithm, and compared Moravec algorithm, Harris algorithm, SUSAN algorithm, SIFT algorithm with SURF algorithm in feature extraction speed and adaptability through experimental study. Experimental results show that the SURF algorithm is better than other algorithm in extraction speed and adaptability.