An approach to scene matching algorithm for UAV autonomous navigation

With the rapid development of image processing technology, image matching has been widely applied in many fields. Speeded-Up Robust Features (SURF) algorithm is a robust and efficient method usually used for the matching of UAV aerial images. However, mismatch may happen when extracting interest points and doing image matching due to complex data features contained in UAV aerial images. In order to solve this problem, this paper mainly studies the factors affecting the image matching effect. It is found that the matching effect may have low quality due to the large different scales between the UAV aerial image and the matching image. By traversing different image scale, this paper finds the best matching scale. Furthermore, Sum of Absolute Differences (SAD) algorithm is used to remove fail matching feature points and to improve the matching accuracy of UAV aerial images. The experimental results show that the proposed method can effectively improve the matching accuracy of UAV aerial images.