A Scale Invariant Feature Transform based matching approach to Unmanned Aerial Vehicles image geo-reference with large rotation angle

A SIFT (Scale Invariant Feature Transform) based image feature extraction and key points matching approach is proposed for the triangle adjustment calculation of the UAV (Unmanned Aerial Vehicles) images with large rotation angle. The aerial triangle experiment of 16 strips with 787 UAV images shows the SIFT based UAV image matching approach can obtain more than 400 stable image matched key points per image so that it can realize robust external orientation parameters with AT(Aerial Triangle) than the traditional AAT(Automatic Aerial Triangle). The GCPs (Ground Control Points) accuracy of AT is less than 0.4m which can meet the requirement of 1:1000 scale map.