Discussion on the mosaic and geometric correction technique of UAV remote sensing image

UAV low-altitude remote sensing develops rapidly. Because of the characteristics such as unstable flight attitude, low-level image automatic matching and large image distortion, there are still just a few successful researches in efficient image stitching and image correction. So this paper takes UAV (Unmanned Aviation Vehicle) image in Jiuyuangou Township, Suide County, Yulin City, Shanxi Province for an example, firstly mosaics images based on SIFT (Scale Invariant Feature Transform) Feature Matching Method, secondly proceeds image geometric correction based on Quadratic Polynomial Method, at last overlays the result image on terrain data of 30 meters resolution, displays three-dimensional map, and views the processing results. Experiments show that SIFT Feature Matching Method and geometric correction based on Quadratic Polynomial Method are feasible for UAV image mosaic and geometric correction.