Aerial Image Stitching Method Based on Feature Transfer and Tile Image

For the problems of slow stitching speed and large memory resource consumption, which caused by the high pixel, high precision and large size of UAV image. In this paper, a method of feature information transfer and image slicing is proposed to improve the stitching speed of images and reduce the memory usage in the stitching process. First, the image is corrected through POS data and reduced image size. Then, the SURF algorithm is used to calculate the feature points and feature vectors of the image, the image stitching is completed by gradually entering and leaving, and the feature points and feature vectors of the previous one is assigned to the stitched image after the stitching is completed. After every N images are stitched, the images are sliced and stored, freeing up system memory. The result is a good performance between speed and memory.