Automatic Satellite Image Stitching Based on Speeded Up Robust Feature

Abstract

The process of merging multiple images those having overlapping areas and covering different views of the same scene into a composite image is called image stitching. Image stitching can be used in variety of applications such as computer vision, medical image analysis, satellite imaging, photogrammetry etc. Image stitching also used for creating mosaic images which is useful for extending the field of view of the image. Image stitch are generally classified into direct image stitching and feature based image stitching. In this paper we proposed a feature-based satellite image stitching using Speeded Up Robust Features (SURF) algorithm. SURF mainly involves two steps, feature point detection and feature description. Detected feature points of input images are compared using nearest neighbourhood matching for identifying the overlapping points. Affine transformation-based image warping is used for aligning the input images to the mosaic frame. Finally, alpha blending is used for making the final mosaic as perfect smooth image.