IMAGE STITCHING USING HARRIS CORNER & SIFT FEATURES

In this project work, the objective is to implement and design an algorithm of image stitching construction with approaches of two types of feature based, which are Harris corner and Scale-Invariant Feature Transform (SIFT) features method. The designed algorithm is implemented along with homography estimation in images which is determined by RANSAC application. This paper also discusses and reviews on various applications in the area of image stitching with different types of feature detection techniques.
The application of image stitching is found widely in many clinical studies, which includes those cancer, where the sections of whole tumour are being overlapped without affecting microscopic resolution. Through image stitching algorithms, high resolution photo of mosaics are produced and used in digital maps as well as satellite photos.Image stitching is known as the process of producing a combined large image with high resolution by the integration of such images with different angle of views. Images are overlapped in order to form a wide field of view called as panorama. Most image stitching techniques require exact overlap of images so that a better result can be formed. This paper demonstrates how different techniques applied in image stitching method result in distinct performances on the outputs.