Image Mosaicing with Invariant Features Detection using SIFT

There are situations where it is not possible to capture larger views with the given imaging media such as still cameras or video recording machines in a single stretch because of their inherent limitations. So to avoid such conditions a term Image Mosaicing comes into play. This Paper presents a complete system for mosaicing a group of still images with some amount of overlapping between every two successive images. Mainly the idea is to wrap up the overlapping areas within the group of images. Detection for the common area is done using common features by the help of feature extraction from the images. In this paper technique used for the feature extraction is SIFT which is used to extract invariant features which are stable in nature. Invariant features are those features of an image which does not change even after the scaling, rotation, or zooming, change in illumination of the image is done. Multiple level filtering and downsampling are the key factors of the SIFT. So the steps involved are feature detection, matching of stable features, wrapping up of features around those feature locations. Mosaicing part consists of two major part and those are transformation matrix and bilinear interpolation. Mosaiced images are full length images which consist of all the group images. GJCST-F Classification: I.4.7 Image Mosaicing with Invariant Features Detection using SIFT Strictly as per the compliance and regulations of: Image Mosaicing with Invariant Features Detection using SIFT