An Advanced Technique of Image Matching Using SIFT and SURF

Abstract: Image matching is a method of identifying an image from the already stored images in the database. Feature detection and feature matching are two important stages of the image matching process. This paper explains the improved method of image matching to enhance the performance of two well known images matching methods SIFT and SURF by considering the color information present in the images and provides an algorithm for reducing the matching time. The behaviour of feature detection algorithms is compared by taking one image of different conditions like illumination, scale, rotate, etc. The performance is evaluated and compare the results of SURF with SIFT by using a dataset of five images. SURF algorithm is better than the SIFT algorithm in terms of speed and will give better matching.

Keywords: SIFT, SURF, Feature detection, Image Matching.