Abstract
Scale Invariant Feature Transform (SIFT) is a powerful tool in image/object matching and recognition. However, with its local nature, global information of images, such as the histogram, is ignored in its original formulation. Since histogram matching is almost a necessary condition for a pair of matching images, such ignorance can be problematic especially when SIFT is used for matching images/scenes. In this paper we propose a novel method based on making use of both SIFT features and the local intensity histograms on the feature points in order to achieve more robust image matching. And many false matches can be rejected by the proposed method. Experimental results on natural scene matching and image retrieval have showed the efficiency of the proposed approach.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
