PERFORMANCE EVALUATION OF POINT MATCHING METHODS IN VIDEO SEQUENCES WITH ABRUPT MOTIONS
Wael Elloumi, Sylvie Treuillet, Rémy Leconge, Aïcha Fonte · 2010
Institut Prisme, PolytechOrleans, 12 rue de Blois, 45067 Orleans cedex2, France [email protected] Keywords: Interest points, local descriptors, match ing, performance evolution, video sequence, abrupt motions. Abstract: In this paper, we compare the performance of matching algorithms in terms of efficiency, rob ustness, and computation time. Our evaluation uses as criterion, for efficiency and robustness, number of inliers a nd is carried out for different video sequences with abru pt motions (translation, rotation, combined). We co mpare SIFT, SURF, cross-correlation with Harris detector, and cross-correlation with SURF detector. Our experiments show that abrupt movements perturb a lo t the matching process. They show also that SURF is the most disturbed, by such motions, and which even fails in cases that present a large rotation unlik e the rest of descriptors as SIFT and cross-correlation.