A Method of Improving SIFT Algorithm Matching Efficiency

Abstract

Due to the good invariance of scale, rotation, illumination, SIFT (Scale Invariant Feature Transform) descriptor is commonly used in image matching. but its algorithm is complicated and computation time is long. To improve SIFT feature matching algorithm efficiency, the method of reducing similar measure matching cost is mentioned. Euclidean distance is replaced by the linearcombination of cityblock distance and chessboard distance, and reduce character point in calculating with results of part feature. The experimental results show that the algorithm can reduce the rate of time complexity and maintain robust quality at the same time, image matching efficiency is improved.

Vision-Based Human Detection Techniques: A Descriptive Review

“…First, the linear combination of cityblock distance and chessboard distance should be used instead of euclidean distance to increase the SIFT feature matching efficiency. Besides, character point should be reduced while evaluating part-feature results [323]. In the next step, SIFT descriptor should be improved with the polar histogram orientation bin.…”