FOMP: a novel preprocessing technique to speed-up the outlier removal from matched points

Abstract

Image matching plays a major role in many applications, including pattern recognition and biomedical imaging. It encompasses three steps: 1) interest point selection; 2) feature extraction from each interest point; 3) features point matching. For steps 1 and 2, traditional interest point detectors/extractors have worked well. However, for step 3 even a few points incorrectly matched (outliers), might lead to an undesirable result. State-of-the-art consensus algorithms present a high time cost as the number of outlier increases. Aimed at overcoming this problem, we present FOMP, a novel preprocessing approach, that reduces the amount of outliers in the initial set of matched points by filtering out the vertices that present a higher difference among their edges in a complete graph representation of the points. The precision of traditional methods is kept, while the time is speed up in 50%. The approach removes, in average, more than 65% of outliers, while keeping over 98% of the inliers.