Consistent Multi-Robot Object Matching via QuickMatch

Abstract: In this work, we present a novel solution and experimental verification for the multi-image object matching problem. We first review the QuickMatch algorithm for multi-image feature matching and then show how it applies to an object matching test. The presented experiment looks to match features across a large number of images and features more often and accurately than standard techniques. This experiment demonstrates the advantages of rapid multi-image matching, not only for improving system performance, but also for use in new applications, such as object discovery and localization.