FuzzyPSReg: Strategies of Fuzzy Cluster-Based Point Set Registration

This article studies the fuzzy cluster-based point set registration (FuzzyPSReg). First, we propose a new metric based on Gustafson–Kessel (GK) clustering to measure alignment of two clouds. Unlike c-means (FCM) in our previous work, GK-based includes orientation properties clouds, thereby providing more information for registration. We then develop quality assessment metric, which is sensitive small misalignments than that FCM-based metric. Next, by effectively combining metrics, design FuzzyPSReg strategies with global optimization. 1) FuzzyPSReg-SS , extends work and aligns similar-sized clouds greatly improved efficiency. 2) xmlns:xlink="http://www.w3.org/1999/xlink">FuzzyPSReg-O2S relatively large difference size can be used estimate pose an object scene. In experiment, use different test compare proposed method state-of-the-art approaches. The results demonstrate advantages effectiveness method.