Skip to Main Content
In this paper, we describe an algorithm that provides both rigid and non-rigid point-set registration. The point sets are represented as probability density functions and the registration problem is treated as distribution alignment. Using the PDFs instead of the points provides a more robust way of dealing with outliers and noise, and it mitigates the need to establish a correspondence between the points in the two sets. The algorithm operates on the distance between the two PDFs to recover the spatial transformation function needed to register the two point sets. The distance measure used is the Cauchy-Schwarz divergence. The algorithm is robust to noise and outliers, and performswell in varying degrees of transformations and noise.