Multi-scale Space-time Registration of Growing Plants
Haolin Pan, Julie Charlaix, Franck Hétroy-Wheeler and David Colliaux
In this paper, we introduce a new method for the space-time registration of a growing plant based on matching the plant at different geometric scales. The proposed method starts with the creation of a topological skeleton of the plant at each time step. This skeleton is then used to segment the plant into its different organs, including its main stem, its branches, etc. Then the organs are further divided into smaller segments that possess simpler geometric structures, for instance, cylinders, rectangular. Those segments are matched between two time steps using a random forest classifier based on their topological and geometric features. Then, for each pair of segments matched, a point-wise registration is devised using a non-rigid registration method based on a local ICP (Iterative Closest Point) algorithm. We applied our method to various types of plants, including arabidopsis thaliana, tomato, and maize. We established three different metrics for 3d point-wise shape correspondence to test the accuracy, continuity, and cycle consistency of the mapping and compared our method with the state-of-the-art. Our result showed that our approach could reach better or similar performance with a shorter running time.