Optimizing Partition Trees for Multi-Object Segmentation with Shape Prior

Emmanuel Maggiori (1, 2) , Yuliya Tarabalka (1, 2) , Guillaume Charpiat (3, 4)

1 TITANE - Geometric Modeling of 3D Environments
2 AYIN - Models of spatio-temporal structure for high-resolution image processing
3 TAO - Machine Learning and Optimisation
4 STARS - Spatio-Temporal Activity Recognition Systems

A partition tree is a hierarchical representation of an image. Once constructed, it can be repeatedly processed to extract information. Multi-object multi-class image segmentation with shape priors is one of the tasks that can be efficiently done upon an available tree. The traditional construction approach is a greedy clustering based on color similarities. However, not considering higher level cues during the construction phase leads to trees that might not accurately represent the underlying objects in the scene, inducing mistakes in the later segmentation. We propose a method to optimize a tree based both on color distributions and shape priors. It consists in pruning and regrafting tree branches in order to minimize the energy of the best segmentation that can be extracted from the tree. Theoretical guarantees help reducing the search space and make the optimization efficient. Our experiments show that we succeed in incorporating shape information to restructure a tree, which in turn enables to extract from it good quality multi-object segmentations with shape priors.

Domaines: Traitement des images [eess.IV]

Emmanuel Maggiori, Yuliya Tarabalka, Guillaume Charpiat. Optimizing Partition Trees for Multi-Object Segmentation with Shape Prior. 26th British Machine Vision Conference, Sep 2015, Swansea, United Kingdom. ⟨hal-01182776⟩