Evaluation of Combinations of Watershed Hierarchies

Abstract

The main goal of this paper is to evaluate the potential of some combinations of watershed hierarchies. We also propose a new combination based on merging level sets of hierarchies. Experiments were performed on natural image datasets and were based on evaluating the segmentations extracted from level sets of each hierarchy against the image ground truths. Our experiments show that most of combinations studied here are superior to their individual counterparts, which opens a path for a deeper investigation on combination of hierarchies.

Evaluation of Hierarchical Watersheds

“…In future work, we plan to study the integration of more complex and relevant visual cue to define watershed hierarchies, such as the ongoing works from [30], [60] on iterative stochastic watershed hierarchies generation or [61] on watershed hierarchies combinations. Another challenge will be to take account for richer multi-scale and oriented gradient information provided by deep learning methods that enabled a large performance boost in COB [58].…”

Higra: Hierarchical Graph Analysis

“…Higra is a new library, and as such, still has a limited base of contributors and users. Some of its core functions have been used for a long time in our research group and led to dozens of publications [11,16,19,32,34,[44][45][46][47][48] (among others) and the usage of the library is spreading in new publications [26,49]. We plan to incorporate all our future research works on hierarchical graph representations in the library and we hope that Higra will benefit from a large adoption from the community leading to external users and contributors.…”