Evaluation of morphological hierarchies for supervised video segmentation
Filipe Tório. L. R. Nhimi, Zenilton K. G. Patrocínio, Benjamin Perret, Jean Cousty, Silvio Jamil F. Guimarães · 2018
This paper aims to evaluate the morphological hierarchies (observation scale, watersheds area and volume) in the context of marked video segmentation. Here, we automatically created a set of markers for each video taking into account its ground-truth. In order to evaluate the hierarchies, we have applied several types of markers and their combinations for simulating supervised segmentation idea. The performance of each segmentation was evaluated in terms of F-measures, and according to our results, the observation scale hierarchies outperform the considered hierarchies when the markers for both background and foreground were skeleton-based.