Hierarchical Motion Decomposition for Dynamic Scene Parsing

Juan-Manuel Pérez-Rúa (1, 2) , Tomas Crivelli (2) , Patrick Pérez (2) , Patrick Bouthemy (1)

1 SERPICO - Space-timE RePresentation, Imaging and cellular dynamics of molecular COmplexes
2 Technicolor [Cesson Sévigné]

A number of applications in video analysis rely on a per-frame motion segmentation of the scene as key preprocess-ing step. Moreover, different settings in video production require extracting segmentation masks of multiple moving objects and object parts in a hierarchical fashion. In order to tackle this problem, we propose to analyze and exploit the compositional structure of scene motion to provide a segmen-tation which is not purely driven by local image information. Specifically, we leverage a hierarchical motion-based partition of the scene to capture a mid-level understanding of the dynamic video content. We present experimental results showing the strengths of this approach in comparison to current video segmentation approaches.

Mots clés: conditional random fields; motion models; Index Terms— Video segmentation; hierarchical motion decomposition

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]; Traitement des images [eess.IV]

Juan-Manuel Pérez-Rúa, Tomas Crivelli, Patrick Pérez, Patrick Bouthemy. Hierarchical Motion Decomposition for Dynamic Scene Parsing . 2016 IEEE International Conference on Image Processing (ICIP) , IEEE, Sep 2016, Phoenix, United States. ⟨hal-01314095⟩