Shape from Silhouette in Dynamic Scenes

The volumetric reconstruction of a scene, from multiple views is one of the interesting problems in computer vision. Currently there is a lack of proper methods for reconstructing visual hull in time varying scenes and most of the available techniques are only concentrated on the static scenes. In this work we have developed new methods for reconstructing the visual hull of a dynamic scene incrementally. For each camera, our method processes the image differences in two consequent time steps and uses the ray casting for updating changes in the last reconstructed visual hull.     «

The volumetric reconstruction of a scene, from multiple views is one of the interesting problems in computer vision. Currently there is a lack of proper methods for reconstructing visual hull in time varying scenes and most of the available techniques are only concentrated on the static scenes. In this work we have developed new methods for reconstructing the visual hull of a dynamic scene incrementally. For each camera, our method processes the image differences in two consequent time steps and...     »