Similarity for Shape Analysis

Опубликовано 27 июля 2016, 19:31

3D shape analysis aims at developing computational tools for reasoning about the properties of the object`s shape. It has numerous applications in computer graphics and computer vision. A fundamental aspect of shape analysis is the similarity between objects or between components of objects. I will begin my talk by briefly describing our work on several similarity-based problems, including retrieval, reconstruction from lines, segmentation and colorization. Then, I will focus on one problem - saliency detection on 3D surfaces. While the detection of the salient regions in images has been extensively studied, relatively few papers have addressed surfaces. I will present a novel algorithm to perform this task and its use in viewpoint selection.