Geometric feature detection in images using a contrario theory

The main concern in designing geometric feature detectors in images is to obtain reliable results (i.e. reduced number of false detections) on various types of images without any parameter tuning. To achieve this goal, we propose an automatic procedure, having a twofold decision-making role: model validation and model selection. For a given set of pixels in a grayscale image, the detector decides if the set supports a feature (model validation), and identifies its type when multiple interpretations are possible (model selection). We describe discrete and continuous validation and model selection criteria based on the contrario theory. 

Updated biography: Viorica Patraucean completed her Ph.D. in January 2012 and now works as a postdoctoral researcher at the Department of Engineering, University of Cambridge. At the time of this seminar, she was a researcher at INRIA Ecole Polytechnique, Paris. Her research with Pierre Gurdjos (INP Toulouse) and Maks Ovsjanikov (Ecole Polytechnique Paris) focused on geometric feature detection in images and joint image and 3D shape analysis. Viorica holds an M.S. and Ph.D. degree in Computer Science from INP Toulouse and a B.S. degree in Computer Engineering from the Military Technical Academy of Bucharest.