Ikhlef Bechar

Postdoc, INRIA Sophia Antipolis

Keywords: Higher-order active contours (HOACs), Phase-field models, Prior knowledge, Shape modeling, Parameter estimation, Bayesian inference

Contact:
Mail: ikhlefdotbecharatinriadotfr
Phone: (33)4-92-38-76-83
Fax: (33)4-92-38-76-43
Postal address: INRIA Sophia Antipolis , 2004, route des Lucioles, 06902 Sophia Antipolis Cedex, France
Webpage: visit!

Abstract:

My research work consists in the usage of the Higher-Order Active Contour (HOAC) framework for object detection/segmentation in images using prior geometric information about objects. Indeed, since we regard an object of interest as a local minimum of some HOAC energy via a proper setting of the parameters, therefore the first part of my work consists in estimating such optimal HOAC parameters for any object. The second part of my work consists in proposing an efficient implementation of the proposed object segmentation approach using phase-field models and applying it on real images such as thermal infrared images. The third part of my work consists in viewing the proposed object segmentation approach as a stochastic framework (Bayesian inference) by allowing a simultaneous estimation of the optimal HOAC parameters and object segmentation using image data and prior knowledge about objects.

Short Bio:

Ikhlef Bechar received a Masters degree in Mathematics, Vision and Learning (MSc MVA) from Ecole Normale Superieure de Cachan in 2004, and a Ph.D. degree in Applied Mathematics with specialization in 3D+T video-microscopic sequence analysis from Universite Rene Descartes Paris 5 in collaboration with INRA Jouy en Josas, Institut Curie Paris and IRISA/INRIA Rennes, in 2007. He had worked during 2008 as a Postdoctoral Research Fellow in the Oxford Centre for Magnetic Resonance Imaging (OCMR) of Oxford University on the segmentation of MR vascular images, and in 2009-2010 as a Postdoctoral Research Fellow in Computer Vision within Pulsar Group, INRIA. Currently he is working as a Postdoctoral Research Fellow in the ARIANA Group, INRIA (Funding of the project by EADS Foundation) and his work concerns the use of Higher Order Active Contours (HOACs) and Phase Field Models for the modeling and the detection/segmentation of objects in images. His research interests lie mainly in the field of application of mathematical methods (deterministic and stochastic) for image/video analysis.

Last publications:
To be updated soon.