Soutenance Darshan Venkatrayappa

Soutenance de THÈSE


Présentée par


Spécialité: I2S informatique


Mise en correspondance d’images avec des filtres tournants


Le vendredi 4 décembre 2015


Salle de conférence, site de Nîmes école des mines d’Alès


Devant le jury composé de :     

Philippe Montesinos             Ecole des mines d’Alès  Directeur de thèse
Daniel Diep       Ecole des mines d’Alès Co-encadrant de thèse
Jenny Benoit-Pineau Université Bordeaux 1(LaBRI) Rapporteur
Frédéric Jurie      Université de Caen (GREYC)  Rapporteur
Atilla Baskurt          INSA (Liris)      Examinateur  
William Puech LIRMM (Montpellier) Examinateur

Résumé :

Nowadays computer vision algorithms can be found abundantly in applications related to video surveillance, 3D reconstruction, autonomous vehicles, medical imaging etc. Im- age/object matching and detection forms an integral step in many of these algorithms.

The most common methods for Image/object matching and detection are based on local image descriptors, where interest points in the image are initially detected, followed by extracting the image features from the neighbourhood of the interest point and finally, constructing the image descriptor. In this thesis, contributions to the field of the image feature matching using rotating half filters are presented. Here we follow three approaches:

first, by presenting a new low bit-rate descriptor and a cascade matching strategy which are integrated on a video platform. Secondly, we construct a new local image patch de- scriptor by embedding the response of rotating half filters in the Histogram of Oriented gradient (HoG) framework and finally by proposing a new approach for descriptor con- struction by using second order image statistics. All the three approaches provides an interesting and promising results.

Mots-clés :

Rotating half filters, local image descriptor, image matching, Histogram of Orientated Gradients (HoG), Difference of Gaussian (DoG).