Robust Descriptors Fusion for Pedestrians' Re-identification and Tracking Across a Camera Network

Ahmed Derbel (1) , Yousra Ben Jemaa (2) , Sylvie Treuillet (3) , Bruno Emile (3) , Raphael Canals (3) , Abdelmajid Ben Hamadou (4)

1 PRISME - Laboratoire Pluridisciplinaire de Recherche en Ingénierie des Systèmes, Mécanique et Energétique
2 Signal and System Research Unit
3 IRAUS - Département Images, Robotique, Automatique et Signal [Orléans]
4 MIRACL - Multimedia, InfoRmation systems and Advanced Computing Laboratory

In this paper, we introduce a new approach to identify people in multi-camera based on AdaBoost descriptors cascade. Given the complexity of this task, we propose a new regional color feature vector based on intra and inter color histograms fusion to characterize a person in multi-camera. This descriptor is, then, integrated into an extensive comparative study with several existing color, texture and shape feature vectors in order to choose the best ones. We prove through a comparative study with the main existing approaches on the VIPeR dataset and using Cumulative Matching Characteristic measurement that the proposed approach is very suitable toidentify a person providing very satisfactory performances.

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Ahmed Derbel, Yousra Ben Jemaa, Sylvie Treuillet, Bruno Emile, Raphael Canals, et al.. Robust Descriptors Fusion for Pedestrians' Re-identification and Tracking Across a Camera Network. nternational Conference on Computer Vision Theory and Applications, Feb 2013, Spain. pp.92-97. ⟨hal-00924706⟩