Using polarimetric imaging for material classification

In this paper we are mainly interested in the classification of images into different clusters, each corresponding to an object material in the imaged scene. For that purpose, the use of a polarimetric-based imaging system is proposed. This system has been developed in order to enlighten the huge potential of the information contained in polarization encoded images. This paper reports the experimental apparatus and its application to the classification process.

• Type: Conference
• Year: 2003
• Where: ICIP
• Authors: Jihad Zallat, Pierre Grabbling, Yoshitate Takakura
• Keywords: Classification Process, Huge Potential, ICIP 2003, Image Processing, Imaged Scene, Polarization Encoded Images