The Palermo-ICAR  research staff comprises research scientists from both the CNR and the University of Palermo. Currently, the total number of researchers involved in the Palermo section activities amounts to 25 units.

The Palermo-ICAR section collaborates extensively with companies, universities and research institututions, in Italy and abroad. Among the others, there are cooperations with DIAI (Univ. of  Palermo, Italy), DIST  (Univ. of Genova, Italy), Agenzia Spaziale Italiana (Italian Space Agency, Italy), LISI-INSA (Lyon, France), Dept. of Engineering (Univ. of Cambridge, UK), ICSI (Berkely, USA), LUCS (Lund University, Sweden), SONY CSL (Paris, France).

A list of publications is available.
 

Expertise and experience of Palermo-ICAR section

The Palermo-ICAR section adopted some main directions in its research activities. They are carried on by specific research groups, as shortly described in the following.

 
• Artificial Intelligence

Research interests of this group are related to hybrid knowledge representation systems and their management. In particular, we are developing decision making techniques for modelling multi-agent systems under uncertainty. Other efforts are related to mobile agents for Internet applications.
 
• Robotic Systems and Artificial Vision

This group faces with strategies for multi-robot environments, perceptive systems for robotics, the RoboCup environment, robotics for space applications. Research interests of this group are also in the areas of 3D vision (3D shape reconstruction by algorithms of shape from shading, shape from stereo and shape from motion), integration of symbolic and subsymbolic processing for image understanding, active vision based on attentive models, dynamic scene understanding, neural networks for computer vision.

• Computer Networks

This group is exploring key issues in the recent technology of Active Networks, mainly for the aspects related to a number of new challenges, such as distributed strategies, resource management, application interactions, efficiency and performance. This recent paradigm allows users to develop specific algorithms to be integrated into the network protocols, and to achieve application-oriented network functions.
Research interests of this group are also in the areas of Routing Protocols. An adaptive routing protocol sensible to the network load is explored and new heuristics for the Steiner Tree Problem in Networks are proposed to be adopted in multicast routing.

• Image Analysis and Visual Databases

Main experiences of this group are in the areas of automatic feature extraction
From images and video, image motion description, content-based indexing and retrieval in multimedia databases, similarity evaluation based on algorithmic distance measures, fuzzy techniques and neural networks.

• Distributed Information Systems

 

 
 
 
 

• Numerical Analysis and Algorithms

The research is devoted to the analysis and development of numerical methods. At the moment interests are in discrete wavelet transform (DWT) and dyadic, M-band and packet transforms are investigated.
Wavelet transforms are studied for solving integral equations regarding electromagnetic problems devoted to compute the electric and magnetic field. Moreover, DWT are also treated for problems of computer vision.
The numerical algorithms studied are also handled for parallel and distributed computational environments.
 
• Innovative Digital Computer Architecture

The research trusts of the INCA (Innovative Digital Computer Architecture) group are along the following two lines:
(i) definition and design of digital neural architectures for object and feature classification, both supervised and  unsupervised;
(ii) design and implementation of fine-grain SIMD computer architectures for digital color image processing.
These researches are developed in tight collaboration with the Dipartimento di Ingegneria Automatica e  Informatica at the University of Palermo, and with the School of Electrical and Computer Engineering at the Georgia Institute of Technology (USA).

• Multimedia Processing on Grid Computing
  

ICAR  Home Page