Particle Filters for Object Tracking: Enhanced Algorithm and Efficient Implementations - Brossura

Abd El-Halym, Howida; Ismail Mahmoud, Imbaby; El-Din Habib, Serag

Brossura

ISBN 10: 3659243485 ISBN 13: 9783659243486

Casa editrice: LAP LAMBERT Academic Publishing, 2013

Riassunto

Particle filters are advanced for building robust object trackers capable of operation under difficult tracking conditions. An Excitation Particle Filter (EPF) is introduced in this book for object tracking. A new likelihood model is proposed. It depends on multiple likelihood functions: position likelihood; gray level intensity likelihood and similarity likelihood. Also, we modify the PF as a robust estimator to overcome the well-known sample impoverishment problem of the PF. The proposed enhanced PF (EPF) is implemented in software and evaluated. Simulation results demonstrated the superior performance of the proposed tracker in terms of accuracy, robustness and occlusion over classical tracking algorithms. Three efficient novel hardware architectures of the Sample Important Resample Filter (SIRF) and the EPF are introduced and implemented on FPGA platform. These architectures feature speed improvement, efficient memory utilization, and/or hardware resource saving.

Su questo libro

Informazioni sull'autore

Lecturer at Department of Computers and Systems Engineering , Faculty of Engineering, Zagazig University. Lecturer at Nuclear Research Center. PhD. and Msc. in Electronic and Communication from Faculty of Engineering, Cairo University 2010 and 2002. Bsc. in Electronic and Communication from Faculty of Engineering, Zagazig University 1993.

Dettagli dell'edizione

• Editore: LAP LAMBERT Academic Publishing
• Data di pubblicazione: 2013
• Lingua: Inglese
• ISBN 10: 3659243485
• ISBN 13: 9783659243486
• Rilegatura: Copertina flessibile
• Numero di pagine: 120
• Contatto del produttore: non disponibile
• Persona responsabile: Buchpark GmbH; gpsr@buchpark.de; ; Krügerweg 1; Trebbin; Brandenburg; 14959; Germania