In this paper we propose to exploit cellular neural networks (CNNs) as a computational tool to obtain real-time compression of video sequences. In particular, we present a CNN-based architecture, which combines object-oriented CNN algorithms and basic coding/decoding MPEG capabilities. The proposed real-time compression architecture has been tested using standard benchmarking video sequences. Simulation results, in terms of compression ratio and peak to signal noise ratio, show that the proposed approach enables CNN-based real-time coding systems with satisfying compression ratios and good visual appearance.
A CNN-based object-oriented coding system for real-time video compression / Di Sciascio, E.; Grieco, Luigi Alfredo; Grassi, G.. - (2004), pp. 407-410. (Intervento presentato al convegno IEEE 6th Workshop on Multimedia Signal Processing, MMSP 2004 tenutosi a Siena, Italy nel September 29 - October 1, 2004) [10.1109/MMSP.2004.1436579].
A CNN-based object-oriented coding system for real-time video compression
Di Sciascio, E.;GRIECO, Luigi Alfredo;
2004-01-01
Abstract
In this paper we propose to exploit cellular neural networks (CNNs) as a computational tool to obtain real-time compression of video sequences. In particular, we present a CNN-based architecture, which combines object-oriented CNN algorithms and basic coding/decoding MPEG capabilities. The proposed real-time compression architecture has been tested using standard benchmarking video sequences. Simulation results, in terms of compression ratio and peak to signal noise ratio, show that the proposed approach enables CNN-based real-time coding systems with satisfying compression ratios and good visual appearance.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.