Object-oriented approach to video compression via Cellular Neural Networks

Video compression technologies have recently become an integral part of the way we create and consume visual information. This paper aims to show that the Cellular Neural Network (CNN) paradigm can be exploited for obtaining accurate video compression. In particular, the paper presents an architecture that combines CNN algorithms and H.264 codec. The compression capabilities of the devised coding system are analyzed using benchmark video sequences, and comparisons are carried out between the CNN-based approach and the H.264 codec working alone. The outcome of the analysis is that the CNN-based approach outperforms the H.264 codec working alone, making perceive the capabilities of the CNN paradigm.