Low bit rate image coding based on vector transformation with neural network approach
Suksmono A.B.a, Karsa K.a, Tjondronegoro S.a, Soegijoko S.a
a ITB, Indonesia
Abstract
Vector transformation is a new methods in unifying vector quantization (VQ) and transform coding. So far, codebook generation that has been applied in this coding is LBG algorithm. With the development of neural network, especially Self Organizing Feature Map (SOFM), there are some advantages that can be use to improve system’s performance. In this paper, we explore the application of SOFM algorithm to generate VTC codebook and compare the result for some coding rate with LBG algorithm.
Author keywords
Self organizing feature map,Vector transform coding