Adaptive image coding based on vector quantization using SOFM-NN algorithm
Suksmono A.B.a, Sastrokusumo U.a, Kondo K.a
a ITB Bandung, Indonesia
Abstract
Vector quantization (VQ) has been applied widely in image coding. The codebook of VQ system can be generated by clusterring algorithm such as K-Means and LBG algorithm or feature mapping algorithm such as SOFM-NN. In this paper, we present performance comparison between LBG and SOFM algorithm in term of PSNR for several bit rate and propose a simple adaptive mechanism for SOFM codebook. The result leads to a conclusion that LBG and SOFM are comparable, but SOFM has more advantage in providing simple adaptation mechanisms.
Author keywords
Clustering algorithm,Codebook,Self organizing feature map neural networks