Title :
Low bit rate image coding based on vector transformation with neural network approach
Author :
Suksmono, A.B. ; Karsa, K. ; Tjondronegoro, S. ; Soegijoko, S.
Author_Institution :
Dept. of Electr. Eng., ITB, Bandung, Indonesia
Abstract :
Vector transformation is a new method in unifying vector quantization (VQ) and transform coding. So far, the codebook generation that has been applied in this coding is the LBG algorithm. With the development of neural networks, especially Self Organizing Feature Maps (SOFM), there are some advantages that can be used to improve a system´s performance. In this paper, we explore the application of the SOFM algorithm to generate the Vector Transform Coding (VTC) codebook and compare the result with some coding rates using the LBG algorithm
Keywords :
image coding; self-organising feature maps; transform coding; vector quantisation; VQ; codebook generation; low bit rate image coding; neural network approach; self organizing feature maps; transform coding; vector quantization; vector transform coding; vector transformation; Bit rate; Discrete wavelet transforms; Image coding; Neural networks; Organizing; Signal processing; Signal processing algorithms; Transform coding; Vector quantization; Video coding;
Conference_Titel :
Circuits and Systems, 1998. IEEE APCCAS 1998. The 1998 IEEE Asia-Pacific Conference on
Conference_Location :
Chiangmai
Print_ISBN :
0-7803-5146-0
DOI :
10.1109/APCCAS.1998.743892
