Title :
Image transmission based on neural network approaches
Author :
Refai, W. ; Zaibi, N. ; Kane, G.
Author_Institution :
Dept. of Electr. Eng., Tulsa Univ., OK, USA
Abstract :
A method for designing the vector quantizer (VQ), called concentric-shell partition vector quantization (CSPVQ), is introduced. Neural networks using the frequency sensitive competitive learning (FSCL) algorithm using the concentric-shell partitioning for VQ is presented. This new technique will show the speed and the simplicity of the neural network while retaining the computational advantages of the image. A high compression ratio is achieved which reduces the channel transmission bandwidth and improves the utilities of the communication system
Keywords :
data compression; image coding; neural nets; vector quantisation; visual communication; CSPVQ; concentric-shell partition vector quantization; high compression ratio; image coding; image transmission; neural network approaches; Bandwidth; Computer networks; Design methodology; Frequency; Image coding; Image communication; Neural networks; Partitioning algorithms; Power capacitors; Vector quantization;
Conference_Titel :
Global Telecommunications Conference, 1992. Conference Record., GLOBECOM '92. Communication for Global Users., IEEE
Conference_Location :
Orlando, FL
Print_ISBN :
0-7803-0608-2
DOI :
10.1109/GLOCOM.1992.276605
