Title :
Post-nonlinear blind image separation algorithm with SOM initialization method
Author :
Junhua Wang ; Mingxiang Wang ; Yong Fang
Author_Institution :
Sch. of Commun. & Inf. Eng., Shanghai Univ., Shanghai
Abstract :
An algorithm for post-nonlinear blind image separation is proposed based on self-organizing maps (SOM). This method can improve the convergence rate and avoid lapsing into the local minima. Simulation results verify its validity.
Keywords :
blind source separation; image processing; self-organising feature maps; SOM initialization method; local minima; post-nonlinear blind image separation algorithm; self-organizing maps;
Conference_Titel :
Wireless, Mobile and Sensor Networks, 2007. (CCWMSN07). IET Conference on
Print_ISBN :
978-0-86341-836-5
