Evolutionary Computation Schemes based on Max Plus Algebra and Their Application to Image Processing

• Published in: 2006 International Symposium on Intelligent Signal Processing and Communications [sic.] : Yonago, Japan, December 12-15, 2006 pp. 538 - 541
• Main Authors: Nobuhara, H., Chang-Wook Han
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2006
• Subjects: Algebra; Evolutionary computation; Genetic algorithms; Image coding; Image processing; Image reconstruction; Learning systems; Multi-layer neural network; Neural networks; Optimization methods
• ISBN: 9780780397323; 0780397320
• DOI: 10.1109/ISPACS.2006.364715

A hybrid genetic algorithm based learning method for the morphological neural networks (MNN) is proposed. The morphological neural networks are based on max-plus algebra, therefore, it is difficult to optimize the coefficients of MNN by the learning method with derivative operations. In order to solve the difficulty, a hybrid genetic algorithm based learning method to optimize the coefficients of MNN is proposed. Through the image compression/reconstruction experiment using test images extracted from standard image database (SIDBA), it is confirmed that the quality of the reconstructed images obtained by the proposed learning method is better than that obtained by the conventional method.