Title :
The Hopfield neural network model for solving affine transformation parameters in the correlation method
Author :
Riyamongkol, Panomkhawn ; Zhao, Weizhao
Author_Institution :
Department of Electrical and Computer Engineering, Naresuan University, Thailand
Abstract :
In this paper, we present the Hopfield neural network model to provide solution for affine transformation parameters in the correlation method. Affine transformation equation is derived to get the mean-square error. The model maps the mean-square error equation with the energy function. When the energy function reaches local minima, the mean-square error is minimized. Outputs of the model will be affine transformation parameters. These parameters are applied in the affine transformation equation to register the corresponding two images.
Keywords :
Application software; Artificial neural networks; Correlation; Equations; Hopfield neural networks; Image registration; Medical diagnosis; Nervous system; Pixel; Predictive models;
Conference_Titel :
Region 5 Conference, 2006 IEEE
Conference_Location :
San Antonio, TX, USA
Print_ISBN :
978-1-4244-0358-5
Electronic_ISBN :
978-1-4244-0359-2
DOI :
10.1109/TPSD.2006.5507421
