This paper presents a texture classification approach based on function link network. Image texture is characterized by the second order Gauss MRF model

and the least square error estimation is employed for the estimation of model parameters. However

these parameters are proved to be inefficient in texture classification. To solve this problem

we introduced a function link network to improve the classification performance. Experiment shows that better classification results can be obtained than traditional euclidean distance approach

and it has the advantage of simple processing procedure and fast convergence speed.