Color Space Transform Based on Fuzzy and Neural Network Identification Algorithm

To improve the precision of color transform in different imaging devices, taking the transform of color space between RGB and CIE <em<L<sup<*</sup<a<sup<*</sup<b<sup<*</sup<</em< as an example, the fuzzy model of color space transform is put forward. It divides the input color space into many subspace and, in the subspace, the input value is turned into output via neural network. The color space transform model based on fuzzy neural system is established by using neural network system to optimize the fuzzy model of color space transform. The analytical results show that the converting accuracy of the proposed model is higher than the single fuzzy model of color space transform, and the problem of training difficulty caused by much more sampling points could be solved.