基于分区的RBF神经网络颜色空间转换模型的研究

Color Space Conversion Model Based on Partition RBF Neural Networks

根据CMYK颜色空间的特点,对其进行了分区处理,利用RBF神经网络最佳逼近性能、全局最优特性和高度非线性转换的优势,构建了基于RBF神经网络每个颜色空间分区转换的正向和反向模型。通过训练样本建模和测试仿真实验,并对最后的实验数据分析处理,最终获得了一个较好的神经网络模型,并且模型的精度非常高。

The CMYK color space was partitioned according to its characteristics.Forward and inverse models of each color space partition conversion are constructed based on RBF neural network,which has the advantages of optimal approximation,global optimization characteristics,and highly nonlinear transformation.A better neural network model was got by sample model training,simulation experiments,and final experimental data analysis and processing.The model has very high precision.