基于饱和度优先的BP神经网络颜色空间转换模型研究

Research on Color Space Conversion Model Based on Saturation Priority BP Neural Network

利用神经网络算法的非线性转换优势,构建了基于BP神经网络的颜色空间转换正向和反向仿真模型,提出了数据仿真饱和度优先的方法。通过训练样本的选取、仿真实验和数据分析,得到了较好的训练效率和转换效果。仿真结果表明,BP神经网络适合于颜色空间转换,转换精度较高。

Saturation priority method for data simulation was put forward. Forward and reverse simulation model of color space conversion based on BP neural network was constructed using the advantages of neural network in nonlinear conversion. Better train efficiency and conversion effects were obtained through test sample selection, simulation exper- iment, and data analysis. The simulation results showed that the BP neural network is suitable for color space conver- sion, and accuracy is higher.