涂料颜色配方预测的人工神经网络模型实验研究

人工神经网络技术近年来在颜色空间变换或色域映射的研究中显示了特有的效果。人工神经网络用于粉末涂料配方的预测,其特点是用BP算法神经网络建立粉末涂料反射样品的标准色度参数与配方浓度参数之间的映射关系。通过对典型颜色样本训练和预测,结果表明基于多隐层BP网的模型可以实现粉末涂料样品的配方浓度空间与标准三刺激值颜色空间的相互映射,即用BP网络实现粉末涂料的配方预测。

The research of computer color matching of powder paints using BP neural networks

Using neural networks method to predict the color recipes for powder paints is proposed.The theory of the multi-layer BP neural networks model for the color recipe mapping is introduced,and the experiments for typical powder paints are carried out.The experimental results show that the method is efficient in the recipes mapping.