隐层改进的BP网络在织物染色配色中的应用

Research on Application of BP Neural Networks in Computer Color Matching for Textile Dyeing Based on Hidden Layer Improvement

为提高BP网络及其改进网络的收敛和泛化能力,依据计算机配色理论,深入研究了BP神经网络的优缺点,在此基础上提出了一种基于隐层输出反馈改进的BP网络训练算法应用于织物染色配色的算法,并按照此算法进行了织物染色的计算机配色实验。实验验证了人工神经网络算法在织物染色配色中应用的可能性和可靠性,为神经网络模型在织物染色配色中选择和构造合适的高性能网络结构提供了参考。

In order to improve the convergence and generalization capacity of BP neural network,the fundamental computer color matching principles based on Kubelka-Munk theory is introduced in this paper.The advantages and disadvantages of neural network in color matching for textile dyeing are discussed.And then a network is designed and trained by a modified training algorithm based on BP neural network,which references the output of hidden layer,for textile dyeing computer color matching.Finally,the possibility and reliability of the modified BP network used in color matching for textile dyeing have been confirmed by experiments.Meanwhile,a reference is provided for choosing a suitable neural network model.