GA-BP神经网络结合PCA的多基色颜色预测模型

于海琦; 刘真; 田全慧

doi:10.3788/fgxb20153606.0711

您当前的位置：

首页 >

文章列表页 >

GA-BP神经网络结合PCA的多基色颜色预测模型

发光学应用及交叉前沿 | 更新时间：2020-08-12

- GA-BP神经网络结合PCA的多基色颜色预测模型
- Multi-color Prediction Model Based on BP-NN Optimized by GA and PCA
- 发光学报 2015年36卷第6期页码：711-717
- 作者机构：
  
  1. 上海理工大学出版印刷与艺术设计学院上海,200093
  2. 上海出版印刷高等专科学校印刷包装工程系上海,200093
- 作者简介：
- 基金信息：
  
  国家自然科学基金(41271446)();上海市研究生创新基金(JWCXSL1402)资助项目()
- DOI：10.3788/fgxb20153606.0711
  中图分类号： TS801.3;TS801.9
- 纸质出版日期：2015-6-3，
  
  收稿日期：2015-3-13，
  
  修回日期：2015-4-24，
扫描看全文
于海琦, 刘真, 田全慧. GA-BP神经网络结合PCA的多基色颜色预测模型[J]. 发光学报, 2015,36(6): 711-717

YU Hai-qi, LIU Zhen, TIAN Quan-hui. Multi-color Prediction Model Based on BP-NN Optimized by GA and PCA[J]. Chinese Journal of Luminescence, 2015,36(6): 711-717
于海琦, 刘真, 田全慧. GA-BP神经网络结合PCA的多基色颜色预测模型[J]. 发光学报, 2015,36(6): 711-717 DOI： 10.3788/fgxb20153606.0711.

YU Hai-qi, LIU Zhen, TIAN Quan-hui. Multi-color Prediction Model Based on BP-NN Optimized by GA and PCA[J]. Chinese Journal of Luminescence, 2015,36(6): 711-717 DOI： 10.3788/fgxb20153606.0711.

摘要

颜色分区法是简化多基色颜色复制的主要方法

本研究以3基色为组将7基色色空间分成6个子空间

结合主成分分析(PCA)对分区内颜色样本的光谱反射率降维

通过3层BP神经网络

建立样本网点面积率与多基色复制色光谱反射率的转换模型

并通过遗传算法(GA)对神经网络权值阈值进行优化

提高多基色复制的颜色预测精度。实验结果表明

在各分区训练样本数为64、检测样本数为216时

GA-BP神经网络模型颜色预测的平均色差(

)为1.669

光谱均方根误差(RMSE)为0.7%

预测精度和稳定性均高于BP神经网络模型和胞元Neugebauer模型。最后

将训练样本数为64的GA-BP模型与训练样本数量为125

216

343的BP神经网络模型(平均

分别为3.267

2.776

2.175

光谱RMSE为0.97%

0.79%

0.76%)进行了比较

结果表明训练样本数为64的GA-BP模型的预测精度与训练样本数量为343的BP神经网络模型相当。GA-BP模型仅需少量样本即可实现高精度的颜色预测

在应用中具有良好的可移植性。

Abstract

Color partition method is the main method of multi-color reproduction. In this paper

the whole color area of 7 primary colors was divided into 6 partitions in group of 3 primary colors

and principal component analysis (PCA) was used to dimensionality reduction of the color spectrum reflectance. A three layers BP neural network was established to describe the transformation model of dot area coverage and spectral reflectance. Genetic algorithm (GA) was adopted for optimizing the weight threshold of BP NN to improve the prediction accuracy of model. Experimental results show that GA-BP model has higher prediction precision and stability compared with BP ANN and cell neugebauer model. When the number of training sample is 64 and test sample is 216 in the partition

the optimized model can predict color with the accuracy of 1.669 mean

and 0.7% spectral RMSE. By comparing with the non-optimized model with the training sample numbers of 125

216

343 (mean

are 3.267

2.776

2.175 and spectral RMSE are 0.97%

0.79%

0.76%)

the prediction accuracy of GA-BP model with the training sample numbers of 64 is equal to the accuracy of the non-optimized model with the training sample number of 343. The results show that GA-BP model can reach high precision color reproduction with small amount of samples

and have good portability in practice.

关键词

BP神经网络遗传算法主成分分析颜色预测模型光谱反射率

Keywords

BP neural networkgenetic algorithmprincipal component analysiscolor prediction modelspectral reflectance

references

Li J C. Theory and Practice of High-fidelity Color Reproduction[D]. Nanjng: Nanjing Forestry University, 2013 (in Chinese).

Tian L, Luo Y, Wang Y. Prediction model of TIG welding seam size based on BP neural network optimized by genetic algorithm [J]. J. Shanghai Jiaotong Univ.(上海交通大学学报), 2012, 47(11):1690-1696 (in Chinese).

Qu W W, Shang L P, Li X X, et al. The quantitative analysis of polycomponent PAHs by netural network based on data synthesize and principal [J]. Spectrosc. Spect. Anal.(光谱学与光谱分析), 2010, 30(10):2780-2783 (in Chinese).

Zhu M, Liu Z, Chen G X. Research on six-color separation model based on subarea neugebauer equations [J]. Acta Optica Sinica (光学学报), 2011, 31(7):0733001-1-10 (in Chinese).

Tzeng D Y, Berns R S. A review of principal component analysis and its applications to color technology [J]. Color Res. Appl., 2004, 29(2):104-110.

Fairman H S, Brill M H. The principal components of reflectances [J]. Color Res. Appl., 2005, 30(2):84-98.

Wu J, Huang F R, Huang C H, et al. Study on near infrared spectroscopy of transgenic soybean identification based on principal component analysis and neural network [J]. Spectrosc. Spect. Anal.(光谱学与光谱分析), 2013, 33(6):1537-1541 (in Chinese).

Zhou C C, Li J, Zhang X G, et al. Predication for EOR by polymer flooding based on artificial neural networkComparison between ANN and quadratic polynomial stepwise regression method [J]. Petroleum Geology & Oilfield Development in Daqing (大庆石油地质与开发), 2008, 27(3):113-116 (in Chinese).

Wang Y, Xu H S. Spectral characterization of scanner based on PCA and BP ANN [J]. Chin. Opt. Lett., 2005, 3(12):725-728.

Lin N, Yang W N, Wang B. Hyperspectral image classification on KMNF and BP neural network [J]. Computer Eng. Design (计算机工程与设计), 2013, 34(8):2774-2777 (in Chinese).

Liu C Y, Ling J C, Kou L Y, et al. Performance comparison between GA-BP neural network and BP neural network [J]. Chin. J. Health Stat.(中国卫生统计), 2013, 30(2):173-176 (in Chinese).

Luo R L, Liu S H, Su C. Garment sales forecast method based on genetic algorithm and BP neural network [J]. J. Beijing Univ. Posts Telecommun.(北京邮电大学学报), 2014, 37(4):39-43 (in Chinese).

Li Y B, Huang X Y. Resolution of chromatographic peaks by radial basis function neural network based on plate model based on a developed sorting genetic algorithm [J]. Chin. J. Anal. Chem.(分析化学), 2001, 29(3):253-257 (in Chinese).

Wu Y, Wan W. Method on designing and training of artificial neural network based on genetic algorithm [J]. J. Infrared Millimeter Waves (红外与毫米波学报), 2007, 26(1):65-68 (in Chinese).

Yu F, Zhao Y S, Li H T. Soil moisture retrieval based on GA-BP neural networks algorithm [J]. J. Infrared and Millimeter Waves (红外与毫米波学报), 2012, 31(3):283-288 (in Chinese).

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

高功率掺Tm³⁺光纤放大器热效应管理的泵浦方式优化理论研究

基于近红外光谱及主成分雷达图特征提取的产品质量稳定性研究

基于MC-UVE、GA算法及因子分析对葡萄酒酒精度近红外定量模型的优化研究

农产品品质光谱成像的空间预测规律

基于数字微镜的光谱可调星模拟器光源系统