为了实现扫描仪在不同光源、不同观察者条件下准确获取颜色信息,最大程度的避免同色异谱现象,本文采用光谱的方法对扫描仪进行特性化处理,通过多项式回归和BP神经网络分别与主成分分析法结合,首先对检测样本的光谱反射率进行主成分分析,提取主成分与主成分系数,通过实验得到主成分系数与多项式回归、BP神经网络结构之间的转换模型,实现了扫描仪低维RGB信号对原始光谱反射率信息的重构,进而实现扫描仪的光谱特性化.实验结果表明,多项式项数为19项时,达到训练样本的均方根误差为1.7％,检测样本的均方根误差为1.9％.而包含15个隐层节点的单隐层BP神经网络结构为比较合理的网络结构,达到训练样本的均方根误差为1.3％,检测样本的均方根误差为1.5％.对彩色扫描仪的特征化处理,采用多项式回归法得到光谱特性化精度较低,采用BP神经网络模型能够实现更高的光谱特性化精度.
参考文献
| [1] | 徐艳芳,刘文耀,左坤隆,朱昊.彩色扫描仪的特征化[J].光学精密工程,2004(01):15-20. |
| [2] | 徐艳芳,黄敏,金杨.基于色度比特性的显示器色度特性化[J].液晶与显示,2008(06):771-777. |
| [3] | Kang H R .Color scanner calibration[J].Journal of Imaging Science and Technology,1992,36(02):162-170. |
| [4] | Johnson T .Methods for characterizing color scanners and digital cameras[J].DISPLAYS,16(04):183-191. |
| [5] | Miaohong Shi;Glenn Healey .Using reflectance models for color scanner calibration[J].Journal of the Optical Society of America, A. Optics, image science, and vision,2002(4):645-656. |
| [6] | Shen HL;Xin JH .Colorimetric and spectral characterization of a color scanner using local statistics[J].Journal of Imaging Science and Technology,2004(4):342-346. |
| [7] | Yong Wang,Haisong Xu.Spectral characterization of scanner based on PCA and BP ANN[J].中国光学快报（英文版）,2005(12):725-728. |
| [8] | Hui-Liang Shen;John H. Xin .Spectral characterization of a color scanner by adaptive estimation[J].Journal of the Optical Society of America, A. Optics, image science, and vision,2004(7):1125-1130. |
| [9] | 张显斗,王强,杨根福,武莉.光谱颜色管理系统关键技术综述[J].中国印刷与包装研究,2013(01):10-17. |
| [10] | Di-Yuan Tzeng;Roy S. Berns .A Review of Principal Component Analysis and Its Applications to Color Technology[J].Color Research and Application,2005(2):84-98. |
| [11] | Guowei Hong;M.Ronnier Luo;Peter A.Rhodes .A Study of Digital Camera COlorimetric Characterization Based on polynomial Modeling[J].Color Research and Application,2001(1):76-85. |
| [12] | Vrhel M J;Trussell H J.Color scanner calibration via a neural network[A].Phoenix,USA:IEEE,1999:3465-3468. |
上一张
下一张
上一张
下一张
计量
- 下载量()
- 访问量()
文章评分
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%