光学学报, 2015, 35 (9): 0933001, 网络出版: 2015-09-01
基于模糊局部信息C均值算法的网目调光谱反射率预测模型研究
Halftone Spectral Prediction Model Based on Fuzzy Local Information C-Means Clustering Algorithm
色彩光学 光谱反射率预测 Lambert-Beer定律 模糊局部信息C均值算法 核心-边缘模型 color optics spectral reflectivity prediction Lambert-Beer law fuzzy local information C-means clustering algorit core-fringe model
摘要
提出了一种网目调光谱反射率预测方法。网目调单个网点由若干个区域组成:网点核心区域,该区域的油墨层厚度均匀,且等于实地印刷的油墨层厚度;其余边缘区域有一个模糊的厚度分布函数,这部分油墨层厚度比实地油墨层厚度薄,且越靠近网点边缘油墨层厚度越薄。采用模糊局部信息C均值算法按像素密度值聚类网目调单个网点,得到了层次分明的网点密度分布并计算各个层次的面积率。以此建立了一个新的光谱反射率预测模型,该模型预测的光谱反射率与实际印刷样张测量得到的光谱反射率具有很好的光谱匹配曲线,证明该算法模型具有较高的预测精度。
Abstract
A spectral prediction model is proposed and applied to halftone print on paper substrate. The single dot of halftone print consists of several regions: core area of the dot has uniform thickness and the same thickness of ink-layer as that of the solid print. The edge regions have fuzzy thickness distribution function, and the thickness of ink-layer is thinner than that of the solid print. The single dots of halftone print are clustered based on the density of pixel by the fuzzy local information C-means clustering algorithm, the well-bedded halftone dots can be achieved by the algorithm, and the fractional surface coverage of each cluster region can be calculated. A new algorithm model for halftone spectral prediction is established, and the spectral reflectivity predicted by the algorithm is well consisted with that measured through the halftone presswork proof, demonstrating that the prediction accuracy of the model is higher.
徐军飞, 周小凡, 石勇. 基于模糊局部信息C均值算法的网目调光谱反射率预测模型研究[J]. 光学学报, 2015, 35(9): 0933001. Xu Junfei, Zhou Xiaofan, Shi Yong. Halftone Spectral Prediction Model Based on Fuzzy Local Information C-Means Clustering Algorithm[J]. Acta Optica Sinica, 2015, 35(9): 0933001.