利用神经网络实现地图数字打样线划宽度控制  

Using Neural Network to Realize Line Width Control in Map Digital Proofing

作　　者：汶建龙[1,2] 史瑞芝[1] 赵秀芳[3] 

机构地区：[1]信息工程大学测绘学院 [2]61363部队 [3]西安理工大学

出　　处：《测绘科学技术学报》2010年第4期294-297,共4页Journal of Geomatics Science and Technology

摘　　要：地图数字打样一直不能完全取代传统打样,一个重要原因就是打样稿与印刷品在线划宽度上存在明显差异,因此如何取得二者之间的非线性关系模型就成为解决问题的关键。在介绍了神经网络结构以及常用算法的基础之上,根据地图数字打样和印刷的特点,经过对神经网络算法进行改进,提出了基于神经网络的地图数字打样线划宽度控制模型,实现了数字打样稿和印刷品在线划宽度上的正确匹配,其应用的可靠性已通过具体实验得到了验证。One of the most important reasons why digital proofing can not replace the traditional proofing in map production field is that there is an obvious difference in line width between digital proofing and printing.So the core to work out the problem is how to model the non-linear causality between digital proofing and printing.Firstly,the structure and common algorithm of neural network were introduced.Then the algorithm was improved according to the characteristics of map digital proofing and printing.And finally,a model of line width control in the digital proofing based on neural network was presented.It has realized correct match in line width between digital proofing paper and printing production.And it has been proved to be reliable in application by experiments.

关 键 词：数字打样 色彩管理 神经网络 BP神经网络 线划宽度 

分 类 号：P288[天文地球—地图制图学与地理信息工程]