基于数据相似度的多维海洋数据交互式集成可视化  

Interactive integrated visualization of multidimensional marine data based on the data similarity degree

作　　者：聂俊岚[1,2] 陈贺敏 张继凯[1,2] 郭栋梁[1,2] 

机构地区：[1]燕山大学信息科学与工程学院,河北秦皇岛066004 [2]燕山大学河北省计算机虚拟技术与系统集成重点实验室,河北秦皇岛066004

出　　处：《海洋通报》2015年第5期586-591,共6页Marine Science Bulletin

基　　金：国家"863"高技术研究发展计划;河北省科技支撑计划项目(1420311D);河北省自然科学基金(F2013203099);燕山大学青年教师自主研究计划课题(13LGB016);河北省教育厅研究重点项目(ZH2012016)

摘　　要：为充分挖掘多维海洋数据中的隐藏信息,提出一种基于数据相似度的多维数据交互式集成可视化方法。引入平行坐标作为交互工具,以颜色条带表示用户感兴趣数据集;设计一种迭代式alpha颜色混合方法,对条带重叠区域的颜色进行处理以实现相关信息的有效表征;引入数据相似度概念并给出形式化定义,通过颜色空间RGB与HSL之间的相互转换,改变混合色亮度实现其区别表征。实验结果表明,基于数据相似度的多维数据集成可视化方法不仅能够将多个相关数据集以集成的方式同时展现,并且还对各组交互数据集的相似度进行了区分表征,细化了信息展现层次,有效提高了可视化结果的信息承载量。To fully exploit the hidden information from multidimensional marine data, an interactive integrated visualization method of multidimensional data based on the data similarity degree is presented. We employ the parallel coordinate as the interactive tool and use color bands to represent datasets that users are interested in; an iterative alpha color blending method to conduct the color blending for band overlaps is designed in order to represent related information effectively;the concept of data similarity degree is introduced, and a formal definition is made, and its distinction representation by changing the lightness of blending color with the interconversion between color space RGB and HSL is realized. The experimental re-sults show that the interactive integrated visualization method of multidimensional data based on the data mutuality degree can not only represent related multiple datasets in an integrated way at the same time, but it can also realize the distinction characterization of the similarity degree between various interactive datasets, which improves the information capacity of the visualization result effectively.

关 键 词：多维数据 集成可视化 平行坐标 颜色混合 数据相似度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]