基于聚类的密集型地理数据可视化改进
首发时间:2018-01-22
摘要:随着物联网技术的发展以及移动设备、传感器设备等的不断普及,如今获取地理数据的手段越来越多,现在已经处于地理大数据时代。地理数据可视化能够以地图方式展现数据,帮助用户以直观有效的方式进行数据分析,目前已有很多研究人员对地理数据可视化方向展开深入研究。但是在处理密集型地理数据可视化问题时,已有方案准确率低、时间复杂度高,为用户后续分析造成了较大的误差。本文针对密集型点数据可视化问题,设计并实现了基于格网划分和K-means算法的点聚类算法,提高了点聚类的准确度,同时结合DBSCAN和K-means两种聚类算法,实现了两阶段边线绑定算法来解决密集型线数据可视化。实验表明这两种方法在时间复杂度没有显著提高的情况下提高了可视化的准确度,能够很好的解决密集型地理数据可视化的问题。
关键词: 计算机应用技术 密集型地理数据可视化 聚类
For information in English, please click here
Clustering-based Dense Geographic Data Visualization Improvement
Abstract:With the development of Internet of Things technology and the continuous popularization of mobile devices and sensor devices, nowadays more and more means of obtaining geographic data are now in the age of geographic big data. Geographic data visualization can display the data in the form of a map to help users analyze the data in an intuitive and effective way. At present, many researchers have conducted an in-depth study on the direction of geographic data visualization. However, when dealing with the problem of visualization of intensive geographic data, the accuracy of the proposed scheme is low and the time complexity is high, which causes large errors for subsequent analysis of users. In order to solve the problem of intensive point data visualization, this paper designs and implements a point clustering algorithm based on grid partition and K-means algorithm to improve the accuracy of point clustering. Combining DBSCAN and K-means two clustering algorithms, Implemented a two-phase edge-bound algorithm to solve intensive line data visualization. Experiments show that these two methods can improve the visualization accuracy without significantly increasing the time complexity and can solve the problem of visualization of intensive geographic data.
Keywords: Computer Application Technology Dense Geodata Visualization Clustering
基金:
引用
No.****
同行评议
共计0人参与
勘误表
基于聚类的密集型地理数据可视化改进
评论
全部评论0/1000