一种基于SOM的多维农业时空数据可视化分析方法
发布时间：2018-12-11 23:42
【摘要】：随着"3S"、物联网等技术逐渐被应用于农业生产过程中,多维农业时空数据大量产生。如何挖掘和分析多维时空数据在时空上的内在价值,对推动现代农业发展有重要意义。针对多维农业时空数据存在维数高,数据关系复杂,难以直接挖掘分析的问题,提出一种基于SOM的多维农业时空数据可视化分析方法。该方法利用SOM算法对多维数据降维,结合数据表、平行坐标系、时空立方体等可视化组件,实现数据的多角度显示和分析,达到深层次挖掘的目的。基于该方法,设计完成可视化分析平台,并以美国农产品数据为例,讨论该方法的可行性。
[Abstract]:With "3s", "Internet of things" and other technologies gradually applied in the agricultural production process, multidimensional agricultural space-time data are produced in large quantities. How to excavate and analyze the intrinsic value of multidimensional spatiotemporal data in time and space is of great significance to promote the development of modern agriculture. In order to solve the problem that multidimensional agricultural spatio-temporal data has high dimension and complex data relationship, it is difficult to mine and analyze directly. A visualization analysis method of multidimensional agricultural spatio-temporal data based on SOM is proposed. This method uses SOM algorithm to reduce the dimension of multidimensional data, and combines the visualization components such as data table, parallel coordinate system, space-time cube and so on, to realize the multi-angle display and analysis of data, so as to achieve the purpose of deep-level mining. Based on this method, the visual analysis platform is designed and the feasibility of this method is discussed by taking the data of American agricultural products as an example.
【作者单位】： 浙江农林大学信息工程学院;浙江省林业智能监测与信息技术研究重点实验室;
【基金】：浙江省自然科学基金(编号:Y5110145) 浙江农林大学科研发展基金(编号:2014FR086);浙江农林大学智慧农林业研究中心项目(编号:2013ZHNL01) 浙江省林业智能监测与信息技术研究重点实验室项目
【分类号】：S126
本文编号：2373449
[Abstract]:With "3s", "Internet of things" and other technologies gradually applied in the agricultural production process, multidimensional agricultural space-time data are produced in large quantities. How to excavate and analyze the intrinsic value of multidimensional spatiotemporal data in time and space is of great significance to promote the development of modern agriculture. In order to solve the problem that multidimensional agricultural spatio-temporal data has high dimension and complex data relationship, it is difficult to mine and analyze directly. A visualization analysis method of multidimensional agricultural spatio-temporal data based on SOM is proposed. This method uses SOM algorithm to reduce the dimension of multidimensional data, and combines the visualization components such as data table, parallel coordinate system, space-time cube and so on, to realize the multi-angle display and analysis of data, so as to achieve the purpose of deep-level mining. Based on this method, the visual analysis platform is designed and the feasibility of this method is discussed by taking the data of American agricultural products as an example.
【作者单位】： 浙江农林大学信息工程学院;浙江省林业智能监测与信息技术研究重点实验室;
【基金】：浙江省自然科学基金(编号:Y5110145) 浙江农林大学科研发展基金(编号:2014FR086);浙江农林大学智慧农林业研究中心项目(编号:2013ZHNL01) 浙江省林业智能监测与信息技术研究重点实验室项目
【分类号】：S126
【相似文献】
相关期刊论文 前2条
1 周清,周斌,张杨珠,王人潮;水稻土SOM含量高光谱模型的母质差异性研究[J];科技通报;2004年06期
2 刘真真;;探讨园艺植物可视化技术的应用[J];现代园艺;2013年16期
相关硕士学位论文 前1条
1 卢艳松;基于SOM的农业时空数据可视化分析模式[D];浙江农林大学;2015年
，本文编号：2373449
本文链接：https://www.wllwen.com/kejilunwen/nykj/2373449.html
