资源科学 ›› 2017, Vol. 39 ›› Issue (3): 408-417.doi: 10.18402/resci.2017.03.03

• • 上一篇    下一篇

基于特征匹配的土地利用数据语义邻近度计算

贾小斌1(), 艾廷华2, 彭子凤3   

  1. 1. 中电科新型智慧城市研究院有限公司,深圳 518026
    2. 武汉大学资源与环境科学学院,武汉 430079
    3. 深圳市规划国土房产信息中心,深圳 518034
  • 收稿日期:2016-03-16 修回日期:2016-04-30 出版日期:2017-03-20 发布日期:2017-03-20
  • 作者简介:

    作者简介:贾小斌,男,河南南阳人,博士,研究方向为土地资源数据整理。E-mail:jiaxiaobin_123@126.com

  • 基金资助:
    数字制图与国土信息应用工程国家测绘地理信息局重点实验室开放基金(GCWD201403)

Computation of semantic proximity in land-use data based on feature-matching

Xiaobin JIA1(), Tinghua AI2, Zifeng PENG3   

  1. 1. The Smart City Research Institute of China Electronics Technology Group Corporation,Shenzhen 518026,China
    2. School of Resource and Environmental Sciences,Wuhan University,Wuhan 430079,China
    3. Shenzhen Municipal Planning and Land Real Estate Information Center,Shenzhen 518034,China
  • Received:2016-03-16 Revised:2016-04-30 Online:2017-03-20 Published:2017-03-20

摘要:

定量化语义关系的判断是地理数据处理的关键,也是地理本体建模与表达的难点,土地利用数据作为典型的专题性地理数据,其语义特征上的邻近关系如何,是土地资源统计、整理、规划、管理和开发需要关注的问题。针对土地利用数据的语义特征,从《土地利用现状分类》的依据出发,结合其权属特征,建立了土地利用数据语义特征的多细节层次表达模型;随后应用特征匹配的方法,通过计算其在权属、覆盖成因、覆盖类型、用途、利用状态、植被类型、附着物性质、利用方式等方面的匹配关系得到土地利用数据语义邻近度的具体度量值,实现土地利用数据从语义建模到邻近度计算的整个过程。在试验中则以具有不同语义特征的土地利用数据实体为例进行语义邻近关系计算,并将试验结果与实际经验判断进行比较,证明该模型具有较强的实用性,计算结果符合人类的认知。

关键词: 土地利用数据, 语义特征, 邻近度, 特征匹配

Abstract:

This research aimed to compute the semantic proximity in land-use data. First,this research established a model to express the layer of details of the semantic characteristics in land-use data,started from the basic situation of land-use category and combined with characteristics of ownership. Second,feature-matching was used to gain concrete metric values by computing the matching relationship among ownership,reason of coverage,land cover type,usage,using state,nature of the stem in vegetation,habits of growth in vegetation,usage of vegetation and land use patterns. This research achieves the whole process from semantic modeling to the semantic proximity in land-use data. Third,this experiment takes 38 entities in land-use data as an example to compute semantic proximity among those with the same characteristic in ownership but different characteristic in land use type,and compares the result of the experiment with the judgment of the practical experience to discover that the result of calculations conforms to human cognition and the methods used in calculating the semantic similarity have strong practicability. Finally,the computation results of semantic proximity were transformed into a spatial measurable value to divide the land-use patch into this have the adjacent relationship in spatial position based on the semantic proximity which the area of is smaller than the threshold when the scale is change smaller in the processing of land-use data in order to demonstrate the practical value of the research findings. The results of this research can be directly applied to statistics,processing,planning,management and development of land resources,and in computing semantic proximity in land-use data to calculate semantic adjacent relationships in other related spatial geographic data.

Key words: land-use data, semantic feature, proximity, feature matching