土地利用数据综合结果的质量评价

发布时间：2018-05-09 10:12

本文选题：土地利用数据 + 地图综合　；参考：《武汉大学》2013年博士论文

【摘要】：以质取胜,以精立业。质量是空间数据的生命,质量问题产生于数据的采集、加工与分析应用等多个环节,旨在对空间数据进行简化处理的地图综合操作是空间数据质量问题产生的重要来源之一,与数据采集、结构化组织、参考系变换等一般性数据处理相比,地图综合对数据质量产生的影响要大得多。合理的地图综合要在空间数据质量损失和质量维护中求得平衡。针对地图综合结果的数据质量评价研究具有重要意义,该问题的研究与通用的空间数据质量评价相比有其独特性,表现在质量评价的对象、参考基准、质量问题的来源与评价模型等方面。地图综合数据对象包括以地形图为目标的通用空间数据,也包括顾及语义特征表达的专题数据,土地利用数据便是专题性空间数据的一种。与通用地图综合关注空间特征不同,土地利用数据综合的质量评价更多的关注专题属性特征,涉及土地利用类型、行政级别、土地实体权属等。目前针对地图综合空间特征质量评价的研究已有不少成果,然而对专题属性特征的质量评价,尚缺乏系统、深入的技术策略和方法研究。随着全国第二次土地调查工作的完成,国土资源和城镇化管理各业务对多尺度土地利用数据的需求日益增长,需要通过地图综合建立不同层次的数据库,如何评价不同尺度的土地利用数据库的可靠性,成为该领域亟需解决的问题。本文以土地利用数据综合的质量评价为选题,针对尺度变换的特点,在通用空间数据质量评价的基础上拓展研究土地利用数据综合产生的质量问题。从地图综合的多因子约束出发,对土地利用数据综合结果的数据质量进行深入分析,设计了质量评价的模型、算法,从量化的角度在微观(单个对象)、中观(要素群组)和宏观(整个地图)三个空间层次给出评价方法。主要研究内容包括： (1)在深入分析土地利用综合结果质量问题的来源、影响因素的基础上提出了空间层次与约束相结合的质量评价指标二维体系。土地利用数据综合结果的质量问题主要体现为与综合前数据在空间位置、拓扑、语义、土地分类结构等方面的不一致及面积失衡问题,本文由质量问题出发,基于综合约束,对土地利用数据综合结果的三种几何形态给出了质量评价指标以及其在约束和空间层次下的二维体系。 (2)针对土地利用数据综合结果中的面要素,将三种空间层次下的综合评价指标从图形、统计、语义方面进行了归纳,并针对每种指标给出了对应空间层次下的量化评价方法。 (3)研究了综合前后面积平衡性保持和土地利用类型空间分布特征保持的评价方法。考虑到土地利用数据与区域特征的极大相关性,本文改进了景观指数分析方法,设计了基于空间关联关系的土地分类权重体系支撑下的面积平衡性评价方法；考虑到土地利用分类所具有的空间自相关特性,本文引入探索性空间数据分析的方法,基于Moran's I指数进行综合前后数据各种地类的空间自相关分析,实现了土地分类空间分布特征保持的量化评价。 (4)归纳了土地利用数据综合中语义质量的四个方面：精度、一致性、完整性和相似性,设计了基于语义隶属度变化的语义精度评价模型、基于面积变化的语义一致性评价模型、基于要素数量变化的语义完整性评价模型和基于信息熵的语义相似性评价模型。考虑到土地利用数据在语义上的层次化,设计了基于语义层次和有序量的混合隶属度计算模型,给出了土地利用现状分类体系的语义隶属度矩阵,实现了三个空间层次下语义质量的量化评价。 (5)设计了土地利用数据综合结果质量评价概念框架,将评价过程分为三步：要素特征的描述、评价指标的描述、评价结果的整合。构建了三个空间层次下的质量评价模型,并给出了顾及尺度与区域特征的评价指标权重矩阵。使用两个区域的第二次土地调查的多尺度综合成果(基于DOMAP综合软件生成)进行了验证,试验证明本文提出的顾及空间分布模式与空间格局的统计评价方法能够真实的反映数据在综合后对土地分类空间分布特征的保持度,本文提出的基于混合隶属度计算模型的语义质量评价方法能够揭示“综合尺度跨度越大,综合结果语义精度损失越多”的特征。本文的研究将通用空间数据质量问题拓展到专业性的数据加工处理中,针对土地利用数据的地图综合与尺度变换的特殊性从多个层次、多个指标进行了研究,提出了三个空间层次下的质量评价指标、评价模型和算法。本研究结合实际应用需求,基于第二次土地资源调查多尺度数据库成果进行了评价指标选择和权重设置,给出了量化的评价方法并进行了试验验证,具有很强的实际意义。对语义质量评价的研究和顾及空间分布模式与空间格局的统计评价方法在理论上是对已有空间数据质量评价研究的补充。
[Abstract]:Quality is the life of spatial data, quality is the life of spatial data. Quality problems arise from data collection, processing and analysis and application. The integrated operation of maps, which aims at simplifying spatial data, is one of the important sources of spatial data quality problems, such as data collection, structured organization, reference system transformation and so on. Compared with the data processing, map synthesis has much greater impact on the quality of data. Reasonable map synthesis should be balanced in the quality loss and quality maintenance of spatial data. The study of data quality evaluation for comprehensive map results is of great significance, and the research of this problem is compared with the general spatial data quality evaluation. The uniqueness lies in the object of quality evaluation, the reference standard, the source of quality problems and the evaluation model.
The map comprehensive data object includes the general spatial data which aims at the topographic map, including the thematic data, which is expressed by the semantic features. The land use data is one of the thematic spatial data. And land use type, administrative level, land entity ownership and so on. At present, there have been many achievements in the study of map comprehensive spatial characteristics quality evaluation. However, the quality evaluation of the characteristics of thematic attributes is still lack of systematic, in-depth technical strategies and methods. With the completion of the second land survey work in China, land and resources and cities and towns The demand for multi scale land use data is growing increasingly. It is necessary to establish different levels of database through map synthesis. How to evaluate the reliability of land use database of different scales has become an urgent problem to be solved in this field.
Based on the characteristics of the scale transformation, the quality of land use data is developed on the basis of the characteristics of the scale transformation. Based on the multi factor constraints of map synthesis, the data quality of the comprehensive results of land use is analyzed, and the data quality of the comprehensive results of land use is analyzed. The quality evaluation model, algorithm, from the quantitative point of view in the micro (individual object), meso (element group) and macro (whole map) three spatial levels of evaluation methods. The main research content includes:
(1) on the basis of in-depth analysis of the source and influence factors of the quality problems of comprehensive land use results, the quality evaluation index system of spatial hierarchy and constraints is proposed. The quality problems of the comprehensive results of land use data are mainly reflected in the spatial position, topology, semantics and land classification structure of the comprehensive data of the land use data. The problem of inconsistency and area imbalance is based on the quality problem. Based on the comprehensive constraints, the quality evaluation index and the two-dimensional system under the constraints and the spatial levels are given for the three geometric forms of the comprehensive results of the land use data.
(2) in view of the surface elements in the comprehensive results of land use data, the comprehensive evaluation indexes under three kinds of spatial levels are summarized from the aspects of graphics, statistics and semantics, and the quantitative evaluation methods under the corresponding spatial levels are given for each index.
(3) the evaluation method of area balance retention and spatial distribution characteristics of land use types before and after comprehensive is studied. Considering the great correlation between land use data and regional characteristics, this paper improves the method of landscape index analysis, and designs an area balance evaluation under the support of soil classification weight system based on Spatial correlation. Price method; taking into account the spatial autocorrelation characteristics of land use classification, this paper introduces the method of exploratory spatial data analysis, and based on the Moran's I index to carry out spatial autocorrelation analysis of various types of data, and realizes the quantitative evaluation of the spatial distribution characteristics of the land classification.
(4) generalizes the four aspects of semantic quality in land use data synthesis: accuracy, consistency, integrity and similarity. The semantic accuracy evaluation model based on the change of semantic membership degree is designed, semantic consistency evaluation model based on area change, semantic integrity evaluation model based on the change of element quantity and information entropy based language The semantic similarity evaluation model. Considering the semantic hierarchy of land use data, a mixed membership degree calculation model based on the semantic level and the ordered quantity is designed, and the semantic membership matrix of the land use status classification system is given, and the quantitative evaluation of the semantic quality under three spatial levels is realized.
(5) the conceptual framework of the comprehensive results quality evaluation of land use data is designed. The evaluation process is divided into three steps: the description of the characteristics of the factors, the description of the evaluation indexes, the integration of the evaluation results. The quality evaluation model under three spatial levels is constructed, and the weight matrix of the evaluation indexes, taking into account the scale and the regional characteristics, is given. Two areas are used. The multi scale comprehensive results of the second land survey (based on the DOMAP integrated software generation) are verified. The experiment proves that the statistical evaluation method, which is proposed in this paper, can reflect the retention of the spatial distribution characteristics of the land classification, which is based on the mixture of spatial distribution patterns and spatial patterns. The semantic quality assessment method of membership computing model can reveal the characteristics of "the larger the comprehensive scale span and the more the semantic precision loss of the comprehensive result".
The research of this paper extends the problem of general spatial data quality to the professional data processing. In view of the particularity of map synthesis and scale transformation of land use data from multiple levels and several indexes, the quality evaluation index, evaluation model and algorithm under three spatial levels are put forward. According to the demand, the evaluation index selection and weight setting are carried out based on the multi scale database results of the second land resources survey. The quantitative evaluation method is given and the experimental verification is carried out. The research on the semantic quality evaluation and the statistical evaluation method of spatial distribution pattern and spatial pattern are in theory. It is a supplement to the quality evaluation of existing spatial data.

【学位授予单位】：武汉大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：P208;F301.2

【参考文献】