A land evaluation method based on multi-granularity computing

The invention discloses a land evaluation method based on multi-granularity computing, On the basis of acquiring the vector unit of the plot, using multi-source spatio-temporal data of natural resources and socio-economy to expand the plot attributes, Structured representation, reduction and rule extraction can realize the mapping transformation from multi-dimensional attribute value to thematic decision value, and then draw the plot-level thematic conjecture map to provide more precise and accurate land planting evaluation information for the government, enterprises and other departments. Thekey techniques of this method are attribute extension, rule extraction and thematic conjecture mapping. The invention can realize the thematic mapping of land suitability evaluation at land parcel level with high accuracy. 本发明公布了种基于多粒度计算的土地评价方法，在获取地块矢量单元基础上，利用自然资源和社会经济的多源时空数据进行地块属性的扩展、结构化表示、约简以及规则提取等步骤，实现从地块"多维属性值"到"专题决策值"的映射转化，进而绘制形成地块级专题推测图，为政府、企业等部门提供更为精细和精准的土地种植评价信息。本方法主要的关键技术为"属性扩展"、"规则提取"以及"专题推测制图"等。本发明可实现较高精度的地块级土