一种室内空间结构约束的热力图生成方法  

A Heatmap Generation Method with Constraints of Indoor Spatial Structures

作　　者：张蓝天 杨剑 王光霞[1] 张凌梅 ZHANG Lantian;YANG Jian;WANG Guangxia;ZHANG Lingmei(Information Engineering University, Zhengzhou 450001, China)

机构地区：[1]信息工程大学,河南郑州450001

出　　处：《测绘科学技术学报》2018年第5期533-539,共7页Journal of Geomatics Science and Technology

基　　金：国家重点研发计划项目(2017YFB0503500)

摘　　要：在绘制室内移动数据热力图时,传统方法会得到热区穿越空间障碍(如墙)的结果,这直接影响了室内人群密度估计的准确性和可靠性。通过分析室内空间结构特征,抽象出两类空间结构模型,以此改进传统核密度分析算法,提出一种可根据室内结构约束自适应调整带宽的热力图生成方法。以北京某大型商场的室内移动数据为例进行算法的实验验证。It becomes a critical problem when applying heatmap generation to indoor mobile data, in which the generated hotspot often cross spatial obstacles such as walls. This makes the estimated population density less accurate and reliable. Firstly, the characteristics of indoor spatial structure are analyzed and are categorized into two types of spatial structure models in the paper, which leads to an improvement of traditional kernel density estimation. Secondly, a heatmap generation is proposed in which the bandwidth is adjusted adaptively according to the indoor spatial structure constraints. Finally, the performance of the method is verified on a real-world indoor mobile data obtained in a shopping mall in Beijing.

关 键 词：热力图 室内空间结构 核密度分析 带宽 室内移动数据 

分 类 号：P208[天文地球—地图制图学与地理信息工程]