摘要
以深度学习为代表的人工智能技术为制图综合智能化水平的提升创造了有力的条件。然而,智能学习模型的应用依赖高质量的样本数据,目前还缺乏适宜制图综合知识学习的案例数据集。以不同比例尺居民地要素综合为例,深入探讨服务于智能化制图综合的样本数据采集建库方法。主要工作包括:依据地图数据组织特点和制图综合过程决策需求,提出了区分特征型案例和变换型案例的样本数据组织体系以及相关的数据结构设计;基于通用GIS平台设计了样本数据采集与管理系统,形成对样本数据采集、存储、修正、更新等全链条的管理维护框架。本研究是针对图形表达的地图要素如何构建案例数据集的有益探索,对推动以制图综合为代表的复杂地图空间知识学习具有积极意义。
Artificial intelligence technology represented by deep learning has created powerful conditions for the improvement of the intelligence level of cartographic generalization. However, the application of intelligent learning model depends on high quality sample data, and there is still a lack of case data set suitable for cartographic generalization knowledge learning. Taking the integration of residential feature at different scales as an example, the method of sample data collection and library construction for intelligent cartographic generalization are discussed in depth in this paper. The sample data organization system of distinguishing characteristic cases and transformation cases and the design of relevant data structure are proposed, according to the characteristics of map data organization and the decision-making requirements of cartographic generalization process. Based on the general GIS platform, a sample data collection and management system is designed to form a management and maintenance framework for the whole chain of sample data collection, storage, revision and update. It is a useful exploration how to construct case data sets based on graphical map elements, which is of positive significance for promoting the learning of complex map spatial knowledge represented by cartographic generalization.
作者
焦洋洋
刘平芝
蒋萌
杨敏
艾廷华
巩现勇
JIAO Yangyang;LIU Pingzhi;JIANG Meng;YANG Min;AI Tinghua;GONG Xianyong(Information Engineering University,Zhengzhou 450001,China;State Key Laboratory of Geo-information Engineering,Xi’an 710054,China;Xi’an Research Institute of Surveying and Mapping,Xi’an 710054,China;National Engineering Research Center for Geographic Information System,China University of Geosciences(Wuhan),Wuhan 430074,China;School of Resources and Environment Science,Wuhan University,Wuhan 430072,China)
出处
《测绘科学技术学报》
CSCD
北大核心
2021年第4期430-434,共5页
Journal of Geomatics Science and Technology
基金
国家自然科学基金项目(42071450,41801396,41571442)
国家973计划资助项目(613317)。
关键词
制图综合
机器学习
居民地
样本分类
样本库设计
cartographic generalization
machine learning
resident
sample classification
sample library design
作者简介
焦洋洋(1989-),男,河南修武人,助理研究员,博士生,研究方向为自动制图综合、空间数据更新等。E-mail:johnpanther@163.com。
