基于多源空间数据融合的城市建成环境识别研究

Research on urban built up environment identification based on multi source spatial data fusion

精准识别建成环境能够为城市空间结构与城市交通互动关系研究提供科学依据,为合理配置城市空间资源提供数据基础。研究了一种基于语义识别的深度学习模型,在融合兴趣点与兴趣面基本分析单元的基础上,提取兴趣面内兴趣点的高维语义特征向量,通过聚类分析识别得到功能用地类型。结果表明该深度学习模型识别精度的准确率达到80%,模型可以为建成环境识别提供新的方法思路。

Accurately identifying the built environment can provide scientific basis for the study of the interaction between urban spatial structure and urban transportation,and provide a data foundation for the rational allocation of urban spatial resources.Based on this,a deep learning model based on semantic recognition is investigated On the basis of fusing the basic analysis units of interest points and interest surfaces,high-dimensional semantic feature vectors of interest points within interest surfaces are extracted,and functional land types are identified through clustering analysis.The results show that the recognition accuracy of the deep learning model reaches 80%,and the model can provide new methodological ideas for built environment recognition..