Quantifying the Color of Signboards in Historic District Based on Machine Learning and Street View Images:A Case Study of Suzhou
The color of signboards is an important component of urban landscape,especially in the historic district where the landscape is primarily controlled.Using machine learning technology and street view image data,this paper constructs a method for evaluating the color of signboards in the historic district of Suzhou based on three categories:"traditional buildings-controlled buildings-modern buildings".The method identifies the dominant colors of signboards of different types of buildings by street and evaluates the color coordination of signboards on each street.By screening the streets with higher and lower coordination,the color characteristics of their signboards are analyzed and the basis for quantitative color control of the signboards in the historic district is proposed.The method can provide technical support for the refined renewal of signboards in historic districts.
street view datasignboard colorcolor evaluationhistoric districtmachine learning
NETL
NSTL
万方数据
