Building Façade Color Distribution, Color Harmony and Diversity in Relation to Street Functions: Using Street View Images and Deep Learning

Building façade colors play an important role in influencing urban imageability, attraction and citizens’ experience. However, the relations between street functions building color distribution, harmony diversity have not been thoroughly examined. We obtained dominant of façades Changning District, Shanghai, utilizing Baidu view images, image semantic segmentation technology K-means algorithm. The variations façades’ colors, across different types functional streets were examined through logistic regression ANOVA analyses. results indicate that, compared to industrial streets, red hues are more common science education residential mixed streets. Business likely green, red–purple. Residential streets’ saturation is overall higher than that In business medium–high value occurs less frequently other Moreover, we found harmonious diversified This study has implications for planning practices. Color should be well considered future planning. also addressed planning, improve existing frameworks as related strategies.