基于街景图像的泉州历史街区街道色彩评价OA

To address the lack of quantitative methods and refined management difficulties for historic street colors,this study constructs a multi-parameter integrated evaluation model.Taking the historic district of Quanzhou as a case study,street view images and semantic segmentation are employed to analyze streetscape features,followed by analysis to quantify the spatial attributes that influence pedes-trians'color evaluations.Results show that the overall street color palette in Quanzhou's historic district is dominated by dark green,warm khaki,and withered-tea hues,accounting for 54.4%,29.5%,and 16.2%,respectively.Street functional type and built-environment spatial characteristics exhibit strong associations with street color composition.Compared to the commercially-dominated West Street,East Street,which is dominated by residential functions,has a 3.0%higher average color harmony score,but a 20.9%lower average color richness score.Furthermore,East Street exhibits greater volatility in color richness,with a 42.8%higher standard deviation.Streetscape environmental elements significantly affect color perception;specifically,sky view factor and green view index are significantly negatively correlated with color richness.