Comparison of thermal environments and classification of physical environments using fisheye images with object-based classification

Abstract

Pedestrian spaces play important roles as passageways as well as promoting health and social interactions among urban dwellers. However, the thermal comfort in cities is deteriorating as the urban heat island is being exacerbated. By evaluating the characteristics of the urban thermal environments and applying the results to urban planning, one can learn how to improve thermal comfort. Here, I analyzed the thermal environmental characteristics that vary according to the physical environment type (Sky view factor (SVF), vegetation view factor (VVF), and artificial material view factor (AVF)) classified by fisheye images with object-based classification at 16 pedestrian spaces at the Changwon National University campus in South Korea. There were significant correlations between both SVF and VVF with all the thermal environment types. Moreover, Pearson's correlation coefficient (R) for SVF and mean radiant temperature (Tmrt) was 0.4361 (p = 0), and VVF was −0.541 (p = 0.001). Therefore, the higher the SVF and the lower the VVF, the lower the thermal comfort. Our results demonstrate the importance of creating adequate physical environments to improve thermal comfort. Moreover, we provided the quantitative evidence necessary for establishing policy for improving thermal comfort in pedestrian spaces during urban planning.