Walk Score is widely used to measure the convenience of surrounding amenities and generally considers the physical characteristics of streets (density of crossings, distances, etc.). This article proposes an improved Walk Score evaluation method considering what pedestrians perceive through hearing and sight in the street environment. By using a traffic noise assessment model and a semantic segmentation model, the improved method incorporates traffic noise and vegetation coverage factors as penalties into the original Walk Score. Based on the proposed method, a case study was conducted in Nanjing, China. Compared with the original Walk Score: (1) the improved Walk Score results present more information about how the street environment affects walkability; (2) areas near downtown generally scored higher, while some traditional downtown areas scored lower, which is mainly because the latter areas have little vegetation along the streets; and (3) areas around urban viaducts and overpasses scored lower due to the additional penalty of traffic noise. With the support of the improved Walk Score evaluation method, urban planning and policymakers can develop more comprehensive solutions to improve urban walkability.

No abstract