THE ONLINE WALKABILITY EVALUATION FOR STREET VIDEOS CLASSIFIED BY IMAGE RECOGNITION

Abstract

For walkable city development, it becomes increasingly important to understand diverse needs of street users in order to develop multi-functional streets. By integrating image recognition and online questionnaires for walkability evaluation, we can make an evaluation for a variety of people and streets in a systematic way. The aim of this study is to construct an online evaluation system for domestic and international streets with their 360-degree videos classified by image recognition, and to analyze its applicability to walkability evaluation. First, the indicators of spatial elements in the street videos were measured by image recognition, and were classified by cluster analysis. An online questionnaire with the videos was constructed to analyze the impacts of street design on walkability evaluation, taking account of social profiles. As a result, the indicators of street design measured by image recognition accounted for the walkability evaluation along with the social profiles.