최규진, 박초롱연구원이 2024 한국생태환경건축학회 추계학술발표대회에서 우수논문발표상을 수상했습니다.
논문제목: 서울시 가로환경의 시각적 특성 분석
Abstract:
This study analyzes the streetscape characteristics of Seoul’s commercial areas using Google street view panorama images,
reflecting the pedestrian’s perspective and road direction. Road directions were calculated using ArcGIS Pro to ensure that the images aligned with actual pedestrian’s field of view. A deep learning-based semantic segmentation model was applied to classify key streetscape elements. The study measured various streetscape characteristics, including DH ratio, openness,greenery, buildings, roads, traffic, and street furniture. The results demonstrated that considering road direction improved the accuracy of streetscape analysis, providing more precise insights for urban planning and commercial area management.
Keywords : Commercial area, Street environment Street view, Semantic segmentation
