Immersive virtual environment 기반 주거지역 경관 만족도 영향 요인 분석

초록

This study uses Immersive Virtual Reality(VR) equipment to analyze residential landscapefactors that affect streetscape satisfaction. Using Google Street View (GSV) images of residentialroads, this study conducted semantic segmentation for GSV images using deep-learningtechniques. After that, the selected GSV image was projected on the immersive virtual realitydisplay and evaluated by experiment participants. Finally, the relationships between thesubjective and objective indicators of the residential landscape image were verified throughstatistical analysis. The analysis results are as follows. First, the qualitative indicators such ascomfort, harmony, naturalness, beauty, uniqueness, cleanliness, and safety show significantassociations with the overall landscape satisfaction. Second, the quantitative indicators such asgreen area, sky, and building show significant positive effects on the overall residentiallandscape satisfaction. Particularly, the greenness, and openness in the residential areas aremost likely associated with ventilation and air quality, solar radiation, and positive landscapeperception. This finding indicates that greenness and openness are the important elements of residential landscape satisfaction. Finally, this study also identifies determinants factors thataffect the level of satisfaction in terms of naturalness, beauty, cleanliness, and safety andsuggests policy implications to improve residential landscape satisfaction. This study contributesto urban landscape evaluation using the emerging technology of immersive virtual reality and suggests the application of virtual reality technology in urban planning and design.

키워드