Seoul street-view database for urban environment research

Le résumé fourni par la source

Street-view imagery provides valuable insights into the physical and social dimensions of urban environments. Seoul, South Korea, known for its rapid urban transformation, presents a distinctive blend of traditional urban centers and newly developed districts. To gain a comprehensive understanding of Seoul’s urban development, we constructed a street-view database spanning 14 years (2010–2023), comprising 423,353 images. To enhance data quality, noise components within the images were removed using advanced object detection and image inpainting techniques. We then segmented objects relevant to the commercial ecosystem and extracted features related to the intensity, shape, and texture for each object. The dataset includes street-view images, object masks, and their associated feature sets, all of which are publicly accessible to the research community. This resource offers a robust foundation for examining dynamic changes in Seoul’s urban environment. By leveraging this dataset, researchers can investigate the unique characteristics of Seoul’s evolving urban landscape and explore their implications for the commercial area.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.