Abstract:Visual perception plays a critical role in regulating emotions, yet limited attention has been paid to older adults in rural or peri-urban contexts. This study investigated the emotional responses of elderly individuals to outdoor environments in four peri-urban villages in Huadu District, Guangzhou. Real-scene images were collected and analyzed using data annotation, semantic segmentation, and deep learning to extract seven visual environmental features. Spatial autocorrelation and geographically weighted regression were applied to explore spatial clustering and identify key influencing factors. Results show that positive emotion responses are strongly associated with green spaces and waterfronts, while negative emotions cluster in industrial areas, along vehicle roads, and in barren lands. Greenness, rural uniqueness, and facility richness positively correlate with positive emotions, whereas enclosure, road accessibility, and traffic congestion show negative associations. Spatial openness of both roadside had contrasting effects in roads and waterfront areas. These findings provide valuable insights for the design and renewal of age-friendly public spaces in rural settings.
