Title: A dual-system framework for rural spatial gentrification assessment using street view imagery
ABSTRACT: Rural spatial gentrification has emerged as a critical phenomenon fundamentally reshaping rural landscapes through material transformations and socio-cultural reconfigurations driven by urban capital expansion and middle-class consumption preferences, yet existing methodologies persistently struggle to accurately identify its dynamics due to limitations in tracking longitudinal community evolution, resolving spatial heterogeneity in visual indicators, and adapting to China's collective land ownership context. Addressing these gaps, this study pioneers a neurocognitively-grounded dual-system framework that innovatively integrates Kahneman's dualprocess theory with deep learning models to analyze street view imagery, developing the Rural Spatial Gentrification Index (RSGI) which quantifies spatial gentrification intensity through complementary computational pathways: System 1 rapidly detects localized urbanizing elements invasion like sorting rubbish bins and cranes while System 2 deliberately segments infrastructural shifts in asphalt roads and sidewalks. When applied across four Guangdong villages, the framework's robustness is rigorously validated through statistical testing and resident interviews that confirm its capacity to reconcile perceptual judgments with spatial diagnostics. This approach significantly advances neurocognitive geography methodologies by enabling fine-grained monitoring of rural-urban transitions through spatially explicit analytics that inform targeted policy interventions balancing revitalization with displacement mitigation, though limitations in narrow-alley data coverage and multimodal integration present valuable avenues for future research expansion.
Keywords: Rural gentrification; Dual-system framework; Street view imagery; Deep learning;
DOI: https://doi.org/10.1016/j.habitatint.2026.103851
