Under rapid urbanization, rural planning in China is often driven by functional efficiency, contributing to ecological fragmentation and landscape homogenization. This paper proposes a modular digital planning framework coupling quantified shanshui (Chinese landscape aesthetics) parameters with GIS-based ecological analytics. Traditional landscape paintings are treated as visual proxies for cultural spatial cognition. Computer vision extracts spatial-organizational principles, formalized as an aesthetic-spatial mapping matrix translating compositional rules into computable ecological constraints and spatial weights. The framework guides the generation of ecological, productive, residential, and cultural modules. A case study in Wufeng Village demonstrates improved ecological performance, landscape-visual quality, and service accessibility relative to a baseline scenario, indicating a reproducible pathway for embedding cultural principles as generative rules in performance-driven planning.