The incorporation of sophisticated artificial intelligence technologies and data mining methodologies is critical for the planning and design of low-carbon landscape architectural environments. To enhance the three-dimensional reconstruction capabilities and visual esthetics of hierarchical planning and design within low-carbon landscape architectural environments, this study introduces a novel methodology predicated on virtual reality (VR) technology. This approach harnesses the diverse pictorial structure feature distribution inherent to low-carbon landscape architectural environments, amalgamated with VR's feature structure reorganization and enhancement techniques. Furthermore, we propose methodologies for joint measurement and multi-level integration to refine the hierarchical planning and design process within these environments. Simulation outcomes demonstrate the efficacy of this method in achieving superior three-dimensional reconstruction and visual representation, thereby augmenting the planning and design quality within the domain of low-carbon landscape architecture.