<p>Urban parks play a critical role in enhancing ecological well-being and psychological restoration in rapidly urbanizing regions. However, accurately evaluating their restorative benefits remains challenging due to limitations in spatial representation and environmental perception. This study proposes an integrated framework combining Virtual Reality (VR) simulation with GIS-supported spatial analysis to assess the environmental restorative effects of urban parks. This study first integrates two types of spatial data sources-remote sensing raster data of parks in the study area and vector data of land use boundaries-on the basis of the geographic information system platform. It then completes the raster quantification and zoning preprocessing of three spatial factors: vegetation coverage, landscape pattern, and environmental suitability. Subsequently, a multi-scale image enhancement and restoration model is constructed to optimize the image quality of the virtual reality scene of the park and to virtually repair damaged landscapes. Finally, the refined landscape information restored by virtual reality is substituted back into the geographic information system layers to calculate the spatial differentiation of the recreational benefits of various types of parks. Empirical results based on multi-scenario urban park datasets show that the proposed framework significantly improves environmental representation accuracy, with vegetation coverage reaching 95.24% and spatial adaptability exceeding 93%. User-based evaluations further confirm enhanced immersion and improved cognitive perception of environmental restoration. The findings demonstrate that integrating VR with GIS-based spatial analysis provides a robust approach for evaluating urban park restorative benefits, offering valuable support for urban environmental planning, landscape management, and evidence-based policy design.</p>