Design of Green Plant Distribution Pattern in Landscape Planning Based on Optimal Grid Scale Estimation Method

This paper presents an innovative approach for the optimization of greenery distribution patterns in urban landscape planning, leveraging the combined capabilities of Generative Adversarial Networks (GANs) and Residual Neural Networks (ResNet). Acknowledging the crucial role of urban greenery in enhancing environmental quality and aesthetic appeal, the methodology begins with the collection of diverse greenery distribution maps through advanced data acquisition techniques, such as satellite imagery, drone photography, and Geographic Information Systems (GIS). At the core of the approach is the utilization of GANs to generate high-resolution images of greenery distribution, establishing a base for detailed simulation and predictive analysis. Following this, the ResNet framework processes these images to facilitate the automatic extraction of essential features for greenery pattern analysis. This dual-approach culminates in the precise estimation of the optimal grid scale for greenery distribution, demonstrating substantial improvements over traditional methods across various performance metrics. The experimental results indicate a significant advancement in the field, particularly in providing detailed insights and accurate predictions for urban landscape planning.