Study on Landscape Architecture Design System Based on Big Data Algorithm

Incorporating information technology, including information systems, remote sensing, parameterization, and digital tools, enhances the planning and design process of urban landscapes. This integration not only boosts efficiency but also elevates the overall quality of the designs and accentuates their human-centered features. To further refine the quality and efficiency of landscape renderings, this study introduces a novel landscape modeling technique rooted in an advanced CNN powered by big data. Addressing the challenge of blurred boundaries in traditional landscape segmentation, we propose a new energy function approach. This method enriches contextual semantics for precise segmentation, optimizes consistency in extracted landscapes, and restores boundary details. Experimental results indicate that our improved CNN-based landscape drawing and optimization algorithm achieves a remarkable 95.95% accuracy, surpassing other methods by approximately 10%.