This study focuses on the application of Stable Diffusion-based generative AI tools in environmental art design education. Through literature review, case studies, and process-based design methods, it explores the application value and optimization space of its Text-to-Image and Image-to-Image functionalities. Results show that the tool can provide instant visual feedback, expand the boundaries of design possibilities, and promote human-AI collaborative innovation, bringing technological tool innovation to environmental art design education and offering new paths for cultivating design talents.