Parametric design of ecological park landscape with DLA-GIS and GANs-GIS based on remote sensing data

Artificial intelligence is gradually becoming an intelligent technical support for geographic research and decision-making services in ecological park landscapes. To overcome the insufficiencies in scientific and systematic aspects of traditional ecological park design methods, the study combines the remote sensing image fusion technology, diffusion-limited cohesive model, generative adversarial network algorithm and geographic information system to parametrically design the landscape of an ecological park in Sichuan. The experimental results show that the research method can provide a valuable reference for landscape planning and management of an ecological park in Sichuan through parametric design. Compared with other eco-park landscape parametric design methods, the research method has the best performance. Under this method, the average travelling time from the park entrance to any attraction or facility is only 18 minutes, and the average travelling distance is controlled within 0.8 km. These data not only reflect the high efficiency of the road network design, but also significantly improve the convenience and satisfaction of visitors. In summary, the research method can provide an innovative and intelligent solution for landscape planning of ecological parks through accurate data analysis and visual simulation, thus providing more intuitive and scientific decision support for decision makers.