AI绘画技术在桥梁效果图生成中的应用探索  

Exploration of the Application of AI Painting Technology in the Generation of Bridge Rendering

作　　者：杨林浩 YANG Linhao(China Railway First Survey and Design Institute Group Co.Ltd.,Xi’an Shaanxi 710000,China)

机构地区：[1]中铁第一勘察设计院集团有限公司,陕西西安710000

出　　处：《铁道建筑技术》2023年第9期24-27,共4页Railway Construction Technology

基　　金：中铁第一勘察设计院集团有限公司重点专项研发项目(2021KY73ZD(ZDZX)-01)。

摘　　要：本研究旨在响应《关于加快场景创新以人工智能高水平应用促进经济高质量发展的指导意见》的呼吁,通过利用人工智能以实现更高效、更经济的桥梁效果图生成过程。以Stable Diffusion模型为基础,利用人工智能(AI)技术创新桥梁效果图的生成。本文的方法涉及生成初步的桥梁效果图,借助Realistic模型进行细化,并用ControlNet插件进行优化。实验结果表明,相较于传统的图形生成方法,本文的模型在效率和质量上均实现了显著提升。这证明了将最新的人工智能(AI)技术与桥梁工程知识相结合,可以形成一个高效的桥梁效果图生成工具。总的来说,本研究展示了人工智能(AI)在桥梁工程设计领域的创新应用,为该领域的发展打开了新的可能性和机遇。This study aims to respond to the call of the“Guidance on Accelerating Scenario Innovation to Promote High-Quality Economic Development with High-Level Application of Artificial Intelligence”,using artificial intelligence to achieve a more efficient and economical bridge rendering process.We base our method on the Stable Diffusion model,utilizing AI technology to innovate the generation of bridge renderings.Our approach involves generating initial bridge renderings,refining them with the Realistic model,and optimizing them with the ControlNet plugin.Experimental results show that compared to traditional graphics generation methods,our model has achieved significant improvements in both efficiency and quality.This demonstrates that the integration of the latest AI technology and bridge engineering knowledge can form an efficient bridge rendering tool.Overall,this study showcases the innovative application of AI in the field of bridge engineering design,opening up new possibilities and opportunities for its development.

关 键 词：AI绘画技术 人工智能 桥梁效果图 Stable Diffusion CONTROLNET 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] U442[自动化与计算机技术—控制科学与工程]