摘要
随着遗迹景观的疾速消逝与破败,系统科学地开展保护与修复研究工作尤为重要。该研究提出基于人工智能方法,以遗址景观卫星遥感影像为基础研究数据,使用对抗生成式神经网络(GAN)以特征学习和特征复原的方式重新修复遗迹景观损坏、缺失段落,为遗迹景观空间布局的研究与后续保护工作提供全新方位解读及理论支持。结果表明,利用计算机深度学习特定数据学习解析能力,可有效学习遗迹景观空间特征数据逻辑关系并进行遗迹景观修复。
With the rapid disappearance and deterioration of monument landscapes,it is particularly important to carry out systematic and scientific research on conservation and restoration.In this study,we propose an artificial intelligence approach based on the remote sensing images of the relic landscape satellite as the basic research data,and use an adversarial generative neural network(GAN)to repair the damaged and missing sections of the relic landscape by feature learning and feature restoration,so as to provide new orientation interpretation and theoretical support for the study of the spatial layout of the relic landscape and subsequent conservation work.The results show that the spatial feature data and logical relationships can be effectively learned and restored in the monument landscape by using the deep learning specific data learning and parsing ability of computer.
作者
王磊
李鹏波
Wang Lei;Li Pengbo
出处
《华中建筑》
2022年第1期42-45,共4页
Huazhong Architecture
基金
国家重点研发计划“乡村生态景观营造模式研究”(编号:2019YFD1100402)
天津艺术科学规划项目(编号:E20007)。
关键词
人工智能
深度学习
文化景观
遗迹景观修复
Artificial intelligence
Deep learning
Cultural landscape
Relic landscape restoration
