| Abstract |
This paper focuses on exploring the application of virtual reality (VR) technology in the digital interaction of rural landscape, with the aim of building a VR art generation model of rural landscape, and showing the rural landscape in a digital and interactive way. By introducing graph neural network (GNN) for VR scene matching, and comparing with traditional convolutional neural network (CNN), it is found that GNN has obvious advantages in dealing with the complex spatial structure of rural landscape and can accurately capture the details of the landscape. User evaluation shows that digital rural landscape has been highly praised in terms of interactivity, content presentation and knowledge learning, showing a broad application prospect. This study provides a new technical path for digital interaction of rural landscape and injects vitality into VR art innovation. Looking forward to the future, the research work will continue to optimize the algorithm, broaden the application fields, help the construction of smart villages/smart cities, and promote cultural inheritance and development. © 2025 The Authors. |
