Semi-Automatic Construction of Virtual Reality Environment for Highway Work Zone Training using Open-Source Tools

VRW(2023)

引用 0|浏览19
暂无评分
摘要
In this paper, we present a computer vision-based approach to assist in creating a virtual reality (VR) environment for training state transportation agency staff that inspect highway work zones. Current VR work zone environments are manually created by developers based on drive-through videos or work zone plans, which require a considerable amount of time. We leveraged two open-source tools, YOLO and Road Architect in Unity, to improve the efficiency of environment creation. The YOLO open-source tool processes the video input and creates an output file that contains work zone features (i.e., roadway geometrics and traffic signs) and their locations. The virtual environment in Unity was created by Road Architect using its user-friendly interface. Our results showed that the proposed method was able to significantly reduce the development time for creating training modules. The semi-automatic method can be applied to other training applications where VR-environment creation is time consuming.
更多
查看译文
关键词
Computing methodologies,Artificial intelligence,Computer vision,Computer vision problems,Object detection,Applied computing,Operations research,Transportation
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要