一种基于视觉扫描和改进空间波法进行室内电磁环境快速建模的新方法

《“十三五”国家信息化规划》前瞻性地提出要“推进电磁频谱管理专项工程建设，构建军民协同合作的电磁频谱监测、检测和探测网系”。频谱管理的重点之一是建立全面准确的电磁环境数据库。传统的电磁数据采集主要依靠人工测试，难以克服成本高、周期长和覆盖面小的缺点。利用建模和仿真分析电磁环境分布，是高效、全面进行电磁环境管理、频谱优化和基站设计的有效手段。而电磁仿真的难点就是对极端复杂的室内电磁特性进行建模。本研究拟引入视觉扫描方法，全自动准确重建室内三维环境；并对空间波法进行改进，解决室内环境中因多径、边缘绕射、透射、以及墙角等引入的误差，使其能够准确测量室内现场材料电磁参数；然后根据坐标推演算法自动将材料与结构贴合，快速得到具有材料电磁参数的精确室内三维模型；研究同时包括实验平台和测试方案设计，并完成验证工作。本项目提出了室内场景自动化精确建模的新方法，实现了大规模标准化建模，提高了精度和效率。

“The 13th Five-year plan for national information technology” proactively points out that it is necessary to enhance the electromagnetic spectrum management and construct the networks of electromagnetic spectrum monitoring, testing and detection by civil-military cooperation. The key point of electromagnetic spectrum management is to build the database of the electromagnetic environment comprehensively and accurately. The traditional method for the collection of the electromagnetic data mainly depends on the manual test with the high cost, long cycle and limited coverage. It is efficient to fulfill the electromagnetic environment management, spectrum optimization and telecommunication base station design by the electromagnetic environment modeling and simulation. However, it is very difficult to model the extremely complex indoor electromagnetic characteristics in the electromagnetic simulation. In this project, the visual scanning method is to be applied to reconstruct the 3D scene automatically with high accuracy in the meantime. In addition, the free space wave method is planned to be improved to deal with the problems caused by the multipath, the edge diffraction and transmission, as well as the corner of the wall in the real indoor scene. The improved free space wave method can accurately measure the electromagnetic parameters of the indoor material. Finally, the indoor scene is to be constructed by the integration of the 3D geometry model with electromagnetic parameters. The research includes the experiment design and the validation of the proposed method. The proposed method provides a new way to construct the indoor scene for large-scale modeling and significantly improves the efficiency.