基于约束关系的石化装置三维重建方法研究

Research of 3D Reconstruction for Petrochemical Plants Based on Constrained

针对现有的三维建模算法存在耗时长,局部结合细节建模效果差等问题,提出了一种基于石化设备模型库的装置三维重建算法。利用计算机视觉"从运动恢复结构"技术初步实现石化装置部件的识别与定位,提取关键帧中局部特征显著的设备信息,实现基准部件识别及位置匹配。提取复杂管道的空间布局信息,利用阀门(组)的识别信息及约束关系来校正已重建的三维场景,最终实现整个石化装置的自动化三维重建。实例证明,采用文中算法进行石化装置三维重建,可以有效地减少建模时间,提高建模精度。

Aimed at the problems of long time - consu- ming and poorly - integrated local detail modeling, the existing three -dimensional modeling algorithms are proposed. A three -dimensional reconstructinn algorithm hased on petrochemical equipment model library is proposed. Using computer vision ＂ from the structure of the motion recovery ＂ technology to initially identify and locate petrochemical plant components, extract key equipment in the local characteristics of signifieant equipment information to achieve the benchmark components recognition and location matching. The spatial layout information of complex pipeline is extracted, and the reconstrueted 3D scene is corrected by using the identification information and constraint relation of valve （Group）. Finally, the automatic 3D reconstruction of the whole petrochemical plant is realized. The example proves that adopting the algorithm in this paper to carry out three -dimensional reconstruction of petrochemical plant can effectively reduce the modeling time and improve the modeling accuracy.