基于图像处理和BIM的预制钢筋混凝土构件钢筋骨架快速检查方法研究

Building industrialization is the tendency of construction industry, which contributes to solving the problem of low efficiency and quality of construction in China. Precast production of reinforced concrete components (precast production for short hereafter) is a crucial element in building industrialization. Moreover the assembling of reinforcing bars (rebars for short hereafter) is the most important and complex step during precast production, while its quality significantly affects the security of the whole structure. Currently the quality check of the reinforced skeleton frame deeply relys on the manual counting, measuring and referring to design drawings by managers in the plants, which is time consuming and error prone..Based on image processing and BIM (Building Information Modeling) technologies, this research proposes a quick method for the quality check of reinforced skeleton frame of precast reinforced concrete components and develops a set of software and device based on the method. Firstly, the shadow removal and the other preprocessing algorithms for raw images of reinforced skeleton frame are proposed. Then, the algorithm for obtaining the 3D models of the reinforced skeleton frame is established based on the proposed 3D reconstruction method, in which the sheltered rebars in the raw images can be repaired. Third, the comparing algorithm of the established 3D models and as designed BIM model for quality check is proposed based on the relevant rules in China. Finally, a set of corresponding software and device is developed and validated. It can be concluded that the efficiency and accuracy of the quality check of reinforced skeleton frame will be dramatically improved by using the method and consequentially the quality management of precast production can be enhanced.

建筑工业化有助于解决建筑工程效率低下、质量低等问题，是我国建筑业发展方向。钢筋混凝土构件预制生产是其重要组成部分，而钢筋骨架的绑扎又是其中最为重要且复杂的工序，其制作质量直接影响到建筑安全。当前钢筋骨架的检查依赖于现场管理人员的人工计数、尺量及图纸对比，效率低且易出错。.本研究基于图像处理及BIM（Building Information Modelling，建筑信息模型）技术，研究预制钢筋混凝土构件的钢筋骨架快速检查方法，并研制检查装置及处理软件。首先提出针对采集的原始钢筋骨架图像数据的阴影去除等预处理算法。其次基于三维重建技术，研究建立钢筋骨架三维模型，并对其中被遮挡的钢筋进行修复的算法。然后基于我国相关质检规范，研究钢筋骨架三维模型与BIM设计模型对比算法。最后研制相应软件及硬件设备并进行验证。该方法可大幅提高钢筋骨架检查效率和精确度，从而提升预制钢筋混凝土构件质量管理水平。

本研究的目标是，基于图像处理及BIM（Building Information Modelling，建筑信息模型）技术，研究预制钢筋混凝土构件的钢筋骨架快速检查方法，并研制检查装置及处理软件。.围绕该目标，本研究按计划进行以下工作：调研并选择高精度传感器以及建立拍照装置；研究钢筋骨架点云精确配准算法；研究钢筋骨架整体点云获取算法；研究钢筋骨架语义设计点云自动生成算法；研究基于BIM模型的钢筋骨架质量自动检查算法；研制相应的软件。.本研究具有创新性，因为国内还没有类似的研究，国外相同领域的研究要么研究对象是简单的钢筋骨架，要么检查内容仅包含钢筋间距检查。.本研究共发表论文2篇，其中JCR1区SCI论文1篇（IF=4.632），国际会议论文1篇。还有4篇论文正在准备过程中。.国际交流方面：接待美国马里兰大学教授学术交流2次，接待德国慕尼黑工业大学教授学术交流1次。参加国际会议1次，并在分会场做报告，得到同行的认可。.人才培养方面：有1名博士生的博士论文是基于本研究进行的；另有2名博士生和5名本科生参与了本研究。.总体上，本研究达到了预期目的。结题后，本研究将继续进行。