摘要
针对易受噪声数据、建筑物复杂程度干扰,高层建筑边缘空间规划模型的重建完整率、正确率低的问题,提出高层建筑边缘空间BIM透视式增强现实重建模型。采用MVS点云算法,采集高层建筑边缘空间的点云集合,剔除集合中的噪声,采取BIM技术,提取高层建筑边缘空间的轮廓特征,将轮廓特征分割成长方体与三角体,采用投影映射,匹配轮廓的三维角点坐标,引入纹理映射法,优化高层建筑边缘空间BIM透视式纹理,完成高层建筑边缘空间重建。实验结果表明,所提方法的重建效果好,重建完整率高、正确率高,重建几何精度较高。
Aiming at the problems of being easily disturbed by noise data and building complexity,and the low reconstruction integrity and accuracy of high-rise building edge space planning model,a BIM perspective augmented reality reconstruction model of high-rise building edge space is proposed.MVS point cloud algorithm is adopted to collect the point cloud set of high-rise building edge space,eliminate the noise in the set,and BIM Technology is adopted to extract the contour features of high-rise building edge space.The contour features are divided into cuboids and triangles.Projection mapping is adopted to match the three-dimensional corner coordinates of the contour,and texture mapping method is introduced to optimize the BIM perspective texture of high-rise building edge space to complete the reconstruction of high-rise building edge space.Experimental results show that the proposed method has good reconstruction effect,high reconstruction integrity rate,high accuracy and high geometric accuracy.
作者
莫民静
MO Minjing(School of Architectural Engineering,Guangxi Vocational&Technical Institute of Industry,Nanning 530001,China)
出处
《粉煤灰综合利用》
CAS
2023年第5期133-139,共7页
Fly Ash Comprehensive Utilization
基金
广西工业职业技术学院科学研究重点课题“EPC+BIM+装配式”新型建造模式的应用研究(桂工业院科研Y2021KY001)
2022年度广西高校中青年教师科研基础能力提升项目“EPC+BIM+装配式”新型建造模式的应用研究(2022KY1289)。
关键词
点云集合
统计滤波
位置判断函数
轮廓透视投影
相对深度
纹理映射
BP neural network model
statistical filtering
location judgment function
contour perspective projection
relative depth
texture mapping