基于透视式增强现实的BIM建筑三维重建仿真

为了提高三维建筑模型的精准度,需要深入研究BIM建筑三维重建方法。当前方法耗时较长,得到的三维建筑模型与实际建筑之间的误差较大,存在效率低和精准度低的问题。将透视式增强现实技术应用到BIM建筑三维重建中,提出基于透视式增强现实的BIM建筑三维重建方法,通过BIM构建初始三维建筑模型,采用直接线性变换算法计算摄像机的内部参数和外部参数,完成摄像机标定。在摄像机标定结果的基础上采用LK光流计算方法得到像素在图像中的光流,根据光流的方向阈值和光流的大小筛选图像中的光流,提取到图像的匹配点,基于初始三维建筑模型针对建筑图像匹配点构成空间三维点云,采用Delaunay方法对空间三维点云进行三角化处理,针对处理后的建筑图像通过贴纹理完成BIM建筑三维重建。仿真结果表明,所提方法的效率高、精准度高。

3D Reconstruction Simulation of BIM Building Based on Perspective Augmented Reality

In order to improve the accuracy of 3D building model,it is necessary to research the 3 D reconstruction method of BIM building in depth.In this article,a new method of three-dimensional reconstruction for BIM buildings based on perspective augmented reality was proposed.Firstly,the initial three-dimensional building model was built by BIM,and then internal and external parameters of camera were calculated by direct linear transformation algorithm,so as to complete the camera calibration.On the basis of the result of camera calibration,the optical flow of pixel in image was obtained by LK optical flow calculation method.According to the direction threshold of optical flow and the size of optical flow,the optical flow in image was filtered and the matching points of image were extracted.Based on the initial three-dimensional building model,the matching points of building image were used to construct the spatial three-dimensional point cloud.Moreover,Delaunay method was used to triangulate three-dimensional point cloud.Finally,BIM building three-dimensional reconstruction was completed by texturizing the processed building images.Simulation results show that the proposed method has high efficiency and accuracy.