虚拟现实环境下破损古建筑场景复原仿真分析

Simulation Analysis of Scene Restoration of Damaged Ancient Building in Virtual Reality Environment

为获取破损古建筑的完整结构,修复古建筑原貌,提出一种虚拟现实环境下破损古建筑场景复原方法。利用扫描、控制和电源装置组成激光扫描系统,通过发射与反射光线的时间差计算扫描点的具体位置,采集古建筑点云数据;使用双边滤波法去除点云噪声,建立三角网格,对网格做八叉树结构划分,缩减点云数据,构建法矢量分割线,分割点云数据,提取点云特征;利用虚拟现实技术图像设计、三维建模等功能,结合点云数据建立尺度变换矩阵,渲染图像轮廓与纹理;根据古建筑的颜色特征,生成破损与完整部分的尺度分割线,通过特征匹配实现场景复原。实验结果表明,所提方法复原出的场景清晰度和完整度更高,结构尺寸与实际尺寸的误差较小。

In order to obtain the complete structure of damaged ancient buildings and restore their original appearance,a restoration method for damaged ancient building scenes in virtual reality environment was proposed.Firstly,scanners,control devices and power supplies were used to construct a laser scanning system.Then,the specific position of the scanning point was calculated by the time difference between the emission and reflection of light.Meanwhile,the point cloud data of the ancient building was collected.Moreover,the bilateral filtering method was used to remove the noise of the point cloud and construct a triangular mesh.Furthermore,the mesh was divided by octree structure,thus reducing the point cloud data.After that,a normal vector segmentation line was built to segment the point cloud data,thus extracting the point cloud features.Finally,the functions of image design and three-dimensional modeling of virtual reality technology were utilized to establish a scale transformation matrix combined with the point cloud data,and thus to render image contours and textures.According to the color characteristics of ancient buildings,the scale segmentation lines of damaged and integral parts were generated,and the scene restoration was realized by the feature matching.Experimental results show that the restored scene has higher clarity and completeness,and the structural size has a smaller error compared to the actual size.