基于SteamVR激光扫描的点云图像界面交互设计

Interactive design of point cloud image interface based on SteamVR laser scanning

三维空间信息是现今测量领域关键数据之一,主要获取手段为激光扫描技术。激光扫描目标获得的点云数据体量较大,要想将其转换为二维点云图像数据应用,需要较高超的界面交互功能,提出基于Steam-VR激光扫描的点云图像界面交互设计方法研究。选取SteamVR激光扫描设备,获取目标点云数据,采用拉普拉斯滤波算法去除点云数据中的噪声,通过主成分分析法提取点云数据的特征点与特征线,构建相应的点云图像模型,利用四叉树空间检索模式检索、调度点云数据,结合曲面细分技术与Instancing技术渲染点云图像,从而实现了点云图像界面的交互功能。实验数据显示:在不同实验工况下,提出方法点云数据检索效率最大值为96%,点云图像绘制质量因子最大值为9.4,点云图像渲染时延最小值为17 s,充分证实了提出方法界面交互效果更佳。

Three dimensional spatial information is one of the key data in the field of measurement,and the main acquisition method is laser scanning technology.The volume of point cloud data obtained by laser scanning target is large.In order to convert it into two-dimensional point cloud image data application,it needs superb interface interaction function.The research on point cloud image interface interaction design method based on SteamVR laser scanning is proposed.Select the SteamVR laser scanning equipment to obtain the target point cloud data,remove the noise in the point cloud data by using the Laplace filtering algorithm,extract the feature points and feature lines of the point cloud data by the principal component analysis method,build the corresponding point cloud image model,retrieve and schedule the point cloud data by using the QuadTree spatial search mode,and render the point cloud image by combining the surface subdivision technology and instancing technology.Thus,the interactive function of point cloud image interface is realized.The experimental data show that under different experimental conditions,the maximum point cloud data retrieval efficiency of the proposed method is 96%,the maximum point cloud image rendering quality factor is 9.4,and the minimum point cloud image rendering delay is 17 s,which fully proves that the interface interaction effect of the proposed method is better.