三维激光扫描点云去噪及在建模中的应用

发布时间：2019-01-22 16:53

【摘要】：三维激光扫描技术在获取空间数据方面的应用日益广泛。用三维激光扫描获取点云数据时,由于观测条件、仪器设备自身及外界环境条件的影响,一般都会扫入一些噪声点。在三维建模时,必须对点云数据进行去噪处理,以提高三维模型的精度。本文在研究现有点云数据噪声分类和噪声处理方法的不足之处的基础上,提出一种基于重心的点云数据去噪算法。该算法是在基于包围盒的点云数据压缩算法的基础上改进的,其技术流程清晰,算法可靠,便于matlab编程实现。同时,为了对比说明该算法的去噪效果,开发了一个增加点云随机噪声的matlab程序。选取两个地物进行实地三维激光扫描。对扫描的点云数据做过拼接等预处理后,用本文程序增加一定量的随机噪声,利用本文算法和程序对增加随机噪声的点云数据进行处理。试验结果表明,基于重心的点云去噪算法在去除随机噪声方面有明显效果,程序运行正确。应用本文的去噪算法和Geomagic Studio软件自带的去噪方法分别对点云数据进行去噪处理并建模。试验表明,两种去噪方法获得的模型精度相当。经过去噪后的建模精度均优于不进行去噪处理的精度,这表明该算法应用于建模方面具有类似于专业软件的有效性,有一定实用价值。
[Abstract]:Three-dimensional laser scanning technology in the acquisition of spatial data is increasingly widely used. In order to obtain point cloud data by 3D laser scanning, some noise points are usually swept in due to the influence of observation conditions, instruments themselves and external environmental conditions. In order to improve the accuracy of the 3D model, the point cloud data must be de-noised. On the basis of studying the shortcomings of existing noise classification and noise processing methods of point cloud data, a barycenter based point cloud data denoising algorithm is proposed in this paper. The algorithm is improved on the basis of point cloud data compression algorithm based on bounding box. Its technical flow is clear, the algorithm is reliable and easy to realize by matlab programming. At the same time, in order to compare the denoising effect of the algorithm, a matlab program is developed to increase the random noise of the point cloud. Two objects were selected for 3D laser scanning. After preprocessing the scanned point cloud data, a certain amount of random noise is added by the program in this paper, and the point cloud data with random noise is processed by the algorithm and program in this paper. The experimental results show that the point cloud denoising algorithm based on barycenter has obvious effect in removing random noise and the program is running correctly. The point cloud data are de-noised and modeled by using the denoising algorithm of this paper and the de-noising method of Geomagic Studio software. The experimental results show that the model accuracy obtained by the two denoising methods is equivalent. The modeling accuracy after denoising is better than that without denoising, which shows that the algorithm is similar to the professional software in modeling and has some practical value.
【学位授予单位】：西安科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P225.2

【参考文献】