Rapid elimination of noise in 3D laser scanning point cloud data

When using a hand-held 3D laser scanner to collect target object data, due to factors such as personnel operation, collection environment and equipment itself, a large number of external noise points are often produced. This will seriously affect the processing and reconstruction accuracy of later point cloud data. According to the data analysis, these external noise points are divided into two categories: flying points and cluster points. Aiming at this phenomenon, a point cloud model noise removal algorithm combining statistical filtering and pass-through filtering is proposed. Firstly, the flying points are eliminated by statistical filtering, and then the remaining large area cluster points are removed by through filtering. The experimental results show that the algorithm can quickly and accurately identify external noise points and eliminate them completely.

I. Introduction

3D laser scanning technology is a technology that uses laser scanning to obtain 3D point cloud data on the surface of a target object. By preprocessing the acquired 3D point cloud data, the 3D model of the target object can be reconstructed^[1].This technology breaks the limitation of traditional human single point measurement. Surface scanning with high speed laser has the advantages of no contact, high efficiency, high resolution and high precision^[2] .It provides a reliable means to realize high precision modeling of 3D point cloud data.