Precise Processing of Point Cloud Data in Omni-Directional Scanning Based on Three-Dimensional Laser Sensor

摘要

In this paper, three-dimensional point cloud map is established after physical scanning and data acquisition using three-dimensional laser sensors. Compared to traditional two-dimensional, the three-dimensional map is more clear and visualized and performs better in scanning and acquiring point cloud data. Therefore three-dimensional map has been extensively applied in various fields. Three-dimensional laser scanning technology is used for establishing high-precision three-dimensional models as it can acquire three-dimensional point cloud data on the surface of the detected objects rapidly, efficiently and accurately without touching the detected objects. Taking domestic service robots as an example and to establish three-dimensional point cloud map with higher preciseness, noisy points were removed from the acquired data manually or using system, and the de-noised data were processed by filtration using plane fitting and surface fitting. The comparison of the two results suggested that surface fitting could more effectively simplify data and precise processing of point cloud data not only improved the preciseness of model and reduced the complexity but also accelerated the speed of data processing. This work is of far-reaching significance to the development of domestic robot technology.