An Improved Low-Noise Processing Methodology Combined with PCL for Industry Inspection Based on Laser Line Scanner

Abstract

This paper introduces a three-dimensional (3D) point cloud data obtained method based on a laser line scanner and data processing technology via a PCL open project. This paper also provides a systematical analysis of the error types of laser line scanner and common error reducing solutions and calibration of the laser line scanner. The laser line scanner is combined with a precision motorized stage to obtain the 3D information of a measurand, and the format of point cloud data is converted via the set of x, y, and z coordinates. The original signal is processed according to the noise signal types of the raw point cloud data. This paper introduced a denoise process step by step combining various segmentation methods and a more optimized three-dimensional data model is obtained. A novel method for industry inspection based on the numerous point cloud for the dimensions evaluation via feature extraction and the deviation of complex surface between scanned point cloud and designed point cloud via registration algorithm is proposed. Measurement results demonstrate the good performance of the proposed methods. An obtained point cloud precision of ±10 μm is achieved, and the precision of dimension evaluation is less than ±40 μm. The results shown in the research demonstrated that the proposed method allows a higher precision and relative efficiency in measurement of dimensions and deviation of complex surfaces in industrial inspection.