The influence of surface topology on the quality of the point cloud data acquired with laser line scanning probe

摘要

Purpose - The purpose of this paper is to investigate the influence of surface topology on the performance of laser line scanning probe and to suggest methodology for 3D digitization of specular surfaces. Design/methodology/approach - Two different molds, one having milled surface and the other with polished surface, were used to identify effect of surface characteristics on the performance of laser line scanning probe mounted on bridge-type coordinate measuring machine. The point cloud data acquisition of two surfaces was carried out using different combinations of laser scanning parameters. The point cloud sets thus obtained were analyzed in terms of completeness, noise and accuracy. The polished mold which exhibited specular reflection was digitized at different scanning angles of laser line scanning probe using the best combination of scanning parameters. Findings - Results confirmed that surface characteristics play important role to determine quality of the reverse engineering (RE) process. The results in terms of completeness, accuracy and noise for point cloud sets have successfully been obtained for milled and polished surfaces. Three-dimensional (3D) comparison analysis suggested larger deviation in cases of polished surface as compared to milled surface. The point cloud set acquired with proposed approach was better in terms of both completeness and noise reduction. Originality/value - There has been an increased demand for measurement of metallic, polished and shiny surfaces in automotive, aerospace and medical industries. These surfaces are very difficult to scan because they exhibit specular reflection instead of diffuse reflection. Laser line scanning probe which is a non-contact method is in great demand for RE. This is due to the fact that it possesses very high data acquisition speed. However,laser scanning is hugely affected by surface characteristics which in turn govern specular reflection. In this paper, it has been shown that a surface that exhibits various degrees of specular reflection can be digitized efficiently if appropriate combination of scanning parameters and positions of laser line scanning probe are used. Also, this paper has attempted to offer a procedure to overcome incompleteness and noise in 3D data as obtained by the laser line scanning probe.