Improvement in accuracy for three-dimensional sensor (faro photon 120 scanner)

Abstract

The ability to provide actual information and attractive presentation, three-dimensional (3D) information has been widely used for many purposes especially for documentation, management and analysis. As a non-contact 3D sensor, terrestrial laser scanners (TLSs) have the capability to provide dense of 3D data (point clouds) with speed and accuracy. However, similar to other optical and electronic sensors, data obtained from TLSs can be impaired by errors coming from different sources. In order to ensure the high quality of the data, a calibration routine is crucial for TLSs to make it suitable for accurate 3D applications (e.g. industrial measurement, reverse engineering and monitoring). There are two calibration approaches available: 1) component, and 2) system calibration. Due to the requirement of special laboratories and tools to perform component calibration, the task cannot be carried out by most TLSs users. In contrast, system calibration only requires a room with appropriate targets. Through self-calibration, this study involved a system calibration for Faro Photon 120 scanner in a laboratory with dimensions of 15.5m x 9m x 3m and 138 well-distributed planar targets. Four calibration parameters were derived from well-known error sources of geodetic instruments. Data obtained using seven scan stations were processed, and statistical analysis (e.g. t-test) shows that all error models, the constant error (8.9mm), the collimation axis error (-4.3”), the trunnion axis error (-11.6”) and the vertical circle index error (8.0”) were significant for the calibrated 3D sensor.