Omni-directional laser radar automatic acquisition system for complex industrial scenes : Taking the pose measurement of hydraulic support as an example

Abstract The dynamic measurement of lidar products in complex and narrow scenes faces many challenges, such as limited space and complex parameter calibration. Thus, this paper develops an omni-directional LiDAR automatic acquisition system for complex industrial environments. The omni-directional lidar automatic acquisition system is first designed. The calibration cost is then reduced using the method of real scene calibration system parameters, which allows to maintain the system error around 2.6 mm. Afterwards, the SIFT feature description algorithm is combined with an enhanced SVDD algorithm to process point cloud data, allowing to increase the operational efficiency and real-time performance of the system. Finally, underground hydraulic support is considered as a case study, and experimental measurements are performed to determine the elongation of the side guard plate. The obtained results show that the proposed omni-directional laser radar automatic acquisition system has high reliability, yielding an absolute error less than 3.8%. This demonstrates that the proposed novel method can accurately measure targets in narrow scenes, and thus it provides an efficient solution for the pose measurement of underground equipment.