High-precision pose measurement of hydraulic supports using LiDAR-binocular camera with cooperative target detection

Abstract During fully mechanized coal mining, hydraulic supports are prone to posture variations such as pitching, tilting, and twisting, which directly impact their support performance, coordination with shearers, and the overall operational stability of the working face. Existing posture detection methods often exhibit limitations in positioning accuracy, long-term stability, and hardware complexity. To address these challenges, this paper proposes an integrated LiDAR-binocular camera detection system. By quantifying the effects of target surface material, distance, angle, and environmental factors on LiDAR performance, a robust target recognition model is developed. Integrating this with binocular distortion correction, image segmentation, and circle detection algorithms, high-precision target localization and posture estimation are achieved. Finally, static and dynamic accuracy tests were conducted on a hydraulic support experimental platform using the MARS motion capture system. The results demonstrate that the proposed method effectively meets the requirements for hydraulic support posture detection in the complex environment of fully mechanized mining faces.