基于视觉测量的液压支架位姿检测方法

Position detection method of hydraulic support based on vision measurement

针对目前综采工作面液压支架位姿测量方法存在实施成本高、精度和实时性较差的问题,提出了一种基于视觉测量的液压支架位姿检测方法。通过在液压支架上设置红外LED标识板,利用安装在采煤机上的防爆相机进行图像采集,基于4个共面特征点的视觉算法对图像进行处理与解算,得到液压支架底座的位姿,根据液压支架的位姿坐标可以得到其在多个方向的直线度。实验结果表明,该方法测量的液压支架位姿检测精度在0.7°以内,直线度精度在20mm以内,满足综采工作面液压支架的检测精度要求。

In view of problems of high cost, poor accuracy and real-time performance in current position measurement methods of hydraulic support used in fully mechanized coal mining face, a position detection method of hydraulic support based on vision measurement was proposed. The infrared LED identification plate is set on the hydraulic support, the explosion-proof camera installed on the shearer is used for image acquisition, and the image is processed and calculated based on the visual algorithm of four coplanar feature points to obtain the position and pose of the hydraulic support base, and the straightness in multiple directions can be obtained according to the position and pose coordinates of the hydraulic support. The experimental results show that the method to measure the detection accuracy of the hydraulic support position is within 0.7°, and the straightness accuracy is less than 20 mm, which meets requirements of accuracy detection of the hydraulic support of fully mechanized coal mining face.