基于Hough变换的带式输送机托辊图像分割方法

由于现阶段对煤矿带式输送机托辊缺乏有效的监测，传统的人工巡检方法存在效率低、自动化程度低等问题，提出一种矿用带式输送机托辊图像分割算法。该算法利用带式输送机沿线巡检装置采集带式输送机图像，对托辊区域图像进行边缘检测，通过霍夫变换和直线拟合方法，确定托辊的4个边界。利用托辊边界制作图像掩膜分割出托辊图像区域。实验表明，该算法对不同位置托辊能够实现准确的自动分割，为利用视觉检测技术监测托辊运动状态提供可靠的技术基础。

Image segmentation method of belt conveyor roller based on Hough transformation

Due to the lack of effective monitoring of the rollers of coal mine belt conveyors at the present stage，the traditional manual inspection method has the problems of low efficiency and low degree of automation.This paper proposed an image segmentation algorithm for the rollers of mine belt conveyors.The algorithm used a belt conveyor inspection device to collect belt conveyor images.The edge of the roller area image was detected，and the four boundaries of the roller were determined by Hough transformation and straight line fitting method.Use the border of the roller to create an image mask to divide the image area of the roller.Finally，experiments showed that the algorithm can achieve accurate automatic segmentation of the rollers at different positions，which provided a reliable technical basis for the use of visual detection technology to monitor the movement status of the rollers.