Abstract: To resolve the problem of belt conveyor’s belt running deviation and distortion, we propose
an intelligent monitoring and warning method of belt conveyor based on visual computing. Firstly,
the video images of belt conveyor in coal mine are preprocessed and the Gaussian Filter is used to
establish the background model. Then, the background subtraction method is applied to get the target
image. Secondly, the Shi-Tomasi algorithm is used to detect the corners of the target image and these
corners are tracked by the method of Pyramid LK optical. Finally, these corners’ coordinates are
converted from the image coordinate system to camera coordinate system according to the principles
of imaging. And we can get the speed of these corners and judge whether the belt conveyor is in a
good condition or not according to the corners’ speed and then send out some warning information in
time when the running belt conveyor is abnormal. The experimental results show that the proposed
method has higher veracity and real time quality.

Key words: background subtraction method, corner detection, optical flow method, warning