A new fault detection method of conveyer belt based on machine vision

A new fault detection and measurement method of conveyer belt based on machine vision is proposed. The conveyer belt used in coal mine transportation usually goes two kinds of faults: joint's elongation and local rust. Under this engineering background, the system focuses on detecting the state of conveyer belt and measuring the fault size. This paper brings forward a modified BP neural network to detect and classify different faults. The new BP algorithm's detecting speed is rapid, and the correct recognition rate of the joint and erosion has a great improvement. The measurements of joint's length and erosion's area are realized on the machine vision platform which built by LabVIEW IMAQ Vision module. And the measurements have a high accuracy. The results demonstrate that the new method is effective and efficiency.