Aiming at the need of real-time monitoring of hob wear in the tunnel boring machine（ TBM）

an online monitoring method based on machine vision system is proposed. Without the need for complex and expensive instruments

cutter wear is monitored by real-time monitoring of large-size rock on the conveyor belt. At the same time

the watershed segmentation algorithm is improved and a method to eliminate invalid background segmentation is also proposed to make it suitable for the image segmentation of rock ballast. The experimental results show that this method can segment the rock quickly and accurately and measure the size of the rock. In general

the algorithm is stable and can meet the monitoring requirements.