Detection of foreign bodies in periodic motion scenes using shape- and size-adaptive descriptors

The detection of foreign bodies in a machining production line is important for guaranteeing its stability and safety. The traditional methods for such detection either rely on manual observation, which is a time-consuming and labor-intensive process. In this paper, we present a machine vision-based method for automatically detecting foreign bodies in a machining production line with periodic motion scenes. The results demonstrate that this method can achieve reasonable accuracy in the detection of foreign bodies in a machining production line with periodic motion scenes.