Foreign matter detection using a 3D convolutional neural network

Abstract

In this study, we propose a method for identifying foreign objects in liquid in a reagent bottle using 3D Convolutional Neural Networks. The video of a reagent bottle containing liquid is divided not only horizontally and vertically but also in the direction of the time axis, and the divided small regions are trained by 3D Convolutional Neural Networks to identify foreign objects by class classification.