Recognizing motion images solid waste jumbled on a neural network with a simple tracking performed in an automatic and robotic recycling line

Abstract

In this paper, we show the vision system in recognition motion images and detection of solid urban waste (SUW) and their integration on an automatic robotic line with a simple tracking algorithm. The detection and image processing are able to detect, identify and calculate the position of the SUW and send the coordinates to a delta robot for selection. The image processing system is previously trained in a neural network. Delta robots are provided by ABB Corporation and have been programmed to select the SUW through a simple algorithm to tracking. We present the integration of these systems and we describe the automatic and robotic machine with the vision system.