Design and construction of a classifier of pieces using visual servoing

It is presented the design and construction of a pieces system classifier through visual control, in which six characteristics of the image are extracted and the four of major discriminative are selected. As a classifier it trains a neuronal artificial network that discriminates against the objects in 6 classes, the average percentage of recognitions was 89.5%. To the pieces positioner element was implemented by a PID controller based on the position of the element to classify regarding to the captured image to have full control of the wished location.