"Verificación dimensional de piezas metálicas de acero, mediante la comparación con una pieza patrón, aplicando un sistema de visión artificial para validar su manufactura"

Abstract

Abstract Arti cial vision deals with the real world description and properties, from one or multiple bidimensional images acquired. An Arti cial vision project is usually multidisciplinary. Project task execution involves acquisition, preprocessing, segmentation, description and interpretation, and requires the knowledge and use of various techniques, some of them are: Picture and optics Digital image processing Pattern recognition Graphic computing This work shows an application of arti cial vision through parts visual inspection, using Matlab software to develop graphic interface, image preprocessing and postprocessing. Image acquisition is performed through two camera models, which main objective is to compare and analyze pro and cons of using each model in the application developed. In this project , many characteristics are analyzed to be used in image processing, such as structure elements, line detection lters, morphological operators, etc., to obtain a correct segmentation and data interpretation that allow part boundaries de nition.