Abstract

LOPEZ-CASTANO, Carlos Alberto; FERRIN, Carlos Diego  and  CASTILLO, Luis Fernando. Computer Vision applied to quality control in manufacturing processes: real-time tracking of refrigerators. Rev.EIA.Esc.Ing.Antioq [online]. 2018, vol.15, n.30, pp.57-71. ISSN 1794-1237.  https://doi.org/10.24050/reia.v15i30.1150.

In this paper a computer vision based system is conceived and implemented for quality assurance of inner accessories in fridge-freezer during assembly process. The computer vision system includes a stage of acquisition of multimodal images (visible, near infrared and depth) from four sensors, followed by a segmentation stage for the extraction of contours both doors and cabinet; these contours are tracked by means of a tracking technique that allows the system to focus its attention on a single refrigerator for individual tests; a final stage allows to estimate the regions of interest of the accessories in the refrigerator under evaluation by means of a homographic mapping technique and the aid of a set of previously registered accessories for each refrigerator model in a database. The computer vision system was tested on a distributed and high performance computing architecture for three different models in a real scenario. Performance and effectiveness of the developed system demonstrated suitability for real time applications.

Keywords : Quality Assurance; Fridge-Freezer; Computer Vision; Distributed System; High Performance Computing.

        · abstract in Spanish | Portuguese     · text in Spanish     · Spanish ( pdf )