Real-time classification of coffee fruits using FPGA

The goal in this work was to design a circuit that could classify objects by color in real-time that can be used for quality improvement. A circuit that performs color analysis of an image and, according to that analysis, classifies the object was designed. A histogram of the Spherical Coordinate Transform of the image is computed and compared to histogram patterns to make a classification decision. The circuit was tested on the classification of coffee fruits in four maturity stages: immature, under-mature, mature and over-mature. The results showed that it is possible to build a system for color object classification that works in real-time and that can be affordable and portable. The designed circuit is implemented on a Field Programmable Gate Array (FPGA), acquires video at 64 frames per second, classifies the coffee fruits at a rate of 25 fruits per second and achieved an average efficacy of 75.7%

• Autores: Montes Castrillón, Nubia Liliana
• Tipo de recurso: Doctoral thesis
• Fecha de publicación: 2015
• Institución: Universidad Nacional de Colombia
• Repositorio: Universidad Nacional de Colombia
• Idioma: spa
• OAI Identifier: oai:repositorio.unal.edu.co:unal/55273
• Acceso en línea: https://repositorio.unal.edu.co/handle/unal/55273; http://bdigital.unal.edu.co/50617/
• Palabra clave: 0 Generalidades / Computer science, information and general works; 51 Matemáticas / Mathematics; 62 Ingeniería y operaciones afines / Engineering; Real-time; Image processing - digital techniques; Color analysis; Coffee grain - classification; Tiempo real; Procesamiento digital de imágenes; Análisis de color; Granos de café - clasificación
• Rights: openAccess
• License: Atribución-NoComercial 4.0 Internacional