Un análisis comparativo de los algoritmos Fast Radial Symmetry Transform y Hough Transform para la detección automática de granos de café en imágenes

 

Authors
León-Sarkis, Marco
Format
MasterThesis
Status
publishedVersion
Description

Proyecto de Graduación (Maestría en Ingeniería en Computación) Instituto Tecnológico de Costa Rica, Escuela de Ingeniería en Computación, 2017.
In this work we present a strategy that contributes to the overall solution of a problem presented by the Costa Rica Co ee Institute (ICAFE). ICAFE owns a set of coffee grains images and needs to nd an automatic way, through computer vision, to detect and count the number of grains in each image in order to increase the e ciency in the process of estimating yield. A strategy to detect co ee grains in images is proposed, by combining the algorithms Fast Radial Symmetry Transform[8] and Hough Transform[19]. Then, this strategy is incorporated in the grain detection process of P-TRAP[13], an open-source tool, to increase the precision in the detection of existing coffee grains. The images are taken with a mobile device in a non-controlled environment in which the grains are not pulled o their natural environment. Likewise, a comparative analisis is done between the P-TRAP version developed in this study and both algorithms running individually. The number of existing grains in an image is determined manually. Then, the cherry detection process is executed over each image and results are collected. Finally, a detailed analisis is done over the results obtained.
Instituto del Café de Costa Rica (ICAFE)

Publication Year
2017
Language
Español
Topic
Café
Eficiencia
Producción
Estrategia
Imágenes
Dispositivos
Research Subject Categories::TECHNOLOGY::Information technology::Computer science::Computer science
Fuente
RepositorioTEC
Get full text
http://hdl.handle.net/2238/9372
Derechos
openAccess
Licencia