Use this link to cite:
http://hdl.handle.net/2183/24873 Estimación automatizada del peso y calibre de aceitunas mediante análisis de imagen
Loading...
Identifiers
Publication date
Authors
Ponce Real, Juan Manuel
Aquino, Arturo
Segura, Francisca
Millán Prior, Borja
Andújar-Márquez, José Manuel
Advisors
Other responsabilities
Journal Title
Bibliographic citation
Ponce Real, J.M., Aquino, A., Segura, F., Millán, B., Andújar Márquez, J.M. Detección de la orientación mediante visión artificial para el control de equilibrio en robots humanoides. En Actas de las XXXIX Jornadas de Automática, Badajoz, 5-7 de Septiembre de 2018 (pp.958-966). DOI capítulo: https://doi.org/10.17979/spudc.9788497497565.0958 DOI libro: https://doi.org/10.17979/spudc.9788497497565
Type of academic work
Academic degree
Abstract
[Resumen] El calibrado y selección de productos agrícolas es una actividad de gran relevancia dentro de la industria agroalimentaria. Este estudio, centrado en el sector del olivo, presenta una solución basada en análisis de imagen que permite la estimación automática y no invasiva del peso y calibre (ejes de simetría mayor y menor) de un conjunto de aceitunas, a partir de una serie de fotografías de las mismas. Utilizando dos variedades distintas de aceituna (Arbequina y Picual), se ha desarrollado un algoritmo de segmentación, a partir del cual se extrae la información necesaria para computar modelos de estimación para cada uno de los parámetros considerados. Una vez aplicados dichos modelos sobre los correspondientes conjuntos de validación, se ha podido comprobar, a través del cálculo de la raíz del error cuadrático medio (RMSE) cometido, la eficacia del método propuesto y su validez como base para el desarrollo de un sistema de calibrado de aceitunas de bajo coste basado en visión artificial.
[Abstract] The sizing and sorting of agricultural commodities is a high relevance activity in food industry. This study, focused on the olive farming sector, presents a solution based on image analysis which allows the automatic and non-invasive estimation of the weight and size (major and minor axis) of a set of olive fruits. Considering two different varieties of olive fruits (Arbequina and Picual), a segmentation algorithm, able to extract from images the needed information to compute the weight and size prediction models, was developed. The effectiveness of the proposed method was assessed by calculating the root-mean-square error (RMSE) produced by the models when applied to the corresponding external validation sets. The measured results show evidences of viability as a base to the development of a low-cost olive fruit grading system based on machine vision.
[Abstract] The sizing and sorting of agricultural commodities is a high relevance activity in food industry. This study, focused on the olive farming sector, presents a solution based on image analysis which allows the automatic and non-invasive estimation of the weight and size (major and minor axis) of a set of olive fruits. Considering two different varieties of olive fruits (Arbequina and Picual), a segmentation algorithm, able to extract from images the needed information to compute the weight and size prediction models, was developed. The effectiveness of the proposed method was assessed by calculating the root-mean-square error (RMSE) produced by the models when applied to the corresponding external validation sets. The measured results show evidences of viability as a base to the development of a low-cost olive fruit grading system based on machine vision.
Description
Editor version
Rights
Atribución-NoComercial 3.0 España


