Pasar al contenido principal

ALGORITMOS DE INTELIGENCIA ARTIFICIAL BASADOS EN APRENDIZAJE PROFUNDO Y REDES GAN PARA EL ANALISIS DE DATOS ESPECTRALES EN PROBLEMAS DE INSPECCION DE FRUTA.

PID2019-107347RR-C33

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2019
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 2
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Detection of Invisible Damages in `Rojo Brillante¿ Persimmon Fruit at Different Stages Using Hyperspectral Imaging and Chemometrics

  • Munera, S|||0000-0003-3064-1186
  • Rodriguez-Ortega, Alejandro|||0000-0003-4781-6368
  • Aleixos Borrás, María Nuria|||0000-0001-6051-3375
  • Cubero, Sergio
  • Gómez-Sanchis, Juan
  • Blasco, José
[EN] The main cause of flesh browning in 'Rojo Brillante' persimmon fruit is mechanical damage caused during harvesting and packing. Innovation and research on nondestructive techniques to detect this phenomenon in the packing lines are necessary because this type of alteration is often only seen when the final consumer peels the fruit. In this work, we have studied the application of hyperspectral imaging in the range of 450-1040 nm to detect mechanical damage without any external symptoms. The fruit was damaged in a controlled manner. Later, images were acquired before and at 0, 1, 2 and 3 days after damage induction. First, the spectral data captured from the images were analysed through an algorithm based on principal component analysis (PCA). The aim was to automatically separate intact and damaged fruit, and to detect the damage in the PC images when present. With this algorithm, 90.0% of intact fruit and 90.8% of damaged fruit were correctly detected. A model based on partial least squares-discriminant analysis (PLS-DA), was later calibrated using the mean spectrum of the pixels detected as damaged, to determine the moment when the fruit was damaged. The model differentiated fruit corresponding correctly to 0, 1, 2 and 3 days after damage induction, achieving a total accuracy of 99.4%., This work is co-funded by the projects AEI PID2019-107347RR-C31, PID2019-107347RR-C32, PID2019-107347RR-C33, IVIA-GVA 51918 and the European Union through the European Regional Development Fund (ERDF) of the Generalitat Valenciana 2014-2020.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Enhancing anthracnose detection in mango at early stages using hyperspectral imaging and machine learning

  • Velásquez, Carlos
  • Cubero, Sergio
  • Gomez-Sanchis, Juan
  • Prieto, Flavio
  • Blasco, José
  • Aleixos Borrás, María Nuria|||0000-0001-6051-3375
[EN] Anthracnose, caused by Colletotrichum sp. infections, poses a significant threat to mango production worldwide, resulting in substantial losses. This devastating disease is challenging to detect and control, primarily due to its ability to spread rapidly. The methods currently used to control anthracnose are primarily corrective, relying on disease detection in the late stages when the infection becomes visible. Hence, there is a need for tools to detect the infection at early stages, before symptoms appear. Hyperspectral imaging systems are promising for developing non-destructive solutions to assess and detect external and internal damage in fruit, including decay caused by anthracnose. These advanced imaging systems make early detection possible before the symptoms are visible, allowing for timely intervention and potentially more effective disease control. This work aims to evaluate the possibility of early detection of anthracnose in two mango cultivars using hyperspectral imaging and machine learning methods. Secondly, to establish correlations between specific wavelengths and the physicochemical symptoms associated with anthracnose. Lastly, to develop a robust model for the spectral detection of anthracnose on mango fruit. Mangoes were inoculated with spores of Colletotrichum gloeosporioides. Hyperspectral images of control and infected fruit were captured in the 450-970 nm spectral range. Five machinelearning models were used to obtain the method that best fits the spectral data. The best model achieved an accuracy = 0.961, recall = 0.961, specificity = 0.992, F1 = 0.961 and Matthews correlation coefficient (MCC) = 0.953 for 'Keitt', and an accuracy = 0.975, recall = 0.976, specificity = 0.995, F1 = 0.975 and MCC = 0.971 for 'Osteen', showing the feasibility to detect early anthracnose infection in mango fruit within 48 h after pathogen inoculation., This work was partially funded by the Ministerio de Ciencia y Tecnologia de Colombia (MINCIENCIAS) through its call "convocatoria 785 para doctorados nacionales 2017", Universidad Nacional de Colombia through its programmes "convocatoria nacional para el fomento de alianzas interdisciplinarias queue articulen investigacion, creacion, extension y formacion en la Universidad Nacional de Colombia 2019-2021" and through MICIN-AEI projects PID2019-107347RR-C31- C32-C33 with ERDF funds of the GVA 2021-2027, and GVA-PROMETEO CIPROM/2021/014. Authors thank Prof Dr Lluis Palou and Veronica Taberner, from the Postharvest Technology Center (CTP) of the IVIA, for providing the inoculum and technical suport.