Skip to main content

ANALISIS POLINICO AUTOMATICO EMPLEANDO REDES NEURONALES CONVOLUCIONALES: APLICACION A LA CLASIFICACION MONOFLORAL DE LA MIEL

PID2019-106800RB-I00

•
Funding agency name Agencia Estatal de Investigación
Funding agency acronym AEI
Program Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Call Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Call year 2019
State Plan 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Beneficiary centre UNIVERSITAT POLITÈCNICA DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 7
Encontrada(s) 1 página(s)

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

Monofloral honey authentication by voltammetric electronic tongue: A comparison with 1H NMR spectroscopy

  • Lozano-Torres, Beatriz
  • Soto Camino, Juan
  • Martínez-Bisbal, M.Carmen|||0000-0002-8526-4413
  • Juan-Borras, María del Sol|||0000-0003-4774-9484
  • Martínez-Máñez, Ramón|||0000-0001-5873-9674
  • Escriche Roberto, Mª Isabel|||0000-0003-0180-0360
[EN] Proton-nuclear-magnetic-resonance-spectroscopy (H-1 NMR) is the widely accepted reference method for monitoring honey adulteration; however, the need to find cheaper, faster, and more environmentally friendly methodologies makes the voltammetric-electronic-tongue (VET) a good alternative. The present study aims to demonstrate the ability of VET (in comparison with H-1 NMR) to predict the adulteration of honey with syrups. Samples of monofloral honeys (citrus, sunflower and heather, assessed by pollen analysis) simulating different levels of adulteration by adding syrups (barley, rice and corn) from 2.5 to 40% (w/w) were analyzed using both techniques. According to the indicators (slope, intercept, regression coefficient-R-2, root mean square error of prediction-RMSEP) of the partial-least-squares (PLS) regression models, in general terms, the performance of these models obtained by both techniques was good, with an average error lower than 5% in both cases. These results support the use of VET as a screening technique to easily detect honey adulteration with syrups., The authors are grateful for financial support from the "Ministerio de Ciencia e Innovacion, Agencia Estatal de Investigacon" of Spain under the project PID2019-106800RB-I00 (2019). Also, we wish to thank the Spanish Government (project RTI2018-100910-B-C41 (MCUI/FEDER, EU)), the Generalitat Valenciana (project PROMETEO 2018/024 B.L-T) and the Spanish "Ministerio de Educacion, Cultura y Deporte" (PhD grant FPU15/02707).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Classification of Honey Pollens with ImageNet Neural Networks

  • López García, Fernando
  • Valiente González, José Miguel|||0000-0003-4055-8976
  • Escriche Roberto, Mª Isabel|||0000-0003-0180-0360
  • Juan-Borras, María del Sol|||0000-0003-4774-9484
  • Visquert Fas, Mario|||0000-0003-4230-1672
  • Atienza-Vanacloig, Vicente|||0000-0001-9017-8707
  • Agustí-Melchor, Manuel|||0000-0002-7280-0360
[EN] The classification of honey pollen grains is performed in order to clas- sify honey according to its botanical origin, which is of great importance in terms of marketing. This visual work is currently done by human specialists counting and classifying the pollen grains in microscopic images. This is a hard, time- consuming, and subject to observer variability task. Thus, automated methods are required to overcome the limitations of the conventional procedure. This paper deals with the automatic classification of honey pollens using five representa- tive Neural Networks coming from the ImageNet Challenge: VGG16, VGG19, ResNet50, InceptionV3 and Xception. The ground truth is composed of 9983 sam- ples of 16 different types of pollens corresponding to citrus and rosemary pollens and its companions. The best result was obtained with the InceptionV3 network, achieving an accuracy of 98.15%, that outperforms the results obtained in previous works., This work is part of the project PID2019-106800RB-I00 (2019) of the Ministry of Science and Innovation (MCIN), State Research Agency MCIN/AEI/https://doi.org/10.13039/501100011033/. It is also part of the AGROALNEXT/2022/043 project, financed by theGeneralitat Valenciana, the Next Generation European Union and the Recovery, Transformation and Resilience Plan of the Government of Spain.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Volatile markers as a reliable alternative for the correct classification of citrus monofloral honey

  • Escriche Roberto, Mª Isabel|||0000-0003-0180-0360
  • Conchado Peiró, Andrea|||0000-0001-5222-5218
  • Peral-Pinto, Ana María
  • Juan-Borras, María del Sol|||0000-0003-4774-9484
[EN] The pollen analysis to classify monofloral honey is an unresolved challenge specially when the pollen is under-represented as the case of citrus honey. Thus, this study assesses the validity of the volatile fraction to differ-entiate types of honey, with special attention to markers compounds of citrus honey that could permit their distinction. Unsupervised analysis (PCA and HCA) showed that the volatile fraction of honey containing Citrus sp. pollen, undoubtedly differentiates it from other types of honey. An OPLS model focused on citrus honey selected 5 volatile compounds (of the 123 found in all samples by GC-MS) as significant predictors of the currently used value of methyl anthranilate obtained by HPLC. The joint detection of 4 lilac-aldehydes and the volatile methyl-anthranilate has the advantage of providing more precise information. Therefore, it could be proposed as a consistent marker to ensure the correct classification of citrus honey, fostering its labelling reliability., This work is part of Spanish project PID2019-106800RB-I00 (2019) with financial support from the Ministerio de Ciencia e Innovacion (MCIN), Agencia Estatal de Investigacion MCIN/AEI/10.13039/501100011033/. It has been also part of the project AGROALNEXT/2022/043, funded by the Next Generation European Union and the Plan de Recuperacion, Transformacion y Resiliencia of the Spanish Government, with the support of Generalitat Valenciana. The authors would like to thank the CRUE-Universitat Politecnica de Valencia for providing the funds for open access publication.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Assisting monofloral honey classification by automated pollen identification based on convolutional neural networks

  • Valiente González, José Miguel|||0000-0003-4055-8976
  • Martín-Osuna, Juan José
  • Peral-Pinto, Ana María
  • Escriche Roberto, Mª Isabel|||0000-0003-0180-0360
[EN] Categorising monoflorality of honey by human experts is labour intensive and time-consuming, resulting in the urgent need for automated techniques to overcome the constraints of traditional method which involves manually sorting of pollen grains. This study constructs a comprehensive dataset of pollen images, labelling them with an enhanced version of HoneyApp application. This Ground Truth termed POLLEN24_SP, comprises 32,285 pollen/particle images (captured by an expert using optical microscopy), covering the 24 most prevalent types of pollen grains found in Spanish honeys. Twelve different pre-existing Convolutional Neural Networks (CNN) were evaluated, achieving an accuracy rate of up to 98.03 % with EfficientNetV2M. The new version of in-house networks (PolleNetV2, PolleNetV2.mobile) was also introduced. The pre-existing CNN models' results were satisfactory in discerning the relative frequencies of different pollen types; nevertheless, our new proposals are promising due to their straightforward architecture as well as their scalability for integration into the HoneyApp ecosystem. Studying the classification errors made by the CNN models confirmed their strong generalization ability. To minimise errors in types such as Leguminoseae, segmenting this family is recommended. In addition, it is vital to continue expanding POLLEN24_SP dataset focusing on the underrepresented types to balance them and include more pollen types to better represent Spanish honeys. Finally, the results of this work will enable the creation of a public dataset to be diffused among the scientific community, following open science principles. The development of a computer application incorporating CNN models, makes these methodologies suitable for efficient classification of monofloral honeys., This work is part of the project AGROALNEXT/2022/043, funded by the Next Generation European Union and the Plan de Recuperacion, Transformacion y Resiliencia of the Spanish Government, with the support of Generalitat Valenciana. It has been also part of Spanish project PID2019-106800RB-I00 (2019) with financial support from the Ministerio de Ciencia e Innovacion (MCIN), Agencia Estatal de Investigacion MCIN/AEI/10.13039/501100011033/.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Automatic pollen recognition using convolutional neural networks: The case of the main pollens present in Spanish citrus and rosemary honey

  • Valiente González, José Miguel|||0000-0003-4055-8976
  • Juan-Borras, María del Sol|||0000-0003-4774-9484
  • Escriche Roberto, Mª Isabel|||0000-0003-0180-0360
  • López García, Fernando
[EN] The automation of honey pollen visual sorting overcomes the limitations of the conventional procedure helping the specialist in this time-consuming task. In this work, a novel and comprehensive Ground Truth of almost 19,000 images (from optical microscopy) of the 16 most abundant types of grains/pollen particles present in citrus and rosemary honey from Spain was constructed. This task was assisted by a HoneyApp (also developed herein) for the labelling and annotation process. Subsequently, the effectiveness of different pre-existing automatic pollen recognizers based on convolutional neural networks (CNN) (VGG16, VGG19, InceptionV3, Xception, ResNet50, DenseNet201, MobileNetV2 and EfficientNetV2M) was tested together with a new network proposed in this paper (PolleNetV1). The extreme complexity of those pre-existing CNN and extensive use of millions of parameters makes this new proposal especially promising. Although with a slightly lower accuracy (average 96%) in determining the relative frequencies of different types of pollen grains/particles, it has considerable advantages such as simplicity and ability to be included in the future functionality to automate pollen recognition in honey. This is the first step to finally achieving an objective tool that allows the correct labelling of any types of pollen in honey, thus contributing to its transparency in the market., This work is part of Spanish project PID2019-106800RB-I00 (2019) with financial support from the Ministerio de Ciencia e Innovacion (MCIN), Agencia Estatal de Investigacion MCIN/AEI/10.13039/501100011033/. It has been also part of the project AGROALNEXT/2022/043, funded by the Next Generation European Union and the Plan de Recuperacion, Transformacion y Resiliencia of the Spanish Government, with the support of Generalitat Valenciana. The authors would like to thank the CRUE-Universitat Politecnica deValencia for providing the funds for open access publication.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

An overview of the challenges when analysing pollen for monofloral honey classification

  • Escriche Roberto, Mª Isabel|||0000-0003-0180-0360
  • Juan-Borras, María del Sol|||0000-0003-4774-9484
  • Visquert Fas, Mario|||0000-0003-4230-1672
  • Valiente González, José Miguel|||0000-0003-4055-8976
[EN] The complexity of the pollen analysis, even for accredited laborat01ies, and the absence of standardized guidelines for a routine method to detennine the monoflorality of a honey is a challenge. To expose the reality of this situation, the aim of this work is to discuss how the info1mation and the quantification of the honey pollen content results are presented, conside1ing the reports provided by four of these laboratoiies. Noticeable differ­ences in how the data is presented is observed; not only in the number of types of pollens (main and accom­panying) which percentage is reponed but also for the ciiteria followed in naniing the sanie pollen, applying different taxonomic levels (faniily, genus, type, specie, etc.). Although there is one common consensus by all laboratories based on discarding the pollens from non-melliferous plants, in general the discrepancies obse1ved are important. The problem is more evident when there is a presence of over-represented pollen and die target pollen is under-represented as is the case of Cimis. Thus, the simultaneous presence of under-represented and over-represented grains of pollen may frequently lead to misleading results. The lack of standardization in pollen analysis to be applied equally by all laborat01ies significantly contributes to die discrepancies. Only by having a common cliterion to interpret the pollen specn-um and, obviously, enstung the adequate experience of the analyst and the knowledge of the related flora is when the results can be considered reliable., The authors are grateful for financial support from the "Ministerio de Ciencia e Innovacion, Agencia Estatal de Investigacion" of Spain under the project PID2019-106800RB-I00 (2019).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Volatile profile of Spanish raw citrus honey: The best strategy for its correct labeling

  • Escriche Roberto, Mª Isabel|||0000-0003-0180-0360
  • Juan-Borras, María del Sol|||0000-0003-4774-9484
  • Visquert Fas, Mario|||0000-0003-4230-1672
  • Valiente González, José Miguel|||0000-0003-4055-8976
  • Asensio-Grau, Andrea
[EN] The objective of this study was to evaluate the presence of specific volatile compounds (analyzed by SPME-GC- MS) in citrus honey, comparing thyme, and sunflower, and to correlate their abundance with the level of methyl-anthranilate (accepted as a true marker in citrus honey), for which a minimum content is mandatory in commercial transactions. Methyl-anthranilate is well correlated with certain volatile compounds such as 1-p- menthene- 9- al (0.903), limonene (0.885), dill-ether (0.842), and ethyl linalool (0.832), and slightly lower with the four lilac aldehydes (0.717¿0.764). However, the latter, together with methyl-anthranilate, are of special interest because these were found in all citrus honey samples. Consequently, the information provided by these five compounds could facilitate the unmistakable classification of this honey. Thanks to these positive results and even though that this method is not commonly practiced to objectively differentiate between monofloral honey it can be considered an interesting analytical tool in the future., The authors are grateful for financial support from the Ministerio de Ciencia e Innovación. Agencia Estatal de Investigación of Spain under the project PID2019-106800RB-I00 (2019).