Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 4
Encontrada(s) 1 página(s)

Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
Publicación Contribución de congreso (ConferenceOutput)

On-site identification of esca-affected vines using hyperspectral imaging

  • León Ecay, Sara
  • Ruiz de Gauna González, Jon
  • López Maestresalas, Ainara
  • Jarén Ceballos, Carmen
  • Arazuri Garín, Silvia
Esca represents one of the greatest threats to modern viticulture as it causes large annual economic losses. At present, there is a lack of effective strategies for disease control, so a technique capable of detecting affected vines would allow annual monitoring of disease incidence in the vineyard leading to a better crop management and decision making. This study evaluates close-range hyperspectral imaging for the detection of esca naturally infected vines. Images of 11 vines of the Tempranillo variety grown on plots in Bodegas Otazu, in Etxauri (Navarre, Spain) were acquired. A Specim IQ snapshot hyperspectral camera was used to record the images on August, 21 2023 on the field under natural light conditions. The camera has a spectral resolution of 7 nm (204 wavelengths) and a spatial resolution of 512 x 512 in the 400 ¿ 1000 nm spectral range (Vis-NIR). An individual image was acquired for each vine, of which 9 were symptomatic and 2 asymptomatic. Three classes were analysed: asymptomatic leaves of asymptomatic vines (Class 1), asymptomatic leaves of symptomatic vines (Class 2) and asymptomatic areas of symptomatic leaves of symptomatic vines (Class 3). A total of 300 pixels were randomly selected, 100 per class, for further analysis. Partial Least Square Discriminant Analysis (PLSDA) was used to classify the pixels into the three categories. An accuracy of 86% was achieved in the cross-validation dataset. Models were externally validated using an image of an asymptomatic vine and an image of a symptomatic vine. The visualisation of the images showed that the majority of the pixels of the asymptomatic vine image were classified as class 1, while most of the pixels of the symptomatic vine image were classified as either class 2 or class 3. Hence, this study demonstrated the potential of close-range HSI for the on-site detection of esca., This research was supported by the Spanish Ministry of Science and Innovation and from the European Union Next-Generation EU/PRTR (TED2021-130364B-I00).
Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
Publicación Artículo científico (JournalArticle)

A low-cost open-cavity moisture content sensor for flour-based matrices in the 5-6.8 GHz range

  • Quemada Mayoral, Carlos
  • Ederra Urzainqui, Íñigo
  • Gonzalo García, Ramón
  • Iriarte Galarregui, Juan Carlos
This study presents the design, implementation and experimental validation of a low-cost and non-destructive water content monitoring sensor for thin corn tortilla samples, intended as a first prototype for potential integration into food production lines. The sensor operates in the 5-6.8 GHz frequency range, with moisture estimation performed at the sensor resonance frequency. This sensor, consisting of an open horn-shaped resonant cavity, represents a first prototype that could potentially be integrated into existing production lines for these food products, highlighting its potential for future industrial implementation. The development of an efficient and thorough design flow, which combines an accurate theoretical calculation of the complex permittivity of the samples with electromagnetic simulations, has allowed the design of a humidity sensor whose simulation model estimates water content in corn tortilla samples with a root mean square error (RMSE) below 0.8 % with respect to the measurements. From the results obtained with this sensor, two moisture estimation models are developed: a simulation-derived regression model, validated using experimental measurements (RMSE < 0.8 %), and a purely experimental regression model. The simulation-derived model achieves a coefficient of determination (R2) of 0.99, while the experimental model yields an R2 of 0.91 and an RMSE below 0.4 %. Due to its easy implementation, low cost, excellent correlation and reduced estimation errors, this prototype constitutes the first step towards the design of a non-destructive and cost-effective moisture sensor at microwave frequencies to be applied in food production lines., The authors of this paper acknowledge the Spanish State Research Agency for funding this research under the PID2023-146064OB-C31 and TED2021-130364B-I00 projects.
Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
Publicación Contribución de congreso (ConferenceOutput)

Early detection of Esca disease in grapevines using in-field hyperspectral proximal sensing

  • López Maestresalas, Ainara
  • Ruiz de Gauna González, Jon
  • Jarén Ceballos, Carmen
  • León Ecay, Sara
  • Arazuri Garín, Silvia
Esca is one of the most destructive vine diseases in the world. It causes significant economic losses, mainly due to reduced grape yield and quality. Currently, the approved methods of controlling esca include preventive methods such as the use of fungicides on plant wounds or the use of planting systems that do not require intensive pruning, among others. It is therefore advisable to monitor the crop to identify those vines that are susceptible to the disease. For this reason, in this study a proximal hyperspectral camera was used for early detection of esca presence in asymptomatic grapevine leaves. Images of 11 vines of the Tempranillo variety grown in Etxauri (Navarre, Spain) were analysed. Hyperspectral images were acquired using a Specim IQ snapshot camera, mounted on a tripod, working in the range of 400¿1000 nm with a spectral resolution of 7 nm (204 bands), and an image resolution of 512 × 512 pixel including an RGB camera (5 Mpix). The images were taken under natural ambient light conditions on August 21, 2023. From the 11 vines selected, 9 showed visual symptoms of esca and the remaining 2 were asymptomatic to the naked eye. A total of 200 pixels were randomly selected from the dataset, 100 from asymptomatic leaves of asymptomatic vines (class 1) and 100 from asymptomatic leaves of symptomatic vines (class 2). Partial Least Square Discriminant Analysis (PLS-DA) was performed to classify the leaves into the two classes. Classification rates of 97% were achieved in the cross-validation dataset. Models were externally validated at pixel-level using one image of an asymptomatic vine and another of a symptomatic vine. The visualisation of the images confirmed the correct classification of the pixels into the two classes, indicating that by using proximal hyperspectral sensing an early identification of the disease is possible., This research was supported by the Spanish Ministry of Science and Innovation and from the European Union Next-GenerationEU/PRTR (TED2021-130364B-I00).
Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
Publicación Artículo científico (JournalArticle)

Portable and low-cost radar sensor for moisture content estimation of flour-based matrices in the 76-81 GHz millimeter-wave range

  • Quemada Mayoral, Carlos
  • Sada Allo, Maider
  • Beruete Díaz, Miguel
  • Angós Iturgaiz, Ignacio
  • Ederra Urzainqui, Íñigo
  • Gonzalo García, Ramón
  • Iriarte Galarregui, Juan Carlos
This article presents the first study of food moisture content (MC) estimation using a portable, cost-effective, and low-power millimeter-wave sensor operating in the 76–81-GHz automotive radar band. Recent advances in fully integrated, off-the-shelf radar transceivers enable the development of compact and affordable sensing solutions suitable for in-line food inspection. The proposed approach estimates the MC of flourbased matrices by correlating the electromagnetic attenuation losses introduced by the sample in a round-trip configuration with gravimetrically determined water content. Experimental validation was carried out on 40 corn tortilla samples covering moisture concentrations between 30 % and 40.4 %. Two regression approaches were evaluated. While a linear model provided a coefficient of determination (R2) of 0.943, residual analysis revealed systematic deviations. To address this limitation, a logarithmic calibration model was implemented and validated using leave-one-out cross-validation (LOOCV), achieving R2 = 0.971, a root mean square error of cross-validation (RMSECV) of 0.857 %, and negligible bias. The logarithmic model reduced the prediction error by 32 % compared to the linear approach and eliminated structural residual trends. These results demonstrate that millimeter-wave, frequency-modulated continuous-wave (FMCW) radar technology enables accurate, real-time, and noncontact MC estimation, paving the way for fully portable and industrially deployable in-line food sensing systems., This work was supported in part by the Spanish State Research Agency under Project PID2023-146064OB-C31 and Project TED2021-130364B-I00 and in part by the Open access funding provided by Universidad Publica de Navarra.