A short note on deep contextual spatial and spectral information fusion for hyperspectral image processing: Case of pork belly properties prediction
View/Open
Publication date
2024-04-18ISSN
0886-9383
Abstract
This study demonstrates a new approach to process hyperspectral images
where both the contextual spatial information as well as the spectral
information are used to predict sample properties. The deep contextual spatial
information is extracted using the deep feature extraction from pretrained
resnet-18 deep learning architecture, while the spectral information was
readily available as the average pixel values. To fuse the information in a
complementary way, a multiblock modeling approach called sequential
orthogonalized partial least squares was used. The sequential model guarantees
that the information learned is complementary from spatial and spectral
domains. The potential of the approach is demonstrated to predict several
physical and chemical properties in pork bellies.
Document Type
Article
Document version
Published version
Language
English
Subject (CDU)
663/664 - Food and nutrition. Enology. Oils. Fat
Pages
9
Publisher
Wiley
Is part of
Journal of Chemometrics
Citation
Mishra, Puneet, Michela Albano‐Gaglio, and Maria Font‐i‐Furnols. 2024. “A short note on deep contextual spatial and spectral information fusion for hyperspectral image processing: Case of pork belly properties prediction”. Journal of Chemometrics, April. doi:10.1002/cem.3552.
Grant agreement number
MICIU/Programa Estatal de I+D+I orientada a los retos de la sociedad/RTI2018-096993-B-I00/ES/CLASIFICACION Y EVALUACION DE LA CALIDAD GLOBAL DE LA PANCETA DE CERDO MEDIANTE TECNOLOGIAS NO DESTRUCTIVAS Y PERCEPCION POR PARTE DE LOS CONSUMIDORES/
Program
Qualitat i Tecnologia Alimentària
This item appears in the following Collection(s)
- ARTICLES CIENTÍFICS [2838]
Except where otherwise noted, this item's license is described as http://creativecommons.org/licenses/by/4.0/