Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products
Hyperspectral imaging (HSI) emerges as a non-destructive and rapid analytical tool for assessing food quality, safety, and authenticity. This work aims to investigate the potential of combining the spectral and spatial features of HSI data with the aid of deep learning approach for the pixel-wise cl...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-09-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/20/18/5322 |
| _version_ | 1797553359992913920 |
|---|---|
| author | Hongyan Zhu Aoife Gowen Hailin Feng Keping Yu Jun-Li Xu |
| author_facet | Hongyan Zhu Aoife Gowen Hailin Feng Keping Yu Jun-Li Xu |
| author_sort | Hongyan Zhu |
| collection | DOAJ |
| description | Hyperspectral imaging (HSI) emerges as a non-destructive and rapid analytical tool for assessing food quality, safety, and authenticity. This work aims to investigate the potential of combining the spectral and spatial features of HSI data with the aid of deep learning approach for the pixel-wise classification of food products. We applied two strategies for extracting spatial-spectral features: (1) directly applying three-dimensional convolution neural network (3-D CNN) model; (2) first performing principal component analysis (PCA) and then developing 2-D CNN model from the first few PCs. These two methods were compared in terms of efficiency and accuracy, exemplified through two case studies, i.e., classification of four sweet products and differentiation between white stripe (“myocommata”) and red muscle (“myotome”) pixels on salmon fillets. Results showed that combining spectral-spatial features significantly enhanced the overall accuracy for sweet dataset, compared to partial least square discriminant analysis (PLSDA) and support vector machine (SVM). Results also demonstrated that spectral pre-processing techniques prior to CNN model development can enhance the classification performance. This work will open the door for more research in the area of practical applications in food industry. |
| first_indexed | 2024-03-10T16:15:17Z |
| format | Article |
| id | doaj.art-17dcc0578f3a482db5c8a4fc5db1a94c |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T16:15:17Z |
| publishDate | 2020-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-17dcc0578f3a482db5c8a4fc5db1a94c2023-11-20T14:05:13ZengMDPI AGSensors1424-82202020-09-012018532210.3390/s20185322Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food ProductsHongyan Zhu0Aoife Gowen1Hailin Feng2Keping Yu3Jun-Li Xu4College of Electronic Engineering, Guangxi Normal University, Guilin 541004, ChinaUCD School of Biosystems and Food Engineering, University College of Dublin (UCD), Belfield, Dublin 4, IrelandSchool of Information Engineering, Zhejiang Agricultural and Forestry University, Hangzhou 310000, ChinaGlobal Information and Telecommunication Institute, Waseda University, Shinjuku, Tokyo 169-8050, JapanUCD School of Biosystems and Food Engineering, University College of Dublin (UCD), Belfield, Dublin 4, IrelandHyperspectral imaging (HSI) emerges as a non-destructive and rapid analytical tool for assessing food quality, safety, and authenticity. This work aims to investigate the potential of combining the spectral and spatial features of HSI data with the aid of deep learning approach for the pixel-wise classification of food products. We applied two strategies for extracting spatial-spectral features: (1) directly applying three-dimensional convolution neural network (3-D CNN) model; (2) first performing principal component analysis (PCA) and then developing 2-D CNN model from the first few PCs. These two methods were compared in terms of efficiency and accuracy, exemplified through two case studies, i.e., classification of four sweet products and differentiation between white stripe (“myocommata”) and red muscle (“myotome”) pixels on salmon fillets. Results showed that combining spectral-spatial features significantly enhanced the overall accuracy for sweet dataset, compared to partial least square discriminant analysis (PLSDA) and support vector machine (SVM). Results also demonstrated that spectral pre-processing techniques prior to CNN model development can enhance the classification performance. This work will open the door for more research in the area of practical applications in food industry.https://www.mdpi.com/1424-8220/20/18/5322hyperspectralspatial-spectral featuresclassificationprincipal component analysisconvolutional neural network |
| spellingShingle | Hongyan Zhu Aoife Gowen Hailin Feng Keping Yu Jun-Li Xu Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products Sensors hyperspectral spatial-spectral features classification principal component analysis convolutional neural network |
| title | Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products |
| title_full | Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products |
| title_fullStr | Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products |
| title_full_unstemmed | Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products |
| title_short | Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products |
| title_sort | deep spectral spatial features of near infrared hyperspectral images for pixel wise classification of food products |
| topic | hyperspectral spatial-spectral features classification principal component analysis convolutional neural network |
| url | https://www.mdpi.com/1424-8220/20/18/5322 |
| work_keys_str_mv | AT hongyanzhu deepspectralspatialfeaturesofnearinfraredhyperspectralimagesforpixelwiseclassificationoffoodproducts AT aoifegowen deepspectralspatialfeaturesofnearinfraredhyperspectralimagesforpixelwiseclassificationoffoodproducts AT hailinfeng deepspectralspatialfeaturesofnearinfraredhyperspectralimagesforpixelwiseclassificationoffoodproducts AT kepingyu deepspectralspatialfeaturesofnearinfraredhyperspectralimagesforpixelwiseclassificationoffoodproducts AT junlixu deepspectralspatialfeaturesofnearinfraredhyperspectralimagesforpixelwiseclassificationoffoodproducts |