Seed Protein Content Estimation with Bench-Top Hyperspectral Imaging and Attentive Convolutional Neural Network Models
Wheat is a globally cultivated cereal crop with substantial protein content present in its seeds. This research aimed to develop robust methods for predicting seed protein concentration in wheat seeds using bench-top hyperspectral imaging in the visible, near-infrared (VNIR), and shortwave infrared...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/2/303 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832587585303609344 |
|---|---|
| author | Imran Said Vasit Sagan Kyle T. Peterson Haireti Alifu Abuduwanli Maiwulanjiang Abby Stylianou Omar Al Akkad Supria Sarkar Noor Al Shakarji |
| author_facet | Imran Said Vasit Sagan Kyle T. Peterson Haireti Alifu Abuduwanli Maiwulanjiang Abby Stylianou Omar Al Akkad Supria Sarkar Noor Al Shakarji |
| author_sort | Imran Said |
| collection | DOAJ |
| description | Wheat is a globally cultivated cereal crop with substantial protein content present in its seeds. This research aimed to develop robust methods for predicting seed protein concentration in wheat seeds using bench-top hyperspectral imaging in the visible, near-infrared (VNIR), and shortwave infrared (SWIR) regions. To fully utilize the spectral and texture features of the full VNIR and SWIR spectral domains, a computer-vision-aided image co-registration methodology was implemented to seamlessly align the VNIR and SWIR bands. Sensitivity analyses were also conducted to identify the most sensitive bands for seed protein estimation. Convolutional neural networks (CNNs) with attention mechanisms were proposed along with traditional machine learning models based on feature engineering including Random Forest (RF) and Support Vector Machine (SVM) regression for comparative analysis. Additionally, the CNN classification approach was used to estimate low, medium, and high protein concentrations because this type of classification is more applicable for breeding efforts. Our results showed that the proposed CNN with attention mechanisms predicted wheat protein content with R<sup>2</sup> values of 0.70 and 0.65 for ventral and dorsal seed orientations, respectively. Although, the R<sup>2</sup> of the CNN approach was lower than of the best performing feature-based method, RF (R<sup>2</sup> of 0.77), end-to-end prediction capabilities with CNN hold great promise for the automation of wheat protein estimation for breeding. The CNN model achieved better classification of protein concentrations between low, medium, and high protein contents, with an R<sup>2</sup> of 0.82. This study’s findings highlight the significant potential of hyperspectral imaging and machine learning techniques for advancing precision breeding practices, optimizing seed sorting processes, and enabling targeted agricultural input applications. |
| format | Article |
| id | doaj-art-998166f1bc5649e4aa201b8907430bc0 |
| institution | Kabale University |
| issn | 1424-8220 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-998166f1bc5649e4aa201b8907430bc02025-01-24T13:48:26ZengMDPI AGSensors1424-82202025-01-0125230310.3390/s25020303Seed Protein Content Estimation with Bench-Top Hyperspectral Imaging and Attentive Convolutional Neural Network ModelsImran Said0Vasit Sagan1Kyle T. Peterson2Haireti Alifu3Abuduwanli Maiwulanjiang4Abby Stylianou5Omar Al Akkad6Supria Sarkar7Noor Al Shakarji8Department of Computer Science, Saint Louis University, Saint Louis, MO 63104, USADepartment of Computer Science, Saint Louis University, Saint Louis, MO 63104, USABayer Crop Science, 800 N Lindbergh Blvd, Creve Coeur, MO 63141, USADepartment of Earth, Environment and Geospatial Sciences, Saint Louis University, Saint Louis, MO 63108, USADepartment of Earth, Environment and Geospatial Sciences, Saint Louis University, Saint Louis, MO 63108, USADepartment of Computer Science, Saint Louis University, Saint Louis, MO 63104, USADepartment of Earth, Environment and Geospatial Sciences, Saint Louis University, Saint Louis, MO 63108, USADepartment of Earth, Environment and Geospatial Sciences, Saint Louis University, Saint Louis, MO 63108, USADepartment of Earth, Environment and Geospatial Sciences, Saint Louis University, Saint Louis, MO 63108, USAWheat is a globally cultivated cereal crop with substantial protein content present in its seeds. This research aimed to develop robust methods for predicting seed protein concentration in wheat seeds using bench-top hyperspectral imaging in the visible, near-infrared (VNIR), and shortwave infrared (SWIR) regions. To fully utilize the spectral and texture features of the full VNIR and SWIR spectral domains, a computer-vision-aided image co-registration methodology was implemented to seamlessly align the VNIR and SWIR bands. Sensitivity analyses were also conducted to identify the most sensitive bands for seed protein estimation. Convolutional neural networks (CNNs) with attention mechanisms were proposed along with traditional machine learning models based on feature engineering including Random Forest (RF) and Support Vector Machine (SVM) regression for comparative analysis. Additionally, the CNN classification approach was used to estimate low, medium, and high protein concentrations because this type of classification is more applicable for breeding efforts. Our results showed that the proposed CNN with attention mechanisms predicted wheat protein content with R<sup>2</sup> values of 0.70 and 0.65 for ventral and dorsal seed orientations, respectively. Although, the R<sup>2</sup> of the CNN approach was lower than of the best performing feature-based method, RF (R<sup>2</sup> of 0.77), end-to-end prediction capabilities with CNN hold great promise for the automation of wheat protein estimation for breeding. The CNN model achieved better classification of protein concentrations between low, medium, and high protein contents, with an R<sup>2</sup> of 0.82. This study’s findings highlight the significant potential of hyperspectral imaging and machine learning techniques for advancing precision breeding practices, optimizing seed sorting processes, and enabling targeted agricultural input applications.https://www.mdpi.com/1424-8220/25/2/303hyperspectral imagingseed composition estimationmachine learning3D CNN modelingattentive models |
| spellingShingle | Imran Said Vasit Sagan Kyle T. Peterson Haireti Alifu Abuduwanli Maiwulanjiang Abby Stylianou Omar Al Akkad Supria Sarkar Noor Al Shakarji Seed Protein Content Estimation with Bench-Top Hyperspectral Imaging and Attentive Convolutional Neural Network Models Sensors hyperspectral imaging seed composition estimation machine learning 3D CNN modeling attentive models |
| title | Seed Protein Content Estimation with Bench-Top Hyperspectral Imaging and Attentive Convolutional Neural Network Models |
| title_full | Seed Protein Content Estimation with Bench-Top Hyperspectral Imaging and Attentive Convolutional Neural Network Models |
| title_fullStr | Seed Protein Content Estimation with Bench-Top Hyperspectral Imaging and Attentive Convolutional Neural Network Models |
| title_full_unstemmed | Seed Protein Content Estimation with Bench-Top Hyperspectral Imaging and Attentive Convolutional Neural Network Models |
| title_short | Seed Protein Content Estimation with Bench-Top Hyperspectral Imaging and Attentive Convolutional Neural Network Models |
| title_sort | seed protein content estimation with bench top hyperspectral imaging and attentive convolutional neural network models |
| topic | hyperspectral imaging seed composition estimation machine learning 3D CNN modeling attentive models |
| url | https://www.mdpi.com/1424-8220/25/2/303 |
| work_keys_str_mv | AT imransaid seedproteincontentestimationwithbenchtophyperspectralimagingandattentiveconvolutionalneuralnetworkmodels AT vasitsagan seedproteincontentestimationwithbenchtophyperspectralimagingandattentiveconvolutionalneuralnetworkmodels AT kyletpeterson seedproteincontentestimationwithbenchtophyperspectralimagingandattentiveconvolutionalneuralnetworkmodels AT hairetialifu seedproteincontentestimationwithbenchtophyperspectralimagingandattentiveconvolutionalneuralnetworkmodels AT abuduwanlimaiwulanjiang seedproteincontentestimationwithbenchtophyperspectralimagingandattentiveconvolutionalneuralnetworkmodels AT abbystylianou seedproteincontentestimationwithbenchtophyperspectralimagingandattentiveconvolutionalneuralnetworkmodels AT omaralakkad seedproteincontentestimationwithbenchtophyperspectralimagingandattentiveconvolutionalneuralnetworkmodels AT supriasarkar seedproteincontentestimationwithbenchtophyperspectralimagingandattentiveconvolutionalneuralnetworkmodels AT nooralshakarji seedproteincontentestimationwithbenchtophyperspectralimagingandattentiveconvolutionalneuralnetworkmodels |