AI-Assisted Cotton Grading: Active and Semi-Supervised Learning to Reduce the Image-Labelling Burden
The assessment of food and industrial crops during harvesting is important to determine the quality and downstream processing requirements, which in turn affect their market value. While machine learning models have been developed for this purpose, their deployment is hindered by the high cost of la...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-10-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/21/8671 |
| _version_ | 1827765394772852736 |
|---|---|
| author | Oliver J. Fisher Ahmed Rady Aly A. A. El-Banna Haitham H. Emaish Nicholas J. Watson |
| author_facet | Oliver J. Fisher Ahmed Rady Aly A. A. El-Banna Haitham H. Emaish Nicholas J. Watson |
| author_sort | Oliver J. Fisher |
| collection | DOAJ |
| description | The assessment of food and industrial crops during harvesting is important to determine the quality and downstream processing requirements, which in turn affect their market value. While machine learning models have been developed for this purpose, their deployment is hindered by the high cost of labelling the crop images to provide data for model training. This study examines the capabilities of semi-supervised and active learning to minimise effort when labelling cotton lint samples while maintaining high classification accuracy. Random forest classification models were developed using supervised learning, semi-supervised learning, and active learning to determine Egyptian cotton grade. Compared to supervised learning (80.20–82.66%) and semi-supervised learning (81.39–85.26%), active learning models were able to achieve higher accuracy (82.85–85.33%) with up to 46.4% reduction in the volume of labelled data required. The primary obstacle when using machine learning for Egyptian cotton grading is the time required for labelling cotton lint samples. However, by applying active learning, this study successfully decreased the time needed from 422.5 to 177.5 min. The findings of this study demonstrate that active learning is a promising approach for developing accurate and efficient machine learning models for grading food and industrial crops. |
| first_indexed | 2024-03-11T11:21:18Z |
| format | Article |
| id | doaj.art-562e2c333fda4b00ae13bbb5fd6848bc |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T11:21:18Z |
| publishDate | 2023-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-562e2c333fda4b00ae13bbb5fd6848bc2023-11-10T15:11:38ZengMDPI AGSensors1424-82202023-10-012321867110.3390/s23218671AI-Assisted Cotton Grading: Active and Semi-Supervised Learning to Reduce the Image-Labelling BurdenOliver J. Fisher0Ahmed Rady1Aly A. A. El-Banna2Haitham H. Emaish3Nicholas J. Watson4Food, Water, Waste Research Group, Faculty of Engineering, University of Nottingham, Nottingham NG7 2RD, UKFood, Water, Waste Research Group, Faculty of Engineering, University of Nottingham, Nottingham NG7 2RD, UKDepartment of Plant Production, Faculty of Agriculture, Saba Basha, Alexandria University, Alexandria 5424041, EgyptDepartment of Soils and Agricultural Chemistry, Faculty of Agriculture, Saba Basha, Alexandria University, Alexandria 5424041, EgyptFood, Water, Waste Research Group, Faculty of Engineering, University of Nottingham, Nottingham NG7 2RD, UKThe assessment of food and industrial crops during harvesting is important to determine the quality and downstream processing requirements, which in turn affect their market value. While machine learning models have been developed for this purpose, their deployment is hindered by the high cost of labelling the crop images to provide data for model training. This study examines the capabilities of semi-supervised and active learning to minimise effort when labelling cotton lint samples while maintaining high classification accuracy. Random forest classification models were developed using supervised learning, semi-supervised learning, and active learning to determine Egyptian cotton grade. Compared to supervised learning (80.20–82.66%) and semi-supervised learning (81.39–85.26%), active learning models were able to achieve higher accuracy (82.85–85.33%) with up to 46.4% reduction in the volume of labelled data required. The primary obstacle when using machine learning for Egyptian cotton grading is the time required for labelling cotton lint samples. However, by applying active learning, this study successfully decreased the time needed from 422.5 to 177.5 min. The findings of this study demonstrate that active learning is a promising approach for developing accurate and efficient machine learning models for grading food and industrial crops.https://www.mdpi.com/1424-8220/23/21/8671machine learningdigital manufacturingcottoncolour vision systemquality assessmentsemi-supervised learning |
| spellingShingle | Oliver J. Fisher Ahmed Rady Aly A. A. El-Banna Haitham H. Emaish Nicholas J. Watson AI-Assisted Cotton Grading: Active and Semi-Supervised Learning to Reduce the Image-Labelling Burden Sensors machine learning digital manufacturing cotton colour vision system quality assessment semi-supervised learning |
| title | AI-Assisted Cotton Grading: Active and Semi-Supervised Learning to Reduce the Image-Labelling Burden |
| title_full | AI-Assisted Cotton Grading: Active and Semi-Supervised Learning to Reduce the Image-Labelling Burden |
| title_fullStr | AI-Assisted Cotton Grading: Active and Semi-Supervised Learning to Reduce the Image-Labelling Burden |
| title_full_unstemmed | AI-Assisted Cotton Grading: Active and Semi-Supervised Learning to Reduce the Image-Labelling Burden |
| title_short | AI-Assisted Cotton Grading: Active and Semi-Supervised Learning to Reduce the Image-Labelling Burden |
| title_sort | ai assisted cotton grading active and semi supervised learning to reduce the image labelling burden |
| topic | machine learning digital manufacturing cotton colour vision system quality assessment semi-supervised learning |
| url | https://www.mdpi.com/1424-8220/23/21/8671 |
| work_keys_str_mv | AT oliverjfisher aiassistedcottongradingactiveandsemisupervisedlearningtoreducetheimagelabellingburden AT ahmedrady aiassistedcottongradingactiveandsemisupervisedlearningtoreducetheimagelabellingburden AT alyaaelbanna aiassistedcottongradingactiveandsemisupervisedlearningtoreducetheimagelabellingburden AT haithamhemaish aiassistedcottongradingactiveandsemisupervisedlearningtoreducetheimagelabellingburden AT nicholasjwatson aiassistedcottongradingactiveandsemisupervisedlearningtoreducetheimagelabellingburden |