Adaptive Thresholding of CNN Features for Maize Leaf Disease Classification and Severity Estimation
Convolutional neural networks (CNNs) are the gold standard in the machine learning (ML) community. As a result, most of the recent studies have relied on CNNs, which have achieved higher accuracies compared with traditional machine learning approaches. From prior research, we learned that multi-clas...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-08-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/12/17/8412 |
| _version_ | 1827667252602732544 |
|---|---|
| author | Harry Dzingai Mafukidze Godliver Owomugisha Daniel Otim Action Nechibvute Cloud Nyamhere Felix Mazunga |
| author_facet | Harry Dzingai Mafukidze Godliver Owomugisha Daniel Otim Action Nechibvute Cloud Nyamhere Felix Mazunga |
| author_sort | Harry Dzingai Mafukidze |
| collection | DOAJ |
| description | Convolutional neural networks (CNNs) are the gold standard in the machine learning (ML) community. As a result, most of the recent studies have relied on CNNs, which have achieved higher accuracies compared with traditional machine learning approaches. From prior research, we learned that multi-class image classification models can solve leaf disease identification problems, and multi-label image classification models can solve leaf disease quantification problems (severity analysis). Historically, maize leaf disease severity analysis or quantification has always relied on domain knowledge—that is, experts evaluate the images and train the CNN models based on their knowledge. Here, we propose a unique system that achieves the same objective while excluding input from specialists. This avoids bias and does not rely on a “human in the loop model” for disease quantification. The advantages of the proposed system are many. Notably, the conventional system of maize leaf disease quantification is labor intensive, time-consuming and prone to errors since it lacks standardized diagnosis guidelines. In this work, we present an approach to quantify maize leaf disease based on adaptive thresholding. The experimental work of our study is in three parts. First, we train a wide variety of well-known deep learning models for maize leaf disease classification, then we compare the performance of the deep learning models and finally extract the class activation heatmaps from the prediction layers of the CNN models. Second, we develop an adaptive thresholding technique that automatically extracts the regions of interest from the class activation maps without any prior knowledge. Lastly, we use these regions of interest to estimate image leaf disease severity. Experimental results show that transfer learning approaches can classify maize leaf diseases with up to 99% accuracy. With a high quantification accuracy, our proposed adaptive thresholding method for CNN class activation maps can be a valuable contribution to quantifying maize leaf diseases without relying on domain knowledge. |
| first_indexed | 2024-03-10T03:06:08Z |
| format | Article |
| id | doaj.art-e81233f38cb1488589d1ff64e370c78a |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T03:06:08Z |
| publishDate | 2022-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-e81233f38cb1488589d1ff64e370c78a2023-11-23T12:39:07ZengMDPI AGApplied Sciences2076-34172022-08-011217841210.3390/app12178412Adaptive Thresholding of CNN Features for Maize Leaf Disease Classification and Severity EstimationHarry Dzingai Mafukidze0Godliver Owomugisha1Daniel Otim2Action Nechibvute3Cloud Nyamhere4Felix Mazunga5Department of Applied Physics and Telecommunications, Midlands State University, Senga Road, Gweru P Bag 9055, ZimbabweFaculty of Engineering, Busitema University, Tororo P.O. Box 236, UgandaFaculty of Engineering, Busitema University, Tororo P.O. Box 236, UgandaDepartment of Applied Physics and Telecommunications, Midlands State University, Senga Road, Gweru P Bag 9055, ZimbabweDepartment of Applied Physics and Telecommunications, Midlands State University, Senga Road, Gweru P Bag 9055, ZimbabweDepartment of Applied Physics and Telecommunications, Midlands State University, Senga Road, Gweru P Bag 9055, ZimbabweConvolutional neural networks (CNNs) are the gold standard in the machine learning (ML) community. As a result, most of the recent studies have relied on CNNs, which have achieved higher accuracies compared with traditional machine learning approaches. From prior research, we learned that multi-class image classification models can solve leaf disease identification problems, and multi-label image classification models can solve leaf disease quantification problems (severity analysis). Historically, maize leaf disease severity analysis or quantification has always relied on domain knowledge—that is, experts evaluate the images and train the CNN models based on their knowledge. Here, we propose a unique system that achieves the same objective while excluding input from specialists. This avoids bias and does not rely on a “human in the loop model” for disease quantification. The advantages of the proposed system are many. Notably, the conventional system of maize leaf disease quantification is labor intensive, time-consuming and prone to errors since it lacks standardized diagnosis guidelines. In this work, we present an approach to quantify maize leaf disease based on adaptive thresholding. The experimental work of our study is in three parts. First, we train a wide variety of well-known deep learning models for maize leaf disease classification, then we compare the performance of the deep learning models and finally extract the class activation heatmaps from the prediction layers of the CNN models. Second, we develop an adaptive thresholding technique that automatically extracts the regions of interest from the class activation maps without any prior knowledge. Lastly, we use these regions of interest to estimate image leaf disease severity. Experimental results show that transfer learning approaches can classify maize leaf diseases with up to 99% accuracy. With a high quantification accuracy, our proposed adaptive thresholding method for CNN class activation maps can be a valuable contribution to quantifying maize leaf diseases without relying on domain knowledge.https://www.mdpi.com/2076-3417/12/17/8412CNNtransfer learningclass activation heatmapadaptive thresholding |
| spellingShingle | Harry Dzingai Mafukidze Godliver Owomugisha Daniel Otim Action Nechibvute Cloud Nyamhere Felix Mazunga Adaptive Thresholding of CNN Features for Maize Leaf Disease Classification and Severity Estimation Applied Sciences CNN transfer learning class activation heatmap adaptive thresholding |
| title | Adaptive Thresholding of CNN Features for Maize Leaf Disease Classification and Severity Estimation |
| title_full | Adaptive Thresholding of CNN Features for Maize Leaf Disease Classification and Severity Estimation |
| title_fullStr | Adaptive Thresholding of CNN Features for Maize Leaf Disease Classification and Severity Estimation |
| title_full_unstemmed | Adaptive Thresholding of CNN Features for Maize Leaf Disease Classification and Severity Estimation |
| title_short | Adaptive Thresholding of CNN Features for Maize Leaf Disease Classification and Severity Estimation |
| title_sort | adaptive thresholding of cnn features for maize leaf disease classification and severity estimation |
| topic | CNN transfer learning class activation heatmap adaptive thresholding |
| url | https://www.mdpi.com/2076-3417/12/17/8412 |
| work_keys_str_mv | AT harrydzingaimafukidze adaptivethresholdingofcnnfeaturesformaizeleafdiseaseclassificationandseverityestimation AT godliverowomugisha adaptivethresholdingofcnnfeaturesformaizeleafdiseaseclassificationandseverityestimation AT danielotim adaptivethresholdingofcnnfeaturesformaizeleafdiseaseclassificationandseverityestimation AT actionnechibvute adaptivethresholdingofcnnfeaturesformaizeleafdiseaseclassificationandseverityestimation AT cloudnyamhere adaptivethresholdingofcnnfeaturesformaizeleafdiseaseclassificationandseverityestimation AT felixmazunga adaptivethresholdingofcnnfeaturesformaizeleafdiseaseclassificationandseverityestimation |