SE-CapsNet: Automated evaluation of plant disease severity based on feature extraction through Squeeze and Excitation (SE) networks and Capsule networks
Diseases in plants have an adverse effect on the quantity of the overall food production as well as the quality of the yield. Early detection, diagnosis and treatment can greatly reduce losses, both economic and ecological, i.e. reduction in the use of agrochemicals due to timely detection of the di...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2021-12-01
|
| Series: | Kuwait Journal of Science |
| Subjects: | |
| Online Access: | https://journalskuwait.org/kjs/index.php/KJS/article/view/10586 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Diseases in plants have an adverse effect on the quantity of the overall food production as well as the quality of the yield. Early detection, diagnosis and treatment can greatly reduce losses, both economic and ecological, i.e. reduction in the use of agrochemicals due to timely detection of the disease, would help in mitigating the environmental impact. In this paper, the authors have proposed an improved feature computation approach based on Squeeze and Excitation Networks, before processing by the original Capsule Networks (CapsNet) for classification. Two SE networks, one based on AlexNet and another on ResNet have been combined with Capsule Networks, for estimating the disease severity in plants. Leaf images for the devastating Late Blight disease occurring in the Tomato crop have been utilized from the PlantVillage dataset. The images, divided into four severity stages i.e. healthy, early, middle and end, are downscaled, enhanced and given as input to the SE networks. The feature maps generated from the two networks are separately given as input to the Capsule Network for classification and their performances are compared with the original CapsNet, on two image sizes 32X32 and 64X64. SE-Alex-CapsNet achieves the highest accuracy of 92.76% and SE-Res-CapsNet achieves the highest accuracy of 94.4% with 64X64 image size, as compared to CapsNet that results in 85.53% accuracy. On the basis of the performances, the proposed techniques can be exploited for disease severity assessment in other crops as well and can be extended to other areas of applications such as plant species classification, weed identification etc. |
|---|---|
| ISSN: | 2307-4108 2307-4116 |