- Version
- Download 18
- File Size 517.94 KB
- File Count 1
- Create Date June 7, 2022
- Last Updated June 7, 2022
Mitigating Disease Infestation Issues of Some Selected Plant Through Early Detection Using Convolutional Neural Network
ABSTRACT
Food security has become a lingering issue that has garnered the interest of farmers, government and researchers from various fields. Factors such as weather, soil degradation, civil unrest, disease infestation, among others have been identified to be the major threats to food security. While other factors are being addressed by governments, economists and researchers; the issue of disease infestation is yet to be fully addressed; even though microbiologists and plant scientists are working with farmers to ensure solutions to disease infestation issues are addressed, advancements in machine and deep learning have proven to also address the issue. To contribute to this, this research aims to propose a solution to plant disease infestation through early detection of plant diseases. To achieve this, a convolutional neural network (ConvNet) with 10 layers was used to train 31,813 images of healthy and infected Tomato, Potato and Pepper bell leaves. The result shows an accuracy of 89.7%, which is a relative improvement on existing systems.
INDEX TERMS ConvNet, Deep learning, Leaves, Pepper bell, Potato, Tomato