To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Deep Learning Based Image Processing For Cotton Leaf Disease And Pests Diagnosis: A Case Of Ethiopia |
|
Author |
Azath M., Melese Zekiwos, Balakumar S. |
| Citation |
Vol. 25 No. 8 pp. 93-101 |
|
Abstract |
Cotton is one of economically significant agricultural product in Ethiopia since Agriculture is backbone of this nation economy. It is believed that Ethiopia is one of suitable nation for several cultivated crops including cotton which is also known as ""White Gold"" and ""The King of Fibers"", that is the world's leading natural textile fiber crop and a significant contributor to the oilseed for consumption. Cotton is known to be affected by different biotic and abiotic constraints occurring on the leaf areas which reduces productivity from 80 to 90% and hard to detect with bare eyes. This study focused to develop a model to boosts performance of identification of cotton leaf disease and pests using Deep Learning technique, CNN. To do so, the researcher used common cotton leaf disease and two pests which are bacterial blight, spider mite, and leaf miner. K-fold cross-validation strategy was used to dataset splitting and boosts generalization of the CNN model. For this research 2400 instances (600 images in each class) are used to train the model. This developed model implemented using python version 3.7.3 and the model is trained on the deep learning package called Keras, TensorFlow backed and Jupyter is used as the development environment. This model achieves an accuracy of 96.4% for identifying the four classes of leaf disease and pest in cotton plants. These all show the feasibility of its usage in real-time applications and the potential need for IT-based solutions to support traditional or manual disease and pests identification. |
|
Keywords |
Image Processing, Deep Learning, Cotton Leaf Disease and Pests, CNN, K-fold cross-validation, Keras. |
|
URL |
