| Autor: |
Thangaraj, P., Krishnakumar, B., Kousalya, K., Mohana, R. S., Kumar, D. Sanjay, Rithik, M., Prasanth, S. |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 2023, Vol. 2764 Issue 1, p1-8, 8p |
| Abstrakt: |
In agriculture the plants are affected by many kinds of diseases which are caused by bacteria, fungus and viruses, identification of these diseases is very essential to control the crop yield and quality. These kind of plant diseases will affect and greatly reduce the yield of the crops. These kinds of diseases can be identified by humans, but classifying the type of disease manually and taking necessary action is harder. Convolutional Neural Network (CNN) based assessments are widely used in the diagnosis and classification of Rice leaf disease. The basic CNN has very low performance for rotated, tilted, and irregular orientation images. So, we are proposing a deep learning framework known as Inception-ResNet-v2 for detecting Rice leaf disease from the image of rice leaf. Jupyter Notebook, Keras, and Tensorflow are used as an implementation tool. The proposed model is applied to the rice leaf disease dataset collected from the Kaggle repository. The system aims at reducing deep learning architecture complexity in classifying Rice leaf disease. The proposed system shows the best validation accuracy of 82% on the rice leaf disease dataset. Our proposed system outperforms some existing state-of-the-art. The image processing time is at an average of 5s. Hence the system proposed can be used by the farmers to classify and take necessary actions to avoid these kinds of Rice leaf disease [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
