PLANT LEAF DISEASE DETECTION USING IMAGE PROCESSING AND DEEP LEARNING

PLANT LEAF DISEASE DETECTION USING IMAGE PROCESSING AND DEEP LEARNING

R Shivani1, Anusha N2, Cleven B3, Manoj N 4, Nischitha V5, Namitha K6, Syed Abrar Ahmed7
1Electronics and Communication student from Presidency University 2 Electronics and Communication student from Presidency University 3Electronics and Communication student from Presidency University 4Electronics and Communication student from Presidency University 5Electronics and Communication student from Presidency University 6Electronics and Communication student from Presidency University

7Assistant Professor in Electronics and Communication Department from Presidency University

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract -Deep learning and image processing are being combined in this study to detect plant diseases, which is important for sustainable agriculture. For reliable classification, it uses ensemble learning, combining deep learning (ResNet-based CNN), Random Forest, and Support Vector Machine algorithms. Using Fastai for dataset preprocessing entails augmentation and scaling for better generalization. High accuracy is achieved by the CNN following transfer learning. For our project's live presentation, we used USN Camera. User-friendly interaction is made possible by Tkinter deployment, which permits real-time picture classification from webcams or uploaded photos.

Instantaneous health status and disease probability are provided by the interface. Keyboard shortcuts speed up the process of evaluating live feeds and aid in making decisions. All things considered, this initiative provides an effective way to identify plant diseases, supporting agricultural sustainability and food security.

Key Words: Deep Learning, CNN, Image Processing, Dataset, Decision making, Ensemble learning.