Overview

A deep learning image classification system that predicts and categorizes plant diseases from leaf images. The model handles both single-symptom classification and multi-symptom identification on a single leaf image.

---

Approach

1. Image Preprocessing — Built a preprocessing pipeline to prepare leaf images for model input, including resizing, normalization, and augmentation
2. Transfer Learning — Utilized pre-trained CNN architectures (ResNet50 and Xception) as feature extractors, fine-tuning the classification head for the plant disease domain
3. Multi-Label Classification — Extended the model to handle cases where multiple disease symptoms appear on a single leaf image
4. Evaluation — Model performance assessed through standard classification metrics including accuracy, precision, recall, and confusion matrices

---

Tech Stack

• Python, TensorFlow / Keras
• ResNet50, Xception (pre-trained CNN architectures)
• NumPy, Matplotlib