This project applies advanced computer vision and deep learning techniques to detect defects in tea leaves. It leverages state-of-the-art models for leaf detection, yield quality prediction, and disease classification, aiming to automate tea quality assessment and assist in crop monitoring.
- Leaf_detection.ipynb – Detects tea leaves in images using YOLOv8.
- Yield_quality_detection.ipynb – Predicts yield quality from tea leaf images using YOLOv8.
- Leaf_disease_detection.ipynb – Classifies tea leaf diseases using VGG16 (work in progress).
- Fertilizer_Deficiency_Detection.ipynb – Identifies nutrient deficiencies in tea leaves (future work).
YOLOv8 is the latest iteration of the YOLO object detection family, designed for fast and accurate real-time detection. Unlike traditional methods, YOLO predicts bounding boxes and class probabilities in a single forward pass, making it highly efficient.
Use in this project:
- Detecting tea leaves in images.
- Predicting yield quality based on visual leaf features.
Advantages: High speed, real-time performance, and robust detection accuracy.
VGG16 is a deep convolutional neural network with 16 layers, developed by the Visual Geometry Group at Oxford. It uses small 3×3 convolutional filters stacked deeply, followed by pooling and fully connected layers for image classification.
Use in this project:
- Classifying healthy vs diseased tea leaves.
- Detecting specific disease patterns for early intervention.
Advantages: Simple yet powerful architecture, effective for image classification tasks with limited data.
- Automated leaf detection for dataset preparation and plantation monitoring.
- Yield and quality assessment to optimize tea production.
- Early disease detection to reduce crop losses.
- Potential extension to nutrient deficiency detection and other crop health monitoring.