// COMPUTER VISION
Custom 3,000+ image dataset and YOLO-based detection model for precision agriculture in Chikkamagaluru.
Curated a custom 3,000+ image dataset from coffee plantations in Kalasa, Chikkamagaluru, capturing real-world variation in lighting, occlusion, and plant structure. Trained a YOLO object detection model for multi-class detection of ripe vs. unripe coffee cherries and leaf structures.
Manual inspection of coffee cherry ripeness is labour-intensive and inconsistent across large plantations. Automated detection enables targeted picking, reducing waste and improving yield quality. The primary challenge is building a robust model from scratch without any existing domain-specific datasets.
Name
Kalasa Coffee Plantation Dataset (Custom)
Metadata Matrix Size
3,000+ high-resolution images
Telemetry Source
Field collection — Kalasa, Chikkamagaluru, Karnataka
Pipeline Preprocessing Steps
Model Architecture
YOLOv8 (fine-tuned)
Technology Stack
Ultralytics YOLOv8 + PyTorch
Key Components
Field Data Collection
3,000+ images captured from Kalasa plantation across varied lighting and angles.
Annotation
Multi-class bounding box labeling using Roboflow annotation tool.
Augmentation
Roboflow-based augmentation pipeline: flips, brightness, mosaic.
Fine-tuning
YOLOv8 fine-tuned from COCO weights on custom dataset.
Evaluation
mAP, Precision, Recall evaluated on held-out test split.
Inference
Real-time detection pipeline for plantation field use.
3,000+
Dataset Size
6
Classes
81.0%
mAP
72.4%
Precision
78.7%
Recall
75.4%
F1-Score