Drone-Based Plant Detection, Tracking & Counting System
🚁 A real-time AI-powered system to detect, track, and count plants from drone-captured agricultural footage using YOLOv8, OpenCV, and a custom object tracker.
📌 Problem Statement
Manual plant counting from aerial drone footage is tedious, inaccurate, and unsustainable at scale.
This system automates plant detection, tracking, and counting using deep learning and computer vision — making agriculture monitoring smarter and faster.
⚙️ Features
✅ Detect plants from drone video using YOLOv8
✅ Assign unique IDs to each plant using a lightweight centroid tracker
✅ Count plants crossing a virtual line
✅ Real-time FPS and plant count overlays
✅ Save output video with all annotations
✅ Clean, readable UI with white overlays
🛠️ Tech Stack
YOLOv8 – Object detection (Ultralytics)
OpenCV – Image & video processing
cvzone – UI overlays (text, boxes, etc.)
Pandas – Dataframe manipulation
Python – Core implementation
Custom Tracker – Centroid-based multi-object tracker