Pumpkin Detection & Tracking System

Project Overview:

PumpkinTrack AI is a computer vision solution designed to detect and track pumpkins in real-time using a custom-trained YOLO12 model. Whether used for agricultural automation, yield estimation, or visual monitoring, this system efficiently identifies pumpkins in video feeds, tracks them across frames, and visually labels them with bounding boxes and unique IDs.

It is built for applications in precision agriculture, smart farming, and harvest automation, offering farmers and researchers a lightweight and real-time visual inspection tool for pumpkins.

Features:

• Pumpkin Detection: Accurately identifies pumpkins using a YOLO12 model (best.pt) trained on pumpkin data.

• Object Tracking: Maintains consistent identity (ID) for each detected pumpkin across frames.

• Visual Feedback: Draws bounding boxes, labels, and object centers for better visualization.

• Mouse Coordinate Debugger: Logs real-time cursor position—useful for calibration and frame analysis.

• Video or Webcam Input: Easily adaptable to any video source—field footage or live camera.

Tech Stack:

Real-World Applications:

• Smart Agriculture: Real-time monitoring of pumpkin growth in the field.

• Automated Harvesting Systems: Detect pumpkins for robotic arms or autonomous harvesters.

• Yield Estimation: Track number of pumpkins in a given area over time.

• Research & Development: Use video analysis to study size, distribution, or ripeness patterns.

Conclusion:

PumpkinTrack AI provides a fast, efficient, and customizable framework for detecting and tracking pumpkins using AI. By combining the power of YOLO12 with OpenCV’s real-time video capabilities, this system empowers agricultural engineers, researchers, and farmers to automate and optimize visual pumpkin analysis with ease.