

🌾🤖 Excited to share AgriVision AI — a professional Agricultural Object Detection, Tracking & Crop Intelligence System built end-to-end in Python!
It watches your field through a webcam, uploaded field footage, or live streams — detecting and continuously tracking crops, fruits, weeds, pests, and machinery with persistent tracking IDs, movement trajectories, and an agronomic event engine.
🧠 Full pipeline:
YOLO detection → ByteTrack / BoT-SORT tracking → object memory → event engine → analytics dashboard → natural-language querying
✨ Highlights:
✅ Real-time multi-object tracking with unique persistent IDs (no double-counting)
✅ Temporal analysis: first/last seen, duration, trajectories, growth proxies
✅ Event engine: weed detected 🌿, pest alert 🐛, missing plants, fruit count increases, density anomalies, tracking lost/recovered
✅ Crop intelligence: density, row occupancy, fruit maturity stages, abnormal-appearance flags
✅ Ask in plain English: “How many fruits were detected?” — answered from structured tracking data, never hallucinated
✅ Instant demo mode: a simulated field auto-starts on launch — zero setup, zero API keys
✅ Built to be fine-tuned on your own agricultural dataset
🛠 Tech stack: Python · OpenCV · Ultralytics YOLO · ByteTrack/BoT-SORT · Streamlit · Plotly · Pandas
Built with an honesty-first design: real model predictions are clearly separated from simulations, and unavailable capabilities are reported as such — never faked.
#ComputerVision #AgriTech #PrecisionAgriculture #SmartFarming #YOLO #ObjectDetection #MachineLearning #DeepLearning #Python #Streamlit #OpenCV #AI
3 Septiembre 2026 à 13h29
Ver en LinkedIn →