Hornet attack detection
| goal / problem solved | 💢Hornet attacks |
|---|---|
| Status | Groomed |
| billing plan | 4-entrance-counter |
| verticality | vertical |
| concept hierarchy path | apiary/hive/box/frame/frame-side/file |
| Blocking | ⛔Close mechanised hive entrance in case of hornet attack |
| Date | |
| Priority | 18 |
| estimate | 5 |
| impact = priority % | 90 |
| Tags | IoTMLbackend |
Background
Hornets, especially asian hornets, are a menace to the bee colony. Without human intervention, colony can die.
Goal
We want to build hornet detector AI model watching the hive entrance and alert the beekeeper in such cases.
Suggested steps
- take an existing dataset and/or model, like one present in roboflowhttps://universe.roboflow.com/use-case-asian-hornet-detection/asian-hornet-detection-a6ael
- alternatively, collect dataset yourself
- train a model
Integration
- We want to run inference on 10 sec mp4 chunks of videos
- We also want to be able to detect hornets in real-time over webcam, for example as part of yolo v8 with cv2
Alerts
- We need to report detections to ⚠️Alerts
- We also want to affect ⛔Close mechanised hive entrance in case of hornet attack
Relevant links
Here is an example from japanese university:
https://satoh.cs.uec.ac.jp/ja/research/bee/images/bee/hornet.mp4