🐝 Honeycomb cell detection & management
🎯 Purpose
Automatically detect honeycomb cell content on uploaded frame photos and convert it into usable hive insights (brood/resources balance) for inspection and decision-making.
📹 Demo
🎭 User Story
- As a beekeeper
- I want the app to analyze frame photos and classify comb cells
- So that I can quickly understand brood pattern and food stores without manual cell counting
- And compare this across inspections over time
🔬 What the Web-App Actually Does
Based on the current app architecture and services:
- Frame photo upload triggers background AI jobs (including
detect_cells) - Cell detections are stored per frame side and linked with inspection context
- The app keeps both raw detections and summary percentages
- Detection summary includes key resource/brood categories used in UI and analytics:
- brood
- capped brood
- drone brood
- eggs
- pollen
- nectar
- honey
- empty/other classes (model-dependent)
🚀 Key Benefits
- Fast frame interpretation: Replace manual visual estimation with automatic detection
- Consistent structure: Every frame side gets comparable resource/brood metrics
- Inspection continuity: Results stay tied to inspection snapshots for historical comparison
- Action support: Helps with decisions like feeding, split planning, or closer follow-up
🧭 Typical Workflow
- Upload frame photos from hive/frame view
- Wait for AI processing to complete
- Open frame side and review detections/percentages
- Create inspection snapshots to preserve state over time
- Compare older vs current inspection resource patterns