PomoloBee

“PomoloBee – Bee Smart Know Your Apple”
PomoloBee is an AI-powered tool that helps farmers estimate apple yield using image or video analysis.
This repository contains the mobile app, backend server, and ML microservice for end-to-end deployment.

PomoloBee Logo


Documentation

✅ For full documentation, see the GitHub Pages Portal


Project Definition PomoloBee Bee Smart Know Your Apple

Goal

Develop an Android app (Kotlin + Android Studio) that allows farmers to estimate apple harvest yield using AI-based video or image analysis. The system will use a cloud-based backend (VPS) to process data and provide accurate results.

PomoloBee App Android Fruit Detection App

PomoloBee is an Android app for image-based fruit detection in orchards.
It lets users capture or upload a photo, tag it with field and row location, and analyze it locally or remotely using a Django + ML backend.

  • Works offline with local model
  • Supports field/row selection via interactive SVG maps
  • Uses Jetpack DataStore, custom config sync (cloud or local), and stores photos in SAF-accessible folders.

️ App Screenshots


Data Flow in PomoloBee

The following diagram illustrates the interaction between the PomoloBee App, Django Backend, and ML Processing Service.


Features Functionalities

1 Mobile App Frontend Android

📱 User Actions:
Record or Upload Video – User walks through the orchard while capturing video.
Take a Picture – Alternative to video for quick analysis.
Mark Orchard Parameters – Farmer defines start and end of a tree row.
Enter Field Data – Total orchard row length, tree count, sample apple size.
Receive Harvest Estimate – Displays apple count and estimated yield.
Local AI Estimation (NEW – Phase 2) – Farmers can analyze images offline using on-device AI.
Manual Override of AI Results (NEW – Phase 2) – Farmers can manually adjust apple count & weight.
Historical Tracking (NEW – Phase 3) – Compare past yield estimations.

🔧 Tech Stack:

  • Language: Kotlin
  • Networking: Retrofit (API calls to VPS)
  • UI: Jetpack Compose
  • Local AI Processing: OpenCV + TensorFlow Lite (Phase 2)

2 Cloud Backend VPS Django or Flask API

🌐 Server Responsibilities:
Receive video/image uploads from the app
Extract key frames from video
Apple Detection & Counting (AI Model)
Calculate Total Yield Estimate
Return Results to the App


Comments

Leave a Reply

Your email address will not be published. Required fields are marked *