Telegram Bot for Trash Classification and Reward System

Overview

This bot promotes environmental awareness by allowing users to classify trash, submit photos and videos, and earn points for responsible waste disposal. Administrators review submissions, and a leaderboard tracks top contributors.

Features

• Trash Classification: Classifies trash using a TensorFlow Lite model. Model can analysie class_names as cardboard, glass, metal, paper, plastic, trash.
• Geolocation Tracking: Users can send their location with waste images.
• Submission Approval: Admins approve or reject user submissions via inline buttons.
• Point System: Users earn points for approved submissions, with a leaderboard to track top contributors.
• Database Integration: Tracks user points and submissions using SQLite.
• Geolacation of trash places

Requirements

• Python check the runtime.txt
• Telegram Bot API Token
• TensorFlow Lite Runtime
• PIL (Pillow)
• SQLite3
• Nest Asyncio
• Python-Telegram-Bot
• folium

Installation

1. Clone the repository:

   git clone https://github.com/Eraly-ml/eco_AI.git
   cd repository

2. Install dependencies:

   pip install -r requirements.txt

3. Set up environment variables:

   export TELEGRAM_BOT_TOKEN=your_token_here
   export ADMIN_CHAT_ID=your_admin_chat_id

4. Run the bot:

   python eco_tg_bot.py

Usage

User Commands

• /start: Register yourself in the system.
• /my: View your current points.
• /stat <user_id>: Check the points of a specific user.
• /top: View the leaderboard.
• /myid: Retrieve your Telegram user ID.

Submission Workflow

1. Send Location: Share the location of the waste.
2. Send Photo: Upload a photo of the waste.
3. Send Video: Upload a video showing proper disposal.
4. Admin Review: Submission is reviewed by an admin.
5. Earn Points: Approved submissions earn points.

Admin Controls

• Approve or reject submissions directly from the chat interface.
• Review photos and videos attached to submissions.

File Structure

repository/
├── eco_tg_bot.py              # Main bot script
├── model.tflite        # TensorFlow Lite model
├── requirements.txt    # Python dependencies
├── bot_database.db     # SQLite database (auto-generated)
└── README.md           # Project documentation

Technologies Used

• Telegram Bot API: For user interaction.
• TensorFlow Lite: For lightweight trash classification.
• SQLite: To store user data and track points.
• Pillow: For image preprocessing.

How It Works

1. Users interact with the bot via commands and share location, photos, and videos.
2. The bot classifies the photo of trash by using a pre-trained MobileNet TensorFlow Lite model.
3. Submissions are reviewed by an admin and sendet to private chat of admins.
4. Approved submissions earn points and update the leaderboard.

information

User data is used exclusively for bot functionality and to evaluate user contributions.

Contributing

Currently, contribution guidelines are informal. Feel free to reach out via Telegram: @eralyf

License

This project is licensed under the Apache 2.0. License. See the LICENSE file for details, also we use liberies like python-telegram-bot from telegram

Contact

For questions or contact the repository owner use telegram @eralyf

Support the Developer

If you'd like to support my work, consider subscribing to my Telegram channel:
🔗 EraJustML

Your support means a lot!