Real-time CCTV footage analysis with AI

The program utilizes OpenCV to capture snapshots from the camera feed, which are then analyzed by a pre-trained AI model (YOLOv10n). If the model detects a person in the image, the program automatically sends an alert to a designated Telegram chat.

Requirements

• Docker
• Telegram

How to use it?

Step 1: Create a Configuration File

You need to create a config.json file that contains a list of cameras with their names and RTSP URLs. Below is an example configuration with three cameras:

{
    "cameras": [
        {
            "name": "Driveway",
            "url": "rtsp://172.16.0.10/stream"
        },
        {
            "name": "Backyard",
            "url": "rtsp://172.16.0.11/stream"
        },
        {
            "name": "Door",
            "url": "rtsp://172.16.0.12/stream"
        }
    ]
}

Step 2: Set Up Telegram Bot

1. Create a Bot: Open Telegram and talk to @BotFather. Follow the instructions to create a new bot and get your bot token.
2. Get Chat ID:
   • Send a message to your newly created bot.
   • Open your browser and visit:
     https://api.telegram.org/bot<YOUR_TOKEN>/getUpdates
     Replace <YOUR_TOKEN> with the token you received from @BotFather.
   • You will receive a JSON response. Look for the following part to find your chat ID:

     "chat": {
         "id": 12345678,

Step 3: Start the Docker Container

Run the following command to start the container:

docker run -d --name analyzecctv \
-v /path/to/config.json:/app/config.json \
-e TG_TOKEN=YOUR_TOKEN \
-e TG_CHAT_ID=YOUR_CHAT_ID \
--restart always \
aleksandergpl/analyzecctv

Replace /path/to/config.json with the actual path to your configuration file. Additionally, substitute YOUR_TOKEN and YOUR_CHAT_ID with your bot’s token and chat ID.

Enviromental variables:

• DELAY - Delay between checking the camera feed
• CONF_TRESH - Confidence treshold for the AI model
• TG_TOKEN - Telegram bot token
• TG_CHAT_ID - Telegram chat id