IP Camera Data

A Node.js application that processes IP camera feeds using TensorFlow.js object detection.

Features

• Connects to multiple IP cameras (up to 4)
• Uses TensorFlow.js with COCO-SSD model to detect objects in camera feeds
• Processes camera feeds in separate worker threads for better performance
• Provides a REST API endpoint to access detection results

Requirements

• Node.js v22 or higher
• NPM or Yarn

Installation

# Clone the repository
git clone https://github.com/yourusername/ip-camera-data.git
cd ip-camera-data

# Install dependencies
npm install

Configuration

Edit the config.js file to set your IP camera URLs and server port:

const config = {
  port: 8080,  // HTTP server port
  CAM1: process.env.CAM1 || 'https://your-camera-url-1',
  CAM2: process.env.CAM2 || 'https://your-camera-url-2',
  CAM3: process.env.CAM3 || 'https://your-camera-url-3',
  CAM4: process.env.CAM4 || 'https://your-camera-url-4',
};

You can also set the camera URLs via environment variables.

Usage

Start the application:

node index.js

Access the camera data via the REST API:

GET http://localhost:8080/camdata

This will return a JSON object with detection results from all cameras:

{
  "CAM1": {
    "image": { "height": 480, "width": 640, "channels": 3 },
    "predictions": [
      { "bbox": [x, y, width, height], "class": "person", "score": 0.8912 },
      ...
    ],
    "timestamp": 1678912345678
  },
  "CAM2": { ... },
  "CAM3": { ... },
  "CAM4": { ... }
}

How It Works

1. The application starts an Express server and creates worker threads for each camera
2. Each worker loads the TensorFlow.js COCO-SSD model for object detection
3. Workers continuously fetch images from the configured camera URLs
4. Object detection is performed on each image
5. Results are sent back to the main thread and stored
6. The REST API endpoint /camdata provides access to the latest detection results

License

ISC