The project has been deployed on Render .
You can visit the deployed website here - https://ranjeetbaraik-imcrypt.onrender.com
imcrypt-preview.mp4
Imcrypt is a full-stack web application that implements image steganography to hide secret text messages within PNG images. The application provides a secure way to store confidential information by embedding it directly into image pixels, making it invisible to the naked eye while maintaining the image's visual integrity.
- 🔐 Secure Text Encoding: Hide any text message inside a PNG image
- 🔍 Message Decoding: Extract hidden messages from encoded images
- 🚫 No Server Storage: Images are processed and immediately deleted for maximum security
- 🌐 User-Friendly Interface: Clean, responsive React-based UI
- ⚡ Real-time Processing: Fast encoding/decoding using optimized algorithms
The application follows a client-server architecture with the following components:
- Single Page Application built with React 18.2.0
- React Router DOM for navigation between encode/decode pages
- Responsive Design with custom CSS styling
- File Upload Interface for image handling
- Real-time Feedback for user interactions
- RESTful API with Express.js framework
- Multer Middleware for handling multipart/form-data file uploads
- CORS Support for cross-origin requests
- Child Process Integration to execute Python scripts
- Temporary File Management with automatic cleanup
- OpenCV for image manipulation and pixel-level operations
- Custom Steganography Algorithm using mathematical patterns
- Lossless Encoding that preserves image quality
- Efficient Decoding with pattern-based extraction
User uploads PNG image + secret message
↓
Multer saves file temporarily with unique filename
↓
Node.js spawns Python child process (encodeScript.py)
↓
Python script uses OpenCV to:
• Load the image as numpy array
• Calculate GCD-based pattern for pixel selection
• Convert message characters to ASCII values
• Embed ASCII values into blue channel (B) of selected pixels
• Save modified image
↓
Server sends encoded image as download
↓
Temporary files are deleted after 3 seconds
User uploads encoded PNG image
↓
Server validates file type (PNG only)
↓
Python script extracts hidden message:
• Uses same GCD pattern to locate embedded pixels
• Reads blue channel values from selected pixels
• Converts ASCII values back to characters
• Reconstructs original message
↓
Server returns decoded message as JSON
↓
Temporary file is immediately deleted
- Pattern Generation: Uses GCD (Greatest Common Divisor) of image dimensions to create a mathematical pattern
- Pixel Selection: Only specific pixels are modified based on the pattern
(i+1 * j+1) % pattern == 0 - Channel Modification: Only the blue (B) channel of RGB pixels is modified to minimize visual impact
- Message Termination: Uses null character (0) to mark end of hidden message
- Capacity: Can hide messages up to the number of selected pixels in the image
- No Server-Side Storage: Images are processed in memory and deleted immediately
- Unique Filenames: Each upload gets a random 8-character identifier
- File Type Validation: Only PNG images are accepted (both client and server-side)
- File Size Limits: 5MB maximum file size to prevent abuse
- Automatic Cleanup: Temporary files are removed within 3 seconds
POST /api/encode- Encode text message into imagePOST /api/decode- Decode hidden message from image
Imcrypt/
├── client/ # React frontend
│ ├── src/components/ # UI components (Navbar, Encode, Decode, etc.)
│ └── public/ # Static assets
├── controllers/ # Business logic
│ ├── encodeController.js # Handles encoding requests
│ ├── decodeController.js # Handles decoding requests
│ ├── encodeScript.py # Python encoding algorithm
│ └── decodeScript.py # Python decoding algorithm
├── routes/ # API route definitions
├── middlewares/ # Multer file upload configuration
└── uploads/ # Temporary file storage
- React.Js
- Node.js
- Javascript
- Python
- OpenCV
- Image Steganography
- Child Process