Matlab Grayscale Steganography

Requirements

• A program to execute .m scripts such as Matlab or the free alternative GNU Octave.
• Only works on .bmp files
• Only works on grayscale images; although to work with color photos its an easy change to multiply the returned array by 3 (# RGB channels)

Usage

• secret_message.txt: Contains the secret message you wish to embed in the target .bmp (bitmap) image.
• encode.m : The main matlab script that utilizes the Least Significant bit algorithm to slightly alter the color channels of the target .bmp file; padding each pixel with an additional number of bytes that comprise the secret message.
  • Limitation: Size of message (in bits ) cannot exceed the (# of pixels * in your target .bmp photo)* 3.
  • 1 pixel = 8 bits
  • #of pixels = image width * height
  • Outputs: encodedImg.bmp with your hidden message

    encode.m

    % Put in path to the image you want to encode
    c = imread('[YOUR_PATH_HERE]\images\earth.bmp');

    % Put in path to secret_message.txt file you want to put encode in image
    id = fopen('[YOUR_PATH_HERE]\secret_message.txt', 'r');

    .
    .
    .

    % write the new image with the hidden image in it to 's' or disk.
    imwrite(s,'[YOUR_PATH_HERE]\encodedImg.bmp','bmp'); 

• decode.m:
• Gets the least significant bit of each pixel in the encodedImg.bmp and combines it to form a binary number then convert to a string.
  • Outputs secret message in file called: decodedMessage.txt

    decode.m

    % Put in path to the encoded image that was the result of encode.m
    s= imread('[YOUR_PATH_HERE]\encodedImg.bmp');

    .
    .
    .

    % Put in path to where you want to save the decodedMessage.txt file
    id = fopen('[YOUR_PATH_HERE]\decodedMessage.txt','w');

Source

Concepts and Languages Used

• Matlab scripting language because it can easily manipulate raw image data and work with lower-level concepts like binary.Python could have also been used to solve this.
• LSB Algorithm: Changing the least significant bit in a binary number is a replacement that overall doesn't largely affect the overall appearance of an image allowing this to be a perfect encoding algorithm.
• Decoding is simple if you know the algorithm used to encode the picture.