This is a simple spread-spectrum based informed watermarking implementation in MATLAB. It is based on the paper Ingemar J. Cox, Joe Kilian, F. Thomson Leighton, and Talal Shamoon, "Secure Spread Spectrum Watermarking for Multimedia," IEEE Trans. on Image Processing, Dec. 1997.
Description: Embeds n watermark sequence into an image file, outputs watermarked image file and writes the watermark sequence into a file with extension '.seq'. If the watermarked image results to the same image with original one after spectrum operations (this is highly related to the n size of watermark sequence and alpha value), it warns the user outputting related message.
Uses: GenerateGaussianSequence.m
Arguments: origfile, markedfile, wmfile, n, alpha
Returns: void
Example Run:
>> SpreadSpectrumEmbed('fruits.bmp', 'fruits_w.bmp', 'wm.seq', 1000, 2)
Description: Extracts watermark sequence (if there exists) from a suspected image file, writes extracted watermark sequence in a file with extension '.extracted', reads the original watermark sequence, computes the similarities of randomly generated n-1 training watermark sequence along with the original one, plots these similarities into a figure file with extension '.png'.
Uses: GenerateGaussianSequence.m, ComputeSimilarity.m
Arguments: suspfile, origfile, wmfile, n, alpha
Returns: void
Example Run:
>> SpreadSpectrumExtract('fruits_w.bmp', 'fruits.bmp', 'wm.seq', 1000, 2)
Description: Generates n sequence of normally distributed pseudo random numbers of the numbers {0, 1}. This is an auxiliary script which is called inside embbedding and extracting programs.
Arguments: n
Returns: watermark_sequence
Example Call:
..
.. watermark_seq = GenerateGaussianSequence(n);
..
Description: Computes the similarities between original watermark sequence and n-1 randomly generated training watermark sequences with the extracted watermark from suspected image.
Arguments: origwm, extwm, n
Returns: origwm_similarity, similarities
Example Call:
..
.. [origwm_sim, similarities] = ComputeSimilarity(origwm, w, n);
..
Working directory for Matlab interpreter must be same for all referenced argument files -- which means if you do not want to specify file paths, just copy the test files into the src/ directory and run embedding and extracting.