Skip to content

Decoding

Ryan Kane edited this page May 3, 2015 · 3 revisions

Steps for Decoding

  1. Partition the image into 8x8 pixel blocks
  2. For each of the 8x8 test blocks Zr’ two things will happen:
  3. Set the LSBs to zero (Hash)
    1. Combine with the entire image X to produce a bitstream of the pixels.
    2. Hash the combined bitstream of pixels together using MD5 to produce Pr.
  4. Extract LSBs this will produce Cr”: (Watermark)
    1. Decrypt Cr” with the private-key to produce the decrypted hash.
  5. The hash (a) is XORed with the watermark (b) to produce Wr”.
  6. Test the difference between the extracted watermark and the original watermark.
  7. Mark the block if the watermarks do not match.
  8. Recombine all the blocks Zr’ to reproduce the original image.

Decoding

Citations

  1. J. Pan et al., "Fragile watermarks" in Intelligent Watermarking Techniques, River Edge, N.J.: World Scientific, 2004, ch. 5, sec. 5.2.2, pp. 18-19.

Clone this wiki locally