Visible Watermarking Algorithm
The steps for watermark insertion are discussed below:
• The original image I (one to be watermarked) and the watermark image W are
divided into blocks of size 8*8. (both the images may not be of equal size).
• For each block of the original image I the mean gray value µ
n
is computed.
• The DCT for each block of the original image are found.
• The DCT of watermark image blocks are found out.
• For AC DCT co-efficient of each original image block, variance σ
n
is found out.
• The block means µ
n
scaled to the range 0.1 to 1.0.
• The log of the variance σ
n
are scaled to the range 0.1 to 1.0.
• The image mean µ is found out which the mean of block means µ
n
.
• Let in denote the nth DCT block of original image I, and w
n
denote the nth DCT
block of watermark image W. Denoting the nth block of watermarked image by
i
n
’, we have,
i
n
’ = α
n
.i
n
+ βn w
n
n = 1,2,…………… eqn. (1)
The α
n
and β
n
are classified and Gaussian random numbers are added
where α
n
and β
n
are scaling and embedding factors respectively for each block
computed as described. They are computed using eqn. (2) and eqn. (3).
(if required).