Watermarking can be done in spatial and transform
domain. The basic problem in watermarking in spatial
domain is that the watermark is more fragile i.e., more
susceptible to attacks than transform domain. The reason for
choosing the transform domain (DCT and DFT is that the
characteristics of human vision system(HVS) are better
captured by spectral coefficients. For eg. Low frequency
coefficients are perceptually significant, which means
alterations to those components might cause significant
distortion to original image On the other hand, high
frequency coefficients are considered insignificant: thus
,processing techniques such as compression tend to remove
high frequency coefficients aggressively.