Towards a universal multiresolution-based perceptual model

Following a recently introduced perceptual model for balanced multiwavelets, we outline, in this paper, an extension of our previous work and propose a new perceptual model for scalar wavelets. The proposed model is derived using multiresolution domain extensions of our previous scheme. Unlike existing models, the proposed one depends only on the image activity and not the filter sets used by the transform. The perceptual redundancy, present in the image, is efficiently quantified through a just-noticeable distortion (JND) profile. In this model, a visibility threshold of distortion is assigned to each wavelet subband coefficient. Therefore, perceptually insignificant subband components can be clearly discriminated from perceptually significant ones. For instance, this discrimination can be constructively used to achieve the imperceptibility requirement often encountered in watermarking and data hiding applications. Furthermore, we illustrate, through simulation, the ability of the proposed model to efficiently capture the salient features of the underlying image regardless of the wavelet filters being used.