A Logarithmic Quantization Index Modulation Data Hiding Using the Wavelet Transform

Abstract

Conventional quantization-based data hiding algorithms used uniform quantization. This scheme may be easily estimated by averaging on a set of embedded signals. Furthermore, by uniform quantization, the perceptual characteristics of the original signal are not considered and the watermark energy is distributed uniformly within the original signal, which introduces visual distortions in some parts of it. Therefore, we introduce a logarithmic quantization-based data hiding method based on the visual model by using the wavelet transform that takes advantage of the properties of Watson's visual model and logarithmic quantization index modulation (LQIM). Its improved robustness is due to embedding in the high energy blocks of original image and by applying the logarithmic scheme. In the detection scheme, we model the wavelet coefficients of image by the generalized Gaussian distribution (GGD). Under this assumption, the bit error probability of proposed method is analytically calculated. Performance of the proposed method is analyzed and verified by simulation. Results of experiments demonstrate the imperceptibility of the proposed method and its robustness.