Image watermarking based on wavelet transform using threshold selection

In this paper, we propose an improved method of multiresolution watermarking scheme based on wavelet transform tuning threshold selection for digital images. To prove the robustness of our watermarking, Gaussian random noise is added to the wavelet coefficients at middle and high frequency band of the discrete wavelet transform. Instead of previous method, the proposed method that allows efficient identification of the embedded a bank of watermarking key without prior notice of it. The proposed detection method aims at ensuring the maximum exactitude in the detection of the owner identification key and minimizing the number of false detection. Simulation results show that the proposed method is robust against filtering and halftoning for detection.