An adaptive watermarking for color images based on venation characteristics

In view of the human visual system(HVS) and by employing the characteristics of multi-resolution analysis of the two-dimension discrete wavelet transform(DWT) and venation characteristics,an adaptive watermark embedding algorithm is proposed in this paper.The wavelet zerotrees are selected carefully and the important wavelet coefficients of the lowest frequency are classified in the algorithm.According to different groups,the watermark was embedded in different ways through the human visual system and the characteristic of texture.The experimental results show that this algorithm is robust against JPEG compression attack,image processing such as noise,filtering,cropping, etc.