LAUSR

This paper presents a dual watermarking technique using discrete wavelet transform (DWT), singular value decomposition (SVD) and set partitioning in hierarchical tree (SPIHT). The method uses second level DWT to transform the host image in to different frequency components. Next, we further transform the selected wavelet component via SVD. Before embedding, the logo watermark is secured via Arnold transform and signature watermark are encoded by hamming code. Finally, we embed both encoded watermarks into the transformed host image via an embedding approach. The watermarked image is further compress by SPIHT scheme along with the location key. With our scheme, maximum PSNR, NC and SSIM are obtained as 36.97 dB, 0.9965 and 0.9974, respectively. However best obtained BER is zero. Experimental results on various images demonstrate the importance of our scheme and superior to competing methods.