A New Singular Value Decomposition Watermarking Algorithm Based on Directionlet

A new singular value decomposition watermarking algorithm based on Directionlet was proposed. Singular Value Decomposition (SVD) is able to resist various geometric attacks when images are subjected to small disturbances. At first, the transform direction of Directionlet transform is positioned by using canny operate detection and Hough line detection technology; Then, selecting the blue region of the color original image to perform the Directionlet transform and where the first frequency subband is obtained to carry out SVD; Finally, the binary watermark image after performing SVD is embedded the singular matrix of the first frequent subband image. When extracting the watermark, we compare the difference between original image and watermarked image to obtain the binary watermark. Experimental results prove that the embedded watermarking color images has a good resistance against to common attacks such as cropping, rotation and noise etc.