Optimizing the Effect of Cropping and Rotation Attacks on Watermarked Images using Back Propagation Neural Network in DWT Domain

Abstract

Hiding an image in another image is the technique used for copy write protection. In this proposed work, the watermark is inserted into blue plane of the cover image, In this watermark extraction and embedding process, the back propagation neural network in conjunction with biorthogonal wavelets is utilized to improve the efficiency. The performance is tested by normalized correlation coefficient. The imperceptibility of the watermark is tested by cropping and rotation attacks effectively.