Digital Image Watermarking Algorithm Based On Genetic Algorithm In DCT Domain

With the progress of society and technology and the development of digital technology and network technology, digital image, digital audio, digital video and other multimedia digital works have been widely used in the government, military, hospitals, schools and other important institutions. However, due to the transparency, easy operation of the digital multimedia works and overly complex network environment for the process of transmission and use bring security risks. Tampering, counterfeiting, and piracy become relatively easy, once these works are maliciously copied or tampered, it will cause important information leakage or error, and significant economic losses. Therefore, how to better protect the copyright of multimedia digital products and information security has become a major challenge we are facing. The digital watermarking can effectively solve these problems.In this thesis, a digital image watermark algorithm based on genetic algorithm in DCT domain is studied. Including two parts:(1)A digital watermarking algorithm based on Arnold scrambling and DCT domain is studied, using Arnold transform and DCT to achieve a better security, robustness and invisibility of the digital image watermarking algorithm. It encrypts watermarking images by applying the periodicity of Arnold transform and embeds watermarking information by changing DCT domain coefficients of the original image.(2)An optimization algorithm on embedding location of digital watermarking based on genetic algorithm is proposed, which improves the traditional method of artificial selecting embedding positions of watermarking. By using global searching ability and intelligence of the genetic algorithm, it uses the peak signal to noise ratio(PSNR) of image as adaptation function and optimizes population individuals(the embedding location) though evolution operation.Experiments prove that the digital watermarking has good invisibility and robustness, because of Arnold transform and selection of embedding location of watermarking, that is achieving the twice encryption, the digital watermarking has strong safety. In addition, after it embeds the watermarking in the embedding position which is optimized though the genetic algorithm, the watermarking image in security and robustness remains good, and the invisibility of it has been obviously improved. Experimental results are better.