Research On Color Image Digital Watermarking Algorithm Based On Discrete Cosine Transform

Nowadays,as a key channel for information exchange and transmission,the Internet technology is increasingly popular.However,due to the openness of Internet information,a series of problems,such as piracy,infringement and tampering arise when various digital products,are transmitted online.Therefore,copyright protection has attracted more attention from scholars.As one of the important approaches and necessary technologies of copyright protection,digital watermarking has made great progress in the past twenty years.The copyright protection under the new environment of 5G network has been considered deeply in this paper.Regarding the copyright protection of color digital images,four new high-performance color image watermarking algorithms are proposed to break through the limitations in many aspects of traditional binary or grayscale watermarking algorithm.The results of a large number of simulation experiments and the comparisons with other advanced algorithms show that the proposed watermarking algorithms have superiority of various performances in invisibility,robustness,real-time,watermark embedding rate,and so on.The main research contents of this paper are as follows:(1)Exploring the relationship between the obtained coefficients of pixel blocks after discrete Cosine transform(Discrete Cosine Transform,DCT),and using different quantization steps to quantify the different coefficients to complete watermark embedding,a color image blind watermarking algorithm based on variable quantization steps in DCT domain are proposed.This algorithm breaks through the limitation of using the same quantization step for information hiding,which effectively improves the invisibility of watermark and solves the problem of poor watermark invisibility caused by the large modification range of pixels of the host image.(2)Analyzing the relationship between the specified coefficients of the frequency-domain coefficient matrix,a new blind image watermarking algorithm based on DCT is designed,which taking into account the performance of watermark's invisibility,robustness and embedding capacity and so on.In this algorithm,DCT is first applied to the selected pixel block,then part of the middle-frequency coefficients are selected,and the selected middle-frequency coefficients is modified by certain rules to complete the embedding and blind extraction of digital watermark.This algorithm balances the performance of the algorithm effectively,which not only has good watermark invisibility,but also has strong robustness,and it is suitable for the occasion of copyright protection of color digital images.(3)Taking full advantage of the energy aggregation property of pixel block,the inter-layer correlation of color image and the spatial property of frequency-domain transformation,a blind image watermarking algorithm combining spatial domain and frequency domain is proposed.In the process of watermark embedding,the obtained direct current(Direct Current,DC)coefficient matrix in spatial domain is partitioned firstly,then the DC coefficient of discrete Hartley transform(Discrete Hartley Transform,DHT)is calculated,and different quantization steps are used to quantify it.The above operation makes the energy of pixel blocks more concentrated,while guarantees the robustness and invisibility,and also improves the real-time performance of the algorithm,which effectively solves the problem of slow running speed of large-capacity color image watermarking algorithm.(4)Combining the advantages of the spatial-domain watermarking algorithm and frequency-domain one,the color digital watermark is embedded and blind extracted in the fusion domain,a new high-capacity spatial-domain image watermarking algorithm is proposed.According to the unique characteristics of DC coefficient of DCT and the correlation principle of DC coefficient of adjacent pixels blocks,different quantization steps are selected to modify the DC coefficient of three channels,and then the watermark embedding and blind extraction are completed in the spatial domain,without the need of real DCT.On the premise of satisfying the other performance of the algorithm,this algorithm improves the embedding capacity of watermark,and solves the problem of insufficient embedding capacity caused by the large amount of information of color image watermark.Through the above research,on the premise of guaranteeing the high watermark invisibility,this paper designs the new algorithm of color digital image watermarking which has strong robustness,high efficiency,and high embedding capacity.This algorithm breaks through the bottleneck of weak robustness of color image blind watermarking technology,solves the time-consuming problem of frequency-domain watermarking technology,and overcomes the disadvantage of small embedding capacity of watermarking algorithm,which can effectively protect the copyright of color images and has important theoretical value and application value.