To search, Click below search items.


All Published Papers Search Service


Digital color image steganography for nonspecific format and secured based on Clustering


Mohamed Tahar Ben Othman, Arshiya Sajid Ansari, Mohammad Sajid Mohammadi


Vol. 19  No. 4  pp. 20-27


We are presenting a more secured and nonspecifically dependent on image formats steganography technique for any bmp or JPEG digital images, based on clustering technique. It is a new color image clustering technique used for robust steganography. It can be defined as the persistence of the embedded steganography under image unrests resulting from attacks. Firstly, the image under consideration is divided into 8x8 blocks and then the clusters of blocks, using the Content Addressable Method (CAM). The cluster of blocks is divided into sub-clusters and Payload data are embedded into next level of cluster i.e. sub clusters. Duplication of the same payload data has been achieved through these sub-clusters which prove the robustness of our proposed steganography technique. Moreover, the distribution rate of image sub-clusters over a cluster it is another parameter that can be tuned to improve this robustness. We used this technique and proved with bmp images in previous works and it showed promised results. In the presented work, this technique has been applied on JPEG image coefficient level. Experimental results indicate that if we have more rate of dispersal of clusters on the image with Block/pixels more uniformly distributed of throughout the clusters, the best robustness of steganography we could obtain. This result comes from the fact that the proposed technique resists to geometric attacks like cropping of stego image.


Steganography, Distribution Rate, Color Images, Content Addressable Method. Cluster Dispersal Rate