An Efficient VQ-Based Data Hiding Scheme Using Voronoi Clustering

In this paper, we propose a vector quantization (VQ) -based information hiding scheme that cluster the VQ codewords according the codewordspsila relation on Voronoi Diagram (VD). The clustering method is fast and easy to implement when projecting the codewords into VD space. The hiding capacities can be adaptive adjusted just modifying the threshold value which didnpsilat need reorganizing the codebook sequence. Besides, the image quality is acceptable when raising the embedding capacity. Experiments demonstrate that the method can efficiently hiding the secret data into VQ indexes by the fast hiding procedure, high hiding capacity and good image quality.