Pixel value differencing steganography is an image steganography that utilizes the difference of the image pixel value to embed the secret message bits. RGB overlapped block-based PVD was introduced by Prasad and Pal which uses the difference value in the pair of RGB color components of a pixel compared to using the difference value of two consecutive pixels. This approach has good performance at increasing capacity especially in images with low pixel variance values. The RGB overlapped block-based PVD algorithm uses a threshold that limits the amount of difference in the color component pairs that are allowed to embed the secret message bits. The use of a global threshold will reduce the potential for optimal capacity utilization of the container image. This study implements an adaptive threshold that uses two different types of thresholds that use the embedding bit limit and the RMSE difference of the pixels before and after the embedding process to the next pixel. This optimization is able to provide a better capacity increase with PSNR degradation from the previous algorithm which is quite low.
Copyrights © 2021