An Efficient Lossless Data Hiding Technique for Palette-Based Images with Capacity Optimization

Recently data hiding over images have drawn tremendous interest, using either lossy or lossless techniques. Although lossy techniques can allow large hiding capacity, host image cannot be recovered with high fidelity. Some applications require exact recovery of the host image as in medicine when personal data are hidden within the medical image. Lossless data hiding techniques suffer from limited capacity as the host image should be kept intact. In this paper a lossless embedding technique is proposed based on image histogram characteristics, zero and peak points are identified and manipulated to embed data. The new technique gives hiding capacity that can reach up to 50% of the host image size for images with large homochromatic regions (cartoons-like). In fact, our study showed that the capacity can vary with appropriate selection of the host image. The capacity is not only affected by the host image size but also by its histogram distribution. The proposed technique showed very good results with cartoon images which have been, previously, considered inappropriate for data hiding because of the high degree of colors uniformity. Simple processing operation is used that does not require high power consumption which is a vital parameter especially in mobile systems. Both the host image and the embedded message are recovered back free of error from the stego image.