Side-match Prediction Scheme in Data Hiding

This paper presents a novel data-hiding scheme based on side-match vector quantization (SMVQ). While encoding a block of an image, a traditional SMVQ data-hiding scheme only considers the neighboring pixels of the block to be encoded to produce a state codebook, and then uses the index of the closest codeword in the state codebook to represent the block. The major problem of the traditional image data-hiding method for SMVQ is image distortion, which subsequently affects the image quality of the stego-image and restricts the embedding capacity. Our scheme utilizes all the pixels of the neighboring blocks to predict whether the pixels of the block to be encoded belong to a monotonic increasing/decreasing pattern, computes the pixels of the block, uses them to generate a state-codebook, and produces a high quality image based on SMVQ step by step. Secret data is then embedded into the image. According to the experimental results, the proposed scheme has good visual quality of stego-image and large embedding capacity.