Novel color image steganography using Vector Quantization Codebook Generation methods LBG, TCEVR and KFCG

Steganography is the science of coverted communication in which information is hidden to provide secrecy. In this hiding technique secret information or message is concealed into other innocent data such as text, audio, image etc. The main objective of Steganography is to preserve covert communication among two parties. Steganography plays crucial role in various applications like modern printers, intellectual property rights, network security. Also it is used by intelligence services. As color image requires large hiding capacity, most of the existing steganography schemes are proposed for gray images. To improve hiding capacity in addition to having minimum alterations to host image, the paper propose a technique of hiding secret color image into cover image with same size as secret image. In the proposed method vector quantization (VQ)is applied to secret color image to get compressed data and that compressed data is concealed into host color image using Least Significant Bit (LSB) replacement technique. Experiment was conducted on 15 secret images, each one embedded into 15 cover images. Experimentation carried out using three different VQ methods, like Linde-Buzo-Gray(LBG), TCEVR (Thepade´s Cosine Error Vector Rotation) and KFCG (Kekre´s Fast Codebook Generation). Experiment shows KFCG performs better followed by TCEVR over conventional method like LBG. In all VQ methods high cluster size gives better quality of reconstructed secret.