A Huffman Code Based Image Steganography Technique

We present here a novel steganographic method based on Huffman coding and the least significant bit substitution in order to provide high embedding capacity, a strong security and imperceptible visual quality to secret message. Every eight bits of the secret image are first encoded by building a Huffman tree. After that those encoded bits of secret image are divided into 4 groups. Each part has a decimal value between 0 to 3. These decimal values determine the location where to embed the message in a particular pixel of cover image. To embed the message we just put a one in the corresponding location in a pixel of the cover image which identified by the decimal values of the secret image. Since Huffman Table reduces the size of the original image, an attacker cannot easily recover from the stego image those fine details of the original image that would enable him to mount a reliable attack. We have got comparable visual quality as the Peak Signal to Noise Ratio values lie between 30 dB to 31 dB.