Improvisation of security in image steganography using DWT, Huffman encoding & RC4 based LSB embedding

Information security is one of the plausible areas of research now-a-days. Steganography plays an effervescent role in the information security. It aims at concealing sensitive information within a host such that the information existence remains confidential. This paper implements image steganography with an objective of raising the security to a higher level while allowing maximum amount of data to be hidden inside. To achieve this, initially the cover image is transformed from spatial domain to frequency domain using 2-D Discrete Wavelet Transform (DWT). The wavelet transformed image allows embedding inside high frequency regions which maintains image quality by shielding the wavelet coefficients in the low frequency region. The secret image bits are compressed using Huffman encoding to increase the capacity of the system to uphold the data. Instead of sequential LSB embedding, RC4 algorithm is used that randomizes the order of embedding secret image message bits inside the cover image. This stochastic embedding of data increases the security as the unauthorized people won't be able to extract the secret image message. A robust steganographic is achieved by using this multi-tier architecture that enhances the security and capacity of the system while restoring the quality.