A Secure DWT based Approach for Image Steganography to Enhance Embedding Capacity and Robustness

Abstract

Image Steganography is one of the essential fields for secure data transmission which hide a secret message inside an array of digital image so that the human examiner cannot perceive it. However, issues such as limited embedding capacity and susceptibility to attacks require more robust and efficient techniques. In this research work, a secure DWT based image steganography approach is proposed to enhance embedding capacity and robustness. Utilizing the multi-resolution features of DWT the method decomposes the cover image into frequency sub-bands and achieves optimal hiding in the zones with excellent invisibility. We incorporate a strong encryption scheme in the embedding process to hide hidden data from an unauthorized extraction. Moreover, adaptive embedding methods are adapted to change data insertion according to image features, resulting in a huge increase in capacity without sacrificing visual quality. Extensive experiments show that the proposed method achieves clear improvements in imperceptibility, capacity and robustness of attacks including noise addition, compression, and cropping. So far, this work takes a significant step forward in the state of the art of secure steganography (i.e., data hiding), as it provides an extremely secure solution that is both high-capacity and robust.