Coverless image steganography based on DWT approximation and pixel intensity averaging

Image steganography uses an image to send a secret message. One way to do it is to modify the image in such a way that the normal eye cannot see it, but the receiver can still extract the information. In traditional embedded steganography even though the human eye cannot detect the manipulation done to the cover image, statistical analysis can reveal the secret information. Coverless image steganography is another approach that solves this problem where instead of embedding secret information we find an image that already consists of a secret message. As no modification is done to the original image no modification traces are found leaving less space for steganalysis. But finding a suitable and optimal data set is a challenge here. Another use of coverless image steganography is to help in protecting the copyright of digital arts or copies of digital records. Since the Internet is not a secure medium these processes must be robust against these attacks to protect the secret message embedded inside the media. In due process of time, many frameworks are devised to increase the capacity of this method and also to increase the robustness against many image attacks. This study has developed a new framework using the Discrete Wavelet Transform of images combined with plane segregation and intensity averaging of sub-images that is robust and has a high capacity to pack information that is to be sent via a communication channel. This proposed method has demonstrated better capacity and robustness compared to previous methods based on DWT.