Design And Research Of Information Steganography Algorithm Based On Compressed Sensing

With the continuous improvement of network technology,the application of digital multimedia has become increasingly popular,and the digital image processing problems have received attention as never before.Especially in the network communication environment with no boundary or low security threshold,safety problems appear frequently associated with digital media,and information security technology emerges at the historic moment,namely information hiding technology for the purpose of spreading secret images.In recent years,compressed sensing technology,which has attracted much attention in image,audio,video and other signal processing,can also be applied to information hiding technology.Focusing on the information hiding technology based on compressed sensing theory,the selection and design of information hiding regions have been studied intensively,and also the pre-processing of secret information and the embedded rules.On this basis,two information hiding algorithms were proposed.The first one is the information hiding algorithm based on compressed sensing and GHM multi-wavelet transform.This algorithm is accomplished by using GHM multi-wavelet transform,wavelet transform,singular value decomposition and other methods for cover image preprocessing operations.Then after wavelet transform,the compressed perception theory and the singular value decomposition of a secret of embedded in the processed image were used to generate the stego images,so as to achieve the goal of secret information security transmission.The experiment shows that the algorithm is effective in the improvement of invisibility and robustness.The Second information hiding algorithm is based on compressed sensing and Arnold scrambling.This algorithm was completed by using the wavelet transform and compressed sensing theory in order to perform preprocessing on the cover image.The secret information handled with discrete cosine transform,Arnold scrambling and compression perception theory with be embedded into the cover image to generated the stego image.The experiment shows that this algorithm is better in invisibility than the first one,and also the better robustness.Above all,the research studies the hiding region,preprocessing of secret images and embedding rules were mainly studied based on compressed sensing,and two new schemes for information hiding were proposed.The proposed algorithms can solves the problem of the robustness improvement of the original secret image produced by the limited number of anti-clutter,and then which can the strict performance requirements of network communication security can be satisfied.