| The problem of information security has become increasingly prominent because of the development of internet technology.Information hiding technology becomes a hot research topic.The information hiding technology based on screen coding stands out in many information hiding algorithms due to its good robustness and concealment.The following research is carried out on the problems of information hiding capacity and low recognition rate in existing algorithms.This thesis proposes an image information compression algorithm based on sparse representation for the problem of low information hiding capacity.Dictionary-based sparse decomposition method,which uses the sparsity of the image,can linearly represent the key information of the image to obtain sparse coefficients.It removes a large amount of redundant information of the image in that process.And it realizes a reduction of the amount of image data.The proposed algorithm solves the problem that the information hiding system cannot meet the increasing demand for capacity.It realizes the purpose of expressing more information per unit of bits.Experimental data verify the effectiveness of the proposed capacity expansion method.In view of the problem that low recognition rate because of the noise when identification of information,this thesis proposes a denoising algorithm based on compressive sensing and morphological methods.The algorithm effectively achieves image denoising while effectively retaining and expanding screen points for reading.At the same time,the coding image is corrected using the slant correction method based on Gabor wavelet.The preprocessing method improves the efficiency and accuracy of information identification.The experiment verify the effectiveness and reliability of the proposed method.In this thesis,the comparison between digital watermark,two-dimensional code,and steganography is conducted relying on four indicators: information hiding capacity,robustness,concealment,and security.The experimental data verify the advantages of screen-based information hiding methods. |
