The Watermarking Algorithms For Medical Images In The Contourlet Transform Domain

In the clinical diagnosis and scientific research of doctors,medical image is an important auxiliary tool.And with the progress of science and technology,medical image data is also increasing rapidly.So we are facing the problem that how to store and transfer medical data and patients' personal privacy information safely and effectively.Thanks to the coming of 5g era,combined with artificial intelligence and cloud tools,smart hospitals are also developing rapidly.Hospitals can choose to store the medical data in the cloud platform.Although cloud platform can provide users with storage space with scale and computing power,there is still a problem of information disclosure.Therefore,how to protect the patient's privacy safely and efficiently under the condition of mass data online storage and dissemination is an urgent problem to be solved.At present,the medical image digital watermarking technology can be used to protect the patient's privacy.The patient's privacy information is hidden in the corresponding medical image as a watermark.The balance of robustness and imperceptibility of the algorithm is always a hot and difficult point.Based on the development of wavelet transform in the field of digital watermarking,the contourlet transform selected in this paper improves the defect that wavelet transform can't approach the edge of image better when it is used in image processing.This paper studies the digital watermarking technology based on Contourlet transform,and the specific research content is the digital zero watermarking algorithm based on Contourlet transform.This paper has done the following research work:A medical image zero watermark algorithm based on Contourlet transform is studied.In this algorithm,firstly,the carrier image is contourlet transformed,and the low frequency part is selected for perceptual hash.In order to match the size of the original encrypted watermark,the 4x8 matrix is selected from the perceptual hash.By comparing with the mean value of the whole matrix,a 32-bit eigenvector is formed by binarization,which advances with the encrypted watermark Row exclusive or calculation completed embedding.This method does not limit the capacity of embedding watermark.Experimental data show that it has strong robustness in both conventional and geometric attacks.This paper studies the medical image zero watermark algorithm based on the nonsubsampling contourlet transform and the direction information intensity.By using the biggest characteristic and advantage of Contourlet transform,the image can be expressed in multiple directions and scales.By calculating the direction information intensity of each direction sub-band,the direction feature information contained in the frequency-domain sub-band is de quantified,that is,the sub-band with the most direction information and the largest intensity is selected to extract the feature value.Then,the sub-band is divided into blocks by SVD,and the 2-norm of each block is calculated.Finally,the feature is extracted by binarization and the embedding is completed.In this chapter,the simulation results show that the algorithm is robust to the rotation attack in geometric attack.