Research And Application Of Image Source Anti-Forensics Technology Based On PRNU Suppression

A digital image is an important carrier of information and plays a very important role in the process of biometric identification and network authentication.Biometric characteristics are closely related to personal identity,and the privacy risks they face are related to all aspects of international livelihood.This thesis focuses on the data and network security aspects of identity identification and cloud authentication,and is committed to solving the problems that(1)open transmission channel images are easy to be stolen and lead to identity disclosure of the photographer;(2)Identity authentication of terminal and server is difficult,and it is difficult to efficiently determine the legitimate collection source in the mass collection equipment.Combined with the research results of digital image source anti-forensics,the author designed and implemented the fingerprint suppression scheme of the image acquisition equipment chip,which mainly includes the following two parts:(1)A PRNU suppression scheme based on discrete wavelet transform is proposed.The scheme implements image identity hiding and forgery algorithms,interferes with biometric images,cuts off the connection between the processed images and the source device,and preserves the biometric utility.Specifically,the original application image is segmented according to the importance level of the image region to the recognition system.Secondly,the high and low-frequency bands of the chip fingerprint were suppressed by layers to ensure a higher success rate of PRNU suppression under illumination conditions.The experimental results show that compared with other fingerprint suppression schemes,the proposed scheme not only has a better fingerprint anonymity effect but also ensures that the output image retains more original details.(2)A reversible PRNU suppression scheme based on integer wavelet transform is proposed.The scheme implements image identity concealment and forgery algorithms and allows users to weigh and choose between image quality and anonymity according to the working scene.Moreover,it advocates that each output image be given reversible characteristics to filter illegal terminal access to the cloud network.Specifically,we first extract the device reference image PRNU and conduct preliminary PRNU anonymity or forgery spoofing.This work mainly includes digital signature decommissioning of fingerprints in hidden images and reversible scrambling storage of edge information.Secondly,we propose to apply reversible information hiding to the field of image source anti-forensics.The work of this part mainly includes pixel data integer processing,high-frequency scrambling function,and embedding method selection.Finally,the embedded data is extracted and the original image is recovered according to the same orthogonal spread spectrum sequence.The experimental results on several different public data sets show that the proposed scheme has strong scalability and applicability,and has certain practical application value.