Font Size: a A A

Research On Digital Image Watermarking Based On Public Key Encryption

Posted on:2020-01-17Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y LiuFull Text:PDF
GTID:1368330599456479Subject:Geographic Information System
Abstract/Summary:PDF Full Text Request
Under the background of taking economic construction as the center,China vigorously develops the Internet economy.The "Internet +" mode has begun to go deep into all walks of life,and has also contributed to the rapid development of the information industry.As the core technology of information industry,network technology and multimedia technology make it easy for people to carry out information transmission,information hiding,data replication and other operations.The huge hidden danger is information security,especially the copyright of digital works.As a means of information hiding,digital watermarking technology can provide copyright protection for digital works.In the case of copyright disputes,we can identify the copyright owner,track the infringement,verify the integrity of digital content,authenticate the authenticity of digital content,and provide effective legal basis for the identification of infringement,so as to deter illegal users and reduce infringement.As a result,this technology has become a commercial tool for digital works protection.With the wide application of digital watermarking technology,many researchers begin to regard it as an important research direction in the field of information security,but some problems still remain unsolved.For example,the comprehensiveness of the anti-attack ability of the watermarking system,the security of the watermarking system,the time consumption of the system,and the accuracy of tampering detection in content authentication,etc.In view of the shortcomings of current digital watermarking algorithms,this paper explores the watermarking generation process,the relationship between image processing and watermarking embedding,and makes the following contributions to the research of image watermarking technology.(1)Traditional digital watermarking algorithms are only robust to common noise interference and image compression processing,but they are powerless against geometric deformation attacks such as image rotation and scale transformation.In this paper,a digital watermarking algorithm based on host image features is proposed.Scale invariant feature transformation is used to form a stable watermark embedding region,which solves the problem that the embedding domain is easy to change with the scale of the image.Firstly,the invariant feature transformation is used to generate a stable watermark embedding domain on the host image,and then the robust watermark is obtained from the host by using image texture feature extraction operator.The obtained feature watermark is close to the texture of the host image,and the data is embedded in the way of image enhancement,which is also conducive to conceal the watermark.This paper is the first attempt to explore the robustness of digital watermarking system from both the robust embedding domain and the robust digital watermark information.The experimental results show that the proposed algorithm has good invisibility without reducing the amount of watermark,and can hide the copyright verification information well.Compared with other related algorithms,the NCC value of this algorithm against Gauss noise is increased by 0.0789,the NCC value against salt and pepper noise is increased by 0.0865,the image rotation attack is increased by 0.1159,and the image scaling attack is increased by 0.2369.Therefore,this algorithm has stronger resistance to noise interference such as Gaussian noise and geometric attacks such as rotation and scaling.(2)In order to ensure the security of digital watermark,the traditional method is to use chaotic mapping technology or symmetric encryption algorithm to scramble the watermark image into cipher text.However,the large amount of image data leads to the time-consuming process of watermark encryption.In this paper,a secure digital watermarking algorithm based on hybrid encryption is proposed.The asymmetric encryption algorithm is used to encrypt chaotic mapping parameters,which solves the time-consuming problem in watermark encryption.First,the watermark information is encrypted by Logistic chaotic mapping,and then the mapping parameters are encrypted by RSA asymmetric encryption algorithm.In this paper,the asymmetric encryption algorithm is applied to the chaotic technology for the first time to protect the security of digital watermark.It not only solves the problem of small and insecure key space in the chaotic map,but also greatly shortens the encryption time of the watermarking image.The experimental results show that the proposed scheme takes into account the main performance indicators in watermarking,guarantees better robustness and larger amount of data embedding.Compared with other related methods,the encryption time of a single watermark image is reduced by 4.5 seconds,which not only improves the time efficiency,but also improves the security of the watermark.(3)Traditional semi-fragile watermarking algorithms are usually only robust to image compression processing and vulnerable to other noise interference,which results in missed detection or false detection in tampering detection due to the impact of non-malicious attacks on watermark.For the first time,this paper adopts the strategy of embedding different parts separately after splitting the original watermark,combines the advantages of spatial domain and transform domain to improve the survivability of watermark information in non-malicious attack environment,and solves the problem of single resistance of semi-fragile watermarking algorithm.In addition,this paper improves the traditional image scrambling method,and proposes a K-Arnold encryption algorithm to encrypt watermark image,which solves the problem of low security of watermark information caused by small secret key space in most chaotic mapping or image scrambling algorithms.Compared with traditional image scrambling algorithms,this algorithm has more secret key space,can better deal with watermark analysis attacks,and has higher security.Experiments show that the algorithm has better resistance to noise interference and image compression in a reasonable range,and can distinguish malicious tampering attacks from conventional image processing.Especially when the intensity of salt and pepper attack is 0.002,the measured NCC value is 0.8415,and the intensity of JPEG compression attack is 10%,the NCC value is 0.8991,which is higher than traditional methods.(4)In most image authentication watermarking algorithms,tamper locating watermark is composed of authentication codes based on hash value,which is sensitive to non-malicious attacks.Due to the inadequate consideration of noise,compression processing and other interference to the locating watermark,there are many false located areas in image tampering detection.In this paper,a new image authentication watermark algorithm based on double watermark embedding is proposed,which uses the even convolution sliding technique to generate stable features from smaller image blocks and solves the problem of false locating.The algorithm first applies integer wavelet transform to the host image,embeds the host-independent gray-scale watermark image into low-frequency wavelet coefficients,and then combines image convolution operation and singular value decomposition technology to obtain salient algebraic features from the host,which can be used for image tamper locating.The experimental results show that the proposed algorithm can better resist the interference of mean filtering and salt and pepper noise,improve the tamper locating accuracy of digital watermarking system.And in the same security level,it takes less time to encrypt watermarking than RSA algorithm.Compared with other three image authentication watermarking algorithms,the average tamper locating accuracy of this algorithm is 98.23%,which is improved by 2.73%.
Keywords/Search Tags:Copyright protection, image authentication, image digital watermarking, robust feature detection, tampering detection
PDF Full Text Request
Related items