Research On Encrypted JPEG Image Retrieval

With the rapid technological development of cloud computing,the users would like to store their multimedia data into the cloud for low cost and convenience.For privacy protection,the users tend to encrypt the multimedia data before transmission to the server,which may impede further processing operations,such as information retrieval.Therefore,it becomes highly desirable to develop retrieval techniques for the purpose of providing privacy-preserving and effective retrieval service in encrypted multimedia databases.This dissertation mainly focuses on image retrieval in encrypted domain and summarizes the results obtained in image encryption,image retrieval,JPEG compression coding.The contributions of this dissertation are listed as follows:1.Encrypted JPEG image retrieval based on statistical characteristics of DCT coefficientsThis work proposes a novel scheme for encrypted JPEG image retrieval,which includes image encryption and unsupervised/supervised retrieval phases.Using this scheme,the encrypted images are produced by permuting DCT coefficients,and transmitted to a database server.With an encrypted query image,although the server does not know the plaintext content,he may get the histogram at each frequency position.After calculating the distances between the histograms of encrypted query image and database image,the server can return the encrypted images with plaintext content similar to the query image according to integrated distances.If a training image set is available,the retrieval results can be also determined by conditional probabilities calculated from a supervised mechanism.2.AC-coefficient histogram-based retrieval for encrypted JPEG imagesThis paper proposes a novel retrieval scheme for encrypted JPEG images.With this scheme,the DC and AC coefficients of JPEG images are encrypted using a stream cipher and scrambling encryption,respectively.Then,the encrypted images are transmitted to and stored in a server,which can also provide retrieval service.When receiving an encrypted query image,the server without any knowledge of the plaintext content may acquire statistically its AC coefficients histogram.By calculating the distances between the histograms of encrypted query image and database image,the server may output the encrypted images closest to the query image to the authorized user.3.Encrypted JPEG image retrieval based on Markov process and multi-class support vector machineThis paper develops a retrieval scheme for encrypted JPEG images based on Markov process and multi-class support vector machine.In image encryption phase,the stream cipher is applied to encrypt VLI code of DCT coefficients for protecting JPEG image content's confidentiality.And thus it is easy for the content owner to achieve the encrypted JPEG images uploaded to a database server.In the image retrieval stage,the server without knowing the original content can extract 768 dimensional features from the transition probability matrices of the AC coefficients of encrypted query image,in which those coefficients are modeled by Markov process.After that,with the multi-class support vector machine(SVM),the feature of the encrypted query image can be converted into a vector with low dimensionality determined by the number of image categories.The encrypted database images are conducted similarly.After low-dimensional vector representation,the similarity between encrypted query image and database image may be measured by calculating the distance of their corresponding feature vectors.At the client side,the returned encrypted images similar to query image can be decrypted to the plaintext images with the help of the encryption key.Experimental results show that the proposed scheme can preserve file compliance and file size for encrypted JPEG images,while efficiently providing privacy-preserving image retrieval.4.New block descriptor-based retrieval for encrypted JPEG imagsThis letter proposes a novel retrieval scheme for encrypted JPEG images.Using the scheme,the content owner obtains encrypted JPEG images mainly by modifying the encoded bit sequences of DCT coefficients and quantization tables,and then transmits them to a database server.When receiving an encrypted query image from the authorized user,the server may extract block-wise feature vectors,each of which is generated from a same feature descriptor newly constructed to capture the local structure of intra-block AC coefficients.Based on the feature distance between blocks,although the server learns nothing about the plaintext content,he can calculate the distances between encrypted query image and database image.After that,the encrypted images ranked by their distances to the query image are returned to the authorized user.The proposed scheme can ensure both format compliance and file size preservation when providing retrieval service in encrypted domain.This dissertation has an extensive study in JPEG images,mainly focusing on compression coding,bit-stream formation,image encryption and image retrieval.Meanwhile,we propose some simple and effective retrieval methods for encrypted JPEG images,which will be helpful for the development of image retrieval in encrypted domain.