“…The main objective in clustering applications is to group samples into clusters of similar features (e.g., the SPNs). Among a wide variety of methods, k-means [23,24] and fuzzy c-means [25][26][27] have been intensively employed in various applications. However, classical k-means and fuzzy c-means clustering methods rely on the user to provide the number of clusters and initial centroids.…”
Section: The Challenges Of Source-oriented Image Clustering and Relatmentioning
We present in this paper an algorithm that is capable of clustering images taken by an unknown number of unknown digital cameras into groups, such that each contains only images taken by the same source camera. It first extracts a sensor pattern noise (SPN) from each image, which serves as the fingerprint of the camera that has taken the image. The image clustering is performed based on the pairwise correlations between camera fingerprints extracted from images. During this process, each SPN is treated as a random variable and a Markov random field (MRF) approach is employed to iteratively assign a class label to each SPN (i.e., random variable). The clustering process requires no a priori knowledge about the dataset from the user. A concise yet effective cost function is formulated to allow different "neighbors" different voting power in determining the class label of the image in question depending on their similarities. Comparative experiments were carried out on the Dresden image database to demonstrate the advantages of the proposed clustering algorithm.
“…The main objective in clustering applications is to group samples into clusters of similar features (e.g., the SPNs). Among a wide variety of methods, k-means [23,24] and fuzzy c-means [25][26][27] have been intensively employed in various applications. However, classical k-means and fuzzy c-means clustering methods rely on the user to provide the number of clusters and initial centroids.…”
Section: The Challenges Of Source-oriented Image Clustering and Relatmentioning
We present in this paper an algorithm that is capable of clustering images taken by an unknown number of unknown digital cameras into groups, such that each contains only images taken by the same source camera. It first extracts a sensor pattern noise (SPN) from each image, which serves as the fingerprint of the camera that has taken the image. The image clustering is performed based on the pairwise correlations between camera fingerprints extracted from images. During this process, each SPN is treated as a random variable and a Markov random field (MRF) approach is employed to iteratively assign a class label to each SPN (i.e., random variable). The clustering process requires no a priori knowledge about the dataset from the user. A concise yet effective cost function is formulated to allow different "neighbors" different voting power in determining the class label of the image in question depending on their similarities. Comparative experiments were carried out on the Dresden image database to demonstrate the advantages of the proposed clustering algorithm.
“…Besides [15], we also compare in Table 2 some other image hiding methods [16][17][18] reported in 2005-2007. We bypass the detail introduction of [16][17][18] to save the paper length.…”
Section: Experiments and Comparisonsmentioning
confidence: 99%
“…We bypass the detail introduction of [16][17][18] to save the paper length. In Tests 1 and 2, since Ref.…”
“…Cryptography, watermarking and steganography technologies are effective tools for communicating secret messages and data protection [3][4][5] . Researchers have proposed various methods to protect biometric data through information hiding techniques [6][7][8][9][10][11][12][13][14][15] . According to the hiding domain, there are spatial domain methods 8 , frequency domain methods 6,11,12 , and combination method of spatial and frequency domain 15 .…”
For secure biometric verification, most existing methods embed biometric information directly into the cover image, but content correlation analysis between the biometric image and the cover image is often ignored. In this paper, we propose a novel biometric image hiding approach based on the content correlation analysis to protect the networkbased transmitted image. By using principal component analysis (PCA), the content correlation between the biometric image and the cover image is firstly analyzed. Then based on particle swarm optimization (PSO) algorithm, some regions of the cover image are selected to represent the biometric image, in which the cover image can carry partial content of the biometric image. As a result of the correlation analysis, the unrepresented part of the biometric image is embedded into the cover image by using the discrete wavelet transform (DWT). Combined with human visual system (HVS) model, this approach makes the hiding result perceptually invisible. The extensive experimental results demonstrate that the proposed hiding approach is robust against some common frequency and geometric attacks; it also provides an effective protection for the secure biometric verification.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.