Topic
Feature hashing
About: Feature hashing is a research topic. Over the lifetime, 993 publications have been published within this topic receiving 51462 citations.
Papers published on a yearly basis
Papers
More filters
01 Dec 2012
TL;DR: A robust and secure perceptual image hashing technique based on random sub-image selection, non-negative matrix factorization and DWT transform where the hashes are extracted in a binary form with a perfect control over the probability distribution of hash bits.
Abstract: This paper proposes a robust and secure perceptual image hashing technique based on random sub-image selection, non-negative matrix factorization and DWT transform where the hashes are extracted in a binary form. The binary hash values are obtained with a perfect control over the probability distribution of hash bits. A key is used randomize the feature vector. Experimentally, the proposed technique has been shown to yield a good performance with respect to robustness against image processing operations including JPEG lossy compression, additive noise and geometric attacks such as rotation, translation and more secure of image hashing scheme.
4 citations
22 Sep 2010
TL;DR: In this paper, the authors propose the creation of hash values that keep similar data stored near to each other in a P2P network, reducing the effort to retrieve similar data.
Abstract: The increasing volume of semantic content available in the Web, generally classified by concept hierarchies or simple ontologies, turns the searching and reasoning upon these data a great challenge. Generally, a search in Semantic Web may not be addressed to a specific document, but to a group of data classified in the same concept. Several structures used to distribute data, e.g. P2P networks, use hash values to identify these data, without maintaining the semantic values of the stored data. This paper contributes by proposing the creation of hash values that keep similar data stored near to each other in a P2P network, reducing the effort to retrieve similar data. The proposed hash values are derived from the data classification based on ontologies, using locality sensitive hashing (LSH) functions.
4 citations
01 Aug 2014
TL;DR: This paper proposes a feature fusion based hashing method which effectively utilize the correlation between two feature models and efficiently accomplish large scale image copy detection.
Abstract: Most of existing approaches use only a single feature to represent an image for copy detection. However, a single feature is often insufficient to characterize the image content. Besides, with the exponential growth of online images, it's urgent to explore a way of tackling the problem of large scale. In this paper, we propose a feature fusion based hashing method which effectively utilize the correlation between two feature models and efficiently accomplish large scale image copy detection. To accurately map images into the Hamming space, our hashing method not only preserves the local structure of individual feature but also globally consider the local structures for all the features to learn a group of hash functions. The experiment results show that the proposed method outperforms the state-of-the-art techniques in both accuracy and efficiency.
4 citations
TL;DR: The hash incorporation into the system is found very sensitive to abnormal image modifications and hence robust to splicing and copy-move type of image tampering and, therefore, can be applicable to image authentication.
Abstract: an efficient image security, image hashing is one of the solutions for image authentication. A robust image hashing mechanism must be robust to image processing operations as well as geometric distortions. A better hashing technique must ensure an efficient detection of image forgery like insertion, deletion, replacement of objects, malicious color tampering, and for locating the exact forged areas. This paper describes a novel image hash function, which is generated by using both global and local features of an image. The global features are the representation of Zernike moments on behalf of luminance and chrominance components of the image as a whole. The local features include texture information as well as position of significant regions of the image. The secret keys can be introduced into the system, in places like feature extraction and hash formulation to encrypt the hash. The hash incorporation into the system is found very sensitive to abnormal image modifications and hence robust to splicing and copy-move type of image tampering and, therefore, can be applicable to image authentication. As in the generic system, the hashes of the reference and test images are compared by finding the hamming or hash distance. By setting the thresholds with the distance, the received image can be stated as authentic or non-authentic. And finally location of forged regions and type of forgery are found by decomposing the hashes. Compared to most recent work done in this area, our algorithm is simple and cost effective with better scope of security. Keywordsmoments, Forgery detection, SHA-1, MD5 Image hash, Salient detection
4 citations
TL;DR: A concentrated hashing method with neighborhood embedding (CHNE) for efficient and effective image retrieval and classification by integrating Cauchy cross-entropy and pair-wise weighted similarity loss and jointly minimize the regression quantization and neighborhood structure reconstruction errors in the loss function to improve the classification accuracy.
4 citations