In this paper, we present a comparison of two key approaches for fingerprint identification. These approaches are based on (a) classification followed by verification, and (b) indexing followed by verification. The fingerprint classification approach is based on a novel feature-learning algorithm. It learns to discover composite operators and features that are evolved from combinations of primi...

Image compression and image encryption are pivotal to proper storage and transmission of images over MANET. Simultaneous image compression and encryption aims at achieving enhanced bandwidth utilization and security at the same time. The Number Theory based Image Compression and Quasigroup Encryption (NTICQE) algorithm employs number theoretic paradigm Chinese Remainder Theorem and Quasigroup E...

Singularity is the special feature of fingerprints for identification and classification. Since the performance of singularity extraction depends on the quality of fingerprint images, image enhancement is required to improve the performance. Image enhancement with various image filters might be more useful than a filter, but it is very difficult to find a set of appropriate filters. In this pap...

Many digital image forensics techniques extracting various fingerprints are dependent on data on digital images from an unknown environment. As often software modifications leave no appropriate traces in images metadata, critical inconveniences and miscalculations of fingerprints arise. This is the problem addressed in this paper. Modeling information noise in image metadata, we introduce a sta...

In this paper we purposed an approach to remove ghosting artifact generated by moving objects and reconstruct the movement of these objects in panorama images. Our approach contains three steps. First is to detect moving objects in the image flow with moving background. Second is to replace the moving objects with corresponding background scenes. Third is to reconstruct the movement of moving o...

We proposed a blind watermarking technique to embed the watermark in JPEG compressed domain. Low frequency DCT coeffi cients are extracted to form an M-dimensional vector. Watermarking is achieved by modifying this vector in order to point to the centroid of a particular cell. This cell is determined according to the extract ed vector, private keys and the watermark. A dual-key system is used t...

This study presents an expandable essential secret image sharing structure showing that some important shared images, named essential ones, are necessary when recovering the secret image and the proposed structure must incorporate other secret image sharing scheme to expand to a new secret image sharing scheme with multiple functions. An (s, t, n) essential secret image sharing structure shares...

Introduction: Echo planar images (EPI) are applied in a variety of diagnostic techniques that rely on fast acquisition of imaged data. EPI applications in common practice range from dynamic-contrast studies used to monitor the passage of injected contrast over time for stroke, tumor evaluation, and cardiac assessment to fMRI applications where a large series of images (typically thousands) are ...

In this paper, we propose a novel framework to represent visual information. Extending the notion of conventional image-based rendering, our framework makes joint use of both light fields and holograms as complementary representations. We demonstrate how light fields can be transformed into holograms, and vice versa. By exploiting the advantages of either representation, our proposed dual repre...

With the growth in the communication, there is an increase in the transmission of data across the various medium; security is the key issue that needs to be addressed. Various steganography techniques are used to hide actual message secretly to provide control over the medium where data is sent. This paper is intended to propose an image steganography technique that will hide four bits of data ...