Biblio

Image retrieval systems have been an active area of research for more than thirty years progressively producing improved algorithms that improve performance metrics, operate in different domains, take advantage of different features extracted from the images to be retrieved, and have different desirable invariance properties. With the ever-growing visual databases of images and videos produced by a myriad of devices comes the challenge of selecting effective features and performing fast retrieval on such databases. In this paper, we incorporate Fourier descriptors (FD) along with a metric-based balanced indexing tree as a viable solution to DHS (Department of Homeland Security) needs to for quick identification and retrieval of weapon images. The FDs allow a simple but effective outline feature representation of an object, while the M-tree provide a dynamic, fast, and balanced search over such features. Motivated by looking for applications of interest to DHS, we have created a basic guns and rifles databases that can be used to identify weapons in images and videos extracted from media sources. Our simulations show excellent performance in both representation and fast retrieval speed.

There are different traditional approaches to handling a large number of computers or workstations in a campus setting, ranging from imaging to virtualized environments. The common factor among the traditional approaches is to have a user workstation with a local hard drive (nonvolatile storage), scratchpad volatile memory, a CPU (Central Processing Unit) and connectivity to access resources on the network. This paper presents the use of block streaming, normally used for storage, to serve operating system and applications on-demand over the network to a workstation, also referred to as a client, a client computer, or a client workstation. In order to avoid per seat licensing, an Open Source solution is used, and in order to minimize the field maintenance and meet security privacy constraints, a workstation need not have a permanent storage such as a hard disk drive. A complete blue print, based on performance analyses, is provided to determine the type of network architecture, servers, workstations per server, and minimum workstation configuration, suitable for supporting such a solution. The results of implementing the proposed solution campus wide, supporting more than 450 workstations, are presented as well.

Cyber-induced dependent failures are important to be considered in composite system reliability evaluation. Because of the complexity and dimensionality, Monte Carlo simulation is a preferred method for composite system reliability evaluation. The non-sequential Monte Carlo or sampling generally requires less computational and storage resources than sequential techniques and is generally preferred for large systems where components are independent or only a limited dependency exists. However, cyber-induced events involve dependent failures, making it difficult to use sampling methods. The difficulties of using sampling with dependent failures are discussed and a solution is proposed. The basic idea is to generate a representative state space from which states can be sampled. The probabilities of representative state space provide an approximation of the joint distribution and are generated by a sequential simulation in this paper but it may be possible to find alternative means of achieving this objective. The proposed method preserves the dependent features of cyber-induced events and also improves the efficiency. Although motivated by cyber-induced failures, the technique can be used for other types of dependent failures as well. A comparative study between a purely sequential methodology and the proposed method is presented on an extended Roy Billinton Test System.