Digital techniques

Video identification or copy detection is a challenging problem and is becoming increasingly important with the popularity of online video services. The problem addressed in this thesis is the identification of a given video clip in a given set of videos. For a given query video, the system returns all the instance of the video in the data set. This identification system uses video signatures based on video tomography. A robust and low complexity video signature is designed and implemented. The nature of the signature makes it independent to the most commonly video transformations. The signatures are generated for video shots and not individual frames, resulting in a compact signature of 64 bytes per video shot. The signatures are matched using simple Euclidean distance metric. The results show that videos can be identified with 100% recall and over 93% precision. The experiments included several transformations on videos.

Video signature techniques based on tomography images address the problem of video identification. This method relies on temporal segmentation and sampling strategies to build and determine the unique elements that will form the signature. In this thesis an extension for these methods is presented; first a new feature extraction method, derived from the previously proposed sampling pattern, is implemented and tested, resulting in a highly distinctive set of signature elements, second a robust temporal video segmentation system is used to replace the original method applied to determine shot changes more accurately. Under a very exhaustive set of tests the system was able to achieve 99.58% of recall, 100% of precision and 99.35% of prediction precision.

High-resolution sonar systems are primarily used for ocean floor surveys and port security operations but produce images of limited resolution. In turn, a sonar-specific methodology is required to detect and classify underwater unexploded ordnance (UXO) using the low-resolution sonar data. After researching and reviewing numerous approaches the Multiple Aspect-Fixed Range Template Matching (MAFR-TM) algorithm was developed. The MAFR-TM algorithm is specifically designed to detect and classify a target of high characteristic impedance in an environment that contains similar shaped objects of low characteristic impedance. MAFR-TM is tested against a tank and field data set collected by the Sound Metrics Corp. DIDSON US300. This thesis document proves the MAFR-TM can detect, classify, orient, and locate a target in the sector-scan sonar images. This paper focuses on the MAFR-TM algorithm and its results.

Provision of complete and responsive solution to healthcare services requires a multi-tired health delivery system. One of the aspects of healthcare hierarchy is the need for nursing care of the patient. Nursing care and observation provide basis for nurses to communicate with other aspects of healthcare system. The ability of capturing and managing nursing practice is essential to the quality of human care. The thesis proposes knowledge based decision making and analyzing system for the nurses to capture and manage the nursing practice. Moreover it allows them to monitor nursing care quality, as well as to test an aspect of an electronic healthcare record for recording and reporting nursing practice. The framework used for this system is based on nursing theory and is coupled with the quantitative analysis of qualitative data. It allows us to quantify the qualitative raw natural nursing language data. The results are summarized in the graph that shows the relative importance of those attributes with respect to each other at different instances of nurse-patient encounter. Research has been conducted by the Department of Computer and Electrical Engineering and Computer Science for the College of Nursing.

The Intermediate Frequency Acoustic Modem (IFAM), developed by Dr. Beaujean, is designed to transmit the command-and-control messages from the top-side to the wet-side unit in ports and very shallow waters. This research presents the design of the turbo coding scheme and its implementation in the IFAM modem with the purpose of meeting a strict requirement for the IFAM error rate performance. To simulate the coded IFAM, a channel simulator is developed. It is basically a multi-tap filter whose parameters are set depending on the channel geometry and system specifics. The simulation results show that the turbo code is able to correct 89% of the messages received with errors in the hostile channel conditions. The Bose-Chadhuri-Hocquenghem (BCH) coding scheme corrects less that 15% of these messages. The other simulation results obtained for the system operation in different shallow water settings are presented.

This work explores the process of model-based classification of speech audio signals using low-level feature vectors. The process of extracting low-level features from audio signals is described along with a discussion of established techniques for training and testing mixture model-based classifiers and using these models in conjunction with feature selection algorithms to select optimal feature subsets. The results of a number of classification experiments using a publicly available speech database, the Berlin Database of Emotional Speech, are presented. This includes experiments in optimizing feature extraction parameters and comparing different feature selection results from over 700 candidate feature vectors for the tasks of classifying speaker gender, identity, and emotion. In the experiments, final classification accuracies of 99.5%, 98.0% and 79% were achieved for the gender, identity and emotion tasks respectively.

Professional imaging systems, particularly motion picture cameras, usually employ larger photosites and lower pixel counts than many amateur cameras. This results in the desirable characteristics of improved dynamic range, signal to noise and sensitivity. However, high performance optics often have frequency response characteristics that exceed the Nyquist limit of the sensor, which, if not properly addressed, results in aliasing artifacts in the captured image. Most contemporary still and video cameras employ various optically birefringent materials as optical low-pass filters (OLPF) in order to minimize aliasing artifacts in the image. Most OLPFs are designed as optical elements with a frequency response that does not change even if the frequency responses of the other elements of the capturing systems are altered. An extended evaluation of currently used birefringent-based OLPFs is provided. In this work, the author proposed and demonstrated the use of a parallel optical window p ositioned between a lens and a sensor as an OLPF. Controlled X- and Y-axes rotations of the optical window during the image exposure results in a manipulation of the system's point-spread function (PSF). Consequently, changing the PSF affects some portions of the frequency components contained in the image formed on the sensor. The system frequency response is evaluated when various window functions are used to shape the lens' PSF, such as rectangle, triangle, Tukey, Gaussian, Blackman-Harris etc. In addition to the ability to change the PSF, this work demonstrated that the PSF can be manipulated dynamically, which allowed us to modify the PSF to counteract any alteration of other optical elements of the capturing system. There are several instances presented in the dissertation in which it is desirable to change the characteristics of an OLPF in a controlled way.

Digital video is being used widely in a variety of applications such as entertainment, surveillance and security. Large amount of video in surveillance and security requires systems capable to processing video to automatically detect and recognize events to alleviate the load on humans and enable preventive actions when events are detected. The main objective of this work is the analysis of computer vision techniques and algorithms used to perform automatic detection of events in video sequences. This thesis presents a surveillance system based on optical flow and background subtraction concepts to detect events based on a motion analysis, using an event probability zone definition. Advantages, limitations, capabilities and possible solution alternatives are also discussed. The result is a system capable of detecting events of objects moving in opposing direction to a predefined condition or running in the scene, with precision greater than 50% and recall greater than 80%.

With augmenting security concerns and decreasing costs of surveillance and computing equipment, research on automated systems for object detection has been increasing, but the majority of the studies focus their attention on sequences where high resolution objects are present. The main objective of this work is the detection and extraction of information of low resolution objects (e.g. objects that are so far away from the camera that they occupy only tens of pixels) in order to provide a base for higher level information operations such as classification and behavioral analysis. The system proposed is composed of four stages (preprocessing, background modeling, information extraction, and post processing) and uses context based region of importance selection, histogram equalization, background subtraction and morphological filtering techniques. The result is a system capable of detecting and tracking low resolution objects in a controlled background scene which can be a base for systems with higher complexity.

Promoting healthcare and wellbeing requires the dedication of a multi-tiered health service delivery system, which is comprised of specialists, medical doctors and nurses. A holistic view to a patient care perspective involves emotional, mental and physical healthcare needs, in which caring is understood as the essence of nursing. Properly and efficiently capturing and managing nursing knowledge is essential to advocating health promotion and illness prevention. This thesis proposes a document-indexing framework for automating classification of nursing knowledge based on nursing theory and practice model. The documents defining the numerous categories in nursing care model are structured with the help of expert nurse practitioners and professionals. These documents are indexed and used as a benchmark for the process of automatic mapping of each expression in the assessment form of a patient to the corresponding category in the nursing theory model. As an illustration of the proposed methodology, a prototype application is developed using the Latent Semantic Indexing (LSI) technique. The prototype application is tested in a nursing practice environment to validate the accuracy of the proposed algorithm. The simulation results are also compared with an application using Lucene indexing technique that internally uses modified vector space model for indexing. The result comparison showed that the LSI strategy gives 87.5% accurate results compared to the Lucene indexing technique that gives 80% accuracy. Both indexing methods maintain 100% consistency in the results.