Application software--Development

In this thesis, a practical solution for drive test data evaluation and a real
application are studied. We propose a system framework to project high dimensional
Drive Test Data (DTD) to well-organized web pages, such that users can visually review
phone performance with respect to different factors.
The proposed application, iVESTA (interactive Visualization and Evaluation
System for driven Test dAta), employs a web-based architecture which enables users to
upload DTD and immediately visualize the test results and observe phone and network
performances with respect to different factors such as dropped call rate, signal quality,
vehicle speed, handover and network delays. iVESTA provides practical solutions for
mobile phone manufacturers and network service providers to perform comprehensive
study on their products from the real-world DTD.

The rapid increase in the amount of data gathered by Autonomous Underwater Vehicles (AUV) leads to a global data management issue. Indeed, this large data collection effort is only interesting if the data collected can be easily retrieved and analyzed by many researchers. The main contribution of this thesis is the design of data management and retrieval schemes useful to the whole AUV community that both simplify the access and treatment of the data collected. This is achieved by the use of a self-describing standard data format (Hierarchical Data Format) and the use of Internet browsers' file download ability. Recent developments in Sun's Java applet technology have been used to provide a user-friendly Graphical User Interface (GUI) so that the user can select data files according to a large number of parameters (what variables have been collected, when and where).

Object segmentation in a video sequence is an essential task in video processing and forms the foundation of content analysis, scene understanding, object-based video encoding (e.g. MPEG-4), various surveillance and 2D-to-pseudo-3D conversion applications. Popularization and availability of video sequences with increased spatial resolution requires development of new, more efficient algorithms for object detection and segmentation. This dissertation discusses a novel neural-network-based approach to background modeling for motion-based object segmentation in video sequences. In particular, we show how Probabilistic Neural Network (PNN) architecture can be extended to form an unsupervised Bayesian classifier for the domain of video object segmentation. The constructed Background Modeling Neural Network (BNN) is capable of efficiently handling segmentation in natural-scene sequences with complex background motion and changes in illumination. The weights of the proposed neural network serve as an exclusive model of the background and are temporally updated to reflect the observed background statistics. The proposed approach is designed to enable an efficient, highly-parallelized hardware implementation. Such a system would be able to achieve real-time segmentation of high-resolution image sequences.