A. James Clark School of Engineering

Permanent URI for this communityhttp://hdl.handle.net/1903/1654

The collections in this community comprise faculty research works, as well as graduate theses and dissertations.

Browse

Subject: search.filters.subject.statistical

Search Results

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Statistical Characterization and Prediction for a Stochastic Sea Environment
    (2012) Chang, Che-yu; Ayyub, Bilal M.; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Designing marine and maritime systems requires the probabilistic characterization of sea waves in the time-history and spectral domains. These probabilistic models include parameters that can be empirically estimated based on limited data in durations, locations and applicability to particular designs. Characterizing the statistical uncertainties associated with the parameters and the models is an essential step for risk-based design methods. A framework is provided for characterizing and predicting the stochastic sea-state conditions using sampling and statistical methods in order to associate confidence levels with resulting estimates. Sea-state parameters are analyzed using statistical confidence intervals which give a clear insight for the uncertainties involved in the system. Hypothesis testing and goodness-of-fit are performed to demonstrate the statistical features. Moreover, sample size is required for performing statistical analysis. Sample size indicates the number of representative and independent observations. Current practices do not make a distinction between the number of discretization points for numerical computations and the number of sampling points, i.e. sample size needed for statistical analysis. Sample size and interval between samples to obtain independent observations are studied and compared with existing methods. Further, spatial relationship of the sea-state conditions describes the wave energy transferred through the wave movement. Locations of interest with unknown sea-state conditions are estimated using spatial interpolations. Spatial interpolation methods are proposed, discussed, and compared with the reported methods in the literature. This study will enhance the knowledge of sea-state conditions in a quantitative manner. The statistical feature of the proposed framework is essential for designing future marine and maritime systems using probabilistic modeling and risk analysis.
  • Thumbnail Image
    Item
    Object Tracking and Mensuration in Surveillance Videos
    (2010) Shao, Jie; Chellappa, Rama; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    This thesis focuses on tracking and mensuration in surveillance videos. The first part of the thesis discusses several object tracking approaches based on the different properties of tracking targets. For airborne videos, where the targets are usually small and with low resolutions, an approach of building motion models for foreground/background proposed in which the foreground target is simplified as a rigid object. For relatively high resolution targets, the non-rigid models are applied. An active contour-based algorithm has been introduced. The algorithm is based on decomposing the tracking into three parts: estimate the affine transform parameters between successive frames using particle filters; detect the contour deformation using a probabilistic deformation map, and regulate the deformation by projecting the updated model onto a trained shape subspace. The active appearance Markov chain (AAMC). It integrates a statistical model of shape, appearance and motion. In the AAMC model, a Markov chain represents the switching of motion phases (poses), and several pairwise active appearance model (P-AAM) components characterize the shape, appearance and motion information for different motion phases. The second part of the thesis covers video mensuration, in which we have proposed a heightmeasuring algorithm with less human supervision, more flexibility and improved robustness. From videos acquired by an uncalibrated stationary camera, we first recover the vanishing line and the vertical point of the scene. We then apply a single view mensuration algorithm to each of the frames to obtain height measurements. Finally, using the LMedS as the cost function and the Robbins-Monro stochastic approximation (RMSA) technique to obtain the optimal estimate.
  • Thumbnail Image
    Item
    Shadow detection in videos acquired by stationary and moving cameras
    (2005-12-09) Trias, Antonio; Chellappa, Rama; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Shadow Detection has become a key issue in object detection, tracking and recognition problems. Object appearances might be completely changed by the effects of shading and shadows. Finding good algorithms for shadow detection and reducing shading effects in order to segment objects from video sequences, will enhance the performance of our detection, tracking and recognition algorithms. In this thesis, we present data, physics and model-driven approaches for detecting shadows and correcting shading effects. The effectiveness of these algorithms in video sequences acquired by stationary surveillance cameras and airborne platforms is illustrated.