Electronic Thesis/Dissertation


Time Order Reconstruction of Video Content Open Access

Downloadable Content

Download PDF

The global current community enjoys the free availability in data rich environment, but sometime suffers from an explosive growth of useless data known as the curse of dimensionality in a digital age. Thus, the scientific community has developed different pre-processing and post-processing to deal with the curse of dimensionality. For example, a smart organization sampling principle called compressive sensing (Gilbert Walter, IEEE/IT 1992, and John Donoho, loc cit 2006). Specifically, when one doesn't need the frequency of those sinusoidal components, one shall not sense it using the Fourier Transform, but with a physical group wave packet called the Wavelet Transform. For post-processing, conventional data compressions or removing redundancy in data such as JPEG, MPEG, Face Detection etc. and dimensionality reduction algorithms, i.e. PCA, ISOMAP, Laplacian Eigenmap, etc. can either reduce the storage requirements of data or keep the original degrees of freedom but in a lower dimensional manifold. To deal with the curse of dimensionality in persistent surveillance, different strategies mentioned above can be applied but in the end result, a potential loss of association among different sensory tracks, i.e. video and audio, may occur and observers cannot answer "who speaks what, when and where." In this dissertation, a spatiotemporal ordering reconstruction is given by a higher order asymmetric graph theory to associate different sensory modalities, such as video and audio. The mathematical approaches and theorems developed in this dissertation can be generalized for different applications such as (i) smart grid infrastructure, (ii) biomedical wellness web, and (iii) hyper-spectral data for precision farming, etc.

Author Language Keyword Date created Type of Work Rights statement GW Unit Degree Advisor Committee Member(s) Persistent URL