Multiresolution Methods in Systems, Signals and Images
Abstract
In this final report we summarize our research accomplishments under this grant: (a) research on our multiresolution estimation methodology to solve a variety of difficult image and large scale data assimilation problems and to extend the theoretical domain of applicability of our methodology considerably to allow for the fusion of multisensor data; (b) the development of several new methods involving synthetic aperture radar (SAR) data including multiresolution segmentation and compression algorithms and an estimation-theoretic approach to moving target SAR image formation; (c) the development of a new nonlinear multiresolution image evolution algorithm that produces extremely robust image segmentations; (d) the development of a new variation on so-called matching pursuit methods with application to robust and stable feature extraction and recognition of objects; and (e) the development of hierarchical and wavelet-based methods for the design of joint detection algorithms for multiple access communication in the presence of highly correlated user signals. In addition a number of significant transitions and interactions resulting from our work are described.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1998
- Accession Number
- ADA341363
Entities
People
- Alan S. Willsky
Organizations
- Massachusetts Institute of Technology