Wavelet-Based Signal and Image Processing for Target Recognition
Abstract
For the initial year of the project new wavelet based signal and image processing algorithms were developed, specifically directed towards usefulness in target recognition applications. Classical spatial and frequency domain image processing algorithms were generalized to process discrete wavelet transform (DWT) data. Results include adaptation of classical filtering, smoothing and interpolation techniques to DWT. From 2003 the research direction changed, in keeping with changes in the direction of ONR's Probability and Statistics Program. A sabbatical leave was devoted to broadening the P.I.'s expertise in aspects of Pattern Recognition. Research was also done on-site at the Naval Surface Warfare Center, Dahlgren, Virginia, where collaborations were formed with NSWC scientists. These resulted, inter alia, in the development of a new tracking algorithm for laser guided weapons. While at NSWC, the P.I. presented tutorial courses and seminars to NSWC scientists. The grant supported 4 graduate students who performed software development and theoretical derivations. During the grant period, 8 peer-reviewed papers were published.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2002
- Accession Number
- ADA430988
Entities
People
- Barry G. Sherlock
Organizations
- University of North Carolina at Charlotte