Underwater UXO classification using matched subspace classifier with synthetic sparse dictionaries
| dc.contributor.author | Hall, John Joseph, author | |
| dc.contributor.author | Azimi-Sadjadi, Mahmood R., advisor | |
| dc.contributor.author | Morton, Jade (Yu), committee member | |
| dc.contributor.author | Kirby, Michael, committee member | |
| dc.date.accessioned | 2016-08-18T23:11:35Z | |
| dc.date.available | 2016-08-18T23:11:35Z | |
| dc.date.issued | 2016 | |
| dc.description.abstract | This work is concerned with the development of a system for the discrimination of military munitions and unexploded ordnances (UXO) from non- UXO's, man-made objects, and other clutter in shallow underwater environments. In this thesis a thorough overview is given of the Matched Subspace Classification (MSC) framework and extensions of this framework when applied to this difficult problem. Acoustic color (AC) features corresponding to calibrated target strength, as a function of frequency and look angle, are generated from the raw sonar returns for munition characterization. Three variations of the signal subspace matching framework when used for classifying AC features are discussed in this work. The systems proposed are then exclusively trained using synthetic sonar data and then tested using real datasets collected from a side-looking sonar system. These real datasets were collected during three different controlled sonar experiments, PondEX09, PondEX10, and the Target and Reverberation Experiment 2013 (TREX13). Classification is performed on the AC features extracted from the all datasets and the performance of the linear sparse variations of the MSC are bench marked against a non-linear kernel form of the MSC. Classification results are presented using standard performance metrics such as Receiver Operating Characteristic (ROC) curve and confusion matrices. It was found that a Locality Preserving variation of the popular K-SVD algorithm (LP K-SVD) provided the best linear subspace matrix for class discrimination across all datasets, with relatively high probability of correct classification even on the most difficult dataset. Future work motivated by this research will also briefly be discussed. | |
| dc.format.medium | born digital | |
| dc.format.medium | masters theses | |
| dc.identifier | Hall_colostate_0053N_13798.pdf | |
| dc.identifier.uri | http://hdl.handle.net/10217/176755 | |
| dc.language | English | |
| dc.language.iso | eng | |
| dc.publisher | Colorado State University. Libraries | |
| dc.relation.ispartof | 2000-2019 | |
| dc.rights | Copyright and other restrictions may apply. User is responsible for compliance with all applicable laws. For information about copyright law, please see https://libguides.colostate.edu/copyright. | |
| dc.subject | sonar | |
| dc.subject | underwater | |
| dc.subject | classification | |
| dc.subject | UXO | |
| dc.subject | Sparse | |
| dc.title | Underwater UXO classification using matched subspace classifier with synthetic sparse dictionaries | |
| dc.type | Text | |
| dcterms.rights.dpla | This Item is protected by copyright and/or related rights (https://rightsstatements.org/vocab/InC/1.0/). You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s). | |
| thesis.degree.discipline | Electrical and Computer Engineering | |
| thesis.degree.grantor | Colorado State University | |
| thesis.degree.level | Masters | |
| thesis.degree.name | Master of Science (M.S.) |
Files
Original bundle
1 - 1 of 1
Loading...
- Name:
- Hall_colostate_0053N_13798.pdf
- Size:
- 912.38 KB
- Format:
- Adobe Portable Document Format