TRACKING INFORMATION IN SAR IMAGE FORMATION AND CLASSIFICATION ALGORITHMS
| dc.contributor.author | Rajagopal, Abhejit | |
| dc.contributor.author | Radzicki, Vincent | |
| dc.contributor.author | Chandrasekaran, Shivkumar | |
| dc.contributor.author | Lee, Hua | |
| dc.date.accessioned | 2018-03-02T23:13:20Z | |
| dc.date.available | 2018-03-02T23:13:20Z | |
| dc.date.issued | 2017-10 | |
| dc.identifier.issn | 0884-5123 | |
| dc.identifier.issn | 0074-9079 | |
| dc.identifier.uri | http://hdl.handle.net/10150/626979 | |
| dc.description.abstract | Traditional target detection pipelines involve two sequential steps: the formation of a range-profile or likely-image, and the classification of likely targets within that image. Although it has been shown that target tracking in the RaDAR image-domain can be unnecessarily noisy, with more accurate and efficient implementations involving a direct analysis of the measured wavefield, image formation remains a desirable output in many applications due to its highly descriptive and interpretable nature. In this paper, we outline a mechanism for formalizing and accelerating this procedure in application-specific use cases. Enabled by recent advances in deep learning, we present a pipeline for automatically selecting an “optimal” filtered back-projection model, forming a likelyimage, and performing target recognition and classification. The architecture allows practitioners to track and optimize the flow of information throughout the pipeline, enabling applications that utilize only intermediate outputs of the algorithm. | |
| dc.description.sponsorship | International Foundation for Telemetering | en |
| dc.language.iso | en_US | en |
| dc.publisher | International Foundation for Telemetering | en |
| dc.relation.url | http://www.telemetry.org/ | en |
| dc.rights | Copyright © held by the author; distribution rights International Foundation for Telemetering | en |
| dc.rights.uri | http://rightsstatements.org/vocab/InC/1.0/ | |
| dc.title | TRACKING INFORMATION IN SAR IMAGE FORMATION AND CLASSIFICATION ALGORITHMS | en_US |
| dc.type | text | en |
| dc.type | Proceedings | en |
| dc.contributor.department | UCSB, Dept Electrical & Comp. Eng. | en |
| dc.identifier.journal | International Telemetering Conference Proceedings | en |
| dc.description.collectioninformation | Proceedings from the International Telemetering Conference are made available by the International Foundation for Telemetering and the University of Arizona Libraries. Visit http://www.telemetry.org/index.php/contact-us if you have questions about items in this collection. | |
| refterms.dateFOA | 2018-04-12T09:44:23Z | |
| html.description.abstract | Traditional target detection pipelines involve two sequential steps: the formation of a range-profile or likely-image, and the classification of likely targets within that image. Although it has been shown that target tracking in the RaDAR image-domain can be unnecessarily noisy, with more accurate and efficient implementations involving a direct analysis of the measured wavefield, image formation remains a desirable output in many applications due to its highly descriptive and interpretable nature. In this paper, we outline a mechanism for formalizing and accelerating this procedure in application-specific use cases. Enabled by recent advances in deep learning, we present a pipeline for automatically selecting an “optimal” filtered back-projection model, forming a likelyimage, and performing target recognition and classification. The architecture allows practitioners to track and optimize the flow of information throughout the pipeline, enabling applications that utilize only intermediate outputs of the algorithm. |
