Linear feature delineation in digital imagery using neural networks
dc.contributor.advisor | Hunt, Bobby R. | en_US |
dc.contributor.author | Sementilli, Philip Joseph, Jr., 1958- | |
dc.creator | Sementilli, Philip Joseph, Jr., 1958- | en_US |
dc.date.accessioned | 2013-04-03T13:09:42Z | |
dc.date.available | 2013-04-03T13:09:42Z | |
dc.date.issued | 1991 | en_US |
dc.identifier.uri | http://hdl.handle.net/10150/278012 | |
dc.description.abstract | A goal of many image analysis, image understanding and computer vision problems is the delineation of linear features. This thesis addresses the specific problem of operator guided road delineation in aerial photographs. Our solution to this problem applies the classical pattern recognition paradigm of feature extraction followed by pattern classification. The feature extraction process merges features extracted at different levels of a multi-resolution image pyramid to obtain a dichotomization of image coordinates into classes of road pixels and not road pixels. The road center line is estimated from this road pixel image using a generalized Hamming distance based decision scheme. An artificial neural network (ANN) architecture is developed which implements the generalized Hamming distance classifier. It is shown that the ANN implementation offers significant throughput improvements over sequential implementations. Results of applying the road delineation algorithm to digitized aerial photographs demonstrate delineation accuracy suitable for computer-aided cartography applications. | |
dc.language.iso | en_US | en_US |
dc.publisher | The University of Arizona. | en_US |
dc.rights | Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author. | en_US |
dc.subject | Geography. | en_US |
dc.subject | Engineering, Electronics and Electrical. | en_US |
dc.subject | Artificial Intelligence. | en_US |
dc.subject | Computer Science. | en_US |
dc.title | Linear feature delineation in digital imagery using neural networks | en_US |
dc.type | text | en_US |
dc.type | Thesis-Reproduction (electronic) | en_US |
thesis.degree.grantor | University of Arizona | en_US |
thesis.degree.level | masters | en_US |
dc.identifier.proquest | 1346430 | en_US |
thesis.degree.discipline | Graduate College | en_US |
thesis.degree.name | M.S. | en_US |
dc.identifier.bibrecord | .b27227091 | en_US |
refterms.dateFOA | 2018-08-27T12:22:28Z | |
html.description.abstract | A goal of many image analysis, image understanding and computer vision problems is the delineation of linear features. This thesis addresses the specific problem of operator guided road delineation in aerial photographs. Our solution to this problem applies the classical pattern recognition paradigm of feature extraction followed by pattern classification. The feature extraction process merges features extracted at different levels of a multi-resolution image pyramid to obtain a dichotomization of image coordinates into classes of road pixels and not road pixels. The road center line is estimated from this road pixel image using a generalized Hamming distance based decision scheme. An artificial neural network (ANN) architecture is developed which implements the generalized Hamming distance classifier. It is shown that the ANN implementation offers significant throughput improvements over sequential implementations. Results of applying the road delineation algorithm to digitized aerial photographs demonstrate delineation accuracy suitable for computer-aided cartography applications. |