Automatic detection of geospatial objects using multiple hierarchical segmentations

Simple item page
dc.citation.epage2111en_US
dc.citation.issueNumber7en_US
dc.citation.spage2097en_US
dc.citation.volumeNumber46en_US
dc.contributor.authorAkçay, H. G.en_US
dc.contributor.authorAksoy, S.en_US
dc.date.accessioned2016-02-08T10:08:28Z
dc.date.available2016-02-08T10:08:28Z
dc.date.issued2008-07en_US
dc.departmentDepartment of Computer Engineeringen_US
dc.description.abstractThe object-based analysis of remotely sensed imagery provides valuable spatial and structural information that is complementary to pixel-based spectral information in classification. In this paper, we present novel methods for automatic object detection in high-resolution images by combining spectral information with structural information exploited by using image segmentation. The proposed segmentation algorithm uses morphological operations applied to individual spectral bands using structuring elements in increasing sizes. These operations produce a set of connected components forming a hierarchy of segments for each band. A generic algorithm is designed to select meaningful segments that maximize a measure consisting of spectral homogeneity and neighborhood connectivity. Given the observation that different structures appear more clearly at different scales in different spectral bands, we describe a new algorithm for unsupervised grouping of candidate segments belonging to multiple hierarchical segmentations to find coherent sets of segments that correspond to actual objects. The segments are modeled by using their spectral and textural content, and the grouping problem is solved by using the probabilistic latent semantic analysis algorithm that builds object models by learning the object-conditional probability distributions. The automatic labeling of a segment is done by computing the similarity of its feature distribution to the distribution of the learned object models using the Kullback-Leibler divergence. The performances of the unsupervised segmentation and object detection algorithms are evaluated qualitatively and quantitatively using three different data sets with comparative experiments, and the results show that the proposed methods are able to automatically detect, group, and label segments belonging to the same object classes. © 2008 IEEE.en_US
dc.identifier.doi10.1109/TGRS.2008.916644en_US
dc.identifier.issn0196-2892
dc.identifier.urihttp://hdl.handle.net/11693/23067
dc.language.isoEnglishen_US
dc.publisherInstitute of Electrical and Electronics Engineersen_US
dc.relation.isversionofhttp://dx.doi.org/10.1109/TGRS.2008.916644en_US
dc.source.titleIEEE Transactions on Geoscience and Remote Sensingen_US
dc.subjectHierarchical segmentationen_US
dc.subjectImage segmentationen_US
dc.subjectObject-based analysisen_US
dc.subjectUnsupervised object detectionen_US
dc.subjectBoolean functionsen_US
dc.subjectFeature extractionen_US
dc.subjectImage processingen_US
dc.subjectInformation theoryen_US
dc.subjectLearning algorithmsen_US
dc.subjectMathematical morphologyen_US
dc.subjectProbabilityen_US
dc.subjectProbability distributionsen_US
dc.subjectRisk assessmenten_US
dc.subjectSet theoryen_US
dc.subjectApplied (CO)en_US
dc.subjectAutomatic detectionen_US
dc.subjectAutomatic labellingen_US
dc.subjectComparative experimentsen_US
dc.subjectConditional probability distributionsen_US
dc.subjectConnected componentsen_US
dc.subjectData setsen_US
dc.subjectDifferent scalesen_US
dc.subjectFeature distributionen_US
dc.subjectGeneric algorithmsen_US
dc.subjectGeo spatial objectsen_US
dc.subjectGrouping problemsen_US
dc.subjectHigh-resolution (HR) imagesen_US
dc.subjectIndividual (PSS 544-7)en_US
dc.subjectKullback leibler divergence (KLD)en_US
dc.subjectMorphological operationsen_US
dc.subjectNew algorithmen_US
dc.subjectNovel methodsen_US
dc.subjectObject classesen_US
dc.subjectObject detectionen_US
dc.subjectObject detection algorithmsen_US
dc.subjectObject modellingen_US
dc.subjectObject-baseden_US
dc.subjectProbabilistic latent semantic analysis (PLSA)en_US
dc.subjectRemotely sensed imagery (RSI)en_US
dc.subjectSegmentation algorithmsen_US
dc.subjectSpectral bandingen_US
dc.subjectSpectral informationsen_US
dc.subjectStructural informationsen_US
dc.subjectStructuring element (SE)en_US
dc.subjectUnsupervised segmentationen_US
dc.subjectObject recognitionen_US
dc.subjectAlgorithmsen_US
dc.subjectImage analysisen_US
dc.subjectInformation retrievalen_US
dc.subjectLabelsen_US
dc.subjectMathematical modelsen_US
dc.subjectRemote sensingen_US
dc.titleAutomatic detection of geospatial objects using multiple hierarchical segmentationsen_US
dc.typeArticleen_US
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Automatic detection of geospatial objects using multiple hierarchical segmentations.pdf
Size:
3.07 MB
Format:
Adobe Portable Document Format
Description:
Full printable version
Download
Collections
Department of Computer Engineering