Transform pre-processing for neural networks for object recognition and localization with sonar

Simple item page
dc.citation.epage128en_US
dc.citation.spage114en_US
dc.citation.volumeNumber5102en_US
dc.contributor.authorBarshan, Billuren_US
dc.contributor.authorAyrulu, Birselen_US
dc.coverage.spatialOrlando, Florida, United Statesen_US
dc.date.accessioned2016-02-08T11:55:38Zen_US
dc.date.available2016-02-08T11:55:38Zen_US
dc.date.issued2003en_US
dc.departmentDepartment of Electrical and Electronics Engineeringen_US
dc.descriptionDate of Conference: 21-25 April 2003en_US
dc.descriptionConference Name: SPIE Aerosense, 2003en_US
dc.description.abstractWe investigate the pre-processing of sonar signals prior to using neural networks for robust differentiation of commonly encountered features in indoor environments. Amplitude and time-of-flight measurement patterns acquired from a real sonar system are pre-processed using various techniques including wavelet transforms, Fourier and fractional Fourier transforms, and Kohonen's self-organizing feature map. Modular and non-modular neural network structures trained with the back-propagation and generating-shrinking algorithms are used to incorporate learning in the identification of parameter relations for target primitives. Networks trained with the generating-shrinking algorithm demonstrate better generalization and interpolation capability and faster convergence rate. The use of neural networks trained with the back-propagation algorithm, usually with fractional Fourier transform or wavelet pre-processing results in near perfect differentiation, around 85% correct range estimation and around 95% correct azimuth estimation, which would be satisfactory in a wide range of applications. Neural networks can differentiate more targets, employing only a single sensor node, with a higher correct differentiation percentage than achieved with previously reported methods employing multiple sensor nodes. The success of the neural network approach shows that the sonar signals do contain sufficient information to differentiate a considerable number of target types, but the previously reported methods are unable to resolve this identifying information. This work can find application in areas where recognition of patterns hidden in sonar signals is required. Some examples are system control based on acoustic signal detection and identification, map building, navigation, obstacle avoidance, and target-tracking applications for mobile robots and other intelligent systems.en_US
dc.description.provenanceMade available in DSpace on 2016-02-08T11:55:38Z (GMT). No. of bitstreams: 1 bilkent-research-paper.pdf: 70227 bytes, checksum: 26e812c6f5156f83f0e77b261a471b5a (MD5) Previous issue date: 2003en
dc.identifier.doi10.1117/12.499753en_US
dc.identifier.issn0277-786Xen_US
dc.identifier.urihttp://hdl.handle.net/11693/27523en_US
dc.language.isoEnglishen_US
dc.publisherSPIEen_US
dc.relation.isversionofhttp://dx.doi.org/10.1117/12.499753en_US
dc.source.titleProceedings of SPIE Vol. 5102, lndependent Component Analyses, Wavelets, and Neural Networksen_US
dc.subjectAcoustic signal processingen_US
dc.subjectArtificial neural networksen_US
dc.subjectDiscrete wavelet transformen_US
dc.subjectFeature extractionen_US
dc.subjectFractional Fourier transformen_US
dc.subjectInput pre-processingen_US
dc.subjectLearningen_US
dc.subjectObject recognitionen_US
dc.subjectPosition estimationen_US
dc.subjectSonar sensingen_US
dc.subjectTarget differentiationen_US
dc.subjectTarget localizationen_US
dc.titleTransform pre-processing for neural networks for object recognition and localization with sonaren_US
dc.typeConference Paperen_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Transform pre-processing for neural networks for object recognition and localization with sonar.pdf
Size:
1.53 MB
Format:
Adobe Portable Document Format
Description:
Full printable version
Download

Collections

Department of Electrical and Electronics Engineering