File Download
  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Learning-based nonparametric autofocusing for digital holography

TitleLearning-based nonparametric autofocusing for digital holography
Authors
Issue Date2018
PublisherOptical Society of America. The Journal's web site is located at https://www.osapublishing.org/optica/home.cfm
Citation
Optica, 2018, v. 5 n. 4, p. 337-344 How to Cite?
AbstractIn digital holography, it is crucial to extract the object distance from a hologram in order to reconstruct its amplitude and phase. This is known as autofocusing, which is conventionally solved by first reconstructing a stack of images, and then the sharpness of each reconstructed image is computed using a focus metric such as entropy or variance. The distance corresponding to the sharpest image is considered the focal position. This approach, while effective, is computationally demanding and time-consuming. To cope with this problem, we turn to machine learning, where we cast the autofocusing as a regression problem, with the focal distance being a continuous response corresponding to each hologram. Therefore, distance estimation is converted to hologram prediction, which we solve by designing a powerful convolutional neural network trained by a set of holograms acquired a priori. Experimental results show that this allows fast autofocusing without reconstructing an image stack, even when the physical parameters of the optical setup are unknown.
Persistent Identifierhttp://hdl.handle.net/10722/259275
ISSN
2017 Impact Factor: 7.536
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorRen, Z-
dc.contributor.authorXu, Z-
dc.contributor.authorLam, EYM-
dc.date.accessioned2018-09-03T04:04:17Z-
dc.date.available2018-09-03T04:04:17Z-
dc.date.issued2018-
dc.identifier.citationOptica, 2018, v. 5 n. 4, p. 337-344-
dc.identifier.issn2334-2536-
dc.identifier.urihttp://hdl.handle.net/10722/259275-
dc.description.abstractIn digital holography, it is crucial to extract the object distance from a hologram in order to reconstruct its amplitude and phase. This is known as autofocusing, which is conventionally solved by first reconstructing a stack of images, and then the sharpness of each reconstructed image is computed using a focus metric such as entropy or variance. The distance corresponding to the sharpest image is considered the focal position. This approach, while effective, is computationally demanding and time-consuming. To cope with this problem, we turn to machine learning, where we cast the autofocusing as a regression problem, with the focal distance being a continuous response corresponding to each hologram. Therefore, distance estimation is converted to hologram prediction, which we solve by designing a powerful convolutional neural network trained by a set of holograms acquired a priori. Experimental results show that this allows fast autofocusing without reconstructing an image stack, even when the physical parameters of the optical setup are unknown.-
dc.languageeng-
dc.publisherOptical Society of America. The Journal's web site is located at https://www.osapublishing.org/optica/home.cfm-
dc.relation.ispartofOptica-
dc.rightsOptica. Copyright © Optical Society of America.-
dc.rightsThis work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.-
dc.titleLearning-based nonparametric autofocusing for digital holography-
dc.typeArticle-
dc.identifier.emailLam, EYM: elam@eee.hku.hk-
dc.identifier.authorityLam, EYM=rp00131-
dc.description.naturepublished_or_final_version-
dc.identifier.doi10.1364/OPTICA.5.000337-
dc.identifier.hkuros288775-
dc.identifier.volume5-
dc.identifier.issue4-
dc.identifier.spage337-
dc.identifier.epage344-
dc.identifier.isiWOS:000430601100003-
dc.publisher.placeUnited States-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats