Lp-TV model for structure extraction with end-to-end contour learning

Simple item page
dc.creatorSong, Chunwei
dc.creatorMaiseli, Baraka J.
dc.creatorZuo, Wangmeng
dc.creatorGao, Huijun
dc.date2019-05-05T11:42:07Z
dc.date2019-05-05T11:42:07Z
dc.date2017
dc.date.accessioned2021-05-03T13:17:00Z
dc.date.available2021-05-03T13:17:00Z
dc.descriptionStructure extraction is important for human perception. However, for various textured images, computers can hardly achieve this goal. Despite a plethora of studies to address the challenge, results from most previous methods contain unwanted artifacts and over-smoothed structures. Therefore, to address the weaknesses, we have proposed a variational model with end-to-end contour learning capability. Our formulation dwells in two observations: likelihood for representation of residual textures may be well abstracted using super Gaussian distribution, and edge metrics with semantic meaning may benefit structure preservation. The augmented Lagrangian method is adopted for optimal computation. Compared with classical approaches, our method offers a higher performance in structure extraction, including situations where the images have significant nonuniformity of the scale features.
dc.identifier978-1-5386-1127-2
dc.identifierhttp://hdl.handle.net/20.500.11810/5196
dc.identifier10.1109/IECON.2017.8217241
dc.identifier.urihttp://hdl.handle.net/20.500.11810/5196
dc.languageen
dc.publisherIEEE
dc.subjectImage edge detection
dc.subjectComputational modeling
dc.subjectTV
dc.subjectSemantics
dc.subjectGaussian distribution
dc.subjectMathematical model
dc.subjectAdaptation models
dc.titleLp-TV model for structure extraction with end-to-end contour learning
dc.typeConference Paper

Files

Collections

Department of Electronics and Telecommunications Engineering