Advances in variational image segmentation using AM-FM models: Regularized demodulation and probabilistic cue integration

Evangelopoulos, G; Kokkinos, I; Maragos, P

dc.contributor.author	Evangelopoulos, G	en
dc.contributor.author	Kokkinos, I	en
dc.contributor.author	Maragos, P	en
dc.date.accessioned	2014-03-01T02:43:06Z
dc.date.available	2014-03-01T02:43:06Z
dc.date.issued	2005	en
dc.identifier.issn	0302-9743	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/31230
dc.subject	Cue Integration	en
dc.subject	Derived Equivalence	en
dc.subject	Evolution Equation	en
dc.subject	Feature Extraction	en
dc.subject	Generic Model	en
dc.subject	Image Segmentation	en
dc.subject	Noise Robustness	en
dc.subject	Unsupervised Segmentation	en
dc.subject	Level Set Method	en
dc.subject.classification	Computer Science, Theory & Methods	en
dc.subject.other	Algorithms	en
dc.subject.other	Demodulation	en
dc.subject.other	Feature extraction	en
dc.subject.other	Level control	en
dc.subject.other	Mathematical models	en
dc.subject.other	Signal filtering and prediction	en
dc.subject.other	Textures	en
dc.subject.other	Variational techniques	en
dc.subject.other	AM-FM models	en
dc.subject.other	Berkeley segmentation	en
dc.subject.other	Complex textured images	en
dc.subject.other	Demodulation algorithm	en
dc.subject.other	Level set methods	en
dc.subject.other	Multiband filtering	en
dc.subject.other	Image segmentation	en
dc.title	Advances in variational image segmentation using AM-FM models: Regularized demodulation and probabilistic cue integration	en
heal.type	conferenceItem	en
heal.identifier.primary	10.1007/11567646_11	en
heal.identifier.secondary	http://dx.doi.org/10.1007/11567646_11	en
heal.language	English	en
heal.publicationDate	2005	en
heal.abstract	Current state-of-the-art methods in variational image segmentation using level set methods are able to robustly segment complex textured images in an unsupervised manner. In recent work, [18,19] we have explored the potential of AM-FM features for driving the unsupervised segmentation of a wide variety of textured images. Our first contribution in this work is at the feature extraction level, where we introduce a regularized approach to the demodulation of the AM-FM -modelled signals. By replacing the cascade of multiband filtering and subsequent differentiation with analytically derived equivalent filtering operations, increased noise-robustness can be achieved, while discretization problems in the implementation of the demodulation algorithm are alleviated. Our second contribution is based on a generative model we have recently proposed [18,20] that offers a measure related to the local prominence of a specific class of features, like edges and textures. The introduction of these measures as weighting terms in the evolution equations facilitates the fusion of different cues in a simple and efficient manner. Our systematic evaluation on the Berkeley segmentation benchmark demonstrates that this fusion method offers improved results when compared to our previous work as well as current state-of-the-art methods. © Springer-Verlag Berlin Heidelberg 2005.	en
heal.publisher	SPRINGER-VERLAG BERLIN	en
heal.journalName	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)	en
heal.bookName	LECTURE NOTES IN COMPUTER SCIENCE	en
dc.identifier.doi	10.1007/11567646_11	en
dc.identifier.isi	ISI:000233133600011	en
dc.identifier.volume	3752 LNCS	en
dc.identifier.spage	121	en
dc.identifier.epage	136	en