Adaptive triangulations

dc.contributor.authorMaizlish, Oleksandr
dc.contributor.examiningcommitteeZorboska, Nina (Mathematics) Thomas, Gabriel (Electrical & Computer Engineering) de Boor, Carl (Computer Science, University of Wisconsin-Madison)en_US
dc.contributor.supervisorKopotun, Kirill (Mathematics)en_US
dc.date.accessioned2014-04-17T17:05:17Z
dc.date.available2014-04-17T17:05:17Z
dc.date.issued2014-04-17
dc.degree.disciplineMathematicsen_US
dc.degree.levelDoctor of Philosophy (Ph.D.)en_US
dc.description.abstractIn this dissertation, we consider the problem of piecewise polynomial approximation of functions over sets of triangulations. Recently developed adaptive methods, where the hierarchy of triangulations is not fixed in advance and depends on the local properties of the function, have received considerable attention. The quick development of these adaptive methods has been due to the discovery of the wavelet transform in the 1960's, probably the best tool for image coding. Since the mid 80's, there have been many attempts to design `Second Generation' adaptive techniques that particularly take into account the geometry of edge singularities of an image. But it turned out that almost none of the proposed `Second Generation' approaches are competitive with wavelet coding. Nevertheless, there are instances that show deficiencies in the wavelet algorithms. The method suggested in this dissertation incorporates the geometric properties of convex sets in the construction of adaptive triangulations of an image. The proposed algorithm provides a nearly optimal order of approximation for cartoon images of convex sets, and is based on the idea that the location of the centroid of certain types of domains provides a sufficient amount of information to construct a 'good' approximation of the boundaries of those domains. Along with the theoretical analysis of the algorithm, a Matlab code has been developed and implemented on some simple cartoon images.en_US
dc.description.noteMay 2014en_US
dc.identifier.urihttp://hdl.handle.net/1993/23529
dc.language.isoengen_US
dc.rightsopen accessen_US
dc.subjectadaptive approximationen_US
dc.subjecttriangulationen_US
dc.subjectimage compressionen_US
dc.titleAdaptive triangulationsen_US
dc.typedoctoral thesisen_US
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
maizlish_oleksandr.pdf
Size:
2.03 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
2.25 KB
Format:
Item-specific license agreed to upon submission
Description: