Acivs 2008 Abstracts
Paper 288: Marked point processes for object detection in high resolution images. Application to Earth observation and cartography
To be announced.
Paper 289: To be announced
To be announced.
Paper 290: Geometrically oriented wavelet transforms using rational dilations
Wavelet transforms (WTs) are used in various image processing applications
including compression, denoising, deconvolution, de-quantization, and
classification. In the dyadic WT, the resolution is doubled from one one level
to the next. For WTs based on rational dilation factors between one and two,
the resolution increases more gradually; however, the design and implementation
of critically-sampled 'rational' WTs can be difficult. This talk will show that
by over-sampling, the design and implementation of rational WTs is greatly
simplified and their performance improved. Perfect reconstruction filters can
be designed to have a specified number of vanishing moments and minimal length,
or can be designed in the frequency domain. In addition, the presented
overcomplete WT (or 'frame') is approximately shift-invariant and compares
favorably with the un-decimated WT yet is less redundant. We will also show
geometrically oriented rational WTs based on the dual-tree complex WT and
steerable-pyramid schemes, with applications in image processing.
This software generating these pages is ©
(not Ghent University), 2002-3205.
All rights reserved.
The data on this page is ©
Acivs 2008. All rights reserved.
The server hosting this website is owned by the department of Telecommunications and
Information Processing (TELIN) of Ghent University.
Problems with the website should be reported to
This page was generated on Tuesday June 22nd, 2021 22:20:05.