🚀 go-pugleaf

SWISS FEDERAL INSTITUTE OF TECHNOLOGY (EPFL)
Department of Electrical Engineering
Signal Processing Laboratory (LTS)

Prof. Dr. Murat KUNT
CH-1015 Lausanne    Switzerland
Phone	 	(41 21) 693 26 26
Secretary	(41 21) 693 26 01 or 26 24
Telex 		454 062  EPFE CH
Telefax 	(41 21) 693 26 03 (direct) or 693 46 60
	



CONFERENCE ANNOUNCEMENT

	
The EPFL Signal Processing Laboratory is happy to announce its next
conference, entitled:


	Morphological Multiscale Segmentation for Image Coding


Speaker:	Dr. Philippe Salembier
		Universite polytechnique de Catalogne
		Polytechnical University of Catalunya
		Department of Signal Theory and Communications
		Barcelona - Spain

Date:		Friday December 18th, 1992 - 11:00 h
Place:		Room ELD 224 - EPFL


			ABSTRACT

This presentation deals with a morphological approach to the problem
of unsupervised image segmentation for image coding. The proposed
technique relies on a multiscale approach which allows a hierarchical
processing of the data ranging from the most global scale to the most
detailed one. At each scale, the algorithm relies on four steps:
preprocessing, feature extraction, decision and quality estimation.
The goal of the preprocessing step is to simplify the original signal
which is too complex to be processed at once. After a brief
introduction about Mathematical Morphology, we will show that
Morphological filters by reconstruction are very attractive for this
purpose because they simplify without corrupting the contour
information. The feature extraction intends to extract the pertinent
parameters for assessing the degree of homogeneity of the regions. To
this goal, morphological techniques extracting flat or contrasted
regions (Top Hat) are very efficient. The decision step defines
precisely the contours of the regions. This decision is achieved by a
watershed algorithm. Finally, the quality estimation is used to
compute the information that has to be further processed by the next
scale to improve the segmentation result. The estimation is achieved
by actually coding the image. The resulting segmentation is very
robust and can deal with very different type of images. Moreover, it
is very efficient in terms of computational load.