Is it an open door to common parallelization strategy for topological operators on SMP machines ? R. MAHMOUDI – A3SI Lab. 1
2 Summary Scientific and technical context PhD Objectives
3 Scientific and technical context (1) Image processingoperators Fourier Transformation Opening Thinning Dynamic redistribution Linear filters Closing Crest restoring Not-linear filters Euclidean Distance Transformation Thresholding Smoothing Attributed Filter Watershed Associated class Topological operators Morphological operators Local operators Point-to-Point operators Global operators R. MAHMOUDI – A3SI Laboratory– 2009 April
4 Scientific and technical context (2) (Associated class) Vs (Parallelizationstrategies) Global operators Topological operators Morphological operators Local operators Point-to-Point operators Sienstra  (2002) Wilkinson  (2007) Meijster   F. J. Seinstra, D. Koelma, and J. M. Geusebroek, “A software architecture for user transparent parallel image processing”.  M.H.F. Wilkinson, H. Gao, W.H. Hesselink, “Concurrent Computation of Attribute Filters on Shared Memory Parallel Machines”.  A. Meijster, J. B. T. M. Roerdink, and W. H. Hesselink, “A general algorithm for computing distance transforms in linear time” . R. MAHMOUDI – A3SI Laboratory– 2009 April
5 PhD Objectives (1) Topological operators Thinning operator  common parallelization strategy Crest restoring  2D and 3D smoothing  Watershed based on w-thinning  Watershed based on graph  Homotopic kernel transformation  Leveling kernel transformation   M. Couprie, F. N. Bezerra, and G. Bertrand, “Topological operators for grayscale image processing”,  M. Couprie, and G. Bertrand, “Topology preserving alternating sequential filter for smoothing 2D and 3D objects”.  G. Bertrand, “On Topological Watersheds”.  J. Cousty, M. Couprie, L. Najman and G. Betrand “Weighted fusion graphs: Merging properties and watersheds”.  G. Bertrand, J. C. Everat, and M. Couprie, "Image segmentation through operators based on topology“
6 PhD Objectives (2) Main Architectural Classes SISD machines SIMD machines MISD machines MIMD Machine : (Execute several instruction streams in parallel on different data) Shared Memory Machine Distributed Memory System CPU1 CPU2 CPU3 CPUn Random Access Memory
7 PhD Objectives (3) Needs Common parallelization strategy of topological operators on multi-core multithread architecture (MIMD Machines – Shared Memory System)? Main Objectives Unifyingparallelizationmethod of topologicaloperators class (Algorithmiclevel) Implementation Methodology and optimization techniques on multi-core multithread architecture (Architecture level).