FEEC - School of Electrical and Computer Engineering Technical Presentation - Overview Leonardo Marques Rocha
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
FEEC - School of Electrical and Computer Engineering
Technical Presentation - OverviewLeonardo Marques Rocha
FEEC - School of Electrical and Computer Engineering
PhD student (2004-2008)- Expected graduation at December 17th- School of Electrical and Computer Engineering – Unicamp- Research area: Data Clustering
MSc degree (2001-2002)- Institute of Computing – Unicamp- Research area: Visualization of 3D images
Bachelor degree (1995-2000)- Department of Electrical Engineering – UFC- Research area: High-perfomance computing
Page 2
Background information
Education
FEEC - School of Electrical and Computer Engineering
Background information
Page 3
TOSHIBA/eiTV/Unicamp (2006-2007)- Design of a setup box the Brazilian
Digital TV Intel/OverPlay/Unicamp (2005-2006)
- Design of a 3D engine for Intel PDAs SBTVD/UNICAMP (2005)
- DSP port of H.264 video codec for Brazilian Digital TV
FAPESP Technical Training (TT5) (2002-2003)- Development of a 3D visualization
software (3D MR images)
Microsoft Research, Redmond (2009)- Research in 3D video
CNPq/UCSB (2008)- Split program at University of California,
Santa Barbara CNPq/Unicamp (2004-2008)
- Research in Data Clustering applied to Images
Capes/Unicamp (2001-2002)- Research in Visualization of 3D images /
Image Processing CENAPAD-NE/UFC (2000)
- Internship in ATM Networks CNPq/UFC (1996 – 1999)
- Research in Parallel Computing
Work Experience Scholarships / Internships
FEEC - School of Electrical and Computer Engineering
Identify natural groups in a database- Unsupervised (no prior knowledge)- Supervised (training phase)- Non-parametric (no shape is assumed)
Clustering based on optimum-path florest of a graph (added to IAPR)- 50x faster than SVM, 100x faster than Neural
Networks for supervised classification Applications in:
- Data mining- Pattern recognition- CBIR – Context based image retrieval- Image segmentation/analysis- Video compression- 3 patents (2 national and 1 international)
Demo avaliable:- http://www.ic.unicamp.br/~afalcao/libopf
Research
Page 4
Data clustering
FEEC - School of Electrical and Computer Engineering
Real-time visualization of 3D images based on shear-warp factorization
Software-based rendering using efficient data structure- Surface and volumetric rendering- Maximum Intensity Projection- Stereo rendering- Voxel classification
Applications in:- Terrain data analysis- Medical Imaging- Mechanical Inspection
Demo avaliable:- http://www.ic.unicamp.br/~afalcao/3dview
Research
Page 5
3D Visualization
FEEC - School of Electrical and Computer Engineering
C/C+ MATLAB OpenGL, OpenGL|ES Tcl/Tk, wxWidgets,VTK
Software
Page 6
Programming languages and toolkits
Development plataforms Linux Windows Embedded systems DSP boards
FEEC - School of Electrical and Computer Engineering
Thank you
Page 7
3D Visualization
Real-time visualization of 3D images based on shear-warp factorization
Software-based rendering based on efficient data structure:- Surface and volumetric rendering- Stereo rendering- Maximum Intensity Projection- Voxel classification
Applications in:- Medical Imaging- Mechanical inspection- Sonic data Visualization Demo avaliable:- http://www.ic.unicamp.br/~afalcao/3dview
contact: [email protected]
Thank you“Obrigado”