Vision-Based Place Recognition for       Autonomous Robots       First Seminar - Project Overview       Team Members:     ...
Prof.Dr Mohamed Roushdy       Dr. Mohamed Abdel Megeed       Dr. Safaa Amin       T.A. Mohamed Fathy       Supervisors    ...
Agenda         Objective                  What        Methodology               How         Theoretical Background        ...
Objective                              Where am I ?                                             4Saturday, December 11, 20...
Theoretical Background       Where are we in the field of computer science ?                                Pattern        ...
Motivation           Interested in robot vision.           Has many applications, help in rescue missions.           Co-op...
Problem Definition                                  Meeting Room   Vision-Based Place Recognition for Autonomous Robot     ...
Problem Definition       SLAM                Simultaneous Localization And Mapping.                Our problem is to focus ...
Conventional PR System Architecture                              Training Phase       Testing Phase                       ...
Related Work       ImageCLEF       ImageCLEF 2010    22-23th September 2010 A yearly contest which focuses on information ...
Related Work                             1st Position                                 CVG          Olivier Saurer, Friedri...
Related Work                                      4th Position                                Centro Gustavo Stefanini    ...
Agenda                                                Methodology         Objective                  What                 ...
Methodology       1. Improve previous work           Combine the pros of each group.           Try to avoid their mistakes...
Methodology       2. Adaptive Multi-Scale Classification                   What is the meaning of an environment?          ...
Methodology       2. Adaptive Multi-Scale Classification              Unrecognized                 Image                   ...
Challenges           Objects’ appearance varies due to                Cluttered background.                Difference in i...
Testing Platforms           1) Bielefeld University’s workbench           2) ImageCLEF’s testing dataset.           3) Bui...
Development Tools                        C++   Matlab                                       19Saturday, December 11, 2010 ...
Agenda         Objective                  What        Methodology                How         Theoretical Background       ...
Time Plan                               2010                               2011                        Sep   Oct   Nov   D...
Our Progress             Survey 1               Survey 2        (Project Overview)      (Project in Depth)        -Problem...
Next Objective       Simple Pattern Recognition System  Image                                      Decision ImageCLEF     ...
References           “The Robot Vision Track at ImageCLEF 2010”Andrzej Pronobis, Marco           Fornoni, Henrik I. Christ...
Contacts                Blog: autovpr.wordpress.com           Ahmed Saher Maher                    a7med.saher@gmail.com  ...
Upcoming SlideShare
Loading in...5
×

VBPR 1st seminar

524

Published on

Published in: Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
524
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
6
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

VBPR 1st seminar

  1. 1. Vision-Based Place Recognition for Autonomous Robots First Seminar - Project Overview Team Members: Ahmed Abd-El Fattah Mohammed Ahmed Saher Maher Mourad Aly Mourad Yasser Hassan Ahmed 1Saturday, December 11, 2010 1
  2. 2. Prof.Dr Mohamed Roushdy Dr. Mohamed Abdel Megeed Dr. Safaa Amin T.A. Mohamed Fathy Supervisors 2Saturday, December 11, 2010 2
  3. 3. Agenda Objective What Methodology How Theoretical Background Improve previous work Motivation Adaptive Multi-Scale Classification Problem Definition Challenges System Architecture Testing Platform Conventional Pattern Recognition System Architecture Development Tools Related Work Time Plan When ImageCLEF Our Progress Top Related Systems Next Objective References 3Saturday, December 11, 2010 3
  4. 4. Objective Where am I ? 4Saturday, December 11, 2010 4
  5. 5. Theoretical Background Where are we in the field of computer science ? Pattern Image Recognition Processing Computer Computer Vision Science Artificial Intelligence 5Saturday, December 11, 2010 5
  6. 6. Motivation Interested in robot vision. Has many applications, help in rescue missions. Co-operation between our university and Bielefeld University. 6Saturday, December 11, 2010 6
  7. 7. Problem Definition Meeting Room Vision-Based Place Recognition for Autonomous Robot What does it mean ? 7Saturday, December 11, 2010 7
  8. 8. Problem Definition SLAM Simultaneous Localization And Mapping. Our problem is to focus on localization issues in most SLAM systems. 8Saturday, December 11, 2010 8
  9. 9. Conventional PR System Architecture Training Phase Testing Phase Sensing Sensing Pre-Processing Pre-Processing Feature Extraction Feature Extraction Training Classification Knowledge Decision Base 9Saturday, December 11, 2010 9
  10. 10. Related Work ImageCLEF ImageCLEF 2010 22-23th September 2010 A yearly contest which focuses on information retrievalusing image processing. It branches to many applications including robot vision. 10Saturday, December 11, 2010 10
  11. 11. Related Work 1st Position CVG Olivier Saurer, Friedrich Fraundorfer, and Marc Pollefeys – “Visual localization using global visual features and vanishing points” - ETH Zurich, Switzerland Pros Cons Focused on feature extraction Used very primitive classification phase developed new feature methods. extraction algorithms. 11Saturday, December 11, 2010 11
  12. 12. Related Work 4th Position Centro Gustavo Stefanini W.Lucetti, E. Luchetti – “Combination of Classifiers for Indoor Room Recognition” - Gustavo Stefanini Research Center - Padua, 23 September 2010 Pros Cons Focused on classification phase Used very primitive feature developed many new extraction methods. combination of classifiers. 12Saturday, December 11, 2010 12
  13. 13. Agenda Methodology Objective What How Theoretical Background Improve previous work Motivation Adaptive Multi-Scale Classification Problem Definition Challenges System Architecture Testing Platform Conventional Pattern Recognition System Architecture Development Tools Time Plan Related Work When ImageCLEF Our Progress Top Related Systems Next Objective References 13Saturday, December 11, 2010 13
  14. 14. Methodology 1. Improve previous work Combine the pros of each group. Try to avoid their mistakes and cons. 14Saturday, December 11, 2010 14
  15. 15. Methodology 2. Adaptive Multi-Scale Classification What is the meaning of an environment? Env 1 Env 2 Env 3 Kitchen,Bathroom LivingRoom,Office BedRoom,Corridor White illumination White illumination Yellow illumination Color White Color Blue Color Brown How can the system differentiate between environments? Differentiation using discriminative features only. 15Saturday, December 11, 2010 15
  16. 16. Methodology 2. Adaptive Multi-Scale Classification Unrecognized Image PR System for Env 1 (Kitchen,Bathroom) Environment PR System Start Current Operating Environment for Env 2 Decision Identifier (Office,Library) PR System for Env 3 (Bedrooms) Simple classification Full-scale PR systems 16Saturday, December 11, 2010 16
  17. 17. Challenges Objects’ appearance varies due to Cluttered background. Difference in illumination. Imaging conditions. Recognition algorithms perform differently with different environments. It’s difficult to find a solution that is both resource efficient and perform with high accuracy, due to the very limited resources of a mobile robot. 17Saturday, December 11, 2010 17
  18. 18. Testing Platforms 1) Bielefeld University’s workbench 2) ImageCLEF’s testing dataset. 3) Build our own data acquisition tool. 18Saturday, December 11, 2010 18
  19. 19. Development Tools C++ Matlab 19Saturday, December 11, 2010 19
  20. 20. Agenda Objective What Methodology How Theoretical Background Improve previous work Motivation Adaptive Multi-Scale Classification Problem Definition Challenges System Architecture Testing Platform Conventional Pattern Recognition System Architecture Development Tools Related Work Time Plan When ImageCLEF Our Progress Top Related Systems Next Objective References 20Saturday, December 11, 2010 20
  21. 21. Time Plan 2010 2011 Sep Oct Nov Dec Jan Feb Mar April May June July Feasibility Study Survey(1)-Project Overview Survey(2)-Project In DepthDeveloping Simple PR System Iterative System Development Deployment Documentation 21Saturday, December 11, 2010 21
  22. 22. Our Progress Survey 1 Survey 2 (Project Overview) (Project in Depth) -Problem definition. -Description of each -Commonly used algorithm mentioned algorithms in pattern in survey 1 recognition. 22Saturday, December 11, 2010 22
  23. 23. Next Objective Simple Pattern Recognition System Image Decision ImageCLEF ( Class 1 Or Data Set Class 2 ) Simple PR The system has theSystem to differentiate between 2 ability classes. 23Saturday, December 11, 2010 23
  24. 24. References “The Robot Vision Track at ImageCLEF 2010”Andrzej Pronobis, Marco Fornoni, Henrik I. Christensen, and Barbara Caputo. “Evaluation of Bayes, ICA, PCA and SVM Methods for Classification”, V.C.Chen. Radar Division, US Naval Research Laboratory. Olivier Saurer, Friedrich Fraundorfer, and Marc Pollefeys – “Visual localization using global visual features and vanishing points” - ETH Zurich, Switzerland W.Lucetti, E. Luchetti – “Combination of Classifiers for Indoor Room Recognition” - Gustavo Stefanini Research Center - Padua, 23 September 2010 24Saturday, December 11, 2010 24
  25. 25. Contacts Blog: autovpr.wordpress.com Ahmed Saher Maher a7med.saher@gmail.com Ahmed Abd El-Fattah ahmed.abdelfattah1@live.com Mourad Aly Mourad mouraad@windowslive.com Yasser Hassan Ahmed yasserhtd@hotmail.com Thanks! 25Saturday, December 11, 2010 25
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×