CIS 9601 Spring 2010
Computer Graphics and Image Processing - Modern Approaches to Make Robots Recognize the World
Instructor: Rolf Lakaemper
Office hours: by appointment (Phone: 215 204 7996, email: lakamper@temple.edu), or just come in (I'm still on sabbatical in Fall 09).
WELCOME!
The course will give a thorough introduction to classical methods of Computer Vision and their applications in robotics (robot vision). The students will first be introduced to image manipulation using methods of linear and non linear signal processing, as well as statistical methods. Further on, the course will introduce approaches to current problems in image analysis, e.g. object segmentation, object recognition and face recognition. Extending classic computer vision, which is typically based on camera images, the course will cover data analysis of different sensors used in robotics (sonar, 3D laser range finders, range cameras). The course will build on theoretical foundation, but will emphasize on hands on experience using MATLAB/JAVA/C as programming language.
Topics:
Assignments:
Office hours: by appointment (Phone: 215 204 7996, email: lakamper@temple.edu), or just come in (I'm still on sabbatical in Fall 09).
Schedule:
Thursday 5:30 - 8 PM Tuttleman Learning Center 1A
Textbook:
- Gonzales/Woods: Digital Image Processing Third Edition, Prentice Hall ISBN-13:978-0-13-168728-8
- Different Research Articles
- COMMENTS about HW2 are here
- COMMENTS about HW3 are here
- COMMENTS about HW4 are here
- COMMENTS about HW5 are here
WELCOME!
The course will give a thorough introduction to classical methods of Computer Vision and their applications in robotics (robot vision). The students will first be introduced to image manipulation using methods of linear and non linear signal processing, as well as statistical methods. Further on, the course will introduce approaches to current problems in image analysis, e.g. object segmentation, object recognition and face recognition. Extending classic computer vision, which is typically based on camera images, the course will cover data analysis of different sensors used in robotics (sonar, 3D laser range finders, range cameras). The course will build on theoretical foundation, but will emphasize on hands on experience using MATLAB/JAVA/C as programming language.
The first part of the course will mostly be in lecture style. The second part will consist of student (group) presentations of projects. Grading will be based on class participation (incl. in class mini quizzes), homework assignments, projects, project presentation and final exam.
You might want to look at previous versions of this course on my website to get a basic idea. The course in Fall 2010 extends the previous versions by current problems of robot vision. You might also want to have a look at the movies on my website - their content will be part of the class, too.
Topics:
- # Week 01 : Introduction to Image Processing & MATLAB. [Chapter 1,2 of Textbook]
- # Week 02 : Intensity Transforms & Spatial Filtering [Chapter 3]
- # Week 03 : Morphological Image Processing [Chapter 9]
- # Week 04 : Image Processing in the Frequency Domain, 1D fourier filtering, applications in Object detection and recognition [chapter 4]
- # Week 05 : Image Processing in the Frequency Domain, 2D fourier filtering
- # Week 06 : Image Restoration and Reconstruction, Color Image Processing [chapter 5,6]
- # Week 07 : Object Segmentation & shape recognition
- # Week 08 : (spring break)
- # Week 09 : Face Recognition
- # Week 10 : Different sensors in robotics: sonar, 3D laser range finders. object recognition in 3D point sets
- # Week 11 : Robot mapping and localization (SLAM, geometric/topologic/semantic mapping) fusing laser range finders & cameras
- # Week 12 : (additional topics of interest emerging from the previous classes)
- # Week 13 : Presentations 1-2, discussion
- # Week 14 : Presentations 3-4, discussion
- # Week 15 : discussion, review for final exam
- Particle Filter Correspondences
- IJCV Particle Filter Correspondence Paper
- Towards Segment Based Robot Mapping
Assignments: