Instructor

Christopher Rasmussen

E-mail: cer@cis.udel.edu

Office: 409 Smith 

Phone: 302-831-2783

Office hours: Wednesdays, 2-4 pm

Lab: 122 E. Delaware Ave.

Web page

www.cis.udel.edu/~cer/arv

Description

A survey of visual sensing algorithms for terrestrial and aerial mobile robots.  Feature detection, tracking, motion and structure estimation, landmark and obstacle recognition, panoramic imaging and mosaicing, mapping and localization, road following, and other key topics will be covered.

Schedule

Tuesdays & Thursdays from 5:00 pm to 6:15 pm in Smith 102A

Prerequisites

Ability to program in C/C++, Matlab, or the like

Thorough familiarity with linear algebra, calculus, probability, and statistics

Format

First 1/3 of classes: Review lectures on computer vision, robotics by professor

Last 2/3 of classes: Seminar-style paper presentations, discussions by students

Grading

10%: Two small programming assignments

30%: Two oral paper presentations + write-ups

10%: Class participation

50%: Major programming project

Academic

policies

Students may discuss problems with one another, but must work independently on writing and coding assignments (except with your partner on the first paper, of course).  This also applies to online and printed resources: consult them freely, but the words and source code you turn in must be yours alone.

Assignments should be e-mailed to me on the calendar day they are due (i.e., by midnight).

Textbook

Computer Vision – A Modern Approach,  D. Forsyth & J. Ponce [F & P]

• Complete 4th draft

• "Cameras" chapter from 3rd draft

Supplemental readings

Multiple View Geometry in Computer Vision, R. Hartley & A. Zisserman [H & Z]

Cambridge University Press, 2000

Introductory Techniques for 3-D Computer Vision, E. Trucco & A. Verri [T & V]

Prentice-Hall, 1998

Papers  

Image stabilization, mosaicing

• "Video Mosaics for Virtual Environments."  R. Szeliski.  IEEE Computer Graphics and Applications, Vol. 16, No. 2, 1996, pp. 22-30.

• "Automatic Mosaic Creation of the Ocean Floor."  N. Gracias & J. Santos-Victor. OCEANS '98 Conf., 1998. 

Tracking

• "Stochastic Road Shape Estimation." B. Southall & C. Taylor.  ICCV 2001, pp. 201-212. 

• "Vision-Guided Flight Stability and Control for Micro Air Vehicles." S. Ettinger, M. Nechyba, P. Ifju, & M. Waszak.  IROS 2002.

Object detection, recognition

• "Robust Car Tracking Using Kalman Filtering and Bayesian Templates."  F. Dellaert & C. Thorpe.  SPIE Conf. on Intelligent Transportation Systems, Pittsburgh, PA, 1997.

• "Vision-based Perception for an Autonomous Harvester."  M. Ollis & A. Stentz.  IEEE/RSJ International Conference on Intelligent Robotic Systems, 1997, pp. 1838-1844. [The order of pages is reversed in this PDF file for some reason]

Avoidance/Homing

• "Vision-based Autonomous Landing of an Unmanned Aerial Vehicle."  S. Saripalli, J. Montgomery, & G. Sukhatme.  ICRA 2002.  

• "Precise Image-based Motion Estimation for Autonomous Small Body Exploration."  A. Johnson & L. Matthies.  ISAIRAS 1999, pp. 627-634.

Mapping, localization

• "Appearance-Based Place Recognition for Topological Localization."  I. Ulrich & I. Nourbakhsh.  ICRA 2000. 

• "Using the Condensation Algorithm for Robust, Vision-based Mobile Robot Localization."  F. Dellaert, W. Burgard, D. Fox, & S. Thrun.  CVPR 1999.

Lecture #3 [ppt]; 

HW 1;
Presentation guidelines

 Lecture #5 [ppt];

Kalman filter preview

(T & V, pp. 328-31)

  F & P: Chapters 6, 11.1, 18;

H & Z: Chapter 8

Lecture #6 [ppt]; 

HW 2 

Review: Miscellaneous

Wei Zhou

Chaitra Gowri Murthy

Guest lecture:

Darius Burschka,

Johns Hopkins

Chris Haase

CLASS CANCELLED

Guest lecture:

David Coombs,

SAIC

Project presentations

Thommen Korah

Bill Ulrich

Joe Kirk

CLASS CANCELLED BY UNIVERSITY DUE TO WEATHER

Project presentations

Wei Zhou

Min Li

Chaitra Gowri Murthy

Qi Li

Chris Haase

Ibrahim Halil Saruhan

Xiuman Zhu