CSE 455: Computer Vision Winter 2005

Instructor: Linda Shapiro
Office: 634 Allen Center, Phone: 206-543-2196, Email: shapiro@cs.washington.edu, Office Hours (for concepts taught in the class, no programming help): 11-12 MW

Teaching Assistant

• Ian Simon, email iansimon@cs.washington.edu. Office Hours (for all help with programming assignments): 3-4 MF and 11:30-12:30 W in 216 Allen Center

Course Details

• Final Exam: Thursday, March 17, 10:30 a.m., Review on March 10.
  Final Review List
• Midterm Exam: Thursday, Feb. 3, Review on Feb 1.
  Midterm Review List
• Purpose: To indroduce the topic of Computer Vision to undergraduates. The course will discuss all three levels of computer vision: low-level vision or early processing, mid-level vision or feature extraction, and high-level vision or analysis. Both 2D and 3D applications will be discussed.
• Text: Computer Vision, L. G. Shapiro and G. C Stockman, Prentice Hall, 2001
  
• Text Errata
  
• Topics:
  • Into to Imaging (Ch. 1-2)
  • Binary Image Analysis (Ch 3)
  • Filtering (selected topics from Ch 5 and Ch 10)
  • Color/Texture and K-means (Ch. 6-7, very beginning of Ch 10)
  • Advanced Segmentation Methods (Ch 10 plus handouts)
  • Lines, Arcs, and Consistent Line Clusters (Ch 10 plus handouts)
  • Content-Based Image Retrieval (Ch 8)
  • Classifiers (Ch 4)
  • Matching in 2D (parts of Ch 11)
  • Motion Analysis (Ch 9 plus handouts)
  • Video Analysis (handouts)
  • Stereo (parts of Ch 12)
  • Camera Calibration and 3D Reconstruction (parts of Ch 13)
  • 3D Object Recognition (parts of Ch 14)
  • Guest Lecture on Tactile Image Analysis (Prof. Ladner)
• Software:
• Image Viewers:
  • XV: a very good viewer for UNIX
  • IrfanView: good viewer for Windows
• Assignments
  
• Notes
  
• CBIR Links
  
• Evaluation: The grade will be based on a combination of the programming assignments, the midterm, and the final. Roughly 60% projects and 40% exams. There will be four programming assignments. The first will be a warmup to learn to work with images. The others will be 2-3 week projects.