Introduction

• What is computer vision and what is it not
• Several application examples
• Some of the challenges of computer vision
  • Engineering challenges
  • AI-ish challenges

Background on image formation

• camera imaging basics
• image representations
  • image formats
  • color spaces
• Light Transport

Image Processing for Computer Vision

• convolution
  • smoothing
  • derivatives
  • texture
  • edge detection
• nonlinear filters
  • median filtering
  • morphology
  • top-hat/bottom-hat

Features and Feature Matching

• Edge detection
  • Types of edges
  • Scale space
  • Canny edge detection
• Texture features
  • Template matching
  • Corner detectors
  • Sift features
• An example: object detection

Segmentation

• Local classification (ad-hoc hacks ..)
  • thresholding
  • edges
• “Local” (could be energy-based) optimization methods
  • k-means (Eric)
  • watershed book section (Dan M.)
  • snakes?
  • level sets**
  • agglomerative methods/region growing
    • Felzenszwalb example
  • mean-shift
• Global (should be energy-based) methods
  • graphical representations
  • graph cuts book section(Dan M.)
  • iterative graph cuts

Imaging geometry and Calibration

• models: [orthographic], [perspective], [affine]
• [camera calibration]
• [Jason Corso’s Camera Calibration Notes]
  • Direct Linear Methods
  • multi-planar
  • self-calibration
• two camera geometry
  • Fundamental matrix

Computational Stereo

• computational stereo
• generalizations to more cameras
  • space carving ideas
  • plane sweep
  • energy-based
• Projects
  • mosaicking
  • 3D reconstruction

Motion Analysis

• Background generation/subtraction
• optical flow vs. motion field
• image constancy equation
• optical flow algorithms
• template tracking
• tomasi-kanade and extensions
• Other tracking methods
  • mean shift
• Projects
  • SLAM

Other Topics

Learning and Vision

• generative
• discriminative

• Applications (2 lectures)
  • Object recognition
  • medicine

Appendices

• Linear Algebra
  • SVD and linear system solution
• Probability and Statistics
• Spatial Transformations
• Optimization
• Experimental Design