55:247 Image Analysis and Understanding

Individual chapters of 4th edition can be purchased separately as E-chapters
What's new in Ed4:
- We retain the same Chapter structure, but many sections have been rewritten or
  introduced as new.
- Among the new topics are the Radon transform, a unified approach
  to image/template matching, efficient object skeletonization (MB and MB2 algorithms),
  nearest neighbor classification including BBF/FLANN, histogram-of-oriented-Gaussian
  (HOG) approach to object detection, random forests, Markov random fields, Bayesian
  belief networks, scale invariant feature transform (SIFT), recent 3D image analysis/vision
  development, texture description using local binary patterns, and several point tracking
  approaches for motion analysis.
- All in all, about 15% of this edition consists of newly written material presenting stateof-
  the-art methods and techniques that already have proven their importance in the field.
- Additionally, the whole text has been edited for currency and to correct a small number
  of oversights detected in the previous edition.

All book references below refer to Edition 4
Midterm will cover portions of the below-referenced chapters that were covered in class lectures
- Chapter 8
- Chapter 7
- Chapter 9 - Boosting, Random Forests
- Chapter 10 (section 10.10 will not be included in midterm)