ENEE 731: Image Understanding

Course Goals:

The goal of this advanced level course is to train the graduate student in the field of image understanding/computer vision. The course takes a mathematical/statistical approach to many problems in understanding image/video content.

Course Prerequisite(s):

ENEE 631 and ENEE 633.

Topics Prerequisite(s):

Textbook(s)

Reference(s):

Markov random fields: Theory and applications, R. Chellappa and A.K. Jain (eds.), Academic Press, 1993.
Robot Vision, B.K.P. Horn, MIT Press, 1986.
Fundamentals of Computer Vision, O.Faugeras, MIT Press.
Object Recognition by Computer, W.E.L. Grimson, MIT Press, 1990

Core Topics:

  1. Optimal edge and shape detection
    Canny edge finder
    Optimal shape detection
  2. Markov Random Field Models
    Representation
    Parameter estimation and hypothesis testing
    Texture synthesis, classification and segmentation using MRFs
  3. Monte Carlo Markov Chain Techniques Theory
    Applications to tracking
  4. Shape from Techniques
    Shape from shading
    • Uniqueness proofs and algorithms
    • Robust algorithms
    Shape from contour
    • Theory and algorithms
    Shape from stereo
    • Human perception of depth
    • Marr-Poggio algorithm
    • Marr-Poggio-Grimson algorithm
    • Micro annealing algorithm
    • Hierarchical matching using truth-maintenance systems
    • Multi-camera reconstruction
  5. Structure from Motion
    Computation of optical flow
    Structure from motion using flow
    Structure from motion using discrete features
    Use of non-linear recursive filters
  6. Object Recognition
    Representation
    Matching algorithms
    Applications to face recognition

Optional Topics:


Course Structure:

Grading Method:

| Dept. of Electrical & Computer Engineering | A. James Clark School of Engineering | University of Maryland |