Rochester Institute of Technology
College of Science
Chester F. Carlson Center for Imaging Science
TITLE: Advanced Digital Image Processing
PREREQUISITES: 1051-782 Introduction to Digital Image Processing or equivalent
1051-726 Computing for Imaging Science
COURSE DESCRIPTION:
Advanced Digital Image Processing investigates algorithms and techniques for a variety of imaging applications. The techniques build on the background that is established in the course SIMG-782, Introduction to Digital Image Processing, which focuses on basic image processing methods. The course is taught using a lecture and group project format, in which the lectures focus on advanced techniques and provide applications of their use in selected applications. The group projects enable the students to work on substantial designs that require the understanding of the task domain, exploration of solution methods by analysis and prototyping, and implementation of a selected approach. Each team presents a preliminary plan, an approach with feasibility analysis, and a final demonstration.
COURSE OBJECTIVES
Understanding of standard advanced image processing algorithms.
Understanding of image processing system development.
Understanding of team design techniques.
Experience in algorithm development and testing.
COURSE OUTLINE:
The following are examples of image processing algorithms that are presented by the course lectures. Emphasis is modulated by the knowledge and skill needs of the team projects.
Image Segmentation
Region-based methods
Boundary-based methods
Image registration and alignment
Feature point detection
Parametric representation of image distortion
Distortion parameter measurement
Feature Detection & Classification
Statistical decision theory & performance metrics
Lines, edges, corners and other geometric elements
Textures
Pixel classification using spectral techniques
Characters, words and document structure
Use of Object Models
Geometric relationships between features
Hierarchical models, pyramid representations
Parametric models and model fitting
Shape description
Non-parametric and learning-based techniques
Analytical Methods
Wavelet analysis
Scale-space techniques
Linear transforms and projections
Fractal representations
EVALUATION:
Homework assignments
Project Reviews
Final Exam
TEXT: No text is required. Current and classical papers and instructor notes are provided for lecture topics. Students are directed to library research for additional material relating to project topics. Recommended reference-Digital Image Processing, Gonzalez and Woods (1992).