ECS 289H ADVANCED IMAGE PROCESSING AND ANALYSIS (4) III
Lecture: 3 hours
Discussion: 1 hour
Prerequisite: None
Grading: Letter; homework (60%), presentation (15%), final project (25%)
Catalog Description:
Survey of current research topics in the automated extraction of high-level information from images generated by cameras, three-dimensional surface sensors, and medical devices.
Expanded Course Description:
Each week students will read papers, attend lectures, and do presentations on a particular current research area in image processing and analysis. Topics will include a selection of recent research on image processing of two-dimensional images, three-dimensional surface data, and volumetric medical images. Examples topics include:
- Local image patch descriptors for 2D
- Segmentation of 2D images
- Parts-based and configural models for object recognition
- Surface mesh pre-processing
- Surface mesh segmentation and classification
- Object detection in cluttered surface data
- Shape representations for biological objects
- Diffeomorphic matching of images and shapes
- Localization of biological objects in volumetric images
- Statistical analysis of populations of shapes
Two papers from a particular topic will be assigned each week. Early in the week, lecture material will provide crucial background information on the week, lecture material will provide crucial background information on the read the other paper and write a brief summary of it which will be turned in.
Textbook:
None. Each week, research papers will be provided that cover the topic for the week. Lecture slides covering background material for the weekly topic will be provided as well.
Final Project:
Final projects will be negotiated with the course instructor. Students will either write a final paper or implement a final programming project. For the paper, students will select 3 to 5 current research papers that have a coherent, focused theme, in consultation with the course instructor. The student will write a paper that summarizes the critical algorithmic issues involved in the underlying research area and the approaches taken in the selected papers. For the programming project, the student will select a cutting-edge algorithm in consultation with the instructor, implement the algorithm, and write a summary of algorithm performance and implementation issues.
Presentation:
Each student will provide a single in-class presentation that provides an overview of one current research paper.
Goals:
Students will read research papers that represent cutting-edge research directions in image processing and analysis and write brief paper summaries to demonstrate understanding of specific papers. Each student will give an in-class presentation that provides a clear overview of a particular paper. Each student will write a final paper or implement a final programming project that demonstrates mastery of a particular current research area or algorithm.
Instructor: O. Carmichael
Prepared by: O. Carmichael, N. Amenta (August 2007)
Overlap Statement:
This course does not have a significant overlap with any other existing course.
11/07