ECS 189H INTRODUCTION TO IMAGE PROCESSING
AND ANALYSIS (4) II
Lecture: 3 hours
Discussion: 1 hour
Prerequisite: Course 110; Mathematics 22A
Grading: Letter; programming assignment (60%), midterm
(15%), final (25%)
Catalog Description:
Techniques for automated extraction of high-level information
from images generated by cameras, three-dimensional surface sensors, and
medical devices. Typical applications include automated construction of
3D models from video footage and detection of objects in various types
of images.
Expanded Course Description:
- Analysis of 2D images (photographs)
- Low-level information extraction: Edge and corner detection, contour
completion, and texture analysis will be discussed in detail. Specific
algorithms to be discussed include Canny edge detection, Harris corner
detection, random-field-based methods for contour continuation, and
texton estimation.
- Multiple-view geometry: Students will learn the mathematical models
that describe the geometry of single cameras and geometric relationships
between 2 or more images. Orthographic, perspective, and affine models
for cameras will be presented. The fundamental matrix, trifocal tensor,
and quadrifocal tensor will be presented as basic mathematical constructs
for describing geometric relationships between pairs, triples, and
quadruples of images.
- Object recognition: Students will receive an overview of methods
for describing and detecting objects in 2D images. Appearance-based
methods based on principal components analysis, convolutional image
filters, and raw image classification will be described. Shape-based
object detection based on constellations of object parts, local edge
features, and alignment to prototype shapes will be presented
- Analysis of 3D surface imagery
- 3D surface parameterization and representation: Representation
of 3D surfaces based on points, parametric surface models, patches,
and geons will be discussed. The effects of noise, partial occlusion,
and sensor artifacts on these surface descriptions will also be described.
- Automated model building from surface data: Semi-automated and
fully-automated procedures for building complete 3D models from collections
of partial 3D sensor data sets will be presented. Semi-automated
techniques based on larndmark placement will be discussed. Fully-automated
techniques based on local surface descriptors, constrained data collection,
and global surface descriptors will be shown as well.
- Object recognition in 3D data: Detection of objects in 3D surface
data sets will be discussed. The effects of partial occlusion and
clutter will be discussed. Techniques based on alignment, local surface
descriptors, and machine learning approaches will be described.
- Analysis of volumetric images
- Low-level processing: Students will learn approaches to correct
for image artifacts found in computed tomography (CT), magnetric
resonance (MR), and functional MR images. Bias field correction,
blowout, and scattering artifacts will be discussed.
- Description and modeling of biological shapes: Methods for mathematically
describing biological objects found in volumetric images will be
presented. Representation of 3D solids will be discussed, and computational
anatomy approaches to representation of populations of shapes will
be presented.
Localization and detection of objects: Principles and algorithms
for localizing biological shapes in volumetric images will be presented.
Shape-model-based techniques for estimating the location of constrained,
expected objects such as the brain will be discussed . Low-level
detection of amorphous, unexpected objects such as tumors and calcifications
will also be presented.
Textbook:
Insight Into Images
ISBN: 1568812175
Author: Terry Yoo
Publisher: AK Peters (July 29, 2004)
Computer Usage:
Three programming projects will be assigned to allow students
to implement algorithms for processing each of the 3 image modalities:
- 2D images: Students will study and implement a selected algorithm for
low-level image processing or multiple-view geometry estimation
- 3D surface data: Students will implement a selected algorithm for detection
of 3D objects in radar image data.
- Volumetric images: Students will implement an alforithm for modeling
or detection of shapes in medical images.
Program Projects:
The programming projects for this class are chosen to enhance the lecture
material in the course.
Engineering Design Statement:
The individual student taking this class will design and document
a set of software modules for automated analysis of a variety of image
data that includes the well-established algorithms described in the course
outline. Examinations will include questions based on design components
of the course.
ABET Category Content:
Engineering Science: 2 units
Engineering Design: 2 units
Goals:
Students will:
- design and implement algorithm for automated processing of photographic,
three- dimensional, and volumetric image data.
- learn the mathematical principles underlying image analysis algorithms
and understand their characteristics when applied to real-world imagery.
Through programming assignments, students will become familiar with leading
image analysis software systems
Instructor: O.
Carmichael
Prepared by: O. Carmichael (May 2006)
Overlap Statement:
There is no significant overlap with any other course
5/06
Back to Course Descriptions