ECE 438 Image Analysis &
Computer Vision
Goals:
- Give the students a general
understanding of the fundamentals of image analysis and computer vision.
- Introduce the student to
analytical tools which are currently used for image analysis and in computer
vision applications and research.
- Develop the students ability
to apply image analysis tools and methods in solving computer vision
problems.
Instructional Objectives:
By Test 1, the students should:
- Understand differences
between computer vision and image processing.
- Know the basic components of
a computer vision system.
- Understand how images are
represented; including optical images, analog images, and digital images.
Understand image types such as binary images, gray-scale images, color and
multi-spectral images.
- Know the various imaging
modalities ; including, electro-magnetic (EM) imaging (visible and outside
visible spectrum), acoustic imaging, electron, laser imaging.
- Know how to use CVIPtools for
image analysis and binary object classification.
- Know the key concepts in
image file formats.
- Understand image acquisition,
concepts regarding how the 3-D world maps to a 2-D image. This includes
both brightness and location.
- Understand the basic concepts
of lens specification, be familiar with the basic parameters and types of
lenses.
- Know the difference between
irradiance and radiance.
- Know the lens equation and
how to use it to estimate blur circle size based on geometry. Understand
the concepts of depth of field, and "focused sufficiently well".
- Understand the tristimlus curves and how they relate to image
sensing.
- Understand how to mitigate
naturally occurring noise by averaging multiple samples.
- Be able to find simple
geometric properties of binary images, including area, position via center
of area, orientation via axis of least inertia, and projections.
- Understand connectivity issues
and labeling of binary images with multiple objects.
- Understand the model for
application of image analysis to computer vision.
- Understand why preprocessing
is performed and know about image geometry, convolution masks, image
algebra and basic spatial filters.
- Understand image quantization
in both the spatial and brightness domains.
- Know about the various types
of edge detection operators; including noise mitigation and edge detector
performance.
- Understand Frei-Chen masks and the concept of basis images and
edge and line subspaces.
- Know how to use the Hough
transform for line finding, and understand its extension to more complex
shapes such as circles.
By Test 2, the students should:
- Know the three basic
categories of image segmentation algorithms, and representative examples
of each.
- Understand morphological
filters as applied to binary, gray-scale and multiband
images; including erosion, dilation, opening and closing.
- Know about the 2-D Fourier
transform; including implied symmetry, phase, vector inner products and
basic filtering.
- Know why log remapping is
necessary for viewing spectral image data.
- Understand feature analysis
and extraction as applied to computer vision.
- Know how to use binary object
features, histogram features, color features, texture features and
spectral features in computer vision.
- Know how to use CVIPtools for
feature extraction
- Know how feature vectors,
feature spaces, distance and similarity measures are applied in computer
vision
- Understand how to preprocess
feature data for pattern classification, including various methods of data
normalization
- Understand basic pattern
classification; including cost function concepts, and the use of training
and test sets.
- Know simple pattern
classification algorithms such as: template matching, nearest neighbor,
K-nearest neighbor, and nearest centroid.
- Acquire a conceptual overview
of Bayesian methods and artificial neural networks for pattern
classification
By the final project in the laboratory:
- Be able to use CVIPtools to
apply image analysis techniques and solve computer vision problems.
- Be able to write simple C
functions using the CVIPtools libraries
- Be able to develop application-specific
algorithms for computer vision applications.
- Have a practical and visual
understanding of morphological operators.
- Have a practical and visual
understanding of various edge detection methods and techniques.