 |
IE 525 - Statistical Pattern Recognition and Machine Vision
for Manufacturing Systems
Fall 2000
|
Textbook:
|
Milan Sonka, Vaclav Hlavac, Roger Boyle, Image Processing,
Analysis and Machine Vision, 1995.
|
|
References:
|
Richard O.Duda, Peter E. Hart, Pattern Classification and Scene
Analysis, 1973.
Rafael C. Gonzales, Richard E. Woods, Digital Image Processing,
1992.
|
|
Coordinator:
|
A. Erçil, Professor of I.E.
|
|
Goals:
|
This course is designed to give first year graduates in industrial
engineering basic concepts in statistical pattern recognition and
machine vision emphasizing automated inspection and classification in
manufacturing.
|
Prerequisites by Topic:
1. Ability to write computer programs in C or Pascal or use packages
like Matlab
2. Basic probability concepts
Topics:
|
1.
|
Introduction to Machine Vision (What is
machine vision, why use vision, tasks for a vision system, relation to
other fields, place of vision in CIM) (2 classes) |
|
2.
|
Image Acquisition (1 class) |
|
3.
|
Digital Image Representation (image formats,
display devices, digitization of images, gray level histogram) (2
classes) |
|
4.
|
Processing of Binary Images (thresholding,
geometric properties, topological properties) (4 classes) |
|
5.
|
Processing of Gray scale images (statistical
operations, spatial operations, Segmentation, edge detection,
morphological operations) (11 classes) |
|
6.
|
Fundamentals pattern recognition systems (1
class) |
|
7.
|
Parametric classifiers (3 classes) |
|
8.
|
Nonparametric classifiers (Nearest Neighbour,
CART, Neural Networks, Genetic Classifiers) (11 classes) |
|
9.
|
Feature extraction/selection (Discriminant
Analysis, Principle Component Analysis) (3 classes) |
|
10.
|
3-D object representation (2 classes) |
Computer Usage:
1. Several homework assignments for the above topics, some requiring C
programming, some requiring use of Matlab, S-Plus programs.
Homeworks will be towards building a real-world application, which will be
decided during the lectures.
Laboratory Projects:
Term project for each student, which involves the implementation of machine
vision techniques for a real-world problem. (10 weeks)
|
|
