Recognition and Image Processing

Lecturer:

Tutor:

Supervisor:

Department of Cybernetics

Synopsis:

Pattern recognition plays a decisive role in interpretation and processing of medical data, that is why this topic will be given the appropriate attention. The second part of the course will cover the basic techniques of digital image processing, namely those most frequently applied in medicine. The excercises will have the form of laboratory projects. Students will be asked to treat a complex problem from the biomedical domain, which cannot be solved without upper mentioned special knowledge. In this way the student will be taught the correct methodology to be used when solving a task in collaborative team approach.

Requirements:

Syllabus of lectures:

1. Pattern recognition, decision making, formalization, Bayes approach

2. Statistical models, namely Gaussian. Parameter estimation

3. Linear classifier. Support vector machine

4. Perceptron, neural networks. Radial basis functions

5. Clustering. EM algorithm. Unsupervised learning

6. Vapnik´s and other learning theories

7. Structual pattern recognition

Syllabus of tutorials:

1. Probabilities. Bayes classifier

2. Matlab. Parameter estimation

3. Perceptron

4. Linear classifiers and Support Vector Machine

5. Clustering. EM algorithm

6. Solution of the practical pattern recognition problem I.

7. Solution of the practical pattern recognition problem II.

Study Objective:

Study materials:

[1] Russell, S., Norvig, P.: Artificial Intelligence. A Modern Approach. Prentice Hall, Englewood Cliffs, New Jersey, 1995

[2] Schlesinger, M.I., Hlaváč, V.: Theory of statistical and structural recognition in 10 lectures, to appear 2002

Note:

Further information:

No time-table has been prepared for this course

The course is a part of the following study plans:

• Biomedical Engineering- structured studies (compulsory course)