Module details
EL844 Fundamentals of Image Analysis (15 credits)
Syllabus
Fundamentals of image analysis provides an introduction to the principles of
identity authentication based on biometrics and introduces the fundamental
underlying theoretical and practical skills required for their development.
- Fundamentals of Image Processing
General introduction to
digital image processing; image acquisition, quantisation and
representation; Affine transforms; image enhancement techniques: contrast
manipulation, binarisation, noise removal (spatial and frequency domain);
edge detection techniques; image segmentation: edge-based, region-based,
watershed; Hough transform; image feature extraction; advanced image
processing: morphological operations, colour image processing, various image
transforms (Fourier, wavelet, etc).
- Fundamentals of Pattern Recognition
Patterns and pattern
classification, and the role of classification in a variety of application
scenarios, including security and biometrics. Basic concepts: pattern
descriptors, pattern classes; invariance and normalisation. Feature-based
analysis. Texture analysis. The classification problem and formal
approaches. Basic decision theory and the Bayesian classifier. Cost and risk
and their relationship; rejection margin and error-rate trade-off. Canonical
forms of classifier description. Estimation of class-conditional
distributions; bivariate and multivariate analysis. Euclidean and
Mahalanobis distance metrics and minimum distance classifiers. Parametric
and non-parametric classification strategies. Linear discriminant analysis.
Clustering approaches, and relationship between classifier realisations.
Practical case studies.
- Implementation essentials
Programming and data analysis using
MATLAB and other software tools as appropriate. Introduction to practical work using MATLAB.
This module is delivered by the School of
Engineering and Digital Arts.
Note
This web page provides advance information about a module due
to run in the coming academic year. We believe the details are
accurate at the time of writing but they may be subject to
change.