ECS3383

Computer image processing and pattern recognition

(ENG)

R A Jarvis and D Suter

3 points + 16 lectures, 9 laboratory hours and extensive reading + Second semester + Clayton + Prerequisites: Level 2 substantially complete

Synopsis: Image processing: image formation and discretisation, pixel point operators, spatial operators - convolutions and non-linear operators, edge detection and segmentation, computer implementation. Pattern recognition: binary vector component decision rules, Bayesian minimal risk formulations, sequential pattern recognition, linguistic pattern recognition, fuzzy logic and stochastic production systems, minimal spanning trees, clustering methodologies, neural networks, application scope.

Assessment: Examination (2 hours): 60% + Assignments: 40%