Authorised by Academic Registrar, April 1996
Objectives After successful completion of this subject the student should understand the relationship of local spatial and temporal filtering techniques, such as first and second order edge detection, to Fourier and other mathematical transforms; know the general principles and applications of morphologic image processing, including the basis representation, skeletonisation and the relationship to the various forms of distance transform; have reviewed the Hough transform and other template matching techniques.
Synopsis Linear and non-linear filtering. Noise characterisation. Edge detection theory. Template matching and optimal filter design. Generalised Hough transform techniques. Distance transform; uniform and gray weighted, local scalar and vector distance metrics. Mathematical morphology, fundamental theory and applications, general representation theory. Digital topology. 2-D image transforms. Clustering algorithms. Statistical classification techniques.
Assessment Examination (2 hours): 50% + Laboratory exam: 50%