After successful completion of the course, students are able to implement an application of structural pattern recognition. This includes the study of the problem, the selection of the possible methods and the proof of correct functioning.
In this lab exercises small problems of structiral pattern recognition have to be solved: building and using a graph pyramid for image segmentation, connected component labelling, distance transform and similarity detection, similarity of graphs, mean and median of a set of graphs, tracking articulated movements.
Methods are presented in the adjoint lecture.
lecture: study the provided material 15h
homework:
specify problem 5h
design and/or select MATLAB code and test data 24h
experimental evaluation 15h
producing illustrative pictures 2h
report 14h
Successful implementation and scientific report of a selected example within the domain of the lecture.
Basics of digital image processing and pattern recognition