Course Description IMT5111 - Image databases (2005/2006)

Level	Postgraduate
ECTS Credits	5
Duration	Autumn
Language	Norwegian alternatively english
Prerequisite(s)
Aim	After completing this course, the students shall understand why image indexing is a challenging problem. The students shall have insight into a major image indexing techiques - especially techniques based on colour histograms. The students shall be capable of implementing image indexing algorithms for a collection of real images.
Content	- different attributes which can be used to index images; these include: colour, geometry and texture, - the problems surrounding image indexing, which include: the inherent high number of dimensionality of the image domain, - image indexing using gamut intersections, - image indexing using colour histograms, - the effect of the chosen colour space and number of bins.
Study Methods	Project Works, Tutoring
Assessment	Oral Exam, individually (counts 50%, evaluated by lecturer and an examiner) Evaluation of Project(s) (counts 50%, evaluated by lecturer and an examiner) Each part must be individually approved of.
Grading System	Alphabetical Scale, A (best) - F (fail)
Course Requirements	Report(s) The student shall present three status reports in oral form