A FisherRao Metric for curves using the information in edges
Maybank, Stephen J. (2016) A FisherRao Metric for curves using the information in edges. Journal of Mathematical Imaging and Vision 54 (3), pp. 287300. ISSN 09249907.

Text
fisherraometricforcurvesJMIVversion2.pdf  Author's Accepted Manuscript Download (2MB)  Preview 
Abstract
Two curves which are close together in an image are indistinguishable given a measurement, in that there is no compelling reason to associate the measurement with one curve rather than the other. This observation is made quantitative using the parametric version of the FisherRao metric. A probability density function for a measurement conditional on a curve is constructed. The distance between two curves is then defined to be the FisherRao distance between the two conditional pdfs. A tractable approximation to the FisherRao metric is obtained for the case in which the measurements are compound in that they consist of a point x and an angle α which specifies the direction of an edge at x. If the curves are circles or straight lines, then the approximating metric is generalized to take account of inlying and outlying measurements. An estimate is made of the number of measurements required for the accurate location of a circle in the presence of outliers. A Bayesian algorithm for circle detection is defined. The prior density for the algorithm is obtained from the FisherRao metric. The algorithm is tested on images from the CASIA Iris Interval database.
Metadata
Item Type:  Article 

Additional Information:  The final publication is available at Springer via http://dx.doi.org/10.1007/s108510150603y 
Keyword(s) / Subject(s):  Bayesian curve detection, CASIA Iris Database, circle detection, Hough transform, Riemannian metric, step edges 
School:  Birkbeck Faculties and Schools > Faculty of Science > School of Computing and Mathematical Sciences 
Depositing User:  Professor Stephen Maybank 
Date Deposited:  01 Dec 2015 10:49 
Last Modified:  09 Aug 2023 12:36 
URI:  https://eprints.bbk.ac.uk/id/eprint/12948 
Statistics
Additional statistics are available via IRStats2.