Efficient Methods for continuous and discrete shape analysis [Elektronische Ressource] / vorgelegt von Frank R. Schmidt

Efficient Methods forContinuous and DiscreteShape AnalysisDissertationzurErlangung des Doktorgrades (Dr. rer. nat.)derMathematisch-Naturwissenschaftlichen Fakult¨atderRheinischen Friedrich-Wilhelms-Universit¨at Bonnvorgelegt vonDipl.-Math. Frank R. Schmidtaus BonnBonn, 2010Angefertigt mit Genehmigung derMathematisch-Naturwissenschaftlichen Fakult¨atder Rheinischen Friedrich-Wilhelms-Universita¨t Bonn1. Gutachter: Prof. Dr. D. Cremers2. Gutachter: Prof. Dr. M. ClausenTag der Promotion: 12. November 2010iiSummaryWhen interpreting an image of a given object, humans are able toabstract from the presented color information in order to really seethe presented object. This abstraction is also known as shape. Theconcept of shape is not defined exactly in Computer Vision and inthis work, we use three different forms of these definitions in orderto acquire and analyze shapes. This work is devoted to improvethe efficiency of methods that solve important applications of shapeanalysis.The most important problem in order to analyze shapes is the prob-lem of shape acquisition. To simplify this very challenging prob-lem, numerous researchers have incorporated prior knowledge intothe acquisition of shapes.