Tracking-gestützte Kantenextraktion aus Produktionsdaten

This diploma thesis shows how edge extraction from product data can be optimizing for a better treatment in edge based tracking systems. More exact results with an edge tracker are obtained with optimized input data. If this is used in Augmented Reality Systems, a robust course is ensures and the right information are shown in the right place in the field of view of the user. Main concept for this optimizing is a modification of an edge tracker, so it can influence the edge extraction process. Different concepts to control the edge extraction will be developed, implemented and analysed. For this, parameters for edge extraction will be identified and criterions for valuation of edges and edge models will be defined.