Feedback augmentation of pde-based image segmentation algorithms using application-specific exogenous data
Karasev, Peter A.
MetadataShow full item record
This thesis is divided into five chapters. The scope of problems considered is defined in chapter I. Next, chapter II provides background material on image processing with partial differential equations and a review of prior work in the field. Chapter III covers the medical imaging portion of the research; the key contribution is a control-based algorithm for interactive image segmentation. Applications of the feedback-augmented level set method to fracture reconstruction and surgical planning are shown. Problems in vision-based control are considered in Chapters IV and V. A method of improving performance in closed-loop target tracking using level set segmentation is developed, with unmanned aerial vehicle or next-generation missile guidance being the primary applications of interest. Throughout this thesis, the two application types are connected into a unified viewpoint of open-loop systems that are augmented by exogenous data.