Composition of Image Analysis Processes Through Object-Centered Hierarchical Planning

OF THE DISSERTATION Composition of Image Analysis Processes through Object-Centered Hierarchical Planning by Leiguang Gong, Ph.D. Dissertation Director: Professor Casimir A. Kulikowski The present thesis describes a new approach to biomedical image interpretation | the knowledge-based composition of image analysis processes through object-centered hierarchical planning. Its computer implementation in the VISIPLAN system has been tested on di erent image analysis tasks using Magnetic Resonance (MR) scans. The approach has been shown to be exible and e ective in handling a reasonably broad class of image analysis problems. The new approach de nes the problem of composing image analysis processes primarily as that of detecting a sequence of reference objects to establish the context for segmentation and recognition of a goal object at the highest level of problem solving. At a lower level it de nes the composition of speci c analysis processes, which are scheduled for each of the objects, together with appropriate processing tasks and data allocations. The present work provides a computer representation and framework for the overall composition of image analysis processes. A new approach has also been developed for pre-planning as a rst phase of the overall composition process, which includes a new method for systematically selecting edge-de ning thresholds for local-binary segmentation. In addition, an important component of the processing method library in VISIPLAN