Guided Volume Editing based on Histogram Dissimilarity

摘要: Segmentation of volumetric data is an important part many analysis pipelines, but frequently requires manual inspection and correction. While plenty volume editing techniques exist, it remains cumbersome errorprone for the user to find select appropriate regions editing. We propose approach improve by detecting potential segmentation defects while considering underlying structure object interest. Our method based on a novel histogram dissimilarity measure between individual regions, derived from structural information extracted initial segmentation. Based this information, our interactive system guides towards defects, provides integrated tools their inspection, automatically generates suggestions resolution. demonstrate that can reduce interaction effort supports in comprehensive investigation high-quality segmentations.