Object Labeling for 3-D Cross-Sectional Data using Trajectory Tracking

Volume visualization from object cross sections has created the need for object labeling in volume data sets. Labeling techniques for 3D objects vary from the memory-intensive connected-component labeling to the less intense overlap technique. In this paper, a novel labeling technique is presented that represents the objects with curves in 3D and then performs labeling by applying 3D curve tracing techniques and mapping the labels back to the 3D objects. The technique thus attempts to replicate aspects of the human visual approach to the task.