Project Description:
Advancements in Micro-CT scanning have unlocked new opportunities for studying biology. AntScan is a database of over 2000 scans of ants used by entomologists to study the diversity and evolution of ant species. Each of these scans was converted into a static 3D model for visualization. These scans and models are incredibly detailed, but the majority of the specimens are in crumpled or unnatural poses. Researchers can create interactive and articulable 3D models to analyze specimens, but they must be done manually, which is a tedious and time consuming process. We built an automated pipeline that converts static ant models into rigged, poseable 3D models directly from cues such as skeleton structure, thickness, and curvature. This transforms a process that might take hours by hand to one that automatically runs in minutes.