skeletonize

Skeletonize is a term used in multiple fields to describe reducing an object to a simplified skeleton-like representation or exposing the underlying framework of a subject.

In biology and forensic science, skeletonization refers to the state in which only skeletal remains remain after decomposition or dissection. In forensic taphonomy, skeletonization timelines help estimate decay rates across environments such as warm, humid, cold, or dry settings. The process can occur naturally after death or be performed artificially in laboratory or museum contexts to study anatomy or for educational display.

In image processing and computer vision, skeletonization is a morphological operation that reduces foreground regions in a binary image to a skeletal remnant preserving the object's topology. The resulting skeleton is typically a one-pixel-wide line or network that runs through the centers of the object’s regions. It is used for shape analysis, handwriting and character recognition, and as a preprocessing step for feature extraction. Skeletonization is often achieved by thinning algorithms that iteratively erase boundary pixels while preserving connectivity, with common approaches including Zhang–Suen thinning and related schemes. The transformation is not unique and can introduce artifacts depending on the algorithm and image noise.

In computer graphics and 3D imaging, skeletonization can refer to extracting a skeletal structure from volumetric data or a mesh to serve as a rig for animation or as a simplifying representation for deformation. Medial axis transform is a related mathematical concept describing a continuous skeleton of a shape.

Limitations include sensitivity to noise, potential for spurious branches or disconnections, and the need for post-processing to ensure biological or geometric plausibility.