Image skeletonization methods used in structural topology optimization

Image skeletonization is the process of turning a binary or grayscale topology optimization result into a thin, connected centerline representation. It's a popular technique for converting density fields into manufacturable beam-like structures, extracting main load paths, and enabling automated reconstruction of clean CAD models. This paper reviews ۱۰ of the most common skeletonization methods that are widely used in structural topology optimization (TO). The key challenges these methods face are issues such as unwanted noisy branches, preserving the original topology, converting the skeleton into smooth parametric geometry, and making the process differentiable and capable of being integrated with gradient-based optimization are discussed in this paper. They are illustrated with examples from recent research and practical implementations.

کلیدواژه ها:

skeletonization ، thinning ، optimization ، reconstruction ، extraction ، medial axis ، topology ، CAD ، centerline