3D morphological reconstruction from magnetic data using deep learning

Abstract

Accurate 3D reconstruction is essential for the design of customized orthotics. Traditional methods rely on expensive optical scanners or complex manual measurements. This work proposes an alternative approach based on the use of magnetic sensors for 3D shape reconstruction. By placing magnetic sensors around the surface of the object, it becomes possible to acquire spatial information without the constraint of direct visibility. An artificial intelligence model is developed to predict and reconstruct the missing 3D point cloud, which can then be used for modeling and 3D printing orthotics. This method aims to make the manufacture of custom orthotics more accessible, portable, and economical.