Extraction of shape-fitted layers from CT-data

Herein we describe a new method to extract layers from computed tomography volume data of objects that have a curved surface where simple slicing does not show the desired layers. For the extraction we first describe the surface of the object with a net of Bézier-patches. After that we extract the layers by sampling the net of Bézier-patches and then shifting the points of the Bézier-patches along their normals. This results in a volume where the object is virtually “planarized”. The object can now be visualized layer by layer so that layer structures and possible defects therein can be analyzed.