NEW FEATURE POINTS BASED ON GEOMETRIC INVARIANTS FOR 3D IMAGE REGISTRATION

We introduce in this paper a new type of feature points of 3D surfaces , based on geometric invariants. We call this new type of feature poin ts the entremalpoints of the 3D surfaces, and we show that the relativ e positions of those 3D points are invariant according to 3D rigid tra nsforms (rotation and translation). We show also how to extract those points from 3D images, such as Magnetic Resonance images (MRI) or Cat- Scan images, and also how to use them to perform precise 3D registrati on. Previously, we described a method, called the Marching Lines algor ithm, which allow us to extract the extremal lines, which are geometri c invariant 3D curves, as the intersection of two implicit surfaces: t he extremal points are the intersection of the extremal lines with a t hird implicit surface. we present an application of the extremal point s extraction to the fully automatic registration of two 3D images of t he same patient, taken in two different positions, to show the accurac y and robustness of the extracted extremal points.