STABILITY OF CORNER POINTS IN SCALE-SPACE - THE EFFECTS OF SMALL NONRIGID DEFORMATIONS

To provide a good basis for the registration of medical images we sear ch for reliable feature points using a scale-space approach. Our main concern is with 2D images: we analyze corner points, defined by differ ential invariants, at increasing scales. The number and position of co rner points change in the scale-extended space, which define moving pa ths or orbits. To extract them we use a fast and reliable algorithm, b ased on iso-surface techniques, which automatically finds the correspo nding singularities in scale space. We then get a representation of or bits that is very convenient both for detection at a coarse scale and localization at a fine scale. We find that the significance of corner points depends not only on their scale-space lifetime but also on how they are related to curvature inflection points. We investigate some t opological changes of orbits which can be observed following image tra nsformations. Afterward we examine whether features, stable at multipl e scales, are stable as well with respect to various types of transfor mations. Thus we can compare the usefulness of different stability cri teria for registration. We then go to present statistical results show ing the dependency on the type of transformation and on the scale para meters. (C) 1998 Academic Press.