Using Top-Points as Interest Points for Image Matching

Abstract

Abstract. We consider the use of so-called top-points for object retrieval. These points are based on scale-space and catastrophe theory, and are invariant under gray value scaling and offset as well as scale-Euclidean transformations. The differential properties and noise characteristics of these points are mathematically well understood. It is possible to retrieve the exact location of a top-point from any coarse estimation through a closed-form vector equation which only depends on local derivatives in the estimated point. All these properties make top-points highly suitable as anchor points for invariant matching schemes. In a set of examples we show the excellent performance of top-points in an object retrieval task.

Partial Similarity of Objects, or How to Compare a Centaur to a Horse

“…We attribute to this broad category rigid, articulated and non-rigid two-and three-dimensional objects, shapes with texture and, as a particular case, binary, gray and color images. While the analysis of two-dimensional (Platel et al 2005) and three-dimensional rigid shapes (Chen and Medioni 1991;Besl and McKay 1992;Zhang 1994) is a well-established field, analysis of non-rigid shapes is an important direction emerging in the last decade in the pattern recognition community and arising in applications of face recognition (Bronstein et al 2003(Bronstein et al , 2005, shape watermarking (Reuter et al 2006), texture mapping and morphing (Bronstein et al 2006b(Bronstein et al , 2007a, to mention a few. In many practical problems, it was shown that natural deformations of non-rigid shapes can be approximated as isometries, hence, recognition of such objects requires an isometry-invariant criterion of similarity.…”

Analysis of Two-Dimensional Non-Rigid Shapes

“…Analysis of such shapes is often encountered in the computer vision literature [35,74,40,62,72,55,67,66,37,57,24,2,38,33], typically as a subset of the more generic problem of image analysis [65,3,42].…”