Incremental circle transform and eigenvalue analysis for object recognition: an integrated approach

摘要

In this paper, a new method to recognize two-dimensional objects by combining the concept of incremental circle transform (ICT) and eigenvalue analysis is presented. The algorithm has two stages. In the first stage, the shape of an object is preserved by the use of ICT (Bein et al. Proc. ICARCV, 1990, 851-855), which maps the boundary of an object into a circle with constant radius. In the second stage, the eigenvalues of the variance and covariance matrix of the computed ICT vector are found and are used to represent compactly the set of objects. Apart of realizing memory reduction, the proposed integrated method is shown to be rotation and translation invariant. Experimental results have been highly encouraging in terms of accuracy in the recognition of target objects.