Shape characterization by using the Gabor transform

This paper introduces a novel framework for 2D shape analysis from its outline by using the Gabor transform (GT). The shape's boundary is represented by a complex signal, which is analyzed by the GT. Three automatic methods for analyzing the GT representation are discussed. Experimental results have shown that the GT can be used in the identification of dominant points as well as contour regions presenting periodic patterns (both deterministic and statistic patterns). Some results for synthetic and real images are presented and discussed.