Geometric feature extraction using the chord transformation

Abstract

Geometric feature extraction can play an important role in the understanding of the content of an image. Methods such as the Hough transform are useful, but are subject to restrictions imposed by scale and object orientation. This paper describes the chord transform, not widely used to date, primarily due to the computational requirements. The chord transform can be used to detect geometric edge structural content within an image. Various adaptations of the chord transform, its implementation, and some practical applications in geometric feature extraction are discussed in this paper.