Edge and Junction Detection with an Improved Structure Tensor

Ullrich Köthe

in: B. Michaelis, G. Krell (eds.): Pattern Recognition, Proc. of 25th DAGM Symposium, Magdeburg 2003, Lecture Notes in Computer Science 2781, pp. 25-32, Heidelberg: Springer, 2003 (note: this article is © Springer-Verlag)
Awarded the main prize of the German Pattern Recognition Society (DAGM) 2003

Abstract

We describe three modifications to the structure tensor approach to low­level feature extraction. We first show that the structure tensor must be represented at a higher resolution than the original image. Second, we propose a non­linear filter for structure tensor computation that avoids undesirable blurring. Third, we introduce a method to simultaneously extract edge and junction information. Examples demonstrate significant improvements in the quality of the extracted features.

Online version: PostScript (604 kb), PDF (582 kb)

Back to Ullrich Köthe's publication page