General description

This package implements the Förstner interest point detector in Matlab. The algorithm extracts junction and circular points from a greyvalue image with subpixel accuracy. It also includes some extensions proposed by Köthe (2003).


Input and Output


  • a single greyvalue image (image with one channel)
  • control parameters (optional)


  • image coordinates of extracted points with subpixel accuracy. The points are classified into junction and circular points.
  • integer image coordinates of window centers. These windows are used for the estimation of the points mentioned above.

So a simple example call would be:

   [windowcenters, junctions, circular_points] = ip_fop(im);

More details and an example can be found in the zip file.



The following zip-file contains the software: (Version 1.0 from 2011-03-21)




Förstner, Wolfgang / Gülch, Eberhard:
A Fast Operator for Detection and Precise Location of Distict Point, Corners and Centres of Circular Features.
In: Proceedings of the ISPRS Conference on Fast Processing of Photogrammetric Data. Interlaken 1987, S. 281-305.

Ullrich Köthe:
Edge and Junction Detection with an Improved Structure Tensor.
In: Lecture Notes in Computer Science (Proceedings of DAGM 2003) Volume 2781, 2003, DOI:10.1007/b12010, p. 25-32, Springer, Heidelberg.