A Hypothesize and Test Technique for Corner Detection

Abstract

In this paper a new technique is proposed for detection of corners in images of polygonal objects. The corner detector works on the principle of ‘hypothesize and test’. Initial hypotheses about the likely corners we formed on the basis of the lines detected in the edge image using simple Hough transform for line detection. Depending on the domain knowledge the hypotheses set is filtered. The remaining corners are verified using template matching. Templates are generated dynamically according to the type of the corners. The detector can accurately localise the corners and correctly measure the included angles as well as orientations of the corners. The detector can also be easily parallelised. Capabilities of the detector are illusterted through its performance on same example scenes.

Indexing terms:

• Feature extraction
• Hough transform
• Object recognition