A Novel and Automated Circle Pattern Recognition Technique for Infra-Red Stereo Camera Calibration

Martin Kefer, Wilfried Woeber, Daniel Szuegyi, and Wilfried Kubinger


Infra-Red Imagery, Automated Circle Detection, Pattern Recognition, OpenCV


Autonomous robots are increasingly used to respond to disaster scenarios. Sensor systems for those robots must deliver reliable 3D information for both day and night operation. Infra-red stereo camera systems are often used to deliver 3D information 24/7. However, reliable calibration of a 3D stereo sensor is both difficult and crucial. This paper presents a novel method to reliably extract circle-like shapes from infra-red images of specific calibration patterns. Based on these circle positions the calibration of an infra-red stereo vision sensor system can be achieved. Experimental evaluations of our approach show that our automated method is both robust and reliable.

Important Links:

Go Back