Dense Depth Maps Using Active Vision


Institute of System and Robotics
Dense Depth Maps 
Using Active Vision


The aim of this work is to examine and develop a depth map acquisition system with real-time characteristics and based on a binocular active vision system . In order to obtain depth information we present a method that combines stereo matching  with mechanical activity, reducing the time spent to perform the correspondence between the image points on the left and right images.
Controlling the cameras' vergence or the baseline distance it is possible to change continuously  the fixation point in the space and, at the same time, to select points with correspondent image projections.  Computing the distance  from those points to the vision system, is possible to obtain a dense relative map of the scene. The correspondence is established based on similarity measures between image regions. This measures are performed by  operators with characteristics that makes this method suitable for parallel implementation. Since the depth information is relative, the calibration of the active vision system is minimal.

VARMA-Active Vision Head
Click to see an iterative binocular camera model

 Space object being sweep and similarity outputs

Related publications

Luis Almeida, Jorge Dias, "Dense Depth Maps using an Active Vision Head", Proceedings of CARS & FOF'99 the 17th International Conference on CAD/CAM Robotics & Factories of the Future, Aguas de Lindoia, SP, Brazil, August 18-20, 1999.

Luis Almeida, Jorge Dias, "Dense Depth Maps using Stereo Vision Head", Proceedings of SIRS99 the 7th International Symposium on Intelligent Robotic Systems, Coimbra, July 20-23, 1999.

L. Almeida, J. Dias, A. Traça de Almeida, "Dense Depth Maps Using Active Vision", 9th Portuguese Conference on Pattern Recognition, Coimbra, Portugal, 20-21st March, 1997.

Vergenge Sweep Demo
Local time is 0:0

and you are user *** since 30-6-98.