Robust recognition of calibration charts

Methods for recovering 3D pose and internal camera parameters from a single view of a planar calibration chart have been reported in the literature. The mathematics of the computation is well understood, but the lack of robust feature detectors prevents off-the-shelf use. We present a fast and robust recognition approach for the detection of a common calibration chart. After establishing the methodological framework, we check the result on a wide range of real data. We show that the approach is robust over a range of scale, illumination and occlusion. We argue that with this capability we should be able to (a) establish camera position with respect to a landmark (b) recognise an object which is tagged with the calibration chart and (c) test any camera calibration and 3D pose estimation routines, thus helping towards future research and applications in mobile robots navigation, 3D reconstruction and stereo vision.