Fitting Parameterized Three-Dimensional Models to Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
Medical augmented reality based on commercial image guided surgery
EGVE'04 Proceedings of the Tenth Eurographics conference on Virtual Environments
| Hi-index | 0.01 |
One of the main challenges in external optical tracking device based augmented reality (AR) is the extrinsic calibration. Such an example is calibration of the endoscope camera with the tracking device to estimate the transformation between the endoscope-mounted marker and endoscope sensor. In this paper, we describe two alternative techniques to the Hand-Eye method for online calibration. First, we describe a direct technique based on estimating rigid transformation from corresponding point sets and our idea of improving the calibration efficiency by collecting corresponding points covering broader area of the endoscope and tracker field-of-view (FOV). Then, we describe a technique based on estimation of position and orientation of a planar object from camera image. The main advantage of these techniques is that they are easily repeatable in applications where a change in the relation between camera sensor and camera-mounted marker is possible during the run-time of the AR application.