Abstract. We present an effective intensity-based method for rigid registration of a patient preoperative CT to its intraoperative situation with a few fluoroscopic X-ray images obtained with a tracked C-arm. It improves upon existing methods and overcomes most of their intrinsic speed, robustness, and accuracy problems. For speed, we generate Digitally Reconstructed Radiographs on small, dynamically selected regions of interest from precomputed ray gray levels in expected viewing directions, and use a multiresolution hierarchy of fluoroscopic X-ray images. For robustness and accuracy, we use a two-step comparison measure: Normalized Cross Correlation followed by Variance Weighted Sum of Local Normalized Correlation. To avoid local minima, we use a genetic search method. Our experiments on simulated, in-vitro, and cadaver data show an overall mean Target Registration Error of 0.8mm (max=1.5mm), 95% of the time, computed in 20-100 seconds under realistic conditions.

No abstract

No abstract

No abstract

Abstract:Head orientation is an important part of many advanced human-machine interaction systems. We present a single image based head pose computation algorithm. It is deduced from anthropometric data. This approach allows us to use a single camera and requires no cooperation from the user. Using a single image avoids the complexities associated with of a multi-camera system. Evaluation tests show that our approach is accurate, fast and can be used in a variety of contexts. Application to gaze detection, with a working system, is also demonstrated.