A Stereo Tracking System For Arbitrary Slicing Of Volumetric Radiological Data

In this paper, we represent a new user interface for slicing from volumetric images. One of the most convenient ways of arbitrary slicing will be using a stylus which indicates orientation of slicing plane. For this reason, orientation information of the stylus is extracted with the aid of computer vision techniques and stereo tracking algorithms. During the stereo tracking process, noise from different sources and motion of the system user are need to be filtered so as to get smooth and stable motion throughout time sequence. To filter out all noises from system and tremble of hand of user, kent distribution based bayesian filter and Kalman filter are applied and comparison of performance of two filters is given.