| Detecting the position and the level of joint impingement is often a key to a computer-aided surgical plan to normalize joint kinematics. So far for ball-and-socket joints most of the existing impingement detection methods are not efficient or only consider the collided points as the detection result. In this thesis, we present two efficient and accurate collision detection systems to detect and evaluate the impingement on ball-and-socket joints.;Through the first system, we rapidly non-uniformly sample the object surface in the spherical coordinate system based on polygon rasterization and then globally check the distance differences between the objects to approximately estimate the impingement level; the second system has a similar design but is based on a uniform sampling method which can sample the object more adequately and reduce the memory cost by 20%.;We achieved the accuracy, the efficiency, and also the feature of providing overall impingement estimation and visualization on the joint surfaces. The first system is further applied in a real-time ROM observation simulation to flexibly check the location and the level of the impingement occurred during the motion. |
