Researches On Collision-Detection In Virtual-Surgery

As a significant part of the computer aided medical, the technique of Virtual-Surgery is playing a more important role in the field of programming and treatment of surgery, the practice of surgery, navigation and wardship during the surgery time. One of the most vital problems in the Virtual-Surgery system and some other Virtual-Reality system is the problem of Collision-Detection, which focuses on the physics feature of the virtual human body and organs and other objects in the virtual scene in order to present people a strong truthful feeling in the virtual environment. However, low-effective real-time collision detecting methods have strongly restricted the continuous development of this advanced technique of Virtual-Surgery.The characteristics of the Collision-Detection problem in the Virtual-Surgery are expatiated in this paper, and meanwhile a collision detecting method with new k-Dops hierarchical Bounding Box model is given after the theoretical analysis. Based on the hierarchical model of the Bounding Box, the binary tree structure of the Bounding Box was devised. Simultaneously, the way of splitting the crunodes in the tree and the overlapping-testing algorithms of two bounding boxes are also presented.The simulation of the collision of the Virtual-Surgery scene is made on the platform of Collision-Detection. The functions of each bounding box model through the results of the simulation are compared and the factors that would influence the performance of the real-time collision detecting are induced. According to the results, a new method of selecting split point by making an area-weighted mean is proposed. As result the algorithm of the bounding box tree is optimized and the efficiency of the collision detection is enhanced.