Development of a vision integration framework for laparoscopic surgical robot

In order to realize intelligent laparoscopic surgical robot, a vision integrated system constitutes one of the fundamental components. The authors have constructed a vision framework in the current version of NCC (National Cancer Center) laparoscopic surgical robot controlled on a real-time OS (RTLinux-Pro, FSMLabs Inc., U.S.A.). Adding vision framework, we have been applying and testing image processing algorithms- edge detection of object for positioning surgical tool, Watersheds for recognizing object. This paper documents the implementation of the framework and preliminary results of the image segmentation using Watersheds algorithm. Finally the real-time processing capability of our vision system is discussed.