Research On Robot Navigation And Marker Visual Location In Neurosurgery Robot System

At present, research on robot's application in surgery is paid more and more attention to by relative researchers. It will not only bring a series of technique reforms, such as improving the quality of operation and reducing the damage to patient, but also change traditional concepts in surgery as well as promote the development of subjects concerned. Based on neurosurgery and selected robot of model MOTOMAN-UP20 as implement, this thesis concentrates on the experimental research on robot's navigation and marker's visual location, which belongs to the seedtime of neurosurgery robot system and can provide technical reference to the system. The navigation with robot is realized by external host control computer through off-line programming which call the functions in the library of data communication software MOTOCOM32. Firstly, operation scheme for the robot and its requirements are given. Secondly, based on the registration and mapping of markers, the coordinate of focus is transformed from the 3D image space in computer to the operation space in robot, and the mapping error of focus is simulated by MATLAB. Thirdly, according to the Euler rotation and homogeneous transformation, mathematic models of operation tool's location (position & orientation) and tools'automatic exchange are built, and the parameters for navigation are calculated. At last, experiments of navigation are done, and the operation errors are measured. With a camera mounted on the flange of robot, the location of markers is monocularly determined based on stereovision method. Firstly, by comparing methods of camera calibration and parameter models of camera, Tsai's two-stage technology and its improvement are used to calibrate the camera. Secondly, high-speed calibration method for the relationship of eye-in-hand is given, and the method is compared with AX=XB method given by Shui. At last, according to stereovision theory, the location of markers in robot coordinate is calculated, and the factors which affect location precision are analyzed. The result of experiments proves that the location error of marker determined by stereovision is less than 1mm and the synthesis location error of focus caused by robot's navigation is less than 2mm.