Design Method And Image Guidance Mechanism Of Stereo Visual System For Minimally Invasive Surgical Robot

Based on the requirements of the surgical robotic vision system,this paper focuses on the consistency between visual space and operation space,aiming at the depth information perception,intuitive operational consistency,image-guided navigation and visual interface interaction.Research on the minimally invasive surgical robotic system,this paper proposes the design method and guidance mechanism of the vision system in the surgical robot.This thesis makes the following creative achievements:1.According to the basic principle of human visual depth perception,the mapping relationship between binocular parallax clues and robotic vision is determined.The configuration of the robotic vision system is reconstructed with the spatial model.Based on the vertical alignment constraint and horizontal parallax consistency constraint in the binocular vision system,the stereoscopic vision system with intuitive perception is proposed to minimize the energy functions.2.In order to improve the depth perception and observation comfort of the robotic vision system,a binocular adaptive convergence accommodation method is proposed.By rotating the virtual optical axis of the parallel stereo camera,the imaging planes converge at the fixation target,and the stereo image pair is reconstructed by the perspective projection and pin-pole geometry.The detection of fixation target in the operating area instead of the global feature point detection is the input parameter of the reconstructing method to avoid a large amount of computation.3.In order to establish the mapping relationship between the visual space and the operation space of the surgical robot,the parameterized hand-eye coordination model of the minimally invasive surgery robot is proposed.In this model,the camera,surgical instruments,and lesion target are descriptors,and each descriptor is an independent geometric factor.In the optimization scheme,each parameter is treated as a separate geometric element representing the associative function between the descriptors.The hand-eye coordination model based on the operation of the patient’s surgical robot system is transformed into the relative position and direction of the descriptor.4.Based on the three-dimensional geometric model of the anatomical tissue and the kinematic model of the multi-degree of freedom surgical tool,the target trajectory of the surgical channel is constructed under the anatomical constraint to realize the robotic image navigation.With this target trajectory,Follow-the-Leader motion algorithm is proposed to improve the operation accuracy.Based on the above theory,this paper applies the robotic image navigation algorithm to the narrow anatomical space operation,providing the surgical solution to the combined third endoscopic ventriculostomy and endoscopic tumor biopsy.5.In the experimental research,the accuracy and effectiveness of the binocular adaptive convergence accommodation algorithm are simulated and analyzed by the endoscopic task.The orthogonal experiment is used to test the effect of the parametric hand-eye coordination model on the accuracy of operational tasks.The accuracy of the trajectory under robotic image navigation is validated with the clinical application of the surgical procedure.