论文部分内容阅读
本文介绍了一种新的智能内窥镜导航系统。该系统用于实现内窥镜的智能导航介入。其中导航方法是基于视觉传感器和触觉传感器的多信息融合的方法。在视觉传感器中采用了一种基于自适应阈值图像分割算法的寻径方法,该方法采用加速的最大类间方差法进行阈值分割。在触觉传感器中采用了基于MEMS技术的硅微触觉传感器,四个硅微触觉传感器被封装在内窥镜头部表面,可以获得内窥镜头部全方向的压力。本文通过实验验证了该导航系统的可行性及可靠性,该方法的应用弥补了传统方法中因暗区不明显而无法引导内窥镜的不足,提高了内窥镜导航的精度和可靠性。
This article describes a new intelligent endoscopic navigation system. The system is used to achieve intelligent navigation of the endoscope intervention. The navigation method is based on the multi-information fusion method of vision sensor and tactile sensor. In the vision sensor, a path-finding method based on adaptive threshold image segmentation algorithm is adopted, which uses the accelerated maximum between-class variance method for threshold segmentation. In the tactile sensor MEMS-based silicon micro-tactile sensor is used, the four silicon micro-tactile sensor is encapsulated in the endoscope head surface, you can get the endoscopic head pressure in all directions. In this paper, the feasibility and reliability of this navigation system are verified through experiments. The application of this method makes up for the shortcomings of the traditional method, which can not guide the endoscope because of the dark area is not obvious, and improve the accuracy and reliability of endoscopic navigation.