基于图像暗区自转方位的双半球胶囊机器人导航方法

The Visual Navigation Method of Dual Hemispherical Capsule Robot Based on Rotation Location of Dark Region

为了实现双半球型胶囊机器人在肠道内转弯行走,提出了一种基于无线传输图像暗区自转方位的胶囊导航方法,该方法利用一种机械式铅垂角度传感器获取胶囊摄像头绕胶囊机器人轴线的自转角度;然后通过一种基于改进的BAT算法与改进的二维OTSU算法相融合的暗区质心识别计算方法提取肠道图像暗区质心,该算法一定程度上克服了肠道内壁灰度值不均匀和肠壁存在褶皱所造成的质心识别干扰;最后结合胶囊摄像头绕机器人轴线的自转角度和图像暗区质心实现对胶囊机器人的导航。

In order to realize steering locomotion of the dual hemispherical capsule robot in the intestine tract,a navigation method based on the rotation location of the dark region obtained from the wirelessly transmitted image was proposed,which first used a mechanical vertical angle sensor to obtain the rotation angle of the camera embedded in the robot around the robotic axis.Then,the centroid of dark region of the intestinal image was extracted by a dark region centroid recognition calculation method based on the improved BAT algorithm and the improved two-dimensional OTSU algorithm,which overcame the interference caused by the unevenness of the gray value and the fold of the intestinal wall to some extent.Finally,the navigation of the robot was realized by combining the rotation angle of the robot camera around the axis of the robot and the centroid of the dark area of the image.