Elbow detection for localization of a mobile robot inside pipeline using laser pointers

Robots for pipe inspection have been studied for a long time and many mobile mechanisms have been proposed to achieve the inspection task within pipelines. Localization is an important factor for inpipe robot to perform a successful autonomous operation. However, because of its unique characteristics of inpipe condition, sensors which have good performance in localization, like GPS and beacon, cannot be used. In this paper, a localization using elbow detection is presented. With the laser light image processing and the angular velocity of IMU, the odometry module of robot determines whether it is on straight pipe or elbow, and minimizes the integration error in orientation. The experiment environment has consisted of two straight pipes and an elbow, and the pipe map has been constructed as the result.