Identifying and approaching for obscured stairs.

We present an autonomous approaching scheme for mobile robot traversing obstacle stairwells, which overcomes the restricted field of vision caused by obstacles. The scheme consists of three parts: stair localization, structural parameter estimation, and nearing the stair. We use LIDAR with full-angle and long-distance measurement capabilities to initially determine the location of the stairs, and estimate staircase parameters using a depth camera with intensive data. Then, we propose an algorithm to optimize the location of stairs for real-time correction of target point locations during approaching. Finally, we conducted tests to verify the accuracy of the estimated stair positions and structural parameters and performed a comprehensive experiment to demonstrate the effectiveness of the scheme, in which the robot approaches autonomously to the front of the stairs around multiple obstacles. • We develop the double scales normal and selection (DSNS) algorithm to identify stair from sparse point clouds. • We develop the multi-plane joint recognition (MPJR) algorithm to estimate the structural parameter. • We propose the sliding window multi-round voting mechanism (SWMRVM) algorithm to optimize the stair position from recognition. [ABSTRACT FROM AUTHOR]