Detection of ascending and descending stairways by surface normal vectors

Persons with walking frames are often limited with respect to visual acuity and they are depending on assistance. To give them the possibility to move free and autonomous in their environment an intelligent assistance system is proposed which can be mounted on a walking frame is used to observe the scene in walking direction, and obstacles have to be detected. Especially, stairways represent a high risk of injury if a collapse occurs. For stairway recognition, an algorithm is proposed which estimates normal vectors by using a covariance matrix and this makes it possible to segment the point cloud data provided by the Kinect sensor. The calculation of surface normal vectors of these regions helps to detect ascending and descending stairways.

[1]  Shuihua Wang,et al.  RGB-D image-based detection of stairs, pedestrian crosswalks and traffic signs , 2014, J. Vis. Commun. Image Represent..

[2]  Michael Hayes,et al.  Altitude control of a quadrotor helicopter using depth map from Microsoft Kinect sensor , 2011, 2011 IEEE International Conference on Mechatronics.

[3]  Wei Wang,et al.  Stairway detection using Gabor Filter and FFPG , 2011, 2011 International Conference of Soft Computing and Pattern Recognition (SoCPaR).

[4]  Radu Bogdan Rusu,et al.  3D is here: Point Cloud Library (PCL) , 2011, 2011 IEEE International Conference on Robotics and Automation.

[5]  Wai Ho Li,et al.  Plane-based detection of staircases using inverse depth , 2012, ICRA 2012.

[6]  Fatih Murat Porikli,et al.  Fast Construction of Covariance Matrices for Arbitrary Size Image Windows , 2006, 2006 International Conference on Image Processing.

[7]  Tom Drummond,et al.  Robust egomotion estimation using ICP in inverse depth coordinates , 2012, 2012 IEEE International Conference on Robotics and Automation.

[8]  Henrik I. Christensen,et al.  Efficient Organized Point Cloud Segmentation with Connected Components , 2013 .

[9]  Philip David,et al.  Ascending stairway modeling from dense depth imagery for traversability analysis , 2013, 2013 IEEE International Conference on Robotics and Automation.