A terrain classification method for UGV autonomous navigation based on SURF

The ability to navigate autonomously in off-road terrain is critical technology needed for unmanned ground vehicle (UGV). This paper presents a vision-based off-road terrain classification method that is robust despite environmental variation caused by weather changes. In order to cope with an overall image brightness variation, we use speeded-up robust features (SURF), and neural network classifier. Experimental results for real off-road images show that proposed method has a better performance than wavelet based one especially in case of large brightness variation.

[1]  Roberto Manduchi,et al.  Autonomous terrain characterisation and modelling for dynamic control of unmanned vehicles , 2002, IEEE/RSJ International Conference on Intelligent Robots and Systems.

[2]  Do-Jong Kim,et al.  Modified component-labeling algorithms applied to grayscale images , 2008, 2008 International Conference on Control, Automation and Systems.

[3]  Roberto Manduchi,et al.  Terrain perception for DEMO III , 2000, Proceedings of the IEEE Intelligent Vehicles Symposium 2000 (Cat. No.00TH8511).

[4]  Binoy Pinto,et al.  Speeded Up Robust Features , 2011 .

[5]  R. Manduchi,et al.  Classification Experiments on Real-World Texture , 2001 .

[6]  Frans C. A. Groen,et al.  Colour based off-road environment and terrain type classification , 2005 .

[7]  Luc Van Gool,et al.  Speeded-Up Robust Features (SURF) , 2008, Comput. Vis. Image Underst..

[8]  Joon Lyou,et al.  Neural Network Based Terrain Classification Using Wavelet Features , 2010, J. Intell. Robotic Syst..

[9]  Roberto Manduchi,et al.  Obstacle Detection and Terrain Classification for Autonomous Off-Road Navigation , 2005, Auton. Robots.