Abstract—Autonomous robot must possess the capability to traverse intelligently on various terrains that may largely affect the performance and controllability of the vehicle. To enable autonomous navigation on natural unknown terrain, the traversal difficulty must be evaluated initially. This is accomplished by real-time assessment and classification of terrain for traversability analysis. A technique for terrain classification for traversability assessment of terrain by extracting the textural features from visual sensing of terrain data using wavelet packet decomposition (WPD) is presented in this paper. Clustering is used to classify the terrain describing its suitability for navigation. And further we have presented a fast and optimum algorithm for path planning of robot on the assessed navigable terrain.

Index Terms—Terrain classification, real-time traversability assessment, autonomous navigation, wavelet packet decomposition, clustering.

K. K. Soundra Pandian is with the PDPM-Indian Institute of Information Technology, Design and Manufacturing Jabalpur(e-mail: soundra.pandiankk@gmail.com).
Priyanka Mathru is with the Ansal Institute of Technology, Gurgaon,India (e-mail: priyanka10@gmail.com).

Cite: K. K. SoundraPandian and Priyanka Mathur, "Wavelet Packet Decomposition Based Traversable Terrain Classification for Autonomous Robot Navigation," International Journal of Computer and Electrical Engineering vol. 4, no. 5, pp. 661-665 , 2012.