Unstructured road detection based on hybrid features

Erke Shang, Xiangjing An, Lei Ye, Meiping Shi, Hongtao Xue
2012 Proceedings of the 2012 2nd International Conference on Computer and Information Applications (ICCIA 2012)   unpublished
Unstructured road detection is a key step of the Unmanned Guided Vehicle (UGV) system for road following. In traditional methods, kinds of features are combined to improve the effect, but few of them explains the reason to choose feature descriptors. In this paper, we try to find a way that how to choose feature descriptors. First, support vector machine technology is used to analyze the importance among these common feature descriptors during the road detection process. A series of features
more » ... analyzed to show their ability to detect road surface from background. Then, a mass of experiment results verify the correct of our analysis. At last, a novel unstructured road detection algorithm based on hybrid features is proposed. Experiments conducted on actual roads illustrate the effective performance of the proposed algorithm.
doi:10.2991/iccia.2012.225 fatcat:x2s4gzvd3ze5xmweyyx3rafblm