The research of obstacle detection based on AK-means clustering algprithm in crosscountry

In order to obtain obstacle information in the cross-county environment for an unmanned ground vehicle (UGV), AK-means clustering algorithm is applied in four-layer laser radar data mining in this paper. The result of clustering serves as candidate obstacles. To overcome the false clustering due to vibration of UGV, weighted Euclidean distance is used to improve Davies-Bouldin Index (DBI). The experimental results show that the proposed obstacle detection algorithm is reliable and robust in low speed driving.