Robustly unstructured road detection based on road distribution model

Unstructured road detection is a key step of the Unmanned Guided Vehicle (UGV) system for road following. However, current vision-based unstructured road detection algorithms are usually affected by continuously changing backgrounds, different road types (shape, color), variable lighting conditions and weather conditions. Therefore, a road distribution model is theoretically analyzed and experimentally generated to help detecting unstructured roads. Global map and global positioning system (GPS) information are used to choose the corresponding road distribution model. Two traditional algorithms, support vector machine (SVM) and k-nearest neighbor (KNN), are carried out to verify the helpfulness of the proposed algorithm. The proposed algorithm has been evaluated by different types of unstructured roads and the experimental results show its effectiveness.