An obstacle detection algorithm used sequential sonar data for Autonomous Land Vehicle

Aiming to the navigation and control of Autonomous Land Vehicle (ALV), a new obstacle detection algorithm is presented, in which sequential sonar data from the sonar sensor pair is combined to improve the accuracy and the robustness. The obstacles´ features, depicted as points or lines, are firstly extracted according to the concentric arcs model of the sonar returns. Then the feature sequences extracted during the movement of ALV, are recognized and localized using a generalized Hough Transform. The algorithm´s validity is approved through a field test.