Comparing two methods of object detection in autonomous vehicle

Autonomous obstacle avoidance in mobile robots has been an active field in image processing and control theory. In this paper, we propose and evaluate two methods for multiple objects detection for a mobile robot in outdoor. The first method uses a block matching algorithm to obtain a disparity map and SURF features detector. The second method uses a semi-global stereo vision environment. Both methods are tested during detecting pedestrian, boxes and cones using a video sequence in validation of a robot.