Qualitative vision-based navigation based on sloped funnel lane concept

Funnel lane is a map-less visual navigation technique that tries qualitatively to follow a path that has been recorded before by a camera. Unlike some other methods, funnel lane does not require any calculation to relate world coordinates to image coordinates. However, the funnel lane has some shortcomings. First, it does not provide any information about the radius of rotation of the robot. This reduces the robot maneuverability of the robots and, on some occasions, does not let the robot to correct its path if a deviation occurs. Second, funnel lane constraints sometimes do not distinguish between forward or turning movement of the robot while the robot is in the funnel lane, and command the robot to go forward. This prevents the robot to follow the desired path and leads to failure of the robot’s mission. This paper introduces the sloped funnel lane technique to address these shortcomings. It sets the rotation radius based on the observed frames. Moreover, it reduces translation and rotation ambiguity. Therefore, the robot can follow any desired path leading to more robust and accurate navigation. Experimental results on challenging scenarios on a real ground robot demonstrate the effectiveness of the sloped funnel lane technique.