Increasing autonomy in agricultural robots:unevenness estimation of the terrain ahead

Natural terrain traversing is playing an increasinglyimportant role in service and field robotics to enablethe achievement of several tasks in the most disparate fieldsof human activities, e.g. planetary explorations, nuclear sitedecommissioning, precision farming and so forth.In this research, a new method for terrain unevenness estimationis proposed in order to increase the ability of an agriculturalbot to move autonomously in a countryside environment. Bythe use of exteroceptive sensors, such as stereoscopic or RGB-Dcameras, soil surfaces can be acquired and then manipulated toobtain a stochastic description of their geometry by means of thePower Spectral Density (PSD)-based analysis, in order to infersoil irregularities. This automatic ground unevenness estimationis performed online during the motion, allowing safe managementof robot activities.