Path Tracking in Underwater Vehicle Navigation – On-line Evaluation of Guidance Path Errors via Vision Sensor

Atlanta

Conferencia; 49th IEEE Conference on Decision and Control (CDC10); 2010

IEEE

In this paper a vision-based algorithm for onlineestimation of position control errors in a guidance systemfor path tracking is presented. The algorithm uses techniquesof pattern recognition, however with a high degree of morphologicalsimplifications. The algorithm detects the path linestretch which is displayed in the actual frame, and estimatesthe slope of the line and its midpoint. Moreover, it identifies aconfidence zone in where the line stretch would be certainly. Adhocexperiments with a subaquatic vehicle in an test tank showthe features of the algorithm proposed under strong conditionsof light perturbation and cloudy water