Multi-tank fuzzy level controller system using Kinect
Date:
This paper presents the development of a level control system using 3D machine vision techniques based on a single level-detector sensor (Kinect), in an arrangement of three tanks. Using image processing algorithms, each one of the tanks is classified within the video sequence as well as each one of the depth measurements is averaged regarding the cross sectional area of the surface of the tanks. This information is used to control the filling level for each tank using a proportional fuzzy controller and a proportional-derivative fuzzy controller. Finally the performance of each controller is compared.