Implementation of machine vision as an input to a fuzzy water level control system

In this paper, a Type I fuzzy logic controller (FLC) is applied to execute machine vision parameters such as object tracking and image processing on a specific color threshold. This system is used to compensate for large power consumption of current PID controllers that are currently being used in industrial tank systems. The aim of the implemented fuzzy controller is to utilize the conventional camera to replace existing level sensors and linear variable differential transformers (LVDT) in monitoring the water level inside a reservoir under a definite tolerance. The camera is integrated in a Roborealm machine vision software which analyses the input parameter. An embedded Arduino microcontroller handles the regulation pump which maintains the specified water level onto user defined conditions. It is found out that the fuzzy system produced a better solution than conventional control techniques.