Fuzzy logic controller for automatic vision parameter adjustment in a robotic dish handling system

Abstract

In this paper, a multilayer MIMO fuzzy logic controller (FLC) is proposed and implemented to realize automatic adjustment of the vision parameters "gain", "offset", and "threshold" to compensate for power fluctuations, changes in ambient light, and camera sensitivity drift in a machine-vision based integrated robotic dish handling system. The aim of the implemented controller is to provide consistent inspection performance while using machine vision for automated inspection of dishes in large commercial dish washing operations. With no dynamic model available for camera performance, FLC is implemented using the histogram of a multiple gray level vision target to control the vision parameters "gain" and "offset" to yield desired minimum and maximum gray levels of the target. The second layer of a two-layer FLC is used to partition the ZERO fuzzy set of the first layer into three subsets, which results in reduced oscillations of the system output. It is shown through experimental studies that the multilayer FLC performs well in controlling vision parameters. The multilayer FLC yields faster response with less overshoot than that of a conventional single layer FLC, and provides excellent camera performance.