PREDICTING LIQUID FLOW USING NON-CONTACT SENSORS

Flow measurement and control are essential in plant process control. This paper
aims to shed some light on predicting liquid flow using computer vision. The
development will be done onto a level process pilot plant. This pilot plant consists
of a calibration tank, a buffer tank, level transmitter and flow meter. A camera and
an iilumination source wiii be installed on top of the calibration tank, based on the
specific requirements. Some of the required tools for this project are MA TLAB
SIMULINK software and Image Processing Toolbox. The development of the
system simulation is based on the Mathematical Model of the level process pilot
plant, which is developed in the early stage of this project. Overall, this project
proposes a computer vision-based technique for measuring the level of a liquid in
a closed container using an optical imaging system and predictive startegy for the
fluid out-flow rate.