Noncontact level sensing technique using computer vision

Various types of sensors have been developed for measuring the depth of liquids in tanks, such as a fuel tank in an aircraft. Almost all of these methods require the sensor to come in contact with the liquid with low-voltage electrical leads carrying the signal from the sensor to an external computer for processing. A particular concern when such in-contact sensors are used in fuel tanks is the potential for an electrical discharge through the sensor leads resulting in an explosion. In this paper, we propose a computer vision-based noncontact technique for measuring the level of a liquid in a closed container using an optical imaging system coupled with an appropriate set of image-processing algorithms. An experimental setup was specially designed and constructed for simulating the environment of vibrating fuel in the gas tank of aircrafts for demonstrating the effectiveness of the technique. The experimental results of this paper may be used for studying the feasibility of implementing an actual system, which could be incorporated in aircrafts as a part of the aircraft's fuel gauge system.