A non-contact level sensing technique using computer vision and artificial neural networks
Abstract
A non-contact level sensing technique that incorporates computer vision and neural networks was developed. This technique was implemented in measurements of water levels in a cylindrical tank. The results showed that the technique developed is capable of measuring levels of liquids with better accuracy than that of the conventional techniques. Next to the improved accuracy, the method offers the advantage of the absence of a direct contact between the measuring instrument and the liquid.