Phosphorescence marking method for liquid level detection and control using computer vision in dark environments

Abstract

The liquid mixture control and level measurement processes were examined in the study using computer images. Liquids of optical characteristics result in low contrast and light reflections in captured camera images. This causes a slowdown in the image's level detection and the processing of mixed images. The method of marking the liquid surface with phosphorescent material in a dark environment was proposed to reduce the light effect on the images. The method's goal is to lessen the noise and light effect in the image that causes measurement errors. A system for experimental measurement and control was set up, comprising liquid tanks, pumps, phosphorus float, camera, and computer. RGB color space, column profile, color moment, filtering, and threshold techniques were used to detected the liquid level from the camera images. When the computer vision results were compared with the real results, the measurement error was calculated as ± 2.595 mm, RMSE = 2.595, and an error rate of 0.66 % for the 400 ml liquid tank. The method simplifies software development procedures for computer, FPGA and microprocessor-based level measurement devices by simplifying intricate image processing algorithms. The computer vision-based phosphorescence marking method can be used in liquid mixture processes for experimental and industrial purposes. It is a liquid level measurement technique that is low cost and easy to apply. Liquids affected by light can employ this method. It has the potential for application in biology, medicine and chemistry laboratories.