In this work, we report of a noncontact level measurement based on machine vision techniques. Data has been acquired with a high-definition webcam @ 30 frames per second. Image-processing algorithms have been developed and optimized using LabVIEW. The measurement may then automate a servo actuated hand-valve. On testing, the system was found to be linear, repeatable, and accurate to ±0.5% of the measured value for 50 measurements carried out. In the scheme are proposed hereof, the levels may just be remotely observed on a LabVIEW SCADA panel and used as a set point to a regulatory control-actuator system. The system is rugged and can be calibrated to 0.05% precision for all rotameter types used in any environment from a simple DM plant to a petrochemical industry for a fixed camera distance.