This work focuses on measuring the velocity profiles of the liquid phase in annular flows by high-speed camera snapshots. This parameter is important for developing flow models and, therefore, improving engineering processes involving heat and mass transfer. This study introduces an image-based technique based on two-dimensional phase correlation to estimate the velocity profiles of liquid films in downward annular gas–liquid flows. High-speed camera images capture the dynamic behavior of disturbance waves and ripples on the liquid film surface. By analyzing sequential images, the method calculates translational offsets, from which velocity profiles are derived. These velocities are input into a mass conservation model to compute the average liquid film thickness, which is then compared with measurements from a conductance sensor. The results show deviations from reference sensor data of approximately 5% for a 26 mm pipe diameter and up to 20% for a 50 mm diameter, indicating reasonable agreement. Additionally, the technique successfully detects and quantifies droplet entrainment from disturbance waves, with deviations within 10% compared to manual measurements.