An image analysis-based method is developed to automatically interpret videos of a transparent Hele-Shaw model in studies of blast furnace hearth drainage in pilot scale. By the approach, information from experiments can be extracted for quantitative assessment of the draining phenomena. The algorithm consists of modules for (1) image pre-processing, (2) interface tracking, and (3) bending-point detection. The pre-processing step makes preparatory estimates for the next module that extracts the liquid–gas and liquid–liquid interfaces. Finally, bending points of the interfaces are detected and parameters quantifying the interface bending are calculated. The algorithm is illustrated by applying it to a video of a draining experiment, demonstrating how information can be extracted. The method can be used to compile information from a large set of experiments to gain a deeper understanding of the complex two-phase flow in the blast furnace hearth and to quantify the findings for a validation of computational models.