Real-Time Slag Ratio Monitoring of Iron Streams Cast from a Blast Furnace via Histogram-Based Image Analysis

This research presents an image-based real-time monitoring system for estimating the slag ratio in iron streams casting out from the blast furnace. The method leverages the emissivity difference between slag and molten iron, which leads to distinct brightness contrasts in thermal images. By applying histogram analysis to the captured images, the system can estimate the proportion of slag present within the iron stream. The underlying algorithm incorporates predictive thermal radiation models for both slag and iron, and further refines the analysis by accounting for the spectral response characteristics of the optical system. Specifically, the transmission spectrum of the optical filters and the quantum efficiency profile of the imaging sensor are integrated into the model to enhance the accuracy of slag-to-iron segmentation and quantification.

In addition to slag ratio estimation, the optical setup is also capable of real-time flow rate measurement by analyzing surface texture patterns within the stream. Live video of the molten iron jet is captured at approximately 300 frames per second and processed using a modified image correlation algorithm to extract key parameters such as velocity distribution and jet cross-sectional dimensions. These measurements are then combined to estimate the casting rate, which is updated every 5 seconds.

The entire system has been integrated into a prototype unit, which is currently installed adjacent to an operational blast furnace. This system enables continuous, non-contact monitoring of the casting process and provides valuable data to support real-time process control and decision-making within the blast furnace operation, contributing to improved efficiency, consistency, and quality control in ironmaking.