A Visual Based Algorithm for Measuring the Speed of Scrap Metal Vibration Feeding

In the field of steel metallurgy, scrap metal logistics in electric arc furnaces are typically conveyed horizontally through inertia-based vibratory conveyors. However, due to the different types, sizes, and appearances of scrap metal, as well as the repeated horizontal vibrations of conveyors, it is difficult to identify specific types of scrap metal for speed measurement. This article proposed a visual based speed measurement algorithm for the vibration feeding scenario of scrap metal. Firstly, we preprocessed the video dataset during the vibration feeding of scrap metal. Then we constructed a feature point extraction and matching model. Finally, based on the length of the feeding area, we estimated the vibration feeding speed by matching the pixel displacement of feature points in two frames in unit time. After on-site application verification, the speed curve measured by the algorithm is reasonable and consistent with the frequency curve of the horizontal vibrator, solving the problem of measurement of vibration feeding speed.