Fig. 4
From: Deep learning revealed statistics of the MgO particles dissolution rate in a CaO–Al2O3–SiO2–MgO slag
Statistical analysis of the particle diameter with time at three different temperatures and different video data sets. At (a) 1450 °C, (b) 1500 °C, and (c) 1550 °C five different video data sets are analysed to understand the particle dissolution over time. The atU-Net prediction for the different videos is indicated as atU-Net-Video 1–5 and for the manual evaluation as M-Video 1–5. The five different videos for each temperature are indicated by different colors (dark grey, red, blue, green and purple). (d) Diagram which shows the mean MgO-particle diameters extracted from the five video data-sets, utilizing the atU-Net prediction and the manual evaluation (indicated with \(\overline{\text{atU}-\text{Net}}\) and \(\overline{\text{Manual}}\), respectively) at each temperature. The shaded area and the error bars around the mean diameters indicate the standard deviation of the atU-net prediction and the manual evaluation, respectively. The results show a reduction in the particle's dissolution time as the temperature increases, as well as good agreement between atU-Net and manual evaluation.
