High-fidelity tomographic reconstruction for infrared video bolometers through physics-based background radiation modeling

Precise control of radiation losses from impurities remains a critical challenge for achieving stable operation in fusion plasmas. Infrared video bolometers (IRVBs) are an important diagnostic tool for visualizing the spatial distribution of radiated power in such plasmas. However, conventional IRVB signal-processing methods exhibit inherent limitations, as they typically assume a linear temporal variation of dynamically changing background radiation, which restricts the accurate reconstruction of radiated power profiles. To overcome this limitation, this study proposes a novel signal-processing algorithm based on the heat balance equation. The distinct feature of the proposed approach is its ability to physically model time-dependent background radiation, thereby effectively isolating its contribution to the measured signal. The proposed algorithm's performance was validated using phantom datasets with known ground truth. Synthetic tests simulating various discharge scenarios demonstrate that the proposed method reconstructs radiated power profiles with significantly higher accuracy (coefficient of determination, R2≥0.99 in all cases) compared with conventional methods. In addition, the algorithm was applied to representative Korea Superconducting Tokamak Advanced Research high-confinement mode discharges, confirming that the reconstructed total radiated power and its temporal evolution remain physically consistent under realistic plasma conditions. These results indicate that the developed algorithm provides a robust tool for enhancing the reliability of impurity transport and power balance analyses in fusion plasmas.