Presentation Information
[10a-PA6-4]Quantitative Comparison of Regularization Methods for Wide-Field AC Current Imaging Using RF-Dressed States of Electron Spins in Diamond
〇Yu Masuda1,2, Yuma Itabashi1,2, Yuki Nagao1,2, Motofumi Fushimi1, Yuichiro Matsuzaki3, Norio Tokuda4, Junko Ishi-Hayase1,2 (1.Keio Univ., 2.Keio CSRN, 3.Chuo Univ., 4.Kanazawa Univ.)
Keywords:
nitrogen-vacancy centers,Quantum sensing,Inverse problem
Nitrogen-vacancy centers in diamond are quantum sensors capable of highly sensitive and spatially-resolution magnetic field measurements at room temperature and atmospheric pressure, and are expected to have applications in areas such as fault inspection of microcircuits. For this to be realized, the inverse problem of reconstructing the current density distribution from the measured magnetic field is crucial. In a previous report, we succeeded in wide-range imaging of AC current by performing an analysis on the AC magnetic field amplitude and phase distribution obtained using RF-dressed states, applying noise-robust L2 regularization. However, there are concerns that the smoothing inherent in L2 regularization may not adequately reproduce the actual current distribution at curved sections and edge of the circuit. On the other hand, L1 regularization, which is commonly used like L2 regularization, is expected to suppress unwanted background components, but a direct quantitative comparison in this field has not yet been reported. Therefore, we quantitatively compare L1/L2 regularization and investigate a more accurate current imaging method.