Numerically Denoising Thermally Tunable and Thickness‐Dependent Terahertz Signals in ErFeO <sub>3</sub> Based on Bézier Curves and <i>B</i> ‐Splines

Xinxi Zeng,Han Zhang,Xiaoqing Xi,Bo Li,Ji Zhou
DOI: https://doi.org/10.1002/andp.202000464
2021-01-01
Annalen der Physik
Abstract:The terahertz (THz) spectral range offers a platform for experimentally collecting electromagnetic pulses to analyze the intrinsic resonances of matter and has become an important field for applying intrinsic responses to future information devices. However, a lack of numerical methods for signal denoising exists in the knowledge of the THz spectral range. In this paper, numerical methods that employ the Bézier curve and its generalization ( B ‐spline) to denoise thermally tunable and thickness‐dependent terahertz signals are proposed. Different thicknesses of ErFeO 3 are chosen and prepared. Both Bézier curves and B ‐splines have the ability to denoise the signals and the effect of the Bézier curves on the noise is close to but slightly less than that of the B ‐splines. Additionally, the B ‐splines show a better noise reduction effect and have better signal strength stability and robustness than do the Bézier curves. Thus, B ‐splines are applied to process the raw data, and quantitative insight into the trend of the resonant frequencies, peak heights, and transmittances of antiferromagnetic resonance and ferromagnetic resonance with temperature is obtained.
What problem does this paper attempt to address?