Laser & Optoelectronics Progress, Volume. 62, Issue 9, 0930003(2025)
High-Resolution Computational Spectrometer Based on Automatic Differentiation Optimization Algorithm
Spectrometers have important applications in scientific research, industrial production, and daily life. Traditional dispersive spectroscopy spectrometers often have a large volume while achieving high-resolution measurements. In recent years, computational reconstruction spectrometers based on single speckle images have attracted much attention due to their advantages of high resolution and compactness. However, this scheme requires iterative optimization during spectral reconstruction to improve measurement accuracy, so the time cost is generally on the order of minutes, which limits its real-time application. This article introduces the automatic differential optimization algorithm into the speckle calculation spectrometer, which significantly reduces the spectral reconstruction time from the minute level to the second level without affecting the measurement accuracy. It achieves spectral measurement with a resolution of 2 pm and over 2000 channels, providing a new solution for rapid calculation and measurement of speckle based spectra.
Get Citation
Copy Citation Text
Yangfan Qi, Junrui Liang, Jun Li, Jiangming Xu, Jinyong Leng, Pu Zhou. High-Resolution Computational Spectrometer Based on Automatic Differentiation Optimization Algorithm[J]. Laser & Optoelectronics Progress, 2025, 62(9): 0930003
Category: Spectroscopy
Received: Aug. 19, 2024
Accepted: Sep. 23, 2024
Published Online: Apr. 23, 2025
The Author Email: Zhou Pu (zhoupu203@163.com)
CSTR:32186.14.LOP241858