标题： 光谱检测中压缩感知降噪的注意点 显示英文标题

标题： Note on the noise reduction in spectroscopic detection with compressed sensing

摘要： 沿延迟时间的光谱采样通常使用均匀的延迟间隔，这必须足够小以满足奈奎斯特-香农采样定理。 采样定理为采样率设定了下限，以确保光谱特征的准确分辨。 然而，通过利用信号的先验知识，如稀疏性，可以放宽这一限制。 压缩感知是一种成功应用于空间测量（例如单像素成像）的欠采样技术，但在光谱测量中，尤其是时间采样方面，尚未得到充分探索。 在本工作中，我们研究了压缩感知在改善时间光谱测量中的能力，以减轻测量噪声和固有噪声。 通过将压缩感知应用于单次泵浦-探测数据，我们证明了其在降噪方面的有效性。 此外，我们提出了一种可行的实验方案，使用数字微镜器件来实现时间采样的压缩感知。 这种方法为光谱学提供了一种有前景的方法，以减少信号噪声和样本测量的数量。 显示英文摘要

摘要： Spectroscopy sampling along delay time is typically performed with uniform delay spacing, which has to be low enough to satisfy the Nyquist-Shannon sampling theorem. The sampling theorem puts the lower bound for the sampling rate to ensure accurate resolution of the spectral features. However, this bound can be relaxed by leveraging prior knowledge of the signals, such as sparsity. Compressed sensing, a under-sampling technique successfully applied to spatial measurements (e.g., single-pixel imaging), has yet to be fully explored for the spectral measurements especially for the temporal sampling. In this work, we investigate the capability of compressed sensing for improving the temporal spectroscopic measurements to mitigate both measurement noise and intrinsic noise. By applying compressed sensing to single-shot pump-probe data, we demonstrate its effectiveness in noise reduction. Additionally, we propose a feasible experimental scheme using a digital mirror device to implement compressed sensing for temporal sampling. This approach provides a promising method for spectroscopy to reduce the signal noise and the number of sample measurements.