论文标题
超快材料识别的稀疏性
Sparsity for Ultrafast Material Identification
论文作者
论文摘要
中红外光谱法通常用于识别材料。使用昂贵的桌面仪器在耗时的过程中测量了数千个光谱点。但是，材料识别是一个稀疏的问题，从理论上讲，只需几个测量方法就可以解决。在这里，我们利用问题的稀疏性，并开发出一种超快速，便携式和廉价的方法来识别材料。在单枪中，中红外摄像头可以根据其光谱特征识别材料。此方法不需要事先校准，因此可以在处理广泛的材料方面具有稳健性和多功能性。
Mid-infrared spectroscopy is often used to identify material. Thousands of spectral points are measured in a time-consuming process using expensive table-top instrument. However, material identification is a sparse problem, which in theory could be solved with just a few measurements. Here we exploit the sparsity of the problem and develop an ultra-fast, portable, and inexpensive method to identify materials. In a single-shot, a mid-infrared camera can identify materials based on their spectroscopic signatures. This method does not require prior calibration, making it robust and versatile in handling a broad range of materials.
