Abstract. Standoff detection of hazardous materials using infrared backscattering spectroscopy shows promise due to its speed of detection, sensitivity, chemical specificity, eye safety, and the ability to perform detection in a stealthy manner. However, infrared diffuse reflectance spectra of trace particles on substrates exhibit a strong dependence on substrate type, particle size, and mass loading. This has a negative impact on the performance of detection algorithms and false alarm rates when using commercially available spectral databases. We present two models that are quickly computed yet capture most of the observed spectral features. Model I is best suited for calculating the diffuse reflectance spectra of trace amounts of particles on relatively smooth substrates, while model II extends the applicability of model I to particles on very rough substrates. The models can be used for algorithm development and training. The main inputs to the models are the analyte and substrate optical constants as a function of wavelength, and the particle size distribution and average mass loading. The accuracy of the models was checked by comparing to experimentally measured diffuse reflectance spectra of several carefully prepared samples.

中文翻译：

---

表面微粒痕迹红外反射特征的测量和建模

摘要。由于其检测速度、灵敏度、化学特异性、眼睛安全性以及以隐蔽方式进行检测的能力，使用红外反向散射光谱对危险材料进行隔离检测显示出前景。然而，基板上痕量粒子的红外漫反射光谱表现出对基板类型、粒径和质量负载的强烈依赖。当使用商用光谱数据库时，这会对检测算法的性能和误报率产生负面影响。我们提出了两个模型，它们可以快速计算但捕获了大部分观察到的光谱特征。模型 I 最适合计算相对光滑基板上微量颗粒的漫反射光谱，而模型 II 将模型 I 的适用性扩展到非常粗糙的基材上的颗粒。这些模型可用于算法开发和训练。模型的主要输入是作为波长函数的分析物和底物光学常数，以及粒度分布和平均质量负载。通过与几个精心准备的样品的实验测量漫反射光谱进行比较来检查模型的准确性。