Performant Multispectral LIF-based Standoff Detection System with Spectral and Temporal Features for the Classification of CBE Hazards

Abstract

Because of its high sensitivity, laser-induced fluorescence (LIF) is a mature and well known technique for the detection of substances and for monitoring chemical processes. Compared to different methods like Raman spectroscopy, an identification is often not possible with LIF, only, and the classification of CBE hazards is commonly not best performing. Improving the technology by recording spectral and in addition temporal fluorescence emission a significant gain of information can be achieved. A detection system based on a multi-wavelength sub-nanosecond laser source is used to acquire spectral and time-resolved data from a standoff distance of 3.5 m. The dataset contains seven different bacterial species (Bacillus atrophaeus, Bacillus thuringiensis, Burkholderia fungorum, Burkholderia pyrrocinia, Micrococcus luteus, Oligella urethralis, Yersinia aldovae) and six types of oil. The data are analyzed by means of machine learning algorithms. First findings show a valuable contribution of time-resolved fluorescence data to the classification of the investigated chemicals and biological agents to their species level with an accuracy of more than 92 %.