Measurement of emissive spectral fingerprinting for material identity by the fourier transform infrared remote sensing technique

Zheng Fang,Yongyu Wang,Tianxu Zhang
DOI: https://doi.org/10.3788/AOS200929s2.0198
2009-01-01
Abstract:A remote sensing infrared spectroscopy equipment made up of a laboratory spectrometer and a Newtonian reflector telescope is introduced. This equipment is used to measure emissive infrared spectrums of five specific specimens composed of respective materials. The research includes system hardware establishment and identifying algorithm software investigation. Experiments are carried out in the standard condition without any infrared source. The signal collected into the facility is spontaneous radiation from the specimens. A database is found to store the spectrum characteristics of each observed substance, to estimate which material the specimen comprises. Objects in normal temperature radiate a small amount of infrared. As a result, the signal is inundated by noise in the spectral plot, which makes it hard to identify the spectral characteristics by observing the plotted lines. Correlating the experimental data with characteristic spectral sequences in the established database is a better way. The maximum correlation value indicates the most possible substance. The experiment shows that the identifying accuracy is more than 90 percent. From the experimental study, it can be concluded that integrating the hardware and software described above, researchers can find an effective way to identify the substance by means of non-contact infrared spectral fingerprinting detection.
What problem does this paper attempt to address?