Abstract:
Owing to industrialization and the rapid development of modern society, the leakage of dangerous chemical gases in industrial production seriously endangers the safety of human life and property. Effectively detecting the presence of contaminated gas and obtaining information on the gas concentration and distribution have become important topics in gas leakage detection. Uncooled Snapshot Infrared Video Spectrometer (USIVS) is an ideal hardware scheme that can directly perceive the existence of dangerous chemical gas from the image and obtain the position of dangerous chemical gas to provide strong support for emergency responses. However, the sensitivity and spectral resolution of commercial lightweight passive infrared spectral imagers are relatively limited, and it is difficult to accurately detect the presence of polluted gases using existing inversion methods. In this study, an infrared video spectral imager is introduced based on an uncooled snapshot and its applicable data-processing technology. The gas concentration inversion method is used to simulate gas at different temperatures and optical path lengths, and the inversion results are relatively accurate, with average errors of 2.88% and 0.61%, respectively. The gas concentration inversion method is tested in laboratory and outdoor settings. The results show that the algorithm has good stability with average errors of 6.18% and 7.47%. The effective combination of USIVS and data processing technology can quickly and accurately detect the presence of polluted gas and provide the gas concentration of each pixel in the image. This in turn can realize gas cloud concentration inversion, providing a reference for the commercialization and practical application of this technology in the future.