Detection and identification of effluent gases by long wave infrared (LWIR) hyperspectral images

Results will be presented concerning the development of an algorithm for the detection, identification and relative quantification of effluent gases emitted in industrial plume stacks using an LWIR hyperspectral remote sensing system. The technique consists of several steps, which are initialized by the localization of critical wavelengths in the spectral signatures and then their integration into an algorithm for the detection of high concentrated gas pixels in the image cube using a correlation coefficient metric. Further mapping of low concentrated pixels is carried out by an iterative Matched Filter (MF) method. Following the mapping of all the gases in the image cube, a least squares method was applied to derive gas content. The algorithm was tested on data cubes acquired in the bay of Haifa with chimneys emitting SO
₂
and CO
₂
gases from distances of 400 and 1700 m away; good results were obtained.