Importance Of Band Selection For Ethene And Methanol Gas Detection In Hyperspectral Imagery

摘要

There are many traditional and innovative methods for remote sensing toxic gases, either pollutant or as a side effect of developing technology and growing industrial applications. One of the increasingly popular methods for remote sensing of gases is hyperspectral data capture and analysis. It is possible to detect and classify by using the spectral signature of the gas and the data taken by medium and long wave infrared hyperspectral sensors operating in the band interval that the sought gas is effective. In this work, a signature-based multi-gas detection method is proposed for detection of ethene and methanol to show the importance of band selection for gas detection in hyperspectral imagery.