Thermal infrared imaging is a field of science that evolves rapidly. Scientists have used for years the simplest tool: thermal
broadband cameras. These allow to perform target characterization in both the longwave (LWIR) and midwave (MWIR)
infrared spectral range. Infrared thermal imaging is used for a wide range of applications, especially in the combustion
domain. For example, it can be used to follow combustion reactions, in order to characterize the injection and the ignition
in a combustion chamber or even to observe gases produced by a flare or smokestack. Most combustion gases, such as
carbon dioxide (CO2), selectively absorb/emit infrared radiation at discrete energies, i.e. over a very narrow spectral range.
Therefore, temperatures derived from broadband imaging are not reliable without prior knowledge of spectral emissivity.
This information is not directly available from broadband images. However, spectral information is available using spectral
filters. In this work, combustion analysis was carried out using a Telops MS-IR MW camera, which allows multispectral
imaging at a high frame rate. A motorized filter wheel allowing synchronized acquisitions on eight (8) different channels
was used to provide time-resolved multispectral imaging of combustion products of a candle in which black powder has
been burnt to create a burst. It was then possible to estimate the temperature by modeling spectral profiles derived from
information obtained with the different spectral filters. Comparison with temperatures obtained using conventional
broadband imaging illustrates the benefits of time-resolved multispectral imaging for the characterization of combustion