Defeating camouflage and finding explosives through spectral matched filtering of hyperspectral imagery

In order to achieve their goal of surreptitious operation within a country, terrorist organizations attempt to hide themselves from public view. In many instances such masking takes the form of simply appearing like the surrounding populace. In others, such as training facilities, standard military camouflaging techniques are used to conceal the group’s equipment and activities. To effectively monitor and suppress activities of terrorist organizations, defeating the groups’ attempt to hide is essential. Although finding individuals hiding within a society is extremely problematic, discovering camouflaged equipment, facilities, and personnel is readily accomplished by proper exploitation of hyperspectral imagery. Camouflage techniques attempt to make an object appear similar to its background, thereby making it difficult to find. Although making an object have similar color to its background is fairly easy, making it have the same spectral appearance is nearly impossible, unless the object is covered in the same material as the background. Even attempting to hide an object by covering it in background material will not work against a spectral imager since the act of moving the background material, e.g., foliage cuttings, changes the material’s spectral characteristics. Hence, by collecting and properly exploiting spectral imagery, camouflaged objects can be readily differentiated from their background. This paper presents development of this technique, and of the MIDIS (multi-band identification and discrimination imaging spectroradiometer) instrument capable of real-time discrimination of camouflaged objects throughout a scene. Spectral matched-filtering of hyperspectral imagery also has the potential to find vehicles or structures which may be laden with explosives. Many explosives contain volatile materials, the release of which can be imaged by viewing appropriate spectral regions. Volatiles from the fuel oil in readily-produced ANFO are an example. If such volatiles were seen emanating from a vehicle or structure where they would not normally be expected, closer inspection would be warranted. Additionally, packing a vehicle with explosives often leaves trace residues on the outside of the vehicle. Spectral imaging and matched filtering can be used to identify these residues. Incorporation of spectral imaging surveillance equipment at probable terrorist targets could avert disasters such as the tragic bombing of the Murrah Federal Building in Oklahoma City. Application of MIDIS technology to explosive identification is also detailed.