Chemical vapor detection and mapping with a multispectral forward-looking infrared (FLIR)

Detection of chemical vapors with a remote sensor is a requirement for both military defense and civilian pollution control. The FLIR is a natural instrument from which to build a chemical sensor since most chemical vapors of interest are spectrally active within its operating wavelength range, it is widely distributed in the battlefield, and it is becoming a standard surveillance tool in police departments. Additionally, the output image provides a 2D concentration map of the cloud which is easily interpreted by a lightly trained operator. A system has been designed to provide the spectral sensitivity in a dedicated instrument or to place a chemical detection capability as an adjunct function in a military thermal imager. In the latter case an additional detector array which is spectrally filtered at the focal plane is added to the imager. Real-time autonomous detection and alarm is a military requirement and is desired for commercial use. A detection system model based on a Gaussian vapor concentration distribution has been the basis for detection algorithms. Image processing and analysis methods have been based upon information theory. These techniques extract the cloud image from a clutter scene and perform a limited vapor classification. These methods are suited to hyperspectral imagery.