Vegetation mapping for landmine detection using long wave hyperspectral imagery

Abstract—We develop a vegetation mapping method using long-wave hyperspectral imagery and apply it to landmine de-tection. The novel aspect of the method is that it makes use of emissivity skewness. The main purpose of vegetation detection for mine detection is to minimize false alarms. Vegetation, such as round bushes, may be mistaken as mines by mine detection algorithms, particularly in synthetic aperture radar (SAR) imagery. We employ an unsupervised vegetation detection algo-rithm that exploits statistics of emissivity spectra of vegetation in the long-wave infrared spectrum for identification. This informa-tion is incorporated into a Choquet integral-based fusion struc-ture, which fuses detector outputs from hyperspectral imagery and SAR imagery. Vegetation mapping is shown to improve mine detection results over a variety of images and fusion models. Index Terms—Blackbody, clustering, decision-level fusion, emissivity normalization, expectation maximization (EM), mine detection, multisensor systems, vegetation mapping. I