Studying the phenomena related to the IR thermographic detection of buried landmines

Abstract

Some theoretical and experimental problems related to the IR detection of buried landmines are discussed to express a rather pessimistic conclusion on the statistical reliability of this technique. Infrared (IR) thermographic detection of buried landmines has been considered as a promising diagnostic tool for many years. The number of related papers is large, including the information about the successful practical use of this technique by the U.S. Army in Kuwait during the “Desert Storm” operation. However, our fragmentary experience in this application field shows that the problem is not yet exhausted, in particular, if to discuss its statistical reliability. It is known that thermal NDT of materials is accompanied by a high level of specific noise that transits in either a high false alarm or a low correct detection. When detecting landmines, this statement is more than true, because, in this case, the number of factors influencing surface temperature gradients is enormous. In this study, we have analyzed some theoretical and practical issues related to the IR detection of buried landmines. A couple of models have been used to evaluate the influence of some non-trivial factors, such as wind, cloudiness, use of an artificial heater, etc. The experiments have been performed on two mine surrogates made of aluminum and paraffin. 2. Advanced numerical of buried landmines The term advanced model used in this paper is confronted to the term classical model that follows from many studies where a buried mine is treated as a “defect” or foreign object in a soil matrix. When modeling the detection of buried landmines, both heat and mass transfer phenomena must be analyzed. The soil moisture content, and, hence, the soil thermal properties, all vary with time, and not always in a straightforward manner. Surface evaporation depends on relative humidity, temperature and wind speed. Evaporation may change surface temperature gradients. The solar radiation that is the thermal stimulus in the detection of buried mines is not a simple cosine function as it is usually expressed in simple modeling. Cloudy periods, the change in the angle of solar radiation, shadows caused by nearby objects, or surface irregularities require modeling solar radiation input as a complex function. Also, surface heat exchange is strongly dependent on wind and both soil and air temperature. Furthermore, the great variety of mine types and designs, such as a mine in a protective aluminum case, leads to unusual defect geometries. Antipersonnel and antitank landmines are typically modeled as homogeneous disks located in uniform soil at a depth from 1 to 15 cm. Thus, the related geometry is cylindrical (two-dimensional - 2D). The soil is heated by solar radiation, in a diurnal cycle simulated by: , where the cyclic frequency ω is related to 24 hr variations. Most researchers believe that the detection of buried mines is possible primarily due to changes in the thermal properties of the soil appearing when installing a landmine.