O. Robert Mitchell

Algorithms for IR-imagery-based airborne landmine and minefield detection

In this paper we revisit and enhance various algorithms for landmine detection, discrimination and recognition. Single- band and multi-band medium wave infrared (MWIR) image data from the May data collection (part of Lightweight Airborne multispectral Minefield Detection-Interim (LAMBD-I) program) is used for the analysis. In particular discrimination based on gray-scale moments is explored and its effectiveness is evaluated for surface mines under IR imaging using receiver operating characteristics (ROC) curves. The discriminatory power of gray-scale moments is compared with the RX and matched fiber based detectors for different terrain (e.g., grass,sand) and different mine types. The performance of single-band (broadband) MWIR imagery is compared with multi- band (short-pass and long-pass) MWIR images. Also direct multi-band detection is compared against fusion of multiple single-band responses. Gray-scale moment based target discrimination at potential target locations, identified by RX or matched fiber detectors, is shown to be computationally efficient and provides better performance in terms of reduced false alarms for comparable probability of detection. An evolutionary framework for minefield identification, in the presence of inevitable false targets, is also presented. Starting from the locations of individual mine targets and false alarms, the evolutionary algorithm is used to identify the underlying structure of the minefield. Issues in the detection of different minefield layouts are discussed. Preliminary implementation shows the promise of this approach in identification of a wide variety of minefields.

Algorithms and architecture for airborne minefield detection

The current minefield detection approach is based on a sequential processing employing mine detection followed by minefield detection. In case of patterned minefield, minefield detection algorithms seek to exploit the minefield pattern (such as linearity) while in case of scattered minefield they utilize the spatial distribution of the mine targets. However, significant challenges remain in adequate modeling and detection of the minefield process especially in the presence of false alarms due to cultured as well as natural clutter. A short review of the literature on spatial point processes is included especially for the case of scattered minefields. It is further noted that, minefields are characterized by as a pattern (or spatial distribution) of similar looking mine-like objects. The sequential mine-detection followed by mine-field detection paradigm fails to exploit this critical aspect of similarity of targets for minefield detection. In this paper we propose a minefield detection scheme that incorporates similarity based clustering of targets in order to improve the performance of minefield detection. This approach can be interpreted as statistics of a marked point process. Some preliminary comparative ROC curves are evaluated for simulated minefield data in order to show the effectiveness of the minefield detection based on the marked point process. An autonomous self-organizing scheme for on-line clustering of mine-targets is also presented.

Combined sensor approach to the detection and discrimination of antipersonnel mine

In this study we investigate the detection of plastic anti- personnel mine simulants of several sizes using radar and IR imaging. The problem is first studied numerically with the finite-difference, time-domain method and an inverse synthetic aperture radar imaging algorithm to estimate the optimum conditions in dielectric contrast between the mine (epsilon) m and the soil (epsilon) s. It is found that to obtain a recognizable shape, the size of the simulant should be about two wavelengths when the dielectric contrast is low. Range of soil permittivity considered is from 2.8 to 6.0 corresponding to dry and wet soil conditions up to a volumetric soil moisture of 10 percent. The diameter of the mines ranged from 6 cm to 12 cm and their height ranges from 3.5 cm to 5 cm. It is found that when a mine diameter is larger than a wavelength, its image is discernible and becomes clearer when its diameter exceeds two wavelengths. It is found that a plastic mine with a dielectric constant of 2.5 embedded in a dry soil medium with a dielectric constant of 2.8 and locally flat surface can generate an image with the correct geometric shape. An experimental study of microwave-IR imaging is also conducted in this study. The key impact on IR imaging is the change in the amount of moisture in the soil medium caused by the presence of the simulant. Thus, different amounts of moisture are being heated by the microwave in regions with and without the simulant. This in turn causes a temperature difference leading to the IR image. Since the simulant contains no appreciable moisture, the temperature over the simulant is generally lower than the region around it. The soil inhomogeneity and heating pattern effects are discernible in the IR images.

Gray-scale moment invariants for airborne mine detection, discrimination and false alarm mitigation

Shape features based on gray-scale moment invariants are presented for airborne mine detection and discrimination. Eleven shape features are obtained by translation, rotation and contrast normalization of the fourth-order gray-scale moments. Mahalanobis distance between an observed and true (average) shape feature vector is used as a shape metric. Covariance matrix corresponding to the average shape feature vector is obtained analytically using an additive and multiplicative noise model for the MWIR image. Effectiveness of gray scale moment invariant shape features for mine discrimination and false alarm mitigation is shown using MWIR imagery collected for LAMD-I program in May 2000. Successful implementation of the features in an airborne detection depends on the consistency of these shape features over time with change in factors such as solar illumination, ageing, clouds and environmental conditions. A study of the variability of gray-scale moment invariant-based shape features with time is conducted using MWIR time-sequenced imagery acquired in June-July 1998 by E-OIR.

Detection of antipersonnel land mines based on waterjet-induced thermal images

The shape and thermal properties of buried objects can result in a variation in the temperature profile on the surface of the ground. IR imaging has been used to exploit this variation to detect the presence of buried objects. The thermal signature in such cases is normally induced by natural means such as diurnal cycles. This method requires observation at specific times of day and has not in general allowed reliable detection and discrimination, especially for small antipersonnel mines. We have developed a process that uses an array of heated waterjets to rapidly induce a thermal signature of buried objects in the region of interest. The high-pressure, small diameter waterjets penetrate the soil but are deflected by a formed buried objects. A temperature profile on the ground surface is formed due to the radiation and conduction of heat from the water blocked and reflected by the surface of the buried object and the heating of the object itself due to heat transferred from the object to a blurred 2D IR image of the surface. Deblurring and other physics-based image processing techniques are used to correct for the heat diffusion and an estimate can be made of the 3D shape of the part of the buried object which is covered by the waterjet. A time history of the thermal profile is also available when several IR images are acquired after the waterjets are applied. This allows further analysis of the nature of the properties of the buried objects. Known properties of land mines can be used to discriminate them from other buried objects. Shape feature properties based on Fourier descriptors have been developed to allow discrimination of objects.

Surface land mine detection in airborne images using the circular harmonics transform

Multi-band medium wave infrared (MWIR) image data collected from the Lightweight Airborne multispectral Minefield Detection-Interim (LAMD-I) program is examined for the detection of surface landmines. Because the orientation of the image acquisition from aircraft with respect to the mine and the minefield is unknown, there is a need to develop an orientation invariant-based approach for landmine and minefield detection. A rotation invariant circular harmonics transform (CHT)-based approach is presented for surface landmine detection. The magnitude information from the CHT is used for finding mine-like regions within the MWIR images. A three-tiered hierarchical thresholding technique provides the basis for highlighting potential surface landmines. Mine shape and size information are used for generating landmine confidence values. Surface landmine detection capability is presented for 82 MWIR broadband images with sand and short and long grass terrain conditions for daytime and nighttime acquired MWIR image data. Receiver operator characteristic (ROC) curves are used for comparing experimental results from this technique with an existing an adaptive multi-band CFAR detector (RX approach).

Characterization of single-waterjet-induced thermal profile for antipersonnel land mine detection and discrimination

IR imaging has been used for landmine detection and discrimination by exploiting the variations in temperature profile on the surface, which may be induced by natural phenomenon such as diurnal cycles or using artificial means such as heated waterjets. While the former method has, in general, not been able to reliably detect and discriminate for small antipersonnel mines, the latter suffers from poor response time. Our previous research has shown that, for waterjet induced thermal images, it takes approximately 15 minutes for the profile of the buried object before it is available on the surface. In this paper we explore the possibility of using thermal profile induced by a single heated water jet when viewed directly into the hole created by the waterjet. A heated waterjet, as it penetrates the ground cover, also digs a hole through which the heat radiates out. The spatial and temporal variation of the heat profile in and around the hole has shown to be rich in information about the buried object. Moreover, the response is much faster when compared to the conduction of heat through the soil to the surface. This paper will present the basic phenomenology and characterize such thermal images induced by single heated waterjet. The spatial and temporal variations are used to detect the presence of an object and its material type. Some possibility to measure the depth of the buried object is also explored.

Acoustic and Doppler radar detection of buried land mines using high-pressure water jets

The goal of the waterjet-based mine location and identification project is to find a way to use waterjets to locate and differentiate buried objects. When a buried object is struck with a high-pressure waterjets, the impact will cause characteristic vibrations in the object depending on the objects shape and composition. These vibrations will be transferred to the ground and then to the water stream that is hitting the object. Some of these vibrations will also be transferred to the air via the narrow channel the waterjet cuts in the ground. Currently the ground vibrations are detected with Doppler radar and video camera sensing, while the air vibrations are detected with a directional microphone. Data is collected via a Labview based data acquisition system. This data is then manipulated in Labview to produce the associated power spectrums. These power spectra are fed through various signal processing and recognition routines to determine the probability of there being an object present under the current test location and what that object is likely to be. Our current test area consists of a large X-Y positioning system placed over approximately a five-foot circular test area. The positioning system moves both the waterjet and the sensor package to the test location specified by the Labview control software. Currently we are able to locate buried land mine models at a distance of approximately three inches with a high degree of accuracy.

Mixed-language Development Tool for HAVNET Research

While the Hausdorff-Voronoi network has demonstrated promising capabilities for three-dimensional object recognition and classification, serious investigation and refinement of this network required something more than the ad hoc programs currently available. To fill this need an object-oriented mixed-language MATLABd and C++ software toolkit has been developed, together with a user-friendly graphical user interface.

Sensor fusion for hand-held mine detection in investigation mode

In case of hand-held mine detection, the operator functions in two distinct modes. Namely the scan mode and investigation mode. In scan mode, the operator scans the area to look for potential targets. On identifying a suspect target location, the operator switches to investigation mode where he/she closely scan the area and tries to identify/discriminate target based on consistency, size and strength of the response. The aim of this paper is to look at the various aspects of sensor fusion in scan and investigation mode to fuse information from a collocated metal detector and ground penetrating radar sensors on a hand-held mine detection unit. Different sensor fusion schemes are compared. It is found that the two sensors are complimentary for a set of mine targets while they are supplementary for other set of mine targets. A better detection performance can be achieved by suitable modification to the sensor fusion scheme based on identified electromagnetic characteristics of detected targets.