Surface Land Mines Research Articles

Modeling the Effect of Vegetation Coverage on Unmanned Aerial Vehicles-Based Object Detection: A Study in the Minefield Environment

An important consideration for UAV-based (unmanned aerial vehicle) object detection in the natural environment is vegetation height and foliar cover, which can visually obscure the items a machine learning model is trained to detect. Hence, the accuracy of aerial detection of objects such as surface landmines and UXO (unexploded ordnance) is highly dependent on the height and density of vegetation in a given area. In this study, we develop a model that estimates the detection accuracy (recall) of a YOLOv8 object’s detection implementation as a function of occlusion due to vegetation coverage. To solve this function, we developed an algorithm to extract vegetation height and coverage of the UAV imagery from a digital surface model generated using structure-from-motion (SfM) photogrammetry. We find the relationship between recall and percent occlusion is well modeled by a sigmoid function using the PFM-1 landmine test case. Applying the sigmoid recall-occlusion relationship in conjunction with our vegetation cover algorithm to solve for percent occlusion, we mapped the uncertainty in detection rate due to vegetation in UAV-based SfM orthomosaics in eight different minefield environments. This methodology and model have significant implications for determining the optimal location and time of year for UAV-based object detection tasks and quantifying the uncertainty of deep learning object detection models in the natural environment.

Deep Learning-Based Real-Time Detection of Surface Landmines Using Optical Imaging

This paper presents a pioneering study in the application of real-time surface landmine detection using a combination of robotics and deep learning. We introduce a novel system integrated within a demining robot, capable of detecting landmines in real time with high recall. Utilizing YOLOv8 models, we leverage both optical imaging and artificial intelligence to identify two common types of surface landmines: PFM-1 (butterfly) and PMA-2 (starfish with tripwire). Our system runs at 2 FPS on a mobile device missing at most 1.6% of targets. It demonstrates significant advancements in operational speed and autonomy, surpassing conventional methods while being compatible with other approaches like UAV. In addition to the proposed system, we release two datasets with remarkable differences in landmine and background colors, built to train and test the model performances.

Utilizing UAV-based hyperspectral imaging to detect surficial explosive ordnance

Across postconflict regions of the world, explosive ordnance (EO), which includes remnant antipersonnel land mines, antivehicle/tank mines, unexploded cluster munitions, improvised explosive devices, and explosive remnants of war (ERW) such as unexploded ordnance and abandoned explosive ordnance, remains a critical humanitarian concern. Clearance and land release efforts anchored on manual geophysical detection and mechanical probing methods remain painstakingly slow, expensive, and dangerous to operators. As a result, postconflict regions impacted by EO contamination significantly lag in social and economic development. Developing, calibrating, and field testing more efficient detection methods for surficial EO is a crucial task. Unpiloted aerial systems featuring advanced remote sensing capabilities are a key technology that may allow the tide to turn in the EO crisis. Specifically, recent advances in hardware design have allowed for effective deployment of small, light, and less power consuming hyperspectral imaging (HSI) systems from small unpiloted aerial vehicles (UAVs). Our proof-of-concept study employs UAV-based HSI to deliver a safer, faster, and more cost-efficient method of surface land mine and ERW detection compared to current ground-based detection methods. Our results indicate that analysis of HSI data sets can produce spectral profiles and derivative data products to distinguish multiple ERW and mine types in a variety of host environments.

Buried and Surface Mine Detection From Thermal Image Time Series

Millions of buried and surface landmines throughout the world will continue to kill and injure people for many decades. Reliable landmine detection systems, especially those that do not require that an operator enters a minefield, are demanded. In such systems, data are captured by sensors and processed with signal-processing techniques. In recent studies, ground-penetrating radars, ultrasound transducers, metal detectors, and infrared sensors have been used to capture data as individual images or image time series. Bare soil and mines have different thermal characteristics, and this difference can be observed on the soil surface with thermal sensors. Since this is a dynamic behavior driven by radiation from the sun and changes with temperature during the day, it can be observed better in thermal image time series. In the proposed method, a two-step approach is adopted: pixelwise classification followed by mine detection. In the first step, a specific spatial filter is applied on images and the filtered image series is classified pixelwise using supervised classification. In the second step, the classified pixels are combined after a smoothing operation to detect mines. The types of the detected mines can also be determined. The method is tested on real data to quantify its classification and detection performance. The results are quite promising. Among the classifiers tested in this paper, the quadratic discriminate function classifier produces the best results.

Signal subspace change detection in averaged multilook SAR imagery

This paper addresses change detection in averaged multilook synthetic aperture radar (SAR) imagery. Averaged multilook SAR images are preferable to full-aperture SAR reconstructions when the imaging algorithm is approximation-based (e.g., polar format processing) or when motion data are not accurate over a long full aperture. We examine the application of a SAR change-detection method, known as signal subspace processing, which is based on the principles of two-dimensional adaptive filtering, and we use it to recognize the addition of surface landmines to a particular area under surveillance. We describe the change-detection problem as a trinary hypothesis testing problem, and define a change signal and its normalized version to determine whether: 1) there is no change in the imaged scene; 2) a target has entered the imaged scene; or 3) a target has exited the imaged scene. A statistical analysis of the error signal is provided to show its properties and merits. Results are presented for averaged noncoherent multilook and coherent single-look X-band SAR imagery.

The impact of false alarm mitigation on surface landmine detection in MWIR imagery

Surface landmine and minefield detection from airborne imagery is a difficult problem. As part of the minefield detection process, anomaly detection is performed to identify potential landmines in individual airborne images. Post-processing is performed on the initial landmines identified to reduce the number of false alarms, referred to as false alarm mitigation. In this research, a circular harmonics transform image processing approach (the CHT method) and a constant false alarm rate technique (the RX approach) are investigated for surface landmine detection and false alarm mitigation in medium wave infrared (MWIR) image data. The false alarm mitigation approach integrates the CHT and RX methods to identify candidate landmine locations with one technique at a given false alarm rate and applies the other technique to confirm landmine locations and eliminate potential false alarms. Individual detector and false alarm mitigation experimental results are presented for 31 daytime and 43 nighttime MWIR images containing 76 and 142 landmines, respectively. At a 0.9 desired probability of landmine detection, experimental results show that false alarm mitigation reduces the false alarm rate by as much as 84.3% and 13.7% for daytime and nighttime images, respectively, maintaining the probability of detection at 0.85 and 0.90, respectively.

Multisensor data fusion for surface land-mine detection

Receiver operating characteristic (ROC) curves have been used to examine a novel target recognition system using a number of knowledge-based techniques to automatically detect surface land mines that are present in 30 sets of thermal and multispectral images. A summary of the results, graphed at a probability of detection greater than or equal to 96%, shows the false-alarm rates (FARs) obtained using various combinations of fusing sensors and neural classifiers averaged over the 30 images. The results show that using two neural-network classifiers on the input textural and spectral characteristics of selected multispectral bands, we obtained FARs of approximately 3%. Using polarization-resolved images only, we obtained FARs of 1.15%. Fusing the best classifier output with the polarization-resolved images, we obtained FARs as low as 0.023%. This result has shown the large improvement gained in the sensor fusion. Also, an improvement is shown by comparing these results with those reported in an existing commercial system.

Using genetic algorithms and neural networks for surface land mine detection

Knowledge based techniques have been used to automatically detect surface land mines present in thermal and multispectral images. Polarization-sensitive infrared sensing is used to highlight the polarization signature of man-made targets such as land mines over natural features in the image. Processing the thermal polarization images using a background-discrimination algorithm, we were able to successfully identify eight of the nine man-made targets, three of which were mines, with only three false targets. A digital camera is used to collect a number of multispectral bands of the test mine area containing three surface land mines with natural and man-made clutter. Using a supervised and unsupervised neural network technique on the textural and spectral characteristics of selected multispectral bands (using a genetic algorithm tool), we successfully identified the three surface mines but obtained numerous false targets with varying degrees of accuracy. Finally, to further improve our detection of land mines, we use a fuzzy rule-based fusion technique on the processed polarization resolved image together with the output results of the two best classifiers. Fuzzy rule-based fusion identified the locations of all three land mines and reduced the number of false alarms from seven (as obtained by the polarization resolved image) to two. Additional experiments on several other images have also produced favorable results at this early stage in testing the algorithm and comparing it with an existing commercial system.

Ultra-wide-band synthetic-aperture radar for mine-field detection

A full-wave model is developed for electromagnetic scattering from buried and surface land mines (both conducting and plastic), taking rigorous account of the lossy, dispersive, and potentially layered properties of soil. The (polarimetric) theoretical results are confirmed via synthetic-aperture radar (SAR) measurements, performed using the US Army Research Laboratory's BoomSAR, with which fully polarimetric ultra-wide-band (50-1200 MHz) SAR imagery is produced. The SAR system is used to acquire a large database of imagery, including a significant distribution of naturally occurring clutter. Several techniques are used for mine detection with such data, including several detectors that are based on target features gleaned from the modeling, as well as a matched-filter-like detector that directly incorporates the target signatures themselves. In addition, the theoretical model is used to predict wave phenomenology in various environments (beyond the limited range of parameters that can be examined experimentally). Since the efficacy of radar-based subsurface sensing depends strongly on the soil properties, we perform a parametric study of the dependence of such on the target RCS, and on possible landmine resonances.

Design and development of a novel flux compression generator for landmine detection applications

The design and development of a novel magnetic flux compressor to generate high power, ground penetrating radar signals is presented and discussed. The high power radar signals are used to detect surface and subsurface landmines. The magnetic flux compressor is housed with other circuitry in a medium caliber (i.e., 30 mm) high-speed projectile to form a compact, kinetic-to-electrical energy converter. The kinetic energy of the high-speed projectile is the flux compressor's primary power source. Flux compressor seed current is externally supplied from a small onboard battery. Computer models are presented that predict the performance of the flux compressor and the force, acceleration, and velocity parameters experienced by the projectile during launch and impact. Generator design guidelines, based on the predictions of the computer model, are presented and discussed. The experimental facility used to launch the projectile to high velocity and test the magnetic flux compression projectile is also discussed. In this facility, a light-gas (i.e., helium) gun accelerates 50 gram (nominal) projectiles to velocities approaching 1000 meters per second. The experimentally measured flux compressor current is compared to that predicted by computer models. Other applications for the flux compression projectile are discussed and include geological surveying, remote sensing, and mining.

Degree of familiarity ART2 in knowledge-based landmine detection

The self-organizing network ART2 is extended to provide a fuzzy output value, which indicates the degree of familiarity of a new analog input pattern to previously stored patterns in the long-term memory of the network. The outputs of the multilayer perceptron and this modified ART2 provide an analog value to a fuzzy rule-based fusion technique which also uses a processed polarization resolved image as its third input. In real-time situations these two classifier outputs indicate the likelihood of a surface landmine target when presented with a number of multispectral and textural bands. Due to the modifications in ART2 this updated alternative architecture (to that of a previous network in [3]) has improved real-time landmine detection capabilities although the registration of all bands is more critical to the accuracy of results in this case. The real-time fuzzy rule-based system in preliminary tests has detected two of the three landmines and the landmine surrogate with two false alarms. Advanced tests on 30 images using the fuzzy rule-based system further confirmed the distinct advantages of fusion and improved detection rates.