September Article Reviews

Jon Featheringill

Detecting and Locating Landmine Fields from Vehicle- and Air-Borne Measured IR Images

Objective:

To design and evaluate a robust, scalable method for detecting landmine candidates and locating minefields using infrared images collected by vehicle- and air-borne sensors.

Results:

The multiscale detector successfully identified both surface-laid and buried mines, as well as man-made landmarks. Detection probabilities ranged from 70–82%, with hitting rates of 50–70%. It outperformed conventional corner detection algorithms (SUSAN, CSS) in noisy IR conditions. The system proved effective for wide-area minefield surveillance.

Approach:

Mathematical approach using a multiscale isotropic bandpass filter detector. The method enhances point-like thermal signatures of mines in noisy, low-resolution IR images. It incorporates automatic scale selection based on local noise and inter-scale position tracing to improve localization. Tested with Swedish Defense Research Agency datasets from vehicle-mounted and airborne IR cameras.

Issues:

This method only identifies candidate objects and cannot reliably discriminate mines from clutter such as stones. Performance depends on image resolution and environmental conditions. Tests were limited to controlled datasets; real-world performance may be less successful. Future improvements require higher-resolution sensors and additional classification methods (This was done in 2001).

Gu, I. Y.-H., & Tjahjadi, T. "Detecting and Locating Landmine Fields from Vehicle- and Air-Borne Measured IR Images," Pattern Recognition, Vol. 35, No. 12, pp. 3001–3014, 2002.

Land Mine Detecting Technology by Using IR Cameras

Objective:

To develop and demonstrate a safe, remote method of detecting buried landmines for humanitarian demining by using infrared cameras.

Results:

Mines buried at 1 cm depth were clearly detected for up to 7 minutes after cooling; at 2 cm depth they were faint but still recognizable. Image enhancement improved visibility. This method proved safer than contact probes or radar, offering effective remote detection in favorable weather conditions.

William Smith

Bentley Hillis

Objective: To conclusively prove that LWIR cameras can be used to detect with legacy minefields in desert environments. With the goal of detecting simulated and real mines. Detecting three locally common types of mines; PMA-3 anti-personnel, PPM-2 anti personnel, and PRB-M3A1 anti-tank landmines. Simulated and real mine fields would be compared to validate data and real-world applications.

Results: The paper concludes that IR cameras attached to drones can be used in the desert to detect landmines. It concludes that there are limitations primarily related to depth weather conditions such as cloud cover and wind. The paper also notes that it is not the only viable method. Particularly regular cameras are better suited when mapping surface mines in an arid environment

Approach:

Experimental study using IR cameras in the 8–12 µm range to detect thermal contrasts between soil and buried mines. The team conducted outdoor tests with mock mines buried at shallow depths (1–2 cm), applied cooling water to enhance contrast, and evaluated detection performance with image processing techniques.

Issues:

Detection depth was limited to very shallow depth mines (1–2 cm). Effectiveness depended on environmental conditions and cooling. The system only provides candidate locations and requires further development for real-world use. No chemical or multi-sensor integration was tested.

Shimoi, N., Takita, Y., Nonami, K., & Wasaki, K. "Land Mine Detecting Technology by Using IR Cameras," Proceedings of the International Conference on Control, Automation and Systems (ICCAS), Cheju National Univ., Jeju, Korea, October 2001.

Land Mines (Landminen)

Von Trescow, A., ‘Land Mines’, Landminen, 1975, p. 4,7-43, https://apps.dtic.mil/sti/tr/pdf/ADA053305.pdf

         This article provides great insight into the specifications of the many different mines developed pre and post WWII.  It discusses geometry, mass, and materials of nearly eighty kinds of mines and provides illustrations.  It also discusses the triumphs and pitfalls of each design and how effective they were in use.  This is useful for our experiment because it gives us clear detail on how to simulate these mines to test how detectable they are with IR cameras from air.  We will not be fully simulating these mines, though the dimensions and materials of the casings provided will allow us to create multiple approximations for testing.  One drawback of this article though is the fact that it’s out of date.  We will need to do more research to ensure that the evolution of landmines has not resulted in any drastic changes in the fifty years since the publication of this document. 

The article begins by introducing a short history of the term “Mine Warfare”.  This referred to the practice of filling tunnels in an advanced position with explosives and detonating them to destroy or disrupt enemy troops.  It also referred to sea mines though this article’s content is limited to anti-tank and anti-personnel mines.

        The article then begins examining anti-tank mines after WWI.  The development of land mines began in earnest as Germany’s attempt to rebuff tank advances near Cambrai.  These attempts were largely unsuccessful due to Germany being unable to solve a problem with the anti-tank mine fuse used at the time.  This often resulted in inadvertent detonation of the mines.

        Then the safety mechanism solutions that other countries found for this problem are given.  Three safety designs, the shear pin safety mechanism, the lever lock safety mechanism, and the ball lock safety mechanism, are described in detail.

         Land mines of this era had another fatal flaw, durability.  These mines weren’t designed to survive for long periods of time exposed to the elements.  The casings failed to prevent moisture and dirt from undermining the effectiveness of the explosives contained inside the land mine.

        The Germans developed a new fuse which solved their inadvertent detonation issue but were faced with another problem at the beginning of WWII.  An increasing lack in raw materials led them to develop a new type of fuse to replace the previously developed fuse.  This combatted the lack of brass, a main component in their previous fuse.

        When the war with the Soviet Union began, most forces were still using the traditional metal casings used in most landmines at that time.  However, several methods to detect these casings had already been developed and were regularly deployed.  That coupled with decreasing resources in some countries, led to many nations seeking an alternative casing solution.  The Soviets were very skilled in this regard creating casings made of wood, oil-soaked paper, and mixtures of turf, to name a few. 

         The article goes on to detail several non-metallic mines, their specifications, and country of origin.

        Then it begins discussing the various anti-personnel mines employed at the time, their specifications, and country of origins.  A common theme among these mines was the mine “jumping” a few feet into the air after being set off, before detonating.  Just as common though were simple fragmentation mines designed to deal just enough damage to remove a soldier from combat, rather than killing him.

         After WWII, the world continued to pursue non-metallic mines.  This trend led to plastics becoming a main component in the construction of mines.  Anti-tank mines had trouble being completely constructed without metal, as the amount of explosives required to deal significant damage to a tank necessitated the use of more durable material.

         The article then goes on to describe several anti-tank landmines that had been developed post WWII and the specifications of said mines.

        Shaped charges were developed around this time allowing the pressure of the explosion to be directed more efficiently.  One way this was utilized was to place a metal plate in the path of the directed pressure, creating a bullet capable of punching a large hole in the hull of a tank.

        The article proceeds to describe a multitude of anti-personnel mines developed after WWII and then comments on the future of mines in warfare.  It says that the use of mines is not going anywhere and will likely increase in the coming years.  This has unfortunately proved to be true.  Following that is a list of seventy-eight of the mines described through the article with illustrations of those mines.

Proof: How Small Drones Can Find Buried Landmines in the Desert Using Airborne IR Thermography

Approach: Both simulated and real-world testing were done in the Sahara Desert in Chad. For controlled field testing a grid of simulated mines was setup in view of a LWIR camera mounted 7m high and thermocouple probes measured data in and around the simulated mines. A FLIR Duo Pro R was used at the control sight whereas the flight tests used A DJI FLIR Zenmuse XT8 thermal/LWIR camera. Both had radiometric data capture. The tests were conducted at multiple time of day with varying weather conditions.

Issues: How this paper quantifies a successful detection is not very technical. To their credit this is admitted in the conclusions of the paper saying, "a more technical discussion of the underlying science and more granular anomaly analysis, comparing theory and reality in more detail." The measurement of outside variables that affect the results is also lacking. Again, the paper states this in the conclusion saying, "a more detailed review of individual variables: natural, weather, object materials, diurnal cycle and geophysical elements from an extensive amount of data in-hand."

Fardoulis, John; Depreytere, Xavier; Gallien, Pierre; Djouhri, Kheria; Abdourhmane, Ba; and Sauvage, Emmanuel (2020) "Proof: How Small Drones Can Find Buried Landmines in the Desert Using Airborne IR Thermography,"The Journal of Conventional Weapons Destruction: Vol. 24 : Iss. 2 , Article 15.

Smart Sensing for Mine Detection Studies with IR Cameras

Objective: This study has the goal of developing a mine detection system using smart sensing technologies. Using a combination of post processing and human observation of an IR image to detect mines.

Results: After 7 minutes, the image becomes indistinct and after 10 minutes, it is no longer possible to distinguish the mine target images. This is because the system being is only effective with a large difference in temperature between the ground and the target mines. Which is why water was used to cool the ground before the test.

Approach: There was a single day of experimentation starting at 9am. Using rain equipment, cold water was sprinkled evenly on the ground surface of the area to be measured. This experiment also used thermocouples to measure the change in temperature of the ground surface in the proximity of the buried mock mines. The detection started immediately after cold water was springled on the ground and the test lasted 15 minutes.

N. Shimoi, Y. Takita, K. Nonami and K. Wasaki, "Smart sensing for mine detection studies with IR cameras," Proceedings 2001 IEEE International Symposium on Computational Intelligence in Robotics and Automation (Cat. No.01EX515), Banff, AB, Canada, 2001, pp. 356-361, doi: 10.1109/CIRA.2001.1013225.

Jacob Black

Issues: While this test required shallow buried mines and water to cool the ground, it was conducted around 2001. So, it is possible that the test was inhibited by the technology available at that time. The test was seemingly only conducted once which limits the variety in its data and in the conditions tested in.

Review on Infrared Imaging Technology

Hou, F., Zhang, Y., Zhou, Y., Zhang, M., Lv, B., and Wu, J., “Review on Infrared Imaging Technology,” Sustainability, vol. 14, Sep. 2022, p. 11161. https://www.mdpi.com/2071-1050/14/18/11161#metrics

This article gives background on infrared imaging technologies and how they are developed. The article also explains what infrared radiation is and how it occurs. This information is beneficial to our senior seminar because it will help understand what infrared imaging actually is and help to understand what our results mean. This knowledge goes hand in hand with measuring the accuracy of infrared cameras ability to pickup infrared signatures due to certain variables. Understanding how infrared cameras work will give insight into what our results should look like and how to deal with challenges involving infrared cameras.

        To begin with, the article introduces infrared radiation which is the spectrum of light just after visible light. Infrared radiation is also an electromagnetic wave. Thermal imaging is the use of optics and infrared detectors to find differences in temperature in different backgrounds. Thermal imaging provides some basic advantages to other techniques such as lasers. Some of the advantages include the ability to get information even through dense fog and clouds as well as in the dark. Thermal imaging requires no contact with its object of observation. There are two main types of thermal imaging devices: cooled and uncooled. Cooled thermal imaging devices allow for high accuracy temperature recording while also being more expensive and requiring more maintenance while uncooled infrared imaging systems are not quite as accurate but are more affordable and require less maintenance. The maintenance required for cooled thermal imaging systems requires periodic rebuilds of the cooling components.

        Infrared thermal imagers are run by 4 main components: optics, detector, processors, and display. The optic focuses the infrared radiation and sends it to the detector which turns the radiation into an electrical signal. The signal is processed and then sent to the display. The quality of the optic and the display of the thermal imager have a major impact on the quality of the results received from the imager. For instance, qualities such as picture resolution, frame rate, identification distance, temperature range, and temperature accuracy.

      Infrared thermal processors have become increasingly smaller which allows them to be used more efficiently. The thermal processors can have some issues that to improve accuracy need to be accounted for. Some of these issues include inconsistent pixel responses, noise, and picture clarity. The inconsistent pixel responses can be fixed by enhanced calibration or higher resolutions. The noise issue can be fixed by certain algorithms added to the processor but comes at the cost of computational complexity. Methods to decrease noise in the system include wavelets, spatial domain filters, and block matching. Image enhancement is an important feature to have on a thermal imaging processor. Adding image enhancement allows for non-important information to be filtered out and for more important information to be made more readable. Image enhancement can be done with a few different methods including histogram equalization, piecewise linear transforms, retinex, and wavelet enhancement.

        Multi/Hyperspectral thermal infrared remote sensing has different classifications which include multispectral and two hyperspectral classifications. Spectral sensors can be differentiated by their number of bands; multispectral sensors have a few bands while hyperspectral have anywhere from hundreds of bands to thousands of bands. These bands affect the quality and accuracy of the temperatures received from the thermal device. The higher accuracy of the increased number of bands requires more computing power to process the information taken in. Due to an increase in information, there is also an increase in noise which must me filtered out to have readable results.

        There are many applications for thermal imagers across many industries. Some of these industries include transportation, medicine, military, electric power, and public security. Thermal imagers used in transportation can inspect temperature of wheels and tires on cars and trains. In the medical industry thermal imagers can allow increased accuracy in diagnosing illnesses by observing temperature changes in the body. Militaries can find many uses for thermal imaging such as finding people hiding behind camouflage or smoke. Other military uses include detecting underground or underwater objects such as ships, submarines, and landmines. In the electric power thermal imagers can be used to inspect equipment that might be damaged. For public safety applications thermal imagers can be used for fire detection and structure inspection.