Issue 7.3, December 2003
Information within this issue may be outdated. Click here to view the most recent issue

Survey of Suspected Mined Areas From a Helicopter

While awaiting the results of airborne remote sensing projects, available in a few years, a simpler solution is recommended. Remote sensing from a manned helicopter for the general survey of minefields and risk-suspected areas has already been developed and was operationally validated in 2002 and 2003 in Croatia.

by Milan Bajić, CROMAC


The ground-based general survey of minefields and of suspected areas is often unaffordable due to limited access. This constraint can be avoided by the use of airborne remote sensing. Airborne remote sensing methods and technology were, in 1999, only promising potential for the needs of humanitarian demining.1 The first project, involving airborne remote sensing aimed at the needs of humanitarian demining, was an excellent means of understanding the complexity of the mine scene.2 There are important lessons to be learned from the project. An interesting approach was the use of an airship as an aerial platform and ultra wide band radar and electro-optical sensors.3 In 2003, airborne remote sensing methods and technology reached maturity, promising their use in wide areas, general surveys and possibly for technical surveys of mined and mine-suspected areas.4, 5 Even though no operational system was available until now, we will know their potential contribution to humanitarian demining in 2004 and 2005.

Figure 1: A small helicopter was approved as suitable for the remote sensing of minefields and risk-suspected areas.

In addition to these projects, which are strictly related to humanitarian demining, there are other projects involving airborne landmine detection that are less public (e.g., Mirage).6 All previously mentioned projects are highly technological and very ambitious. However, the period from initial development to possible deployment for operational use could be longer than four or five years. While awaiting the results of these projects, the discrepancy between needs and available operational airborne technology for general surveys can be compensated for by the use of sustainable and already developed simpler solutions. In Figure 1, we present an application of remote sensing from a manned helicopter for the general survey of minefields and suspected areas, which was developed and operationally validated in 2002 and 2003 in Croatia.7, 8, 9

The example in Figure 1 shows the potential to immediately (rather than in four to five years) establish sustainable aerial general survey of minefields and suspected areas. More importantly, this solution is feasible in many countries contaminated by landmines and UXO, at a cost that is at least five to 10 times less than the cost of one of the projects that were mentioned previously. In the following section, we present the basic assumptions for a general survey from a manned helicopter and we outline the system and its operational validation.

Basic Assumptions

The basic assumptions for the successful development and deployment of the described aerial general survey included the following:

Multisensor Acquisition System On Board Helicopter Bell-206

Figure 2: Three digital sensors, installed on-board helicopter Bell-206, were used for vertical imaging. VNIR and TIR sensors were used in imaging mode, whereas HSLS was used only in spatial sampling mode.

This system uses digital, electro-optical sensors with computer-controlled acquisition and GPS-based navigation. The sensors are (See Figure 2):

(Above) Figure 3: The pilot and co-pilot see the real time position of the helicopter on the large screen, while the background can be a map, ortho photomap or geocoded satellite image. On the map is an area of interest—planned and realized flight routes.
(Above) Figure 4: A view of the flight route on the topographic map of the scale 1:25,000. This serves as an example of the aerial survey of the electricity high tension network (state of the towers, vegetation) and access field roads from the asphalt road on the left side to the network corridor. (For further explanation see Figure 5.) The circle in the upper left part marks an area that is presented by the mosaic in Figure 6.
(Above) Figure 5: The output of the aerial survey is flight routes in vector form. The example shows access roads (1-1, 2-2, 3-3, 4-4 and 5-5) and a high-tension network (18-24) for the area in Figure 4.
(Above) Figure 6: The output of the survey is non-geocoded mosaics produced by registering images to images. The example shows mosaics of five thermal infrared images collected by THV-1000 from 500 m above terrain. Trenches (T) are easily detected—they appear as waved bright or dark lines in the thermal infrared images. They are reliable indicators of mined areas.

Each sensor has an independent acquisition system (frame grabber, acquisition software, personal computer) and can be used separately. For VNIR and HSLS, sensors were developed with acquisition software that enables different modes of imaging (VNIR) or of spatial sampling (HSLS). For TIR, imaging was applied to the original software of the camera. VNIR imaging has several options: four channels (each eight bits), three channels (each 10 bits) or single channel (eight or 10 bits). During the acquisition, images are stored in specific formats that provide the highest throughput, while they are exported in tagged information file format (TIFF) for further use. For each image, basic data is provided (time of recording, gain of each channel and exposition time) that enables synchronization and geo-referencing. The trade-off among spatial coverage, spatial overlapping, and resolution and spectral resolution is accommodated by the selection of optical objectives having different focal lengths, the selection of a number of channels, and the selection of radiometric range and height above the terrain.

The lowest height of flight above terrain for Bell-206 is 130 m as specified by the helicopter safety rules (“dead man curve”). The minimum velocity of the flight at the lowest height is desired to reduce blurring of images. For Bell-206, and based on the digital VNIR and TIR sensor, a velocity of 20 m/s was approved as the lowest and most suitable velocity. In comparison to aerial video imaging by television cameras, digital sensors are much more resistant to blurring caused by radial speed and vibrations. This enables the use of unstabilized gimbals and provides surprisingly good images.

A digital moving map, for measurement and recording of the real-time position of the platform by means of the GPS, is applied for the mission planning, flight control and reporting (Figure 3, Figure 4). The flight planning begins with a statement of needs defined by the MAC—specific information that was required by the MAC about the given area of interest (AOI). The next step is planning the routes in accordance with characteristics of the terrain, sun position, expected wind directions and selected modes of imaging.

After acquisition, images and data are exported to the interpretation computers. The flight route, data and the logs of images are combined. This enables geo-referencing of images. The next step in processing is derivation of the flight routes (Figure 5) and of mosaics of images. Mosaics are used for the assessment of the completeness of spatial coverage—registering image to image (Figure 6) can produce them or, if needed, geocoding can follow this process (Figure 7). Furthermore, interpretation is performed on original images if spectral information is more important than spatial information. Interpretation of mosaics occurs in the opposite case. Basic kinds of output of the survey are:

Figure 7: Output of the survey is the geocoded mosaic. A circle on Fig. 4 marks an example of the mosaic of an area. The mosaic is overlaid over the map of the scale 1:5,000. The aim of the aerial survey was to provide information about the status of the access road that was out of use for more than 10 years. Figure 8: Output of the survey in raw images. The example of the raw image in visible wavelengths was acquired at a height 130 m above terrain. The image shows a demined access field road—on both sides are mined areas. Calibration aluminum cross has arms one m long; width of arms is 20 cm.

Typical Tasks

The aerial general survey of minefields and suspected areas is indeed an intelligence-gathering, processing and dissemination system—not a cartographic system. Therefore, it is not always necessary to geocode mosaics or single images (raster); if vector data are needed as an output, it is enough to geocode output vector. Which solution to apply depends on the requirements of MACs. The following are examples of typical tasks of an aerial survey:

Operational Validation

The described general aerial survey of minefields and risk-suspected areas was operationally validated in several missions over flat terrain in 2002 (electricity high-tension networks Drenov bok, Dubica, 14 km; set of networks near Ernestinovo, 167 km) and in difficult mountainous terrain in 2003 (Tulove grede, Velebit). Validation is underway. A cost-benefit analysis shows that the aerial survey is efficient for cases of corridor-like objects and for wide suspected areas and minefields that have limited access. It can provide missing information or increase completeness, accuracy and reliability of information on minefield indicators and reference points of mine records over large areas in a short time. The aerial remote system was used for researching minefields and suspected areas over large areas and different types of terrain and in varying climates. This system should be used much more.


The airborne multisensor system, on board helicopter Bell-206, was directed primarily toward minefield research for the general survey (as a complementary process to the ground-based general survey). But it is not limited to these purposes. Other applications are possible. This is the first fully digital airborne multisensor system in Croatia. Its development was motivated and enabled by the scientific project ARC, funded by the European Commission, and it was realized through the potential, experience and human resources already existing in Croatia in the field of airborne remote sensing.

*All graphics courtesy of author.


  1. Engelhardt F.R. The 1999 Workshop on Remote Sensing of Anti-Personnel Landmines, Final Report, Ottawa, Canada: 19–20 June 1999.
  2. Van Genderen J.L., Maathuis B.H.P. The 1999 Airborne Minefield Detection: Pilot Project, Final Report, (Reg./661–97/2, Volume 1, Volume 2) ITC, Enschede The Nederlands: 31 December 1999.
  3. Cramer E. A., (2003), “The Mineseeker Airship,” Journal of Mine Action, Issue 5.1, 2000,
  4. “EC IST SMART.” Space and airborne Mined Area Reduction Tools—SMART, EC IST, 2000-25044, 6 March, 2003
  5. “EC IST ARC.” Airborne Minefield Area Reduction—ARC, EC IST-2000-25300,
  6. Burke, S. “The U.S. Department of Defense Humanitarian Demining Research and Development Program,” Journal of Mine Action, Issue 7.1, (2003): 6–9.
  7. Bajić, M., Tadic, T. “Airborne Remote Sensing for the General Survey of Damaged and Mined High Voltage Network,” GIS forum, International Geographical Information Systems Conference and Exhibition, Croatia, Split, Trogir, Korcula, Mljet and Dubrovnik: 2–6 September 2002.
  8. Bajić, M., Gold H., “Contribution of the Airborne Remote Sensing to Demining of the Mountains: Case Study Tulove Grede—Velebit,” GIS forum, International Geographical Information Systems Conference and Exhibition, Croatia, Split, Trogir, Korcula, Mljet and Dubrovnik: 2–6 September 2002.
  9. Bajić, M., Gold, H., Franjkovic D. “Airborne Remote Sensing System of the Faculty for Traffic Sciences, University of Zagreb,” Bulletin of Scientific Council for Remote Sensing and Photo Interpretation of the Croatian Academy of Sciences and Arts, 15–16, (2001) 57–70.

Contact Information

Milan Bajić, Assoc. Prof., D. Sc.
Geodesy University of Zagreb
Scientific Council of CROMAC
Ulica grada Vukovara 226 c
10000 Zagreb
Tel: ++385 98 460 917