Hyperspectral Imaging

[ravicv]

[quote name='ramana' date='04 July 2010 - 04:03 PM' timestamp='1278287700' post='107305']

On BR Monitor/SRR there is an article by Ved talking about HSI from space.

[/quote]



This is a pretty decent introduction, written in "plain English", to initiate people:



http://forums.bharat-rakshak.com/viewtop...ral#p62952



INTRODUCTION



1. Today, the Indian military have access to surveillance facilities which are a major step ahead of those available a few years ago; however, our capabilities are still not commensurate with the state-of-the-art in cutting edge military operations and intelligence gathering, even considering our late as well as modest entry into the space arena. The introduction of high-resolution commercial imagery and advances in spectral analysis has greatly expanded the military’s use of commercial imagery.



2. Three limiting factors which affect optical surveillance are cloud cover, darkness and camouflage; the use of Hyper Spectral Imaging (HSI) and Synthetic Aperture Radar (SAR) would circumvent these limitations.



3. There are many applications where the combination (‘combo’) of HSI as well as SAR from a satellite platform could enhance the execution of various military functions, ranging from making it possible to accurately ‘guesstimate’ the payload from analyzing the plume of a rocket; to obtain real-time Battle Damage Assessment (BDA) after a strike mission; detecting and pinpointing an explosion involving an aircraft, like a crash situation; spotting enemy radars SAGW units or other formations in dense camouflage, or detecting and processing, in real time, data of reheat-assisted takeoffs from many airfields.



4. Perhaps as a result of riding piggy-back on the driving force of civilian applications of space technology till comparatively recently, the tools available to the Indian military have largely, hitherto, restricted themselves to the visible wavelengths and the more traditional forms of image analysis. While today’s possibilities in the area of visible imagery are considerable (the latest in the Keyhole series of satellites with resolutions in the order of 2.5 cms, or even less, spring to mind), imagery using visible wavelengths is only one of a number of options, and suffers from a number of serious limitations like clouds and darkness. The Defence Intelligence Agency (DIA) has, recently, articulated the requirements and expectations of the Services through a Space Vision document, the first of its kind, which is under consideration by various agencies including ISRO. Beyond this document, however, is a need to evolve specific requirements; the need for a payload exploiting HSI and SAR (either individually or in ‘combo’ vehicle) is one such requirement. I feel that optical imagery (meaning human analysis of optical pictures) is on the way out, and the new technology consists of HSI platforms (recently launched by the US, and hitherto undeveloped by India) augmented by SAR (under development).



5. India has proved herself capable of pioneering results even in a comparatively new area like space technology, indigenously fabricating comparatively advanced components, sometimes to the surprise of other more advanced powers! There is no reason, therefore, to shy away from a concept just because it has not been attempted (or has been rejected) by other more experienced powers; neither is there a reason to accept limits and barriers encountered by others when there appears to be a reasonable chance of overcoming the same through indigenous effort.



Hyperspectral Imaging



6. Hyperspectral imaging, also known as imaging spectroscopy, is on the way to becoming a key element in remote sensing. It is a type of multispectral imaging that records many tens of bands of imagery at very narrow bandwidths. In hyperspectral imaging, each pixel of the image records a wavelength. In addition to two spatial dimensions, the hyperspectral image contains a third dimension: radiant intensity. Measuring the energy that is reflected (or emitted) by targets over a variety of different wavelengths results in a spectral response for that object. By comparing the response patterns of different features we may be able to distinguish between them, where we might not be able to, if we only compared them at one wavelength. For example, water and vegetation may reflect somewhat similarly in the visible wavelengths but are almost always separable in the infrared.

7. All significant military structures, machines or devices have well known quantities of specific elements in their composition; thus, combat aircraft, for example, have a significant element of aluminum and titanium alloys, certain plastics and composite materials in their construction. Hyperspectral imaging involves acquiring and analyzing images simultaneously at multiple electromagnetic wavelengths. The wavelength information can be interpreted as spectral "signatures" of materials on the ground, and thus analysts can identify them remotely . Consider: if the chemical nature and hence identity of each component of the spectral image were to be instantly analysed, it would be possible to classify the source of such radiations. Based on the deductions of such an analysis, automated sequences could be triggered. Static targets like camouflaged or buried structures can also be detected by hyperspectral sensors on spaceborne platforms, which promise to significantly improve surveillance, target identification, terrain mapping and visualization capabilities for both land and sea operations.



8. Another significant advantage hyperspectral analysis has over normal spectroscopic analysis is the degree of intelligence which can be gleaned from even ‘bad’ spatial resolutions. Hyperspectral systems can potentially identify an object even if the image has only a few (or even one) picture elements (or pixels) covering the object. For normal imaging systems, this would be seen only as a blob in the image with insufficient spatial resolution to identify the object. Spectral analysis of each pixel provides additional information about the materials in the pixel. For hyperspectral systems, if the blob has a unique spectral signature, the analyst can identify it by this signature.



9. Like multispectral or optical imagery, HSI also requires the target to be illuminated by a light source; however, the angle of the light is not as critical in the case of HSI as in other cases, as analysis of the imagery would not be based on the visual appearance as seen by human analysts. In effect, this implies that the temporal window available for HSI operations would be very much more enlarged, with obviously greater flexibility of usage. Certain satellites such as the Defence Meteorological Satellite Programme are able to tune the amplifiers of their scanning devices so as to operate at night under very low illumination conditions. However, the technology in this area being relatively new, the integration of this facility with military applications has yet to materialize.

The Warfighter-1 Programme



10. The first such programme ever, the Warfighter-1 program is an advanced technology demonstration program that will provide hyperspectral imagery and related technology and services as a part of the ORBView-4 (by Orbital Imaging Corp (Orblmage)) high-resolution imaging satellite. Initially, the plan was to use ORBView-3, and OrbImage's OrbView-4 (OV-4) is being modified to incorporate the Warfighter-1 (WF-1) hyperspectral sensor. The program will also include the evaluation and validation of hyperspectral technologies, development of a mobile ground station, related image processing algorithms and software for assessment of tactical utility for military applications.

11. To a hyperspectral camera, neither the leaves on a tree or a decoy look anything like camouflage netting or paint, or a real vehicle or aircraft, respectively, because they are made from different materials. In fact, hyperspectral imaging can detect the subtle differences between two types of similarly colored paint - one applied to a friendly military vehicle and the other used by the enemy. That very well could spell the end of camouflage as we know it, as camouflage paint will no longer keep an enemy hidden. In the short wave infrared portion of the spectrum, many paints are transparent, so these bands can be analysed to look at materials beneath the paint. A typical hyperspectral image, after processing, looks like an ordinary picture except that special colors can be added to identify objects of interest. For example, all the non-armored material in a suspected tank harbor can be made to look red, clearly exposing decoys and camouflage. With a resolution of 8 meters, Warfighter 1 will be able to spot large targets of military interest.



12. Some specific attributes of hyperspectral imagery described below could be exploited for military operations.



Detection of Missile Launches



13. Relevance. Analyses of missile plumes would permit an accurate assessment of their trajectory, nature of rocket engine and hence even the nature of the payload. Such a capability would provide a realistic early warning of hostile missile launches, thereby enhancing the range and quality of response, which would largely depend on State policy. In the case of ballistic missiles, even assuming that there would be no deviation from the present ‘ No First Use’ policy, such early warning would enable better and more credible response. The application of such a facility would not be confined to the nuclear context alone; the Ghauri and Shaheen, it must be remembered, also follow a ballistic path, like the Prithvi series, and so early warning of these weapons would be relevant in a conventional conflict as well. Admittedly, hyperspectral analysis of the plume would not indicate whether the warhead were nuclear or conventional, a limitation we would have to live with for the foreseeable future!



Monitoring Aircraft Movements



14. The detection of high-flying aircraft is also an area which is being studied; the TEAL EMERALD project consists of detection of aircraft from space, including the determination and discrimination of the spectral signatures of ICBMs/SLBMs from strategic aircraft as observed from a space borne sensor. The current state-of-the-art looks at the detection of aircraft flying in the higher reaches of the atmosphere; however, the application is still in its infancy, and has obvious potential.



15. We can take our imagination further; after-burning combat aircraft routinely feature a long flame from the exhaust at the time of take-off. This flame is of the average temperature of 800-840º C, and, when compared to the ambient atmosphere, would stand out in stark contrast. Detection of a number of such flames within a given time period, for example, could well give early warning of an impending air strike. Accurate activity patterns from potentially hostile airfields could also be monitored in real time by satellite.



Detection of Air Crashes



16. Aircraft use a number of materials in their construction, materials which are typical to aviation structures. The detection of these elements by spectral analysis would provide instant warning in the event of crashes, especially when dealing with remote and inhospitable terrain as found in the Asian subcontinent. This, even in peacetime, would be a valuable feature and one which problem area which is currently being addressed.



Battle Damage Assessment (BDA)



17. Effective BDA could be carried out by employing the attributes of hyperspectral analysis. It would be possible to analyse images of the target which reveal recent moving of earth or evidence of damaged structures like broken masonry, runway craters etc, as well as similar evidence to interest the Army and Navy like freshly prepared minefields and underwater shore defenses.



Identification of Mobile Radars and SAGW Sites



18. When deployed, mobile radars and SAGW sites are dispersed and camouflaged, making them difficult to spot and target. Using HSI, such protection will be denied to these targets, the concept of optical or thermal camouflage being now rendered obsolete! During darkness or bad weather, SAR would immediately bring into sharp relief the existence of all metallic objects.



Spectroscopic Analysis of the Battlefield: the Concept of ‘Mission Space’



19. The knowledge of 'mission space', which would include specific details of geographic terrain, is a central premise for decision making in a tactical scenario. There are many applications of HSI which would enhance military operations; some of them have been listed at Appendix B, and all fall within the generic categories mentioned earlier in this document. A common view of the mission space is that it is based on a geospatial framework that includes imagery data, elevation data, and feature data. Geospatial information must therefore be an integral part of the overall defense information infrastructure. Spectral analysis would be an indispensable tool in such operations. Increasingly in the future, operational needs will be partially fulfilled by commercial satellites for both land and sea operations. The scope of surveillance would be immense; for example, even soil disturbance can be an indicator of changes such as those produced by foot, vehicles and buried mines, such changes being detected by physical characteristics of the soil like mineralogical composition, particle size, particle coating, lignin and cellulose content etc. Hyperspectral products (the term ‘product’ is one commonly used by PIs to denote a ‘scene’ of imagery which has been analysed and annotated by trained PIs) would consist of a change detection map that would locate areas which have been disturbed by military activities. This information would be merged with other data sources within a GIS system for intelligence / strategic planning. For land-sea operations, knowledge of near shore bathymetry would be critical for a variety of applications like landing beach assessment, input variables for various modeling algorithms (acoustic prediction, tide and surf forecasting, weapon systems), planning charts (command and control, mission planning, tactical decision aids) etc. These map layers could be integrated with other sources of information (i.e. topography, hydrology, roads, weather, radar) within a Geographical Information System (GIS), a common system which combines aspects of mapping, graphical displays and database management. The system will provide predictive information on the effects of terrain on military operations and will support decision making and planning of operations. Some general applications of HSI are listed at Appx A.



SYNTHETIC APERTURE RADAR



20. Unlike electro-optical systems, radar satellites can see through clouds, rain, and fog in order to detect targets on the ground or underground, and in or under the ocean. In addition, SAR satellites are extremely useful in tracking moving targets, and can be useful in satisfying military mapping requirements. Chinese engineers have been examining SAR satellites as a means to track enemy submarines in shallow waters.



21. One of the most significant aspects of SAR is the extremely strong echo which results from metallic targets, thus immediately negating the effect of camouflage or concealment measures. For example, the barbed-wire fence along the LOC, when viewed from an airborne SAR camera at 40,000 feet, appeared as a very prominent white line, even eclipsing adjacent objects of much larger physical dimensions or greater colour contrast in the visible spectrum. While the intensity of the return may even obscure the outline of the target it could still form the basis of accurate positioning; for example, the ‘fixing’ of coordinates for location of a mobile radar whose approximate position had been triangulated by means of ELINT.



22. Radar has special military value because, using the right wavelengths, this active system can "see through" clouds and can operate at night. The U.S's Lacrosse series consists of a SAR sensor mounted on a very large (reputed to be school bus sized) platform that ties to extended solar arrays. No Lacrosse/Vega images are available on an Internet Search suggesting the secretive and classified nature of this system. The current Lacrosse can probably achieve 1 meter or better resolution. Missions have included providing imagery for bomb damage assessments of the consequences of Navy Tomahawk missile attacks on Iraqi air defense installations in September 1996, monitoring Iraqi weapons storage sites and troop movements. Vega photographed the Shifa Pharmaceutical Plant in Sudan that was hit in the U.S. retaliatory strikes after the Embassy bombings in 1998. The TESAR (Tactical Enhanced SAR) program produces high resolution images like that of the most famed military building in the world - the Pentagon outside Washington. D.C. As an example of the capabilities of BDA by a satellite platform, Appx S shows the results of BDA by satellite following an air strike on a Serbian barracks during the Bosnian conflict.







BENEFITS OF A ‘COMBO’ PAYLOAD



23. Considering the benefits of both HSI and SAR, a combination (‘combo’) payload combining both facilities would yield the following payoffs:-



(a) Since HSI is not as demanding of light conditions as multispectral or normal optical imaging, a larger window of utilization would be available; assuming eight usable hours of daylight and a 100 min period per orbit, it would be possible to extract many more orbits per day during daylight hours than is currently possible with visual imagery satellites, where manual analysis in the visible wavelength demands a sun synchronous orbit aimed at close to noon over the target area.



(<img src='http://www.india-forum.com/forums/public/style_emoticons/<#EMO_DIR#>/cool.gif' class='bbc_emoticon' alt='B)' /> The satellite could be gainfully used during hours of darkness by employing the SAR sensor.



© As an aid to HSI would be the definition of form and outline, SAR would be able to provide an input in this area, which may be integrated with the HSI data by suitable software developed for the purpose.



(d) During daylight hours, both HSI and SAR could be used to augment each other; this may save the wastage which would occur in case of the target being obscured by heavy clouding.



24. Would a ‘combo’ payload be possible? There could be various aspects of size, weight or power consumption which could render the proposal prohibitively expensive or even unworkable. To adequately address this point the issue would need to be studied by ISRO. In the event that a ‘combo’ arrangement were not possible, separate satellites would need to be used for each payload; even so, the gains from each would be substantial and would be justifiable in their own rights.







Antenna Size



25. For the longer wavelengths beyond visible light, a significant problem could be that of antenna size. Astronomical radio telescopes have no constraints regarding antenna size, which could go into hundreds of meters; however, the same problem applied to a satellite, with the attendant complexities of stowage and subsequent deployment, could prove to be a taxing issue where the longer wavelengths are involved. One alternative could be an arrangement which unfolds in numerous sections to deploy as a large antenna (certain US projects are reported to feature plans for a 100m antenna ); whatever the chosen solution may be, at this stage of expertise the Indian space industry needs to prove itself capable of effectively addressing the problem, of which, in the author’s opinion, there is no doubt, as the requirements will be more in the area of innovation and mechanical dexterity rather than technology. However, this would require a concerted thrust in this area, which, again, would only come about only from a direction from a sufficiently high level.





DEPLOYMENT AND ORBITS



Timeliness



26. While the technology itself may permit such applications, associated measures required to ensure its effectiveness may considerably increase the total cost of the option. The relevance of the time factor in which the data is provided to the end-user is all-important, beyond which the data would cease to be of any value. In a typical air operation, for example, BDA would be required in order to assess the damage to the target, which would determine the need for follow-up attacks, perhaps the same day; in case this information were delayed beyond a certain point, the value of the data would cease to be relevant. Thus the timeliness of the delivery of remote-sensing data from the spacecraft to the user directly affects its utility. This delay could extend from days to weeks or more, and is a consequence of relatively long revisit time, relatively limited capabilities to look off to the side of the satellite’s ground track, and relatively low throughput for image-processing systems on the ground. Overcoming these limitations is expensive, both in equipment cost and personnel hours. Thus, there would be a need for a mechanism to prioritise various requirements, especially relevant in our context where the number and type of platforms are, as yet, restricted. Timeliness, therefore, justifies sufficient reason to invest in a significant effort to provide the means to convey the data from the satellite to the warfighter/operational planner within a timeframe which is of relevance to the end user; this could effect decisions affecting the number of satellites or types of orbits.



Timing of BDA Passes



27. There is no fixed response to the question of how soon after the strike BDA missions would be required, as it would depend on many factors such as whether the target were important enough to warrant a second strike if the first one didn’t have the desired effect, whether it were protected well enough to

make aircraft losses a good possibility, etc.



28. BDA missions in our context would, in the foreseeable future, probably continue to comprise dedicated manned aircraft augmented by UAVs for important targets. As a comparison, however, BDA during Gulf War II was rarely less than a couple of hours following the strike, frequently within the hour. It must be noted, however, that the US philosophy apparently mandates the launch of a constellation of satellites to cover specific operations. The schedule of the passes by US satellites were frequently within the hour, almost always within two to three hours.



29. In the Indian context, while the warfighter would probably like BDA as soon as he could get it, within minutes of the strike if possible, financial constraints would be unlikely to support such a luxury every time; however, the requirement for timely BDA would exist, nonetheless; a possible method to meet this demand is discussed subsequently.



Limited Constellations



30. Ideally, HSI and SAR sensors, if placed on a geostationary platform, would provide the optimum coverage in all areas; however, the state of the technology at the present time would be unlikely to support this requirement due to difficulties of both HSI and SAR applications to perform at orbital ranges exceeding circular or polar orbits. Such being the case, a number of maneuverable satellites would ensure that the specific requirements are met, while emergent demands could be catered for by vectoring the satellite as required. With such an arrangement, short term requirements of BDA, for example, could be met.





FUTURE DEVELOPMENTS



Micro - Satellites



31. The current emphasis in the satellite industry is on replacing large satellite platforms with one or more smaller satellites, built at lower costs, yet able to accomplish similar mission objectives. In this context, there is increasing interest in the potential capabilities and applications of so-called "micro-satellites" satellites of 10-100 kg. However it is recognized that such small satellites pose severe constraints on payload volume, mass and power. Thus, they would appear to be inappropriate for missions such as synthetic aperture radar (SAR) imaging, where payloads have significant size and power demands - specifically the large SAR antenna and high-power radar transmitter. The primary reason for the high transmit power requirement is that traditional SAR systems use backscatter, which is weak from most terrain types as most energy is scattered in the forward direction. Thus, if it were possible to gather this forward scattered element, then the transmit power requirements could drop significantly, potentially making it feasible for installation on a micro-satellite. This research is based on this principle of collecting to the forward scattered element - a novel method by which two micro-satellites 'fly' in a specific formation to accomplish a SAR imaging mission bi-statically. The transmitting satellite will be the master, with the receiver satellite slaved off it for synchronization. The satellites view a swath of 30x30 km, at a ground resolution of 30 m, from an altitude of 700 km. The constellation geometry proposed requires minimal orbit control resources, and allows for the resolution of the left-right ambiguity. The satellite design is based on the Surrey Satellite Technology, Ltd. enhanced micro- satellite, with a mass of 100 kg, and a standard volume of 1x1m base and a 0.6 m height.





State-of-the-Art as an Aiming Point



32. Operation Enduring Freedom, the US operation in Afghanistan post 9/11, may not have had much opposition as far as a battle goes, but was an effective demonstration of the state-of-the-art use of the space medium. To quote from a documented account……





Quote:

‘Before the first soldier, sailor, airman, or Marine was placed in harm’s way — and well before the first unmanned aerial vehicle was deployed — we used satellites to scan hundreds of thousands of square miles of Afghanistan’s rugged terrain. This information gave us a feel for the terrain, for the weapons that potentially could be employed against us, and for an initial set of targets to be attacked with cruise missiles and high-altitude bombers. We used satellites to collect electronic and signals intelligence on the enemy. Satellites fed constant data about cloud cover and moisture into weather forecasting programs. Satellites with spectral imagers were used to detect changes in terrain features indicating potential use by the enemy. Satellites were also available to detect the infrared signature of a missile launch if the terrorists had possessed that capability. Satellites were our first “eyes on target” operating 24 hours a day, during day and night and in all weather. As the decision neared to deploy forces into theater, digital terrain data provided by satellites were used to develop 3-D images of terrain and streets and even to give military planners an idea of the view from a terrorist’s window. This proved to be a boon for pilots flying low-altitude missions through rugged mountains and for special operations forces carrying out covert raids... [/i]





33. The space requirements of the Indian military has, hitherto, been molded more by the dictates of civilian requirements, rather than military considerations. This has resulted in certain wavelengths being exploited only as far as required for civilian applications; the classic case is that of IR, where development has remained restricted to mainly agricultural, marine and geological functions of remote sensing. The development of the Synthetic Aperture Radar (SAR), as yet on an aircraft platform, is a welcome step; even so, the first space-borne application of SAR is not expected to be launched for at least the next two years.



34. Today, bodies like the Space Commission, organisationally under the Prime Minister do not contain any representation of the Services or the MOD. As major decisions regarding the utilization of space are invariably shaped at these forums, it is hardly surprising that military considerations were rarely, if ever, taken into account.





Hyper Spectral Imagery: Upsetting Old Paradigms



35. HSI will trash many old concepts and perceptions; like the X-Ray eyes of the legendary Superman of fiction, HSI will enable its users to look below layers of paint to see the nature of the structure below; to differentiate a wooden mock-up, however realistic, from the ‘real thing’; to spot locations of freshly laid minefields or beaches with booby-traps just by scanning the terrain, and that too, not after hours of painstaking effort by human PIs, but by almost immediate, colour coded computerized images. The possibilities presented by HSI go far beyond those presented by optical imagery.



36. Conventional optical imagery, for which the mechanism is matured, will continue to be of significance, but only till HSI becomes available. Another crucial aspect, SAR, is already under development. However, the time is ripe for the Services to look beyond conventional optical imagery and consider the wealth of information available through HSI and SAR. Development in the space area is as slow as in any other field, and gestation times are large. The Services have clearly articulated their concept of development in the years to come; there is a need to focus on the development of HSI, as the technology for tomorrow.





RECCOMMENDATION



37. As a logical follow-up of the Space Vision, the Services need to strongly project a need for the development of a crucial cutting edge technology, ie, Hyper Spectral Imaging (HSI), operating from either a ‘combo’ space vehicle incorporating an SAR sensor, or from a dedicated maneuverable satellite.





CONCLUSION



38. The utilization of space by the Indian military is still in its infancy; the significance of space being first recognised a little more than two decades ago, the requirements of the Services were first addressed by minor modifications to existing civilian applications. It is not surprising, therefore, that even today the Services remain largely confined to optical imagery, with the use of SAR yet to be operationalised. There is a need to move beyond optical imagery and focus on one of the cutting edge technologies today; spectral analysis through HSI promises a wide variety of revolutionary functions and benefits which would significantly enhance our IMINT capability.