Spie Press BookThermal Infrared Characterization of Ground Targets and Backgrounds, Second Edition
|Format||Member Price||Non-Member Price|
- 1 Sensor systems/11
- 1.1 Active sensor systems/11
- 1.2 Passive sensor systems/13
- 1.3 Active versus passive sensor systems/16
- 2 Radiation terminology and units/17
- 2.1 Definitions and spatial relationships/17
- 2.2 Intrinsic radiation terms/20
- 2.3 Atmospheric propagation/22
- 2.4 Range-dependent radiation terms/23
- 2.5 Target-to-background contrast/24
- 3 Introduction to target detection/26
- 3.1 IR detection process/26
- 3.1.1 Target-to-background radiation contrast/26
- 3.1.2 Attenuation processes/27
- 3.1.3 IR systems/27
- 3.1.4 Detection system/30
- 3.2 Point target detection/31
- 3.3 Extended target detection/32
- 3.4 Signature variations/34
- 3.5 Thermal contrast considerations/37
- 4 Theory of heat and mass transfer/39
- 4.1 Surface-atmospheric boundary layer/39
- 4.2 Heat and mass transfer/41
- 4.3 The heat balance equation/45
- 4.3.1 Solar heating/46
- 220.127.116.11 Solar absorption coefficient/46
- 18.104.22.168 Solar irradiance Esun/51
- 4.3.2 Long wave radiation exchange/54
- 22.214.171.124 Long wave absorption coefficient/55
- 126.96.36.199 Long wave sky irradiance Esky/55
- 4.3.3 Surface emittance Es/58
- 4.3.4 Convective heat exchange Qc/58
- 4.3.5 Heat exchange by evaporation/condensation Qec/63
- 4.4 First principles modeling/65
- 4.4.1 Model definition/66
- 4.4.2 Sensitivity analysis/69
- 4.4.3 Pros and cons/79
- 5 Meteorological and atmospheric parameters/83
- 5.1 Meteorological sensors and measurements/83
- 5.2 SCORPIO infrared sky radiance distribution/87
- 6 Infrared calibration procedures/95
- 6.1 Calibration methodology/97
- 6.2 Calibration parameters/100
- 6.3 Calibration procedures guidelines/103
- 6.3.1 Blackbodies and target in the same image/103
- 6.3.2 Blackbodies and target at different ranges/105
- 6.3.3 Blackbodies and target in different images/107
- 6.4 Practical example/109
- 7 Infrared signature characterization/116
- 7.1 Target signatures/116
- 7.1.1 Field measurement of the long wave reflection coefficient/117
- 7.1.2 Measurement of thermal target signatures in the field/121
- 188.8.131.52 CHAR-II experiment/121
- 184.108.40.206 IRIS trials/126
- 220.127.116.11 Land-mine detection experiments/128
- 7.1.3 IR target modeling/132
- 18.104.22.168 T62 tank/133
- 22.214.171.124 Land mines/136
- 7.2 IR background characterization/139
- 7.2.1 Field measurement of the shortwave reflection coefficient/142
- 7.2.2 Measurement of background apparent temperatures/145
- 126.96.36.199 CARABAS measurement system configuration/147
- 188.8.131.52 Practical examples/156
- 7.3 Background temperature statistics/157
- 7.3.1 Background temperature distributions/158
- 7.3.2 Statistical temperature differences between 3-5 um and 8-12 um/163
- 7.3.3 Background temperature curve fitting/165
- 184.108.40.206 Radiometer data/166
- 220.127.116.11 Images/174
- 7.3.4 Pros and cons/181
- 8 Signature management/183
- 8.1 Target and background signature analysis/183
- 8.2 Thermal signatures of materials/188
- 8.3 Mobile camouflage systems/194
- 8.4 Thermal camouflage of personnel/196
A number of physical phenomena can be used to observe or, more generally, to detect an object in a background. Usually (see Fig. 1), such a detection process has the following elements: Target-to-background contrast (with or without camouflage), atmospheric attenuation, and sensor performance (detector and signal processing). It is essential that there be a difference, i.e., a contrast, between at least one object and background feature, such as a radiation contrast or a temperature contrast (difference). Furthermore, this contrast must generate a detector signal S that significantly differs from the noise N spectrum of the sensor system, thus S/N > 1. If this is the case, then detection in principle is possible. NATO operations of today are deployed worldwide, exposing man and machine to a wide variety of climatological and environmental conditions. Consequently, target-to-background contrasts vary accordingly, and may reach considerable values at some extreme locations.
Figure 1 Detection scenario.
Various imaging and nonimaging sensor systems have been developed to detect special features such as color, temperature, sound, smell, shape, and such.Detection-system designers select features and system parameters to achieve the highest possible detection probability for a large variety of sensor-target-background scenarios under various weather conditions. On the other hand, military-equipment designers search for construction methods and materials that minimize the detection probability for such detection systems. So, for both worlds it is of utmost important to have detailed knowledge of target signatures under various target and environmental conditions.
Detection systems are subdivided into three categories: 1. Active sensor systems This type of system needs an active source to create or enhance a feature difference in order to detect it. Sometimes source and detector (receiver) are integrated into one system, such as a laser range finder or a Synthetic Aperture Radar (SAR) system. At other times they are separate, as in a situation in which a forward observer illuminates the target as a beacon for homing devices.
2. Passive sensor systems In the same context, passive systems do not need an active source, but utilize the existence of natural features, such as a thermometer, to measure temperatures or a thermal imager measuring emitted radiation of an object (the sun and moon are considered natural passive sources).
3. Semiactive sensor systems Detecting alien, foreign active sources, such as radar (RWR), laser (LRF), and electronic support measures (ESM). Some sensors that are, strictly speaking, active sensors are considered passive when the detection principle is based on signals that exist in a background. A microphone picking up the sound of a bird is an example of a natural sound, but also when it picks up the sound of an artificial source, like a helicopter. Another example is the human eye. The human eye-brain system is the best known active sensor system and, in terms of performance, is probably the best system that exists today. It operates in a very small spectral region, the visible part of the electromagnetic (EM) spectrum, as shown in Fig. 2 (source: www.temple.edu/biomed/spectrum.gif). However, its practical use is limited to daytime and its performance decreases seriously under adverse weather conditions. Figure 2 shows that the EM spectrum offers many more frequency regions, which could be utilized in either a passive or active way.
Gradually, more and more parts of the EM spectrum were used in imaging sensor systems, comprising the ultraviolet (UV), near-infrared (NIR), thermal infrared (TIR), and radar systems. Operational radar systems were introduced during World War II and contributed largely to the defeat of the German U-boat fleet. With the enormous revolution in technology during the post-war years, new detector materials became increasingly available, opening up more spectral bands to be used in sensor systems.
Figure 2 Electromagnetic spectrum.
Imaging sensor systems in the UV region, frequently used for missile (plume) detection, have become available. Especially, the spectral window from 0.25-0.39 um is used for this purpose because the natural background does not contain solar UV radiation (solar blind window). Also, sensor systems have been developed in the NIR 1.1-3 ?m, based on new sensor technology, such as quantum well detectors. Although reflected natural NIR radiation can be used for detection, spotlights such as a NIR lamp or laser may be needed to enhance contrast during adverse weather conditions, as with laser detection and ranging (LADAR) systems.
One promising option was the thermal infrared (TIR) spectral region, ranging from 3-50 um. Since all surfaces emit electromagnetic radiation in this spectral region, it can be utilized for passive detection. The explosive development of modern electronics and ongoing development of new detector materials have resulted in high-performance IR imaging systems. In this context, "high performance" means that the characteristic system parameters, such as NETD (noise equivalent temperature difference), IFOV (instantaneous field of view), MRTD (minimum resolvable temperature difference), etc. can be optimized theoretically to take nearly any desired value for a given choice of hardware components. Over the last years, uncooled IR detector technology emerged, offering a cheaper alternative to cooled systems, but providing less performance. Computer computational power also has grown enormously, offering powerful image- processing tools that can analyze imagery almost in real time. The consequence is that in the detection process, more target and background features can be analyzed quickly and automatically, which makes targets even more vulnerable. Modern sensors use many parts of the EM spectrum to optimize their performance for specific applications. Figure 3 shows an overview of spectral ranges in which modern detection systems operate.
Figure 3 Overview of spectral ranges, in which modern detection systems operate.
In addition to system hardware specifications, system performance should be expressed in terms of the system's ability to perform the task for which it was designed. This could be expressed thus: "With this system, it is possible to detect an armored personnel carrier parked in front of a tree line from a distance of 3 km, during a clear middle European night." The performance of that system is determined not only by its hardware specifications, but also by atmospheric propagation and, probably to the largest extent, the momentary IR contrast between the object and the local background. This is why all components, as given in Fig. 1, must be taken into account to accurately judge the detection performance of a given sensor system.
Analyses of atmospheric transmission enhance the understanding of atmospheric effects, such as aerosols, dust, and smoke particles, which have an impact on the propagation of IR radiation through the atmosphere. Many short- and long-range measurements have been conducted under various meteorological conditions. These efforts have culminated in a number of sophisticated semi-empirical models, such as MODTRAN and HITRAN. These models have been updated continuously over the last 15 years as new measurements and theory became available. Today, these models are reliable tools for calculating the atmospheric propagation and only require a small number of meteorological input parameters. Other than the use of MODTRAN, atmospheric transmission will not be discussed in this text.
The next part of the observation scenario is the characterization of the IR contrast between the object and the background. For simple object geometry, IR signatures can be calculated by solving the heat balance equation for the different object facets. However, with increasing object complexity, calculations soon become unreliable because mathematical representations of some energy flows are not valid (or not known) for small, arbitrarily oriented object facets.
Measurements are a good alternative because on many occasions object conditions can be controlled and the environmental conditions can more or less be selected (e.g., waited for). Databases exist comprising more than 10,000 images of military targets alone, taken at various target conditions and observation angles.
The characterization of background radiation is the most complex issue. This complexity is probably one reason it was (until recent years, when the lack of background information became the most limiting factor in the detection process) given more attention. In spite of this increase of effort, especially in the modeling area, little progress was made in the development of models that produced results accurate enough to be used in detection and recognition studies. This mainly is because, on the one hand, backgrounds are difficult to model as a result of their complex geometrical structure and, on the other hand, because the mathematical description for some transfer processes, such as those that exist in a vegetation layer, are not yet known accurately enough. The temporal variability of IR background radiation can be described satisfactorily with the aid of semi-empirical models. The characterization of the spatial distribution of IR background radiation will not be discussed explicitly in this Tutorial Text.
The main part of this Tutorial Text deals with the characterization of the thermal infrared (3-12 um) radiation contrast between ground targets and backgrounds. Sensor systems will be discussed only as far as necessary to explain the topic under discussion. For more detailed information on sensor systems, the reader is referred to one of the many textbooks available on this topic,1,2,3 or to specific literature on IR detectors.4 Passive radar systems use differences in reflected cosmic radiation between target and background in the higher frequency bands. Generally, detection ranges are relatively short (a few kilometers) and strongly weather dependent. This topic will not be further addressed. Not only is the EM spectrum used to gather information using active and passive detection, but acoustic, magnetic, and seismic phenomena can also be used. Systems based on these principles will not be discussed any further here.
In summary, many sensor systems that can be used for detection purposes are available, offering a wide range of deployment options: active/passive, day/night, long-/short-range, good/bad weather, and such.