The thermal hyperspectral sensor Hyper-Cam was mounted on a light aircraft and measured continuous releases of several atmospheric tracers from a height of 2 km. A unique detection algorithm that eliminates the need for clear background estimation was operated over the acquired data with excellent detection results. The data-cubes were acquired in a "target mode", which is a unique method of operation of the Hyper-Cam sensor. This method provides multiple views of the plume which can be exploited to enhance the detection performance. These encouraging results demonstrate the utility of airborne LWIR hyperspectral imaging for efficient detection and mapping of effluent gases for environmental monitoring.
Along with rising concerns about the global warming and its long term consequences, the need for a better global
radiative balance model increases. While the global impact of the greenhous1e trace gases is well understood, the
radiative forcing of the various natural and manmade aerosols remains uncertain, especially in the IR spectral band.
Studying the optical properties of large scale dust loadings in the atmosphere directly is difficult due to the vast
uncertainties about their composition and size distributions. Furthermore, the chemical composition of a dust grain is
linked to its size. One of the methods to bypass these inherent difficulties is to study anticipated radiative effects with a
clearly defined simulant that is well characterized both chemically and by its particles size distribution. In this
presentation we show results from spectral and spatial measurements of such aerosol plumes composed of silicone oil
droplets. These measurements expand and improve our knowledge of the spectral signature of aerosol clouds obtained in
the IR spectral band. Our previous work presented measurements carried out with a non-imaging spectro-radiometer only
near the release point. In this article, we show experimental data obtained by a hypesrspectral sensor which enabled us,
for the first time to perform a simultaneous measurement of an aerosol cloud, both in the spectral and the spatial
domains. These results were compared to a radiative transfer model, and yielded an excellent agreement between the
predicted and the measured spectral signatures. The proposed model can be used for the prediction of the optical
properties of dust clouds in the atmosphere as well as assessing more accurately their impact on global climate change.
In the previous article, a comparison between the statistical and the spatial properties of IR images of rural and urban background was presented. Analyzing the characteristics of these two backgrounds yielded noticeable differences in most of the extracted parameters and their diurnal patterns. Furthermore, the experimental data pose a remarkable deviation from those predicted by the most accepted model for desert terrain IR images (Ben-Yosef model). The effect that might be responsible for these discrepancies is the local scene topography, which is clearly enhanced in an urban environment. To investigate this hypothesis, we introduce a simple modification of the Ben-Yosef model that incorporates a simulated urban background. The comparison between the simulation and experiment shows good agreement. We conclude that the scene's topography in an urban background is the most important parameter that governs its statistical and spatial characteristics in the IR band.
Machine vision of specific objects on natural backgrounds in the IR is an extensively studied subject. Characterizing the clutter is essential in order to evaluate a sensor's performance under various conditions. The Ben-Yosef model is the main one used for the characterization and parameterization of rural background IR images in terms of image statistics and texture. However, to the best of our knowledge, no such parameterization of urban images has been established. The aim of this work is a comparison between statistical and spatial characteristics of urban and rural scenes in the IR and their diurnal dynamics. We conclude that the Ben-Yosef model cannot fully describe the urban scene characteristics, mainly due to the model assumptions regarding the uniform spatial structure of the emissivity and of the magnitude of the solar flux over the scene. Experimental results show that, although daytime urban scenes have high variance in the IR, they have a less complex spatial structure than nighttime images, which are characterized by much lower variance.
Dissemination of SF6 and tracking its dispersion in the atmosphere is a well-known technique used to predict how pollutant affects the environment. Remote thermal imaging of the atmospheric tracer plume is one of the methods employed to detect and track its dispersion. However, remote detection of SF6 plumes in a stable boundary layer of the atmosphere (SBL) with a multispectral infrared sensor is a challenging task. At SBL conditions the tracer cloud tends to disperse very slowly and therefore its temporal signature is well mixed with the natural temperature variations over the background scene. Furthermore, SBL conditions are frequent during nighttime when the thermal contrast between the air and the background scene is very low. In this article we propose an efficient method to overcome these difficulties. The local temperature variance of the clean background is compared to the variance measured at the same position during the cloud presence in the field of view. The local temperature variance is modified by passage of radiation through the absorbing cloud. The distinctive spectral signature of the atmospheric tracer is expressed in the relative strength of the different spectral band of the IR sensor. The proposed technique is demonstrated with actual data collected during field test in an urban area. Urban background is particularly suitable for applying this method due to its inherent large thermal variance consisted of buildings, streets, parks etc. We demonstrate the usefulness of this detection method for accurate quantitative estimation of the tracer cloud density and its form.
SF6 is the most widely used atmospheric tracer in field tests held for performance evaluation of passive remote sensing systems. The most advanced systems are based on 2D arrays of MCT photovoltaic detectors operating in the LWIR (8-12μ) spectral band. Increasing the number of elements in an array is inevitably associated with a decrease in the detector responsivity cutoff wavelength. This might pose a significant difficulty in the use of SF6 due to its sole narrow absorption band centered at 10.55μ, which falls within a spectral region of poor FPA sensitivity. A search for a more suitable tracer for advanced detector arrays has yielded that Trifluromethane (CHF3) is the optimal alternative. The paper describes the considerations that led to its selection and surveys its main physical properties, safety, availability and costs. It also presents initial results from a field test that demonstrates its utility.
A Physical for the relation between reflective and thermal IR images of natural ground scenes is presented and discussed. The model enables the formulation of the joint distribution density function for the emitted radiance and its diurnal behavior. An approximated analytic expression for the multiband correlation coefficient is derived from the explicit form of the joint distribution function is calculated and presented. The effect of local scene topography and heat transfer mechanisms of vegetation on the obtained correlation coefficient is presented and analyzed. The main effect of the local scene topography is by introduction shaded areas which emit low radiance both in the thermal l and the reflective images. Vegetative objects regulate their temperature through evapotranspiration during daytime and hence usually becomes the coldest objects in the scene as their appearance in a visible image. Local scene topography effect can be estimated through a reasonable quantitative assessment of surface roughness characteristics and the relative geometry of the sun and the observation point. The resistance of vegetation to the external heat load is modeled via daytime and nighttime effective heat conduction. The effective heat conduction represents the unique heat transfer results which back the proposed model are presented. The model can be readily extended in order to describe the mutual relation between multispectral reflective images and thermal images. Furthermore, it enables us to interpret the effect of the addition of the thermal data on principal component analysis. The practical application of the model and its derived conclusions in the fields of remote sensing and data fusion are discussed.
A physical model for the prediction of the radiant statistics over thermal images of ground desert terrain landscapes and their temporal behavior had been fully established by Ben Yosef et al. This model can be further developed in order to formulate the joint radiant statistics of reflective and thermal infrared images over the same type of landscapes. However, it fails to predict the actual measured correlation between the images in the two bands, and hence, a modification of the joint radiant density function in order to consider the influence of the local ground topography over the scene is introduced. The prediction of the modified joint density function and correlation coefficient is consistent with the experimental data acquired over a rough desert landscape. The effect of local scene topography on thermal image properties is not negligible, especially when sun's elevation angle is low. In such cases, shaded areas are generated in the scene, occupying a substantial portion of it. Analysis of the temporal dynamics of the correlation coefficient between thermal and reflective images can infer about the relative importance of the topography contributed variance over the thermal image and as well as the clutter characteristics of the thermal image. Scene topography also introduces errors in the production of thermal inertia maps for remote sensing applications and limits its utility.
Principal Component Analysis is a well-known statistical method which is commonly applied in the analysis of multispectral images. This paper presents some of the results that have been received by this method of multispectral images of natural background terrain at high spectral and spatial resolution in the spectral range of 0.4 - 1.05. The results show that images at the visible band and near IR are highly correlated within each band, but poorly correlated between bands. However, PC analysis shows that they are not independent spectral bands, since they have high correlation or anti-correlation with the main principal components. Another important finding is a `neutral wavelength,' which shows very little spectral difference between bare soil and vegetation. This wavelength can be used as an indicator for vegetation types and seasonal changes, and for spectral enhancement at remotely sensed images in real time.
The statistical behavior of irradiance over a ground-based IR sky image is dependent on the angle of elevation at which the image was made. Images made at different elevation angles cannot be compared unless a correction is made for the different viewing angles. The angle-dependent radiance caused by the difference in the optical path length over the field of view of the image must also be corrected, otherwise incorrect conclusions will result. Once the corrections have been made, the image can be analyzed and the true statistical parameters of the image can be obtained.
The statistical and spatial properties of infrared images of
ground terrain in the LWIR band were studied extensively, and
some generalised models were developed. In this paper we show
that some of these models can applied also to the SWIR band. The
experimental results, combined with simulation results,
demonstrate that there is not any significant difference in the
radiative variances of ground terrain images in the object plane,
between the two band. In the image plane the results are not
compatible with the above stipulations. This dicrepancy is caused
bythe differences in signal, atmospheric transmission and system
response between the two spectral bands. Thus7 if these effects
are not minimized and taken into account, the image obtained will
represent reality with a low contrast image.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.