Background-Subtraction in Thermal Imagery Using Contour Saliency

We present a new contour-based background-subtraction technique to extract foreground objects in widely varying thermal imagery. Statistical background-subtraction is first used to identify local regions-of-interest. Within each region, input and background gradient information are combined to form a Contour Saliency Map. After thinning, an A ^∗ path-constrained search along watershed boundaries is used to complete and close any broken contour segments. Lastly, the contour image is flood-filled to produce silhouettes. Results of our approach are presented for several difficult thermal sequences and compared to alternate approaches. We quantify the results using manually segmented thermal imagery to demonstrate the robustness of the approach.     

Fusion of visible and infrared imagery for night color vision

A combined approach for fusing night-time infrared with visible imagery is presented in this paper. Night color vision is thus accomplished and the final scene has a natural day-time color appearance. Fusion is based either on non-negative matrix factorization or on a transformation that takes into consideration perceptual attributes. The final obtained color images possess a natural day-time color appearance due to the application of a color transfer technique. In this way inappropriate color mappings are avoided and the overall discrimination capabilities are enhanced. Two different data sets are employed and the experimental results establish the overall method as being efficient, compact and perceptually meaningful.     

Early fire detection using non-linear multitemporal prediction of thermal imagery

This paper presents a sub-pixel thermal anomaly detection method based on predicting background pixel intensities using a non-linear function of a plurality of past images of the inspected scene. At present, the multitemporal approach to thermal anomaly detection is in its early development stage. In case of space-borne surveillance the multitemporal detection is complicated by both spatial and temporal variability of background surface properties, weather influences, viewing geometries, sensor noise, residual misregistration, and other factors. We use the problem of fire detection and the MODIS data to demonstrate that advanced multitemporal detection methods can potentially outperform the operationally used optimized contextual algorithms both under morning and evening conditions.     

Infrared and visible image fusion using fuzzy logic and population-based optimization

This paper presents a new wavelet-based algorithm for the fusion of spatially registered infrared and visible images. Wavelet-based image fusion is the most common fusion method, which fuses the information from the source images in the wavelet transform domain according to some fusion rules. We specifically propose new fusion rules for fusion of low and high frequency wavelet coefficients of the source images in the second step of the wavelet-based image fusion algorithm. First, the source images are decomposed using dual-tree discrete wavelet transform (DT-DWT). Then, a fuzzy-based approach is used to fuse high frequency wavelet coefficients of the IR and visible images. Particularly, fuzzy logic is used to integrate the outputs of three different fusion rules (weighted averaging, selection using pixel-based decision map (PDM), and selection using region-based decision map (RDM)), based on a dissimilarity measure of the source images. The objective is to utilize the advantages of previous pixel- and region-based methods in a single scheme. The PDM is obtained based on local activity measurement in the DT-DWT domain of the source images. A new segmentation-based algorithm is also proposed to generate the RDM using the PDM. In addition, a new optimization-based approach using population-based optimization is proposed for the low frequency fusion rule instead of simple averaging. After fusing low and high frequency wavelet coefficients of the source images, the final fused image is obtained using the inverse DT-DWT. This new method provides improved subjective and objectives results as compared to previous image fusion methods.     

Pedestrian detection and tracking in infrared imagery using shape and appearance

In this paper, we present an approach toward pedestrian detection and tracking from infrared imagery using joint shape and appearance cues. A layered representation is first introduced and a generalized expectation-maximization (EM) algorithm is developed to separate infrared images into background (still) and foreground (moving) layers regardless of camera panning. In the two-pass scheme of detecting pedestrians from the foreground layer: shape cue is first used to eliminate non-pedestrian moving objects and then appearance cue helps to locate the exact position of pedestrians. Templates with varying sizes are sequentially applied to detect pedestrians at multiple scales to accommodate different camera distances. To facilitate the task of pedestrian tracking, we formulate the problem of shot segmentation and present a graph matching-based tracking algorithm that jointly exploits the shape, appearance and distance information. Experimental results with both OSU Infrared Image Database and WVU Infrared Video Database are reported to demonstrate the accuracy and robustness of our algorithm.     

Multiple camera people detection and tracking using support integration

This paper proposes a method to locate and track people by combining evidence from multiple cameras using the homography constraint. The proposed method use foreground pixels from simple background subtraction to compute evidence of the location of people on a reference ground plane. The algorithm computes the amount of support that basically corresponds to the “foreground mass” above each pixel. Therefore, pixels that correspond to ground points have more support. The support is normalized to compensate for perspective effects and accumulated on the reference plane for all camera views. The detection of people on the reference plane becomes a search for regions of local maxima in the accumulator. Many false positives are filtered by checking the visibility consistency of the detected candidates against all camera views. The remaining candidates are tracked using Kalman filters and appearance models. Experimental results using challenging data from PETS’06 show good performance of the method in the presence of severe occlusion. Ground truth data also confirms the robustness of the method.     

Saharan dust in nighttime thermal imagery: Detection and reduction of related biases in retrieved sea surface temperature

Guided by radiative transfer modelling of the effects of Saharan dust (aerosol) on brightness temperatures at 3.9, 8.7, 11 and 12 μm measured by the Spinning Enhanced Visible and Infrared Imager (SEVIRI), we propose an indicator of the presence of Saharan dust in nighttime (infrared-only) SEVIRI imagery. Radiative transfer modelling is performed using a fast atmospheric transmittance model and a delta-Eddington approximation for scattering. Using aerosol single scattering properties appropriate to simulation of Saharan dust gives qualitatively good simulations of dust-affected brightness temperatures. From these simulations, we show that aerosol-free brightness temperatures are tightly clustered around the principal axis (explaining 99% of variance) in a 3-dimensional brightness-temperature difference space (the axes being 3.9 μm minus 8.7 μm, 3.9 μm minus 12 μm, and 11 μm minus 12 μm); in contrast, brightness temperatures affected by aerosol are simulated to be significantly off-axis when transformed into this space. Although the detailed effects of aerosol on brightness temperatures are not quantitatively reproduced by our simulations, this off-axis characteristic is empirically found to be also true of Saharan-dust-affected brightness temperatures in observations. The second principal component of the brightness-temperature difference space is identified as a useful index for Saharan dust. Comparisons with independent satellite measurements of aerosol optical depth (AOD, at 0.55 μm) show that the new Saharan dust index (SDI) is loosely correlated with AOD (r = 0.34). The correlation is loose because the SDI is also sensitive to the height distribution of aerosol, which affects the aerosol's thermal emission. The SDI is used to develop an empirical correction scheme for SST retrievals affected by Saharan dust. Application of the SDI correction scheme removes an independent error of 0.2 K from SEVIRI SSTs validated against buoys from all latitudes.     

An improved algorithm for the detection of plumes caused by natural or technological hazards using AVHRR imagery

Fires caused by natural or technological disasters emit large amounts of smoke which, once formed into plumes, may affect the human health and the environment. Satellite remote sensing data provide an effective tool to achieve detection and monitoring of these plumes over large areas on a routine basis. Discrimination of plumes on satellite images is a prerequisite to study and retrieve physical, chemical and optical properties of emitted smoke. An improved algorithm for the detection of plumes caused by natural or technological hazards using AVHRR imagery is presented in this study. The method is based on a multi-temporal and multi-spectral change detection algorithm. It is performed in two main steps: a) appropriate spectral and spatial filters are applied on the images acquired before and after a fire event in visible and near-infrared ranges in order to extract the core of the plume; b) a criterion on spectral information is defined as an homogeneity measure that enables, through a modified version of the region-growing method, the spatial expansion of the detected core to include the complete area covered by the plume. Through this approach, a pixel is identified as a plume pixel if it is “close” to the core plume pixels in both spatial and spectral spaces. The algorithm was developed and calibrated using AVHRR images acquired over Spain before and during a major forest fire event on July 16, 2005. It was applied using past events of natural and technological hazards in several locations to ensure its global applicability and robustness. The algorithm produced accurate results in all cases of plumes, either in natural or in technological fire events. Three application cases are presented in this study: A major fire in an industrial installation in London (December 11, 2005), a major fire in Baghdad during the recent war in Iraq (April 1, 2003) and a forest fire in California (September 29, 2005).     

Detection and monitoring of African vegetation fires using MSG-SEVIRI imagery

An operational procedure is presented that allows detecting active fires based on information from Meteosat-8/SEVIRI over Africa. The procedure takes advantage of the temporal resolution of SEVIRI (one image every 15 min), and relies on information from SEVIRI channels (namely 0.6, 0.8, 3.9, 10.8 and 12.0 µm) together with information on illumination angles. The method is based on heritage from contextual algorithms designed for polar, sun-synchronous instruments, namely NOAA/AVHRR and MODIS/TERRA-AQUA. A potential fire pixel is compared with the neighboring ones and the decision is made based on relative thresholds as derived from the pixels in the neighborhood.An overview is provided of results obtained for January and July 2007, respectively over Northern Africa (NAfr) and Southern Africa (SAfr), paying special attention to the spatial and temporal distribution of active fires. In both NAfr and SAfr, two types of vegetation clearly dominate in terms of fire activity, namely tree-covered areas, containing 40% of total fires observed, and shrub-covered areas, with 25% (19%) of total fires in NAfr (SAfr). However, marked differences were also to be found between the two regions; more than two-thirds (70%) of fires in SAfr were observed in land cover classes dominated by trees but the proportion is much lower (40%) in the case of NAfr. The duration of active fires in both regions tends to follow two-parameter generalized Pareto distributions, with both the scale and the shape parameters presenting very similar values for NAfr and SAfr.An assessment of the robustness of the algorithm, consistency of results and added value of the product was made by studying the daily cycle of fire activity over two regions located in northern and southern hemisphere Africa and by means of systematic comparisons against fire incidence reported in previous works and against hot spots extracted from the global daily active fire product developed by the MODIS Fire Team. The observed fire incidence by land cover class compares well with the results reported in previous works and it is shown that there is an overall coherence between results obtained from SEVIRI and MODIS when adequate spatial and temporal scales are chosen when performing the comparison. Data from MODIS and SEVIRI may be viewed as complementary, the latter having the added value of providing a much finer temporal resolution that allows uncovering certain aspects of fire behavior, namely the characterization of daily fire cycles.     

Identification and reconstruction of chaotic systems using multiresolution wavelet decompositions

A new modelling framework for identifying and reconstructing chaotic systems is developed based on multiresolution wavelet decompositions. Qualitative model validation is used to compare the multiresolution wavelet models and it is shown that the dynamical features of chaotic systems can be captured by the identified models providing the wavelet basis functions are properly selected. Two basis selection algorithms, orthogonal least squares and a new matching pursuit orthogonal least squares, are considered and compared. Several examples are included to illustrate the results.     

多源时序国产卫星影像的森林火灾动态监测

四川省木里县及周边林区是全国林火最为高发和易发区之一,近两年连续发生了扑火人员重大伤亡的事件。利用时序国产卫星影像、无人机影像和现场勘查数据等,从监测火灾蔓延时空过程的角度,对该区林火热点进行了动态监测,并分析了重点火场火灾发展过程,结果表明：以国产GF-4卫星影像为主,辅助以2 m/8 m光学卫星星座影像,可较好地监测林火热点;研究提出林火热点判定阈值为白天亮温值T≥360 K或夜间亮温值T≥330 K;监测发现了该区3月30日至4月6日间共6处火场的25次林火事件,并重点反演了①号木里和②号西昌火灾发展的时空过程。通过将卫星监测热点与现场勘查热点、无人机影像解译热点对比,表明在火灾早期和中期卫星林火热点监测精度可达89%。建议利用时序国产多源卫星影像对该区林火进行持续监测,并结合权威部门现场勘查数据适时发布预警信息,避免造成重大生命财产损失。