Visible Infrared Research Articles

Noble metal-free plasmonic Cu/WO3-x@ZNC bi-functional composite for ciprofloxacin photocatalytic degradation and photothermal water evaporation via visible-infrared exposure

Herein, coupling of noble metal-free plasmonic copper nanoparticles with tungsten suboxide and supporting on zeolite nanoclay (Cu/WO3-x@ZNC) composite will be introduced for bi-functional photocatalytic ciprofloxacin (CIP) degradation and water photothermal evaporation under visible/infrared (Vis/IR) exposure. Reduced band-gap of WO3-x via oxygen vacancies creation and localized surface plasmon resonance (LSPR) formation by Cu nanoparticles contributed significantly the extension and intensification of composite's photo-absorption range. Furthermore, small mesoporous structure of ZNC enhanced CIP adsorption and charge carriers separation where the reported photocatalytic efficiencies were 88.3 and 81.7% upon IR and Vis light exposure respectively. It was evidenced that plasmonic hot electrons (e−.s) and hydroxyl radicals (OH•-) performed the basic functions of the photocatalytic process. At the other side, oxygen vacancies existence, plasmonic effect, and confining thermal characteristics of WO3-x, Cu, and ZNC correspondingly induced water photothermal evaporation with efficiencies up to 97.5 and 72.8% under IR and Vis illumination respectively. This work introduces synthesis of a novel bi-functional photocatalytic-photothermal composite by metal sub-oxide and non-noble metal plasmonic coupling and supporting on naturally-derived carrier for water restoration under broad spectral exposure.

P-NiO/n-ZnO heterojunctions for visible-blind UV and IR photodetectors: A low-temperature fabrication approach

We investigate the design and characterization of stacked semiconductor structures, specifically p-NiO on n-ZnO, for achieving a multispectral photo-response. This device is fabricated by depositing a thin film of NiO on ZnO using a simple, low-temperature spray pyrolysis technique. The resulting heterojunction device, formed on a glass substrate at 250 °C and annealed at 500 °C, exhibits visible-blind characteristics and demonstrates remarkable stability for ultraviolet (UV) and infrared (IR) detection with a low dark current of 22 nA. The Ag/p-NiO/n-ZnO/Ag device shows enhanced photo-response only at wavelengths of 253 nm and 900 nm with a quick response time (Tr = 4.6 ms and Td = 20.3 ms). The observed photo-current is attributed to defect-mediated charge transport at the interface. These promising results highlight the potential of p-NiO/n-ZnO heterojunction-based device for applications demanding robust performance as visible blind IR and UV photodetectors.

Visible-infrared compatible and independent camouflage with multicolor patterns and tunable emissivity

Abstract With the rapid development and wide application of visible (VIS) and infrared (IR) detections, it is necessary to explore visible-infrared (VIS-IR) compatible camouflage. Here, we report a VIS-IR compatible and independent camouflage device which is composed of the upper IR-transparent VIS-color-patterned layer and the lower electrochromic IR layer. The upper layer has amorphous photonic structure of polystyrene nanospheres (PSNSs). By customizing the PSNS size, various colors can be realized for VIS camouflage. The lower electrochromic IR layer takes advantage of multiwall carbon nanotubes (MWCNTs) as the electrode as well as the IR active material. Experimental results reveal that different colors (including blue, green, and purple) have been obtained, and the IR emissivity can be electrically regulated from 0.43 to 0.9. Moreover, the prototype also exhibits good electrical stability as well as hydrophobic characteristic (the water contact angle of the outmost surface exceeds 120°). These output performances demonstrate the success of our design strategy for promoting the finding applied in camouflage fields as well as energy conservation fields.

Electrochromic and thermal-control broadband stealth device based on AZO/Ag/AZO configuration electrodes

Existing stealth technologies have limitations including single bands and functions, and static and passive applications, so that targets cannot remain stealthy if multi-source detection approaches are used. With the wide application of visible-infrared (Vis-IR) detection and guidance technologies, there is an urgent need to develop advanced techniques for visible smart stealth and Vis-IR compatible stealth. In view of this, an electrochromic and thermal-control broadband stealth device based on AZO/Ag/AZO configuration electrodes was developed. The influencing mechanisms of microstructural characteristics of the conducting layer in the AZO/Ag/AZO configuration on the visible transmission spectra and infrared (IR) reflection spectra were revealed. Furthermore, a conducting layer enabling IR stealth with high visible transmittance was optimized, on which a V2O5 multi-color electrochromic film layer was prepared via self-assembled deposition. In this way, the device integrates multiple optical mechanisms including the visible-induced transmission, IR radiation suppression, and electrochromism to modulate optical properties in multiple bands. A technical route that integrates visible smart stealth and active suppression of IR radiation was established, which is conducive to improving the adaptive fusion ability under a multi-domain background and multi-band compatible optical stealth.

HyperFace: A Deep Fusion Model for Hyperspectral Face Recognition.

Face recognition has been well studied under visible light and infrared (IR) in both intra-spectral and cross-spectral cases. However, how to fuse different light bands for face recognition, i.e., hyperspectral face recognition, is still an open research problem, which has the advantages of richer information retention and all-weather functionality over single-band face recognition. Thus, in this research, we revisit the hyperspectral recognition problem and provide a deep learning-based approach. A new fusion model (named HyperFace) is proposed to address this problem. The proposed model features a pre-fusion scheme, a Siamese encoder with bi-scope residual dense learning, a feedback-style decoder, and a recognition-oriented composite loss function. Experiments demonstrate that our method yields a much higher recognition rate than face recognition using only visible light or IR data. Moreover, our fusion model is shown to be superior to other general-purpose image fusion methods that are either traditional or deep learning-based, including state-of-the-art methods, in terms of both image quality and recognition performance.

Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b

Context. WASP-76 b has been a recurrent subject of study since the detection of a signature in high-resolution transit spectroscopy data indicating an asymmetry between the two limbs of the planet. The existence of this asymmetric signature has been confirmed by multiple studies, but its physical origin is still under debate. In addition, it contrasts with the absence of asymmetry reported in the infrared (IR) phase curve. Aims. We provide a more comprehensive dataset of WASP-76 b with the goal of drawing a complete view of the physical processes at work in this atmosphere. In particular, we attempt to reconcile visible high-resolution transit spectroscopy data and IR broadband phase curves. Methods. We gathered 3 phase curves, 20 occultations, and 6 transits for WASP-76 b in the visible with the CHEOPS space telescope. We also report the analysis of three unpublished sectors observed by the TESS space telescope (also in the visible), which represents 34 phase curves. Results. WASP-76 b displays an occultation of 260 ± 11 and 152 ± 10 ppm in TESS and CHEOPS bandpasses respectively. Depending on the composition assumed for the atmosphere and the data reduction used for the IR data, we derived geometric albedo estimates that range from 0.05 ± 0.023 to 0.146 ± 0.013 and from <0.13 to 0.189 ± 0.017 in the CHEOPS and TESS bandpasses, respectively. As expected from the IR phase curves, a low-order model of the phase curves does not yield any detectable asymmetry in the visible either. However, an empirical model allowing for sharper phase curve variations offers a hint of a flux excess before the occultation, with an amplitude of ~40 ppm, an orbital offset of ~ −30°, and a width of ~20º. We also constrained the orbital eccentricity of WASP-76 b to a value lower than 0.0067, with a 99.7% confidence level. This result contradicts earlier proposed scenarios aimed at explaining the asymmetry observed in high-resolution transit spectroscopy. Conclusions. In light of these findings, we hypothesise that WASP-76 b could have night-side clouds that extend predominantly towards its eastern limb. At this limb, the clouds would be associated with spherical droplets or spherically shaped aerosols of an unknown species, which would be responsible for a glory effect in the visible phase curves.

Enhanced visible–infrared person re-identification based on cross-attention multiscale residual vision transformer

Visible–infrared (VI) person re-identification (Re-ID) is a critical identification task that involves retrieving and matching images of an individual using both infrared and visible imaging modalities. To improve the performance, researchers have developed methods to obtain implicit feature information; however, this degrades with fewer discriminative features. To address this issue, we propose a weighted fused cross-attention multi-scale residual vision transformer (WF-CAMReViT) approach to re-identify the appropriate person from visible–infrared modality images by integrating the cross-attention multi-scale residual vision transformer architecture with Opposition-based Dove Swarm Optimization (ODSO). The proposed framework aims to bridge the domain gap between the visible and infrared modalities and significantly improve the re-identification performance. RGB (visible) and infrared (IR) images of persons are gathered from standard datasets, subjected to a cross-attention multi-scale residual vision transformer network to collect features, and then fuse using minimal weight. We also propose Opposition-based DSO to find the minimal weight. The weighted fused features are then subjected to the final decoder layer of CAMReViT to perceive the characteristics of each modality. In this study, model-aware enhancement (MAE) loss is develop to improve the modality information capacity of modality-shared features. Then, the experimental results on the SYSU-MM01 and RegDB datasets are compared with state-of-the-art transformer-based visible–infrared person Re-ID tasks to verify the efficacy of the proposed model.

Statically Multiple Colors and Dynamically Infrared Emissivity Modulation Compatible Electrochromic Devices via Simple Fabry–Perot Photonic Structures

AbstractThe development of electrochromic devices (ECDs) that can dynamically modulate infrared (IR) emissivity while maintaining a wide color gamut distribution for visible (VIS)‐IR compatible spectral manipulation has been a long‐standing challenge. Current ECDs are limited by simultaneous changes in IR emissivity and VIS colors when subjected to external potential, restricting their potential applications. This work presents, for the first time, VIS–IR compatible ECDs capable of displaying VIS multicolor and dynamically modulating IR emissivity via a Ge/HfO2 Fabry–Perot cavity structure. The all‐solid‐state VIS‐IR compatible ECD with the structure of ITO/NiO/Ta2O5/Li/WO3/ITO/Ge/HfO2 achieves not only emissivity modulation of 0.20 at potentials of −2.5 V and +2.5 V in the wavelength range of 2.5–25 µm, but also, due to the presence of Fabry–Perot cavity, the novelty compatible device structure achieves static multicolor modulation, thus decoupling VIS and IR emissivity modulation. Additionally, a flexible VIS‐IR compatible ECD based on polyaniline is also constructed. This novel approach offers a promising possibility for multispectral compatible manipulation, VIS multicolor, and IR signal processing applications.

KTFEv2: Multimodal Facial Emotion Database and its Analysis

In recent years, the focus of human emotion analysis has gradually shifted towards not only using visible information, but also thermal infrared (IR) information. This requires a great deal of facial emotion data both in visible and thermal IR information. However, most existing databases contain either visible information or posed thermal IR information only. For these reasons, we propose and establish a multimodal facial emotion database including both natural spontaneous visible and thermal IR videos. Beside updating more thermal infrared information, the built dataset in this study also enhances the information of intensity emotions. In which, each emotion is classified into three levels (low, medium, and high). Seven spontaneous emotions from thirty subjects are recorded in the database. Audio and visual stimuli were used to elicit emotions during the experiment. After the standard procedure of collecting data finished, the database has been through the careful annotation and verification procedure. Furthermore, the built database is analyzed by using modern machine learning models such as CNN, ResNet50, YOLO, and using a combination of different models to analyze the dataset. The obtained results are feasible and show that this dataset is useful for use in practice. The results of thermal data analysis provide us with a promising idea for future research on estimating human emotion.

Structural, electrical, magnetic & optical properties of Nickel, cobalt doped and Co-doped wurtzite GaN: A first-principle investigation

The structural, electronic, magnetic, and optical properties of Co (6.25%) and Ni (6.25%) monodoped and co-doped GaN were calculated by the DFT+U method using a 2 × 2 × 2 supercell of a total of 32 atoms. The formation energy was negative under both Ga-rich and N-rich conditions for all structures. N-rich condition is found to be favorable for GaN growth. The electronic and magnetic properties changed after doping due to the transition of the 3 d states of Ni, Co and hybridization of 2p of N. Hence monodoped GaN is ferromagnetic whereas the co-doped one is ferrimagnetic due to partial cancellation by the opposite spin of Ni and Co. The doped and co-doped absorption peaks are blue-shifted in the ultraviolet (UV) and redshifted in the visible-infrared (VIS-IR) region from the peaks of pure GaN. A pronounced absorption of VIS-IR is observed in monodoped structures. Both dielectric constant and refractive index increased due to co-doping whereas monodoping had enhanced the properties significantly. This theoretical study suggests the potential application of the Co, Ni monodoped in VIS- IR photonic and co-doped GaN in spintronic devices.

Person Tracking by Detection Using Dual Visible-Infrared Cameras

We study the problem of cross-modality person reidentification (ReID) and tracking with dual visible-infrared (VI) cameras, while most existing efforts on tracking-by-detection have been paid on single-modality visible ReID which is inapplicable for poor-light environments. The major difficulties for cross-modality (e.g., VI) ReID stem from the large modality gap between three-channel visible images and one-channel infrared images and such unknown environmental factors as background clutter, occlusions, etc. To tackle these issues, we propose to enrich the diversities of visible and infrared images for intra- and cross-modality matching by using both the channel-aware data augmentation (DA) techniques (e.g., channel exchanged augmentation and random occlusions) and standard DA techniques. On top of these DA techniques, we incorporate ResNet50 and vision transformer (ViT) into the feature extraction backbone network and apply the dynamic weight average (DWA) strategy for learning loss weights by regarding the minimization of identity loss and triplet loss as a multitask learning problem. We then apply the proposed ReID approach for person tracking in the field of interests. The experiments on two public data sets, i.e., RegDB and SYSU-MM01, show that our approach can improve the performance of state-of-the-art rank-1, mAP, and mINP for cross-modality matching. In addition, the experiments on our data set show that tracking by VI-ReID using dual VI cameras can achieve an accuracy of around 0.24 m.

Improving optical properties of wurtzite GaN with C and Fe co-doping: A DFT+U study

Density Functional Theory with Hubbard U parameter (DFT + U) was used to study the impact of C (6.25%) and Fe (12.5%) mono- and co-doping on wurtzite GaN, which modified the structural, electrical, magnetic, and optical properties. Under conditions of abundant N, the doping effect led to an increase in thermodynamic formability. For all the dopant combinations, a reduction in the band gap was seen. GaN becomes a ferrimagnetic material because of co-doping. The performance of the C and Fe co-doped GaN was the best due to the redshift of the absorption edge, which resulted in the improved absorption of near ultraviolet (UV) and visible-infrared (VIS-IR) photonic energies for both monodoping and co-doping structures. The doping caused an increase in the refractive index and dielectric constant. For C and Fe co-doped GaN, the maximum static dielectric constant and refractive index were 19.58 and 4.45, respectively. The findings of this work therefore point to the possible use of C, Fe mono- and co-doped GaN in UV, IR optoelectronic and photonic devices.

Glass formation and optical properties of Sn modified GeS2-Ga2S3-CsCl chalcogenide glasses

Chalcogenide glass is emerging as important IR materials for advanced athermal infrared (IR) optical systems because of its low thermal coefficient of refractive index. New chalcogenide glasses based on 75GeS2-15Ga2S3-10CsCl were developed through the substitution of GeS2 by SnS2. The introduction of heavier tin (Sn) results in the red shift of transmission window, both of visible and long-wavelength IR cut-off edges, the elevation of refractive index dispersion curves, and the decreasing of glass transition temperature. The formation of [SnS4-xClx] structural units, instead of [GeS4-xClx], should be responsible for the variation of these structural properties. In addition, homogeneous Sn modified GeS2-Ga2S3-CsCl chalcogenide glass of Φ30 mm was also prepared successfully, showing it of high potential for commercial applications.

Estimation of leaf nitrogen levels in sugarcane using hyperspectral models

ABSTRACT: Sugarcane is a good source of renewable energy and helps reduce the emission of greenhouse gases. Nitrogen has a critical role in plant growth; therefore,estimating nitrogen levels is essential, and remote sensing can improve fertilizer management. This field study selects wavelengths from hyperspectral data on a sugarcane canopy to generate models for estimating leaf nitrogen concentrations. The study was carried out in the municipalities of Piracicaba, Jaú, and Santa Maria da Serra, state of São Paulo, in the 2013/2014 growing season. The experiments were carried out using a completely randomized block design with split plots (three sugarcane varieties per plot [variety SP 81-3250 was common to all plots] and four nitrogen concentrations [0, 50, 100, and 150 kgha-1] per subplot) and four repetitions. The wavelengths that best correlated with leaf nitrogen were selected usingsparse partial least square regression. The wavelength regionswere combinedby stepwise multiple linear regression. Spectral bands in the visible (700-705 nm), red-edge (710-720 nm), near-infrared (725, 925, 955, and 980 nm), and short-wave infrared (1355, 1420, 1595, 1600, 1605, and 1610 nm) regions were identified. The R² and RMSE of the model were 0.50 and 1.67 g.kg-1, respectively. The adjusted R² and RMSE of the models for Piracicaba, Jaú, and Santa Maria were 0.31 (unreliable) and 1.30 g.kg-1, 0.53 and 1.96 g.kg-1, and 0.54 and 1.46 g.kg-1, respectively. Our results showed that canopy hyperspectral reflectance can estimate leaf nitrogen concentrations and manage nitrogen application in sugarcane.

An efficient and hybrid pulse coupled neural network - based object detection framework based on machine learning

The objective of fusing Infrared (IR) and Visible Image (VI) is to obtain essential information and reproduce an image with high reliability for human vision. Existing fusion methods are characterized by loss of information in fusion process thereby leading to lack of precision. To preserve the information, a novel fusion method is proposed in this research work by utilizing a pulse coupled neural network-based image fusion methodology. Proposed work integrates the visible image and IR image and generates a fused image with enhanced information. Further, the fused image and non-image data are used to detect the query objects like human and other objects using a convolutional neural network model. The proposed work is apt and suitable for surveillance applications, to analyze the scene in an effective manner. Experimental results reflect a superior performance as compared against benchmark methods in extracting the target information whilst preserving the visible background image.

Use of Sentinel-2 Satellite Data for Windthrows Monitoring and Delimiting: The Case of “Vaia” Storm in Friuli Venezia Giulia Region (North-Eastern Italy)

On the 29th of October 2018, a storm named “Vaia” hit North-Eastern Italy, causing the loss of 8 million m3 of standing trees and creating serious damage to the forested areas, with many economic and ecological implications. This event brought up the necessity of a standard procedure for windthrow detection and monitoring based on satellite data as an alternative to foresters’ fieldwork. The proposed methodology was applied in Carnic Alps (Friuli Venezia Giulia, NE Italy) in natural stands dominated by Picea abies and Abies alba. We used images from the Sentinel-2 mission: 1) to test vegetation indices performance in monitoring the vegetation dynamics in the short period after the storm, and 2) to create a windthrow map for the whole Friuli Venezia Giulia region. Results showed that windthrows in forests have a significant influence on visible and short-wave infrared (SWIR) spectral bands of Sentinel-2, both in the short and the long-term timeframes. NDWI8A and NDWI were the best indices for windthrow detection (R2 = 0.80 and 0.77, respectively) and NDVI, PSRI, SAVI and GNDVI had an overall good performance in spotting wind-damaged areas (R2 = 0.60–0.76). Moreover, these indices allowed to monitor post-Vaia forest die-off and showed a dynamic recovery process in cleaned sites. The NDWI8A index, employed in the vegetation index differencing (VID) change detection technique, delimited damaged areas comparable to the estimations provided by Regional Forest System (2545 ha and 3183 ha, respectively). Damaged forests detected by NDWI8A VID ranged from 500 m to 1500 m a.s.l., mainly covering steep slopes in the south and east aspects (42% and 25%, respectively). Our results suggested that the NDWI8A VID method may be a cost-effective and accurate way to produce windthrow maps, which could limit the risks associated with fieldwork and may provide a valuable tool to plan tree removal interventions in a more efficient way.

Shortwave infrared vegetation index-based modelling for aboveground vegetation biomass assessment in the arid steppes of Algeria

Selecting the appropriate vegetation index for accurate biomass estimation is a prerequisite before and during the ecosystem management project. This study, aims to compare Vegetation Indices (VIs) that are combining both Visible and Near Infrared OLI bands (VNIR-VIs), Visible and Short Wave Infrared OLI bands and also NIR and Short Wave Infrared OLI bands (SWIR-VIs) in order to accurately model the Aboveground Biomass (AGB) of three widely-located study sites over the arid steppe lands in Algeria. The Simple Linear Model (SLM) and Support Vector Machine (SVM) were utilised as statistical learning techniques on data; firstly, from each study site separately, and secondly, from all study sites (pooled data). In all study sites, SVM improves R² with a mean of 4.5% and decreases the Root Mean Squared Error (RMSE) and Percentage of Error (PE), respectively, with 15.50 (kg DM ha−1) and 1.33% on average. In all cases, the SWIR-VIs outperforms the VNIR-VIs with an improvement rate of 40% of R² and an average reduction of 362.36 kg DM ha−1 and 25% of RMSE and PE, respectively. The principal main improvement was found to involve the pooled data-based model utilising normalised difference VI form, which combines OLI2(0.452–0.512 μm) with OLI7(2.107–2.294 μm), (R² = 0.840, p < 0.0005).

Hypothetical Product Generation of Geostationary Ocean Color Imager Bands Over the Yellow Sea and Bohai Sea Using Deep Learning Technique

This study presents a deep learning model, using the conditional generative adversarial nets (CGAN) technique, that can produce daytime visible (VIS) band information, mimicking a narrow band sensor, by combining VIS and infrared (IR) broadband measurements by different sensors. The real-observed datasets of the Geostationary Ocean Color Imager (GOCI) and meteoritical imager (MI) sensors onboard the Communication, Ocean, and Meteorological Satellite were used for training and testing our CGAN model over the Yellow Sea and Bohai Sea. The trained and tested CGAN model was then applied to generate daytime GOCI VIS and near IR (NIR) bands (0.412 to 0.865 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">μ</i> m) using daytime MI VIS (0.675 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">μ</i> m), shortwave IR (3.75 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">μ</i> m), and longwave IR bands (10.8 and 12.0 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">μ</i> m) and the differences between them as input data. GOCI and MI data were collected from January 2017 to December 2018 using 705 images of 256 × 256 pixels for the training and 44 images for the model test. The results are statistically favorable (i.e., bias = −0.013 (in a reflectance unit from 0 to 1), root-mean-square error = 0.112, mean absolute error = 0.076, agreement index = 0.945, and correlation coefficient (CC) = 0.809 for daytime reflectance in the GOCI VIS 0.49- <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">μ</i> m band) between the real GOCI VIS band observation and the CGAN-generated simulation. Our CGAN-based model showed high CC and favorable results in the GOCI VIS and NIR bands. Consequently, our study demonstrates the possibility of applying a deep learning technique to improve the temporal resolution for ocean color studies using the GOCI sensor.

Effect of wood surface roughness on prediction of structural timber properties by infrared spectroscopy using ANFIS, ANN and PLS regression

Predicting the properties of structural timber using a rapid and reliable non-destructive method is a critical quality control task in production. This study investigates using infrared spectroscopy for predicting the density, moduli of elasticity (MOE) and rupture (MOR) of two structural softwoods. Because the produced timber is sometimes planed during manufacturing resulting in a smooth surface finish, the effect of wood surface roughness on prediction accuracy of timber properties was also investigated. Accordingly, infrared spectroscopy experiments were carried out on Douglas-fir and Western hemlock specimens having a rough and smooth surface. In addition, the effect of the infrared spectroscopy range on the predictive models was studied. Data in the visible infrared (VIS), near-infrared (NIR), and the combined VIS and NIR range were used for properties prediction. The acquired infrared data were processed using principal component analysis (PCA) for data reduction and feature selection. The output of PCA was then fed into an adaptive neuro-fuzzy inference system (ANFIS), multilayer perceptron (MLP) neural network (NN) and partial least square (PLS) regression model. The results showed that the wood surface finish, range of infrared data, and the type of machine learning model impact the prediction accuracy. ANFIS showed superior performance to MLP NN and PLS model for properties prediction. In addition, NIR data acquired from rough surface resulted in better prediction accuracy, which suggests that an infrared spectroscopy test should be performed prior to surface planing. The proposed models yield better accuracy for predicting the MOE and MOR than wood density.

Cloud Detection From Paired CrIS Water Vapor and CO₂ Channels Using Machine Learning Techniques

Accurate cloud detection using infrared (IR) data is very challenging due to the limitations and uncertainties from many aspects in the satellite IR remote sensing. This article proposes an end-to-end cloud detection method for the Cross-track IR Sounder (CrIS) using machine learning (ML) techniques. The brightness temperatures from paired CrIS channels in the longwave and midwave water vapor bands and the longwave and shortwave CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> bands are used. After obtaining the linear regression coefficients for each of the selected channel pairs, a complete set of CrIS full spectral resolution (FSR) cloud detection index (FCDI) is derived from the temperature difference between the regression and observation for each channel pair. It is shown that FCDI captures cloud location and structure well by comparing with the cloud products (CPs) from the Visible IR Imaging Radiometer Suite (VIIRS). After collocating FCDI with VIIRS CP, ML techniques such as the extreme learning machine, support vector machine, and multilayer perceptron are used to train the collocated FCDIs for cloud detection. Simulation results show that the accuracy of FCDI cloud detection is slightly above 80%. Moreover, the results encourage the use of water vapor bands in FCDI, in addition to CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> bands.