Intensity Imagery Research Articles

Tonal noise case study: Locating the source of tones on HVAC equipment

The overall noise level measured from chiller units is controlled by primarily tonal noises from (1) the fans, (2) the compressors, and (3) the associated pipework. Three large chiller units were measured using AHRI 370 methodology to quantify the overall sound level spectra and A-weighted sound level from different chiller units in a range of environmental conditions. Whilst this method can identify a general area of tonal emission to within a couple of meters across a unit’s surface that is emitting discrete tonal noises, the method is limited when there are multiple potential tonal sources. Sound intensity imagery was used on two units to quantify the overall sound power emanating from the source, but also to offer a more advanced method of localizing the source of discrete tones to aid in the application of mitigation treatments. Results presented include visually displaying and quantifying the location of noise emission from fan blade pass frequencies and associated harmonics, localization of problematic tones emanating from individual pipework junctions and valves, and the reductions in sound power emissions achieved from lagging compressors and pipework.

Image-to-Height Domain Translation for Synthetic Aperture Sonar

Synthetic aperture sonar (SAS) intensity statistics are dependent upon the sensing geometry at the time of capture. Estimating bathymetry from acoustic surveys is challenging. While several methods have been proposed to estimate seabed relief via intensity, we develop the first large-scale study that relies on deep learning models. In this work, we pose bathymetric estimation from SAS surveys as a domain translation problem of translating intensity to height. Since no dataset of coregistered seabed relief maps and sonar imagery previously existed to learn this domain translation, we produce the first large simulated dataset containing coregistered pairs of seabed relief and intensity maps from two unique sonar data simulation techniques. We apply four types of models, with varying complexity, to translate intensity imagery to seabed relief: a shape-from-shading (SFS) approach, a Gaussian Markov random field (GMRF) approach, a conditional Generative Adversarial Network (cGAN), and UNet architectures. Each model is applied to datasets containing sand ripples, rocky, mixed, and flat sea bottoms. Methods are compared in reference to the coregistered simulated datasets using L1 error. Additionally, we provide results on simulated and real SAS imagery. Our results indicate that the proposed UNet architectures outperform an SFS, a GMRF, and a pix2pix cGAN model.

Near real-time hurricane rainfall forecasting using convolutional neural network models with Integrated Multi-satellitE Retrievals for GPM (IMERG) product

Artificial Intelligence and Machine Learning (AI/ML) techniques are powerful tools, which can be applied to forecast spatial imagery in sequence. In this study, we applied a Convolutional Neural Network (CNN) model to predict the next few hours of rainfall observation imagery from NASA's Global Precipitation Measurement mission and its IMERG Early Run product in assist of hurricane predictions over the U.S.-Mexico Gulf coastal region. The IMERG Early Run is a satellite-based rainfall retrieval technique, providing rainfall measurement with a half-hourly temporal resolution, near-global scale coverage, and 4 h observation latency rainfall intensity imagery. The goal of this study is to extend the observation latency so that the IMERG Early Run product could potentially be used for real-time hurricane prediction and flash flood warning. In this study, we first build a sub-dataset from the IMERG Early Run product containing total 37 past hurricane events that hit the contiguous U.S. (CONUS) bordering the Gulf of Mexico from 2002 to 2019. Then, two CNN models with different model structures are built and tested to forecast the 37 hurricane events in a retrospective manner. Sensitivity experiments are conducted to select the optimal hyperparameters of the two CNN models, i.e., the number of convolution layers, filter size, minibatch size, number of filters, and pooling size. The prediction results show that the two employed CNN models generally provide satisfactory performance, showing averaged Accuracy in the categorical metrics is above 90% and averaged NSE in the continuous metrics is above 0.5. We found that the forecasting performance of the two CNN models is not significantly different; however, the CNN model without pooling layers always shows slightly better performance than the CNN model with pooling layers. We also found that both CNN models can predict the spatial range of hurricane rainfall well, but there are limitations to underestimate in larger rainfall rates. This study will serve as a novel investigation on how AI/ML models could help with real-time hurricane predictions, and provide a future reference for selecting the optimal CNN model parameters for processing spatial rainfall data.

Detection of Collapsed Bridges from Multi-Temporal SAR Intensity Images by Machine Learning Techniques

Bridges are an important part of road networks in an emergency period, as well as in ordinary times. Bridge collapses have occurred as a result of many recent disasters. Synthetic aperture radar (SAR), which can acquire images under any weather or sunlight conditions, has been shown to be effective in assessing the damage situation of structures in the emergency response phase. We investigate the backscattering characteristics of washed-away or collapsed bridges from the multi-temporal high-resolution SAR intensity imagery introduced in our previous studies. In this study, we address the challenge of building a model to identify collapsed bridges using five change features obtained from multi-temporal SAR intensity images. Forty-four bridges affected by the 2011 Tohoku-oki earthquake, in Japan, and forty-four bridges affected by the 2020 July floods, also in Japan, including a total of 21 collapsed bridges, were divided into training, test, and validation sets. Twelve models were trained, using different numbers of features as input in random forest and logistic regression methods. Comparing the accuracies of the validation sets, the random forest model trained with the two mixed events using all the features showed the highest capability to extract collapsed bridges. After improvement by introducing an oversampling technique, the F-score for collapsed bridges was 0.87 and the kappa coefficient was 0.82, showing highly accurate agreement.

Generation of Horizontally Curved Driving Lines in HD Maps Using Mobile Laser Scanning Point Clouds

This paper presents the development of a semiautomated driving line generation method using point clouds acquired by a mobile laser scanning system. Horizontally curved driving lines are a critical component for high-definition maps that are required by autonomous vehicles. The proposed method consists of three steps: Road surface extraction, road marking extraction, and driving line generation. First, the points covering road surfaces are extracted using the curb-based road surface extraction algorithms depending on both the elevation and slope differences. Then, road markings are identified and extracted according to a variety of algorithms consisting of georeferenced intensity imagery generation, multithreshold road marking extraction, and statistical outlier removal. Finally, the conditional Euclidean clustering algorithm is employed, followed by the cubic spline curve-fitting algorithm and equidistant line-based driving line generation algorithms for horizontally curved driving line generation. Our method is evaluated by six MLS point cloud datasets collected from various types of horizontally curved road corridors. Quantitative evaluations demonstrate that the proposed road marking extraction algorithm achieves an average recall, precision, and F1-score of 90.79%, 92.94%, and 91.85%, respectively. The generated driving lines are assessed by overlaying them on the manually interpreted reference buffers from 4-cm resolution unmanned aerial vehicle orthoimagery, and a 15 cm level navigation and localization accuracy is achieved.

Multi-epoch Direct Imaging and Time-variable Scattered Light Morphology of the HD 163296 Protoplanetary Disk

We present H-band polarized scattered light imagery and JHK high-contrast spectroscopy of the protoplanetary disk around HD 163296 observed with the HiCIAO and SCExAO/CHARIS instruments at Subaru Observatory. The polarimetric imagery resolve a broken ring structure surrounding HD 163296 that peaks at a distance along the major axis of 0.65 (66 AU) and extends out to 0.98 (100 AU) along the major axis. Our 2011 H-band data exhibit clear axisymmetry, with the NW- and SE- side of the disk exhibiting similar intensities. Our data are clearly different than 2016 epoch H-band observations from VLT/SPHERE that found a strong 2.7x asymmetry between the NW- and SE-side of the disk. Collectively, these results indicate the presence of time variable, non-azimuthally symmetric illumination of the outer disk. Based on our 3D-MCRT modeling of contemporaneous IR spectroscopic and H-band polarized intensity imagery of the system, we suggest that while the system could plausibly host an inclined inner disk component, such a component is unlikely to be responsible for producing the observed time-dependent azimuthal variations in the outer scattered light disk of the system. While our SCExAO/CHARIS data are sensitive enough to recover the planet candidate identified from NIRC2 in the thermal IR, we fail to detect an object with a corresponding JHK brightness estimated from the atmospheric models of Baraffe et al. 2003. This suggests that the candidate is either fainter in JHK bands than model predictions, possibly due to extinction from the disk or atmospheric dust/clouds, or that it is an artifact of the dataset/data processing. Our SCExAO/CHARIS data lower the IR mass limits for planets inferred at larger stellocentric separations; however, these ALMA-predicted protoplanet candidates are currently still consistent with direct imaging constraints.

A Simple Offset “Calibration” Method for the Accurate Geographic Registration of Ship-Borne X-Band Radar Intensity Imagery

X-band radar remote sensing of the ocean surface using ground-based installations and radar intensity imagery is a mature technology for the determination of ocean wave properties. In order to transition the analytical methods to data collected from a moving vessel the accurate geo-registration of radar images must be performed. However, finite offsets in azimuth, range and time are generated by the physical installation of equipment aboard a vessel that may not be measurable. This paper details a simple, yet robust method to “calibrate” data recorded by an arbitrary equipment installation to allow accurate geo-registration of the radar imagery. Time-integrated radar images are generated using a set of time-stamped radar intensity images and high-frequency, high-accuracy vessel heading and position information (from global satellite navigation systems and inertial navigation systems). The time-integrated image sharpness is found to be maximal for correctly determined angular, range, and time offsets. The requirements to form an operational system from the proposed method are also discussed.

The protoplanetary system HD 100546 in Hα polarized light from SPHERE/ZIMPOL

HD 100546 is one of the few known pre-main-sequence stars that may host a planetary system in its disk. We analyze new VLT/SPHERE/ZIMPOL polarimetric images of HD 100546 with filters in H$\alpha$ and the adjacent continuum. We have probed the disk gap and the surface layers of the outer disk, covering a region < 500 mas (< 55 au at 109 pc) from the star, at an angular resolution of ~ 20 mas. Our data show an asymmetry: the SE and NW regions of the outer disk are more polarized than the SW and NE. This can be explained from a preferential scattering angle close to 90$^o$, consistent with previous polarization images. The outer disk extends from 13 $\pm$ 2 to 45 $\pm$ 9 au, with a position angle and inclination of 137 $\pm$ 5$^o$ and 44 $\pm$ 8$^o$. The comparison with previous estimates suggests that the disk inclination could increase with the stellocentric distance, although the different measurements are still consistent within the error bars. In addition, no direct signature of the innermost candidate companion is detected from polarimetry, confirming recent results based on intensity imagery. We set an upper limit to its mass accretion rate < 10$^{-8}$ M$_{\odot}$/yr for a sub-stellar mass of 15M$_{Jup}$. Finally, we report the first detection (> 3$\sigma$) of a ~ 20 au bar-like structure that crosses the gap through the central region of HD 100546. It is tentatively suggested that the bar could be dust dragged by infalling gas that radially flows from the outer disk. This could represent an exceptional case in which a small-scale radial inflow is observed in a single system. If this scenario is confirmed, it could explain the presence of atomic gas in the inner disk that would otherwise accrete on to the central star on a timescale of a few months/years, as previously indicated from spectro-interferometric data, and could be related with additional (undetected) planets.

A three-class change detection methodology for SAR-data based on hypothesis testing and Markov Random field modelling

ABSTRACTThis study presents a new automatic change detection process chain based on bi-temporal co-registered and calibrated Sentinel-1 level-1 Interferometric Wide Ground Range Detected C-band synthetic aperture radar intensity imagery. The whole processor contains three main components: firstly, a preprocessing step is used to perform geometrical and radiometrical calibration. Secondly, an automatic coarse detection step is applied based on a statistical hypothesis test to obtain an initial classification. Thirdly, a post-classification step is introduced to optimize the initial classification result in the form of minimizing a global energy function defined on a Markov Random Field. In this study, a graph-cut algorithm is applied iteratively to solve the global optimization problem. At each iteration, the data energy function for the current classification is set by the logarithmic probability density function. The relevant parameters are estimated by the method of logarithmic cumulants. Supplemental data is presented to explain the formulae used in this study. Experiments are performed using a flood event which occurred in 2015 along the coastline of Greece near Kavala region and the Evros River at the border between Greece and Turkey. The proposed method shows a satisfying classification result with overall accuracy above 95% and kappa coefficient () above 0.87.

Using remote sensing to detect the polarized sunglint reflected from oil slicks beyond the critical angle

AbstractThe critical angle at which the brightness of oil slicks and oil‐free seawater is reversed occurs under sunglint and is often shown as an area of uncertainty due to different roughness and surface Fresnel reflection parameters. Consequently, differentiating oil slicks from the seawater in these areas using optical sensors is a challenge. Polarized optical remote sensing techniques provide complementary information for intensity imagery with different physical properties and, thus, possess the ability to resolve this difficult problem. In the polarized reflectance model, the degree of linear polarization (DOLP) of sunglint depends on accurately knowing the Stokes parameter for the reflected light, and varies with the refractive index of the surface layer and viewing angles. For the polarized detection of oil slicks, the highest sensitivity of the DOLP to the refractive index is located within the specular reflection direction where the sum of the solar and sensor zenith angles is 82.6°. The modeled results clearly indicate that the DOLP of oil slicks is weaker in comparison with oil‐free seawater under sunglint. Using measurements from the space‐borne Polarization and Anisotropy of Reflectances for Atmospheric Sciences coupled with Observations from a Lidar (PARASOL) over the Deepwater Horizon oil spill in the Gulf of Mexico, we illustrate that the PARASOL‐derived DOLP difference between the oil spill and seawater is obvious and is in accordance with the modeled results. These preliminary results suggest that the potential of multiangle measurement and feasibility of deriving refractive index of ocean surface from space‐borne sensors need further researches.

Discriminative Features Based on Two Layers Sparse Learning for Glacier Area Classification Using SAR Intensity Imagery

Accurate and instant information about changes of snow and glaciers covered areas plays a vital role in hydrological and climatological research and implications. Among all the observation methods, spaceborne remote sensing has a great advantage in monitoring the glaciers located in cold high-altitude regions and inaccessible areas on a large scale. Unlike optical sensors, the synthetic aperture radar (SAR) sensor can obtain images with low limitations in terms of weather phenomena and illumination as some glaciers frequently located in cloudy regions. In this paper, we propose a multiclasses classification method for large area glacier using spaceborne single-polarimetric SAR intensity image. The proposed method takes advantage of the discrimination ability of sparse representations of features, based on which a linear classifier called supervised neighborhood embedding is constructed. Finally, we develop a gradient descent method to alternatively update the dictionary and projection matrix. Two study areas are chosen to represent the discriminative characteristics of glaciers. In the Taku glacier in Alaska, compared to the state-of-the-art methods, our proposed method achieved a suitable performance with the overall classification accuracy of ${\text{90.34}}\%$ , and especially for bare ice of ${\text{91.38}}\%$ . In the Baltoro glacier in Karakoram characterized by high-relief topography and thick debris cover, the overall accuracy of ${\text{72.63}}\%$ and debris accuracy of $\text{90.14}\%$ are obtained.

Complex Spiral Structure in the HD 100546 Transitional Disk as Revealed by GPI and MagAO

Abstract We present optical and near-infrared high-contrast images of the transitional disk HD 100546 taken with the Magellan Adaptive Optics system (MagAO) and the Gemini Planet Imager (GPI). GPI data include both polarized intensity and total intensity imagery, and MagAO data are taken in Simultaneous Differential Imaging mode at Hα. The new GPI H-band total intensity data represent a significant enhancement in sensitivity and field rotation compared to previous data sets and enable a detailed exploration of substructure in the disk. The data are processed with a variety of differential imaging techniques (polarized, angular, reference, and simultaneous differential imaging) in an attempt to identify the disk structures that are most consistent across wavelengths, processing techniques, and algorithmic parameters. The inner disk cavity at 15 au is clearly resolved in multiple data sets, as are a variety of spiral features. While the cavity and spiral structures are identified at levels significantly distinct from the neighboring regions of the disk under several algorithms and with a range of algorithmic parameters, emission at the location of HD 100546 “c” varies from point-like under aggressive algorithmic parameters to a smooth continuous structure with conservative parameters, and is consistent with disk emission. Features identified in the HD 100546 disk bear qualitative similarity to computational models of a moderately inclined two-armed spiral disk, where projection effects and wrapping of the spiral arms around the star result in a number of truncated spiral features in forward-modeled images.

The Shadow Knows: Using Shadows to Investigate the Structure of the Pretransitional Disk of HD 100453

Abstract We present Gemini Planet Imager polarized intensity imagery of HD 100453 in Y, J, and K1 bands that reveals an inner gap (9–18 au), an outer disk (18–39 au) with two prominent spiral arms, and two azimuthally localized dark features that are also present in Spectro-Polarimetric High-contrast Exoplanet REsearch (SPHERE) total intensity images. Spectral energy distribution fitting further suggests that the radial gap extends to 1 au. The narrow, wedge-like shape of the dark features appears similar to predictions of shadows cast by an inner disk that is misaligned with respect to the outer disk. Using the Monte Carlo radiative transfer code HOCHUNCK3D, we construct a model of the disk that allows us to determine its physical properties in more detail. From the angular separation of the features, we measure the difference in inclination between the disks (45°) and their major axes, PA = 140° east of north for the outer disk, and 100° for the inner disk. We find an outer-disk inclination of 25° ± 10° from face-on, in broad agreement with the Wagner et al. measurement of 34°. SPHERE data in J and H bands indicate a reddish disk, which indicates that HD 100453 is evolving into a young debris disk.

Extraction and Classification of Road Markings Using Mobile Laser Scanning Point Clouds

This study aims at building a robust method for semiautomated information extraction of pavement markings detected from mobile laser scanning (MLS) point clouds. The proposed workflow consists of three components: 1) preprocessing, 2) extraction, and 3) classification. In preprocessing, the three-dimensional (3-D) MLS point clouds are converted into radiometrically corrected and enhanced two-dimensional (2-D) intensity imagery of the road surface. Then, the pavement markings are automatically extracted with the intensity using a set of algorithms, including Otsu's thresholding, neighbor-counting filtering, and region growing. Finally, the extracted pavement markings are classified with the geometric parameters by using a manually defined decision tree. A study was conducted by using the MLS dataset acquired in Xiamen, Fujian, China. The results demonstrated that the proposed workflow and method can achieve 92% in completeness, 95% in correctness, and 94% in F -score.

Estimation of glacier surface motion by robust phase correlation and point like features of SAR intensity images

For monitoring of glacier surface motion in pole and alpine areas, radar remote sensing is becoming a popular technology accounting for its specific advantages of being independent of weather conditions and sunlight. In this paper we propose a method for glacier surface motion monitoring using phase correlation (PC) based on point-like features (PLF). We carry out experiments using repeat-pass TerraSAR X-band (TSX) and Sentinel-1 C-band (S1C) intensity images of the Taku glacier in Juneau icefield located in southeast Alaska. The intensity imagery is first filtered by an improved adaptive refined Lee filter while the effect of topographic reliefs is removed via SRTM-X DEM. Then, a robust phase correlation algorithm based on singular value decomposition (SVD) and an improved random sample consensus (RANSAC) algorithm is applied to sequential PLF pairs generated by correlation using a 2D sinc function template. The approaches for glacier monitoring are validated by both simulated SAR data and real SAR data from two satellites. The results obtained from these three test datasets confirm the superiority of the proposed approach compared to standard correlation-like methods. By the use of the proposed adaptive refined Lee filter, we achieve a good balance between the suppression of noise and the preservation of local image textures. The presented phase correlation algorithm shows the accuracy of better than 0.25 pixels, when conducting matching tests using simulated SAR intensity images with strong noise. Quantitative 3D motions and velocities of the investigated Taku glacier during a repeat-pass period are obtained, which allows a comprehensive and reliable analysis for the investigation of large-scale glacier surface dynamics.

CONSTRAINING THE MOVEMENT OF THE SPIRAL FEATURES AND THE LOCATIONS OF PLANETARY BODIES WITHIN THE AB AUR SYSTEM

ABSTRACT We present a new analysis of multi-epoch, H-band, scattered light images of the AB Aur system. We use a Monte Carlo radiative transfer code to simultaneously model the system’s spectral energy distribution (SED) and H-band polarized intensity (PI) imagery. We find that a disk-dominated model, as opposed to one that is envelope-dominated, can plausibly reproduce AB Aur’s SED and near-IR imagery. This is consistent with previous modeling attempts presented in the literature and supports the idea that at least a subset of AB Aur’s spirals originate within the disk. In light of this, we also analyzed the movement of spiral structures in multi-epoch H-band total light and PI imagery of the disk. We detect no significant rotation or change in spatial location of the spiral structures in these data, which span a 5.8-year baseline. If such structures are caused by disk–planet interactions, the lack of observed rotation constrains the location of the orbit of planetary perturbers to be &gt;47 au.

RAPID INSPECTION OF PAVEMENT MARKINGS USING MOBILE LIDAR POINT CLOUDS

Abstract. This study aims at building a robust semi-automated pavement marking extraction workflow based on the use of mobile LiDAR point clouds. The proposed workflow consists of three components: preprocessing, extraction, and classification. In preprocessing, the mobile LiDAR point clouds are converted into the radiometrically corrected intensity imagery of the road surface. Then the pavement markings are automatically extracted with the intensity using a set of algorithms, including Otsu’s thresholding, neighbor-counting filtering, and region growing. Finally, the extracted pavement markings are classified with the geometric parameters using a manually defined decision tree. Case studies are conducted using the mobile LiDAR dataset acquired in Xiamen (Fujian, China) with different road environments by the RIEGL VMX-450 system. The results demonstrated that the proposed workflow and our software tool can achieve 93% in completeness, 95% in correctness, and 94% in F-score when using Xiamen dataset.

RAPID INSPECTION OF PAVEMENT MARKINGS USING MOBILE LIDAR POINT CLOUDS

This study aims at building a robust semi-automated pavement marking extraction workflow based on the use of mobile LiDAR point clouds. The proposed workflow consists of three components: preprocessing, extraction, and classification. In preprocessing, the mobile LiDAR point clouds are converted into the radiometrically corrected intensity imagery of the road surface. Then the pavement markings are automatically extracted with the intensity using a set of algorithms, including Otsu’s thresholding, neighbor-counting filtering, and region growing. Finally, the extracted pavement markings are classified with the geometric parameters using a manually defined decision tree. Case studies are conducted using the mobile LiDAR dataset acquired in Xiamen (Fujian, China) with different road environments by the RIEGL VMX-450 system. The results demonstrated that the proposed workflow and our software tool can achieve 93% in completeness, 95% in correctness, and 94% in F-score when using Xiamen dataset.

ESTIMATING SEA ICE PARAMETERS FROM MULTI-LOOK SAR IMAGES USING FIRST- AND SECOND-ORDER VARIOGRAMS

The spatial structures revealed in SAR intensity imagery provide essential information characterizing the natural variation processes of sea ice. This paper proposes a new method to extract the spatial structures of sea ice based on two spatial stochastic models. One is a multi-Gamma model, which characterizes continuous variations corresponding to ice-free area or the background. The other is a Poisson line mosaic model, which characterizes the regional variations of sea ice with different types. The linear combination of the two models builds the mixture model to represent spatial structures of sea ice within SAR intensity imagery. To estimate different sea ice parameters, such as its concentration, scale etc. We define two kinds of geostatistic metrics, theoretical first- and second-order variograms. Their experimental alternatives can be calculated from the SAR intensity imagery directly, then the parameters of the mixture model are estimated through fitting the theoretical variograms to the experimental ones, and by comparing the estimated parameters to the egg code, it is verified that the estimated parameters can indicate sea ice structure information showing in the egg code. The proposed method is applied to simulated images and Radarsat-1 images. The results of the experiments show that the proposed method can estimate the sea ice concentration and floe size accurately and stably within SAR testing images.

ESTIMATING SEA ICE PARAMETERS FROM MULTI-LOOK SAR IMAGES USING FIRST- AND SECOND-ORDER VARIOGRAMS

Abstract. The spatial structures revealed in SAR intensity imagery provide essential information characterizing the natural variation processes of sea ice. This paper proposes a new method to extract the spatial structures of sea ice based on two spatial stochastic models. One is a multi-Gamma model, which characterizes continuous variations corresponding to ice-free area or the background. The other is a Poisson line mosaic model, which characterizes the regional variations of sea ice with different types. The linear combination of the two models builds the mixture model to represent spatial structures of sea ice within SAR intensity imagery. To estimate different sea ice parameters, such as its concentration, scale etc. We define two kinds of geostatistic metrics, theoretical first- and second-order variograms. Their experimental alternatives can be calculated from the SAR intensity imagery directly, then the parameters of the mixture model are estimated through fitting the theoretical variograms to the experimental ones, and by comparing the estimated parameters to the egg code, it is verified that the estimated parameters can indicate sea ice structure information showing in the egg code. The proposed method is applied to simulated images and Radarsat-1 images. The results of the experiments show that the proposed method can estimate the sea ice concentration and floe size accurately and stably within SAR testing images.