Block-Active ADMM to Minimize NMF with Bregman Divergences.

Over the last ten years, there has been a significant interest in employing nonnegative matrix factorization (NMF) to reduce dimensionality to enable a more efficient clustering analysis in machine learning. This technique has been applied in various image processing applications within the fields of computer vision and sensor-based systems. Many algorithms exist to solve the NMF problem. Among these algorithms, the alternating direction method of multipliers (ADMM) and its variants are one of the most popular methods used in practice. In this paper, we propose a block-active ADMM method to minimize the NMF problem with general Bregman divergences. The subproblems in the ADMM are solved iteratively by a block-coordinate-descent-type (BCD-type) method. In particular, each block is chosen directly based on the stationary condition. As a result, we are able to use much fewer auxiliary variables and the proposed algorithm converges faster than the previously proposed algorithms. From the theoretical point of view, the proposed algorithm is proved to converge to a stationary point sublinearly. We also conduct a series of numerical experiments to demonstrate the superiority of the proposed algorithm.

Omnidirectional-Sensor-System-Based Texture Noise Correction in Large-Scale 3D Reconstruction.

The evolution of cameras and LiDAR has propelled the techniques and applications of three-dimensional (3D) reconstruction. However, due to inherent sensor limitations and environmental interference, the reconstruction process often entails significant texture noise, such as specular highlight, color inconsistency, and object occlusion. Traditional methodologies grapple to mitigate such noise, particularly in large-scale scenes, due to the voluminous data produced by imaging sensors. In response, this paper introduces an omnidirectional-sensor-system-based texture noise correction framework for large-scale scenes, which consists of three parts. Initially, we obtain a colored point cloud with luminance value through LiDAR points and RGB images organization. Next, we apply a voxel hashing algorithm during the geometry reconstruction to accelerate the computation speed and save the computer memory. Finally, we propose the key innovation of our paper, the frame-voting rendering and the neighbor-aided rendering mechanisms, which effectively eliminates the aforementioned texture noise. From the experimental results, the processing rate of one million points per second shows its real-time applicability, and the output figures of texture optimization exhibit a significant reduction in texture noise. These results indicate that our framework has advanced performance in correcting multiple texture noise in large-scale 3D reconstruction.

Image Quality Assessment for Realistic Zoom Photos.

New CMOS imaging sensor (CIS) techniques in smartphones have helped user-generated content dominate our lives over traditional DSLRs. However, tiny sensor sizes and fixed focal lengths also lead to more grainy details, especially for zoom photos. Moreover, multi-frame stacking and post-sharpening algorithms would produce zigzag textures and over-sharpened appearances, for which traditional image-quality metrics may over-estimate. To solve this problem, a real-world zoom photo database is first constructed in this paper, which includes 900 tele-photos from 20 different mobile sensors and ISPs. Then we propose a novel no-reference zoom quality metric which incorporates the traditional estimation of sharpness and the concept of image naturalness. More specifically, for the measurement of image sharpness, we are the first to combine the total energy of the predicted gradient image with the entropy of the residual term under the framework of free-energy theory. To further compensate for the influence of over-sharpening effect and other artifacts, a set of model parameters of mean subtracted contrast normalized (MSCN) coefficients are utilized as the natural statistics representatives. Finally, these two measures are combined linearly. Experimental results on the zoom photo database demonstrate that our quality metric can achieve SROCC and PLCC over 0.91, while the performance of single sharpness or naturalness index is around 0.85. Moreover, compared with the best tested general-purpose and sharpness models, our zoom metric outperforms them by 0.072 and 0.064 in SROCC, respectively.

UAV hyperspectral remote sensor images for mango plant disease and pest identification using MD-FCM and XCS-RBFNN.

To diminish disease transmission together with promoting effective management techniques, it is crucial to monitor plant health and detect pathogens earlier. The initial part in reducing losses sourced from plant diseases is to make an accurate and earlier identification. Thus, the usage of unmanned aerial vehicle (UAV) hyperspectral imaging (HSI) sensors for surveying and assessing crops, orchards, and forests has rapidly elevated over the last decade, particularly for the stress management like water, diseases, nutrition deficits, and pests. Using Minkowski Distance-based Fuzzy C Means (MD-FCM) clustering and Xavier initialization-adapted Cosine Similarity-induced Radial Bias Function Neural Network (XCS-RBFNN) techniques, a UAV HS imaging remote sensor for Spatial and Temporal Resolution (STR) of mango plant disease and pest identification is proposed in this scheme. Collecting the input UAV source (image or video) is eventuated initially along with the Region of Interest (ROI) calculated which is followed by preprocessing. Leaf segmentation is eventuated using Logistic U-net after preprocessing. Next, MD-FCM performs clustering to cluster the diseased leaves and pests individually. The disease and pest characteristics are then retrieved separately and classified further. The requisite features are then chosen from the retrieved features utilizing the Levy Flight Distribution-produced Butterfly Optimization Algorithm (LFD-BOA). Finally, the XCS-RBFNN classifier is utilized to categorize the various diseases together with pests found in the UAV input source using the chosen features. The proposed framework's experimental findings are then compared to some prevailing schemes, with the results revealing that the proposed work outperforms other benchmark techniques.

Estimation of Lens Stray Light with Regard to the Incapacitation of Imaging Sensors-Part 2: Validation.

Recently, we developed a simple theoretical model for the estimation of the irradiance distribution at the focal plane of commercial off-the-shelf (COTS) camera lenses in case of laser illumination. The purpose of such a model is to predict the incapacitation of imaging sensors when irradiated by laser light. The model is based on closed-form equations that comprise mainly standard parameters of the laser dazzle scenario and those of the main devices involved (laser source, camera lens and imaging sensor). However, the model also includes three non-standard parameters, which describe the scattering of light within the camera lens. In previous work, we have performed measurements to derive these typically unknown scatter parameters for a collection of camera lenses of the Double-Gauss type. In this publication, we compare calculations based on our theoretical model and the measured scatter parameters with the outcome of stray light simulations performed with the optical design software FRED in order to validate the reliability of our theoretical model and of the derived scatter parameters.

UAV-Based Hyperspectral Monitoring Using Push-Broom and Snapshot Sensors: A Multisite Assessment for Precision Viticulture Applications.

Hyperspectral aerial imagery is becoming increasingly available due to both technology evolution and a somewhat affordable price tag. However, selecting a proper UAV + hyperspectral sensor combo to use in specific contexts is still challenging and lacks proper documental support. While selecting an UAV is more straightforward as it mostly relates with sensor compatibility, autonomy, reliability and cost, a hyperspectral sensor has much more to be considered. This note provides an assessment of two hyperspectral sensors (push-broom and snapshot) regarding practicality and suitability, within a precision viticulture context. The aim is to provide researchers, agronomists, winegrowers and UAV pilots with dependable data collection protocols and methods, enabling them to achieve faster processing techniques and helping to integrate multiple data sources. Furthermore, both the benefits and drawbacks of using each technology within a precision viticulture context are also highlighted. Hyperspectral sensors, UAVs, flight operations, and the processing methodology for each imaging type' datasets are presented through a qualitative and quantitative analysis. For this purpose, four vineyards in two countries were selected as case studies. This supports the extrapolation of both advantages and issues related with the two types of hyperspectral sensors used, in different contexts. Sensors' performance was compared through the evaluation of field operations complexity, processing time and qualitative accuracy of the results, namely the quality of the generated hyperspectral mosaics. The results shown an overall excellent geometrical quality, with no distortions or overlapping faults for both technologies, using the proposed mosaicking process and reconstruction. By resorting to the multi-site assessment, the qualitative and quantitative exchange of information throughout the UAV hyperspectral community is facilitated. In addition, all the major benefits and drawbacks of each hyperspectral sensor regarding its operation and data features are identified. Lastly, the operational complexity in the context of precision agriculture is also presented.

Spinal cord injury alters microRNA and CD81+ exosome levels in plasma extracellular nanoparticles with neuroinflammatory potential.

Extracellular vesicles (EVs) have been implicated mechanistically in the pathobiology of neurodegenerative disorders, including central nervous system injury. However, the role of EVs in spinal cord injury (SCI) has received limited attention to date. Moreover, technical limitations related to EV isolation and characterization methods can lead to misleading or contradictory findings. Here, we examined changes in plasma EVs after mouse SCI at multiple timepoints (1d, 3d, 7d, 14d) using complementary measurement techniques. Plasma EVs isolated by ultracentrifugation (UC) were decreased at 1d post-injury, as shown by nanoparticle tracking analysis (NTA), and paralleled an overall reduction in total plasma extracellular nanoparticles. Western blot (WB) analysis of UC-derived plasma EVs revealed increased expression of the tetraspanin exosome marker, CD81, between 1d and 7d post-injury. To substantiate these findings, we performed interferometric and fluorescence imaging of single, tetraspanin EVs captured directly from plasma with ExoView®. Consistent with WB, we observed significantly increased plasma CD81+ EV count and cargo at 1d post-injury. The majority of these tetraspanin EVs were smaller than 50 nm based on interferometry and were insufficiently resolved by flow cytometry-based detection. At the injury site, there was enhanced expression of EV biogenesis proteins that were also detected in EVs directly isolated from spinal cord tissue by WB. Surface expression of tetraspanins CD9 and CD63 increased in multiple cell types at the injury site; however, astrocyte CD81 expression uniquely decreased, as demonstrated by flow cytometry. UC-isolated plasma EV microRNA cargo was also significantly altered at 1d post-injury with changes similar to that reported in EVs released by astrocytes after inflammatory stimulation. When injected into the lateral ventricle, plasma EVs from SCI mice increased both pro- and anti-inflammatory gene as well as reactive astrocyte gene expression in the brain cortex. These studies provide the first detailed characterization of plasma EV dynamics after SCI and suggest that plasma EVs may be involved in posttraumatic brain inflammation.

Characterization of Dielectric Oil with a Low-Cost CMOS Imaging Sensor and a New Electric Permittivity Matrix Using the 3D Cell Method.

In this paper, a new method for characterizing the dielectric breakdown voltage of dielectric oils is presented, based on the IEC 60156 international standard. In this standard, the effective value of the dielectric breakdown voltage is obtained, but information is not provided on the distribution of Kelvin forces an instant before the dynamic behavior of the arc begins or the state of the gases that are produced an instant after the moment of appearance of the electric arc in the oil. In this paper, the behavior of the oil before and after the appearance of the electric arc is characterized by combining a low-cost CMOS imaging sensor and a new matrix of electrical permittivity associated with the dielectric oil, using the 3D cell method. In this way, we also predict the electric field before and after the electric rupture. The error compared to the finite element method is less than 0.36%. In addition, a new method is proposed to measure the kinematic viscosity of dielectric oils. Using a low-cost imaging sensor, the distribution of bubbles is measured, together with their diameters and their rates of ascent after the electric arc occurs. This method is verified using ASTM standards and data provided by the oil manufacturer. The results of these tests can be used to prevent incipient failures and evaluate preventive maintenance processes such as transformer oil replacement or recovery.

On-Orbit Geometric Calibration from the Relative Motion of Stars for Geostationary Cameras.

Affected by the vibrations and thermal shocks during launch and the orbit penetration process, the geometric positioning model of the remote sensing cameras measured on the ground will generate a displacement, affecting the geometric accuracy of imagery and requiring recalibration. Conventional methods adopt the ground control points (GCPs) or stars as references for on-orbit geometric calibration. However, inescapable cloud coverage and discontented extraction algorithms make it extremely difficult to collect sufficient high-precision GCPs for modifying the misalignment of the camera, especially for geostationary satellites. Additionally, the number of the observed stars is very likely to be inadequate for calibrating the relative installations of the camera. In terms of the problems above, we propose a novel on-orbit geometric calibration method using the relative motion of stars for geostationary cameras. First, a geometric calibration model is constructed based on the optical system structure. Then, we analyze the relative motion transformation of the observed stars. The stellar trajectory and the auxiliary ephemeris are used to obtain the corresponding object vector for correcting the associated calibration parameters iteratively. Experimental results evaluated on the data of a geostationary experiment satellite demonstrate that the positioning errors corrected by this proposed method can be within ±2.35 pixels. This approach is able to effectively calibrate the camera and improve the positioning accuracy, which avoids the influence of cloud cover and overcomes the great dependence on the number of the observed stars.

Simultaneous evaluation of multiple microarray surface chemistries through real-time interferometric imaging.

Surface chemistry is a crucial aspect for microarray modality biosensor development. The immobilization capability of the functionalized surface is indeed a limiting factor for the final yield of the binding reaction. In this work, we were able to simultaneously compare the functionality of protein ligands that were locally immobilized on different polymers, while on the same solid support, therefore demonstrating a new way of multiplexing. Our goal was to investigate, in a single experiment, both the immobilization efficiency of a group of reactive polymers and the resulting affinity of the tethered molecules. This idea was demonstrated by spotting many reactive polymers on a Si/SiO2 chip and depositing the molecular probes on the spots immediately after. As a proof of concept, we focused on which polymers would better immobilize a model protein (α-Lactalbumin) and a peptide (LAC-1). We successfully showed that this protocol is applicable to proteins and peptides with a good efficiency. By means of real-time binding measurements performed with the interferometric reflectance imaging sensor (IRIS), local functionalization proved to be comparable to the classical flat coating solution. The final outcome highlights the multiplexing power of this method: first, it allows to characterize dozens of polymers at once. Secondly, it removes the limitation, related to coated surfaces, that only molecules with the same functional groups can be tethered to the same solid support. By applying this protocol, many types of molecules can be studied simultaneously and immobilization for each probe can be individually optimized.

Estimation of Lens Stray Light with Regard to the Incapacitation of Imaging Sensors.

We present our efforts on estimating light scattering characteristics from commercial off-the-shelf (COTS) camera lenses in order to deduce thereof a set of generic scattering parameters valid for a specific lens class (double Gauss lenses). In previous investigations, we developed a simplified theoretical light scattering model to estimate the irradiance distribution in the focal plane of a camera lens. This theoretical model is based on a 3-parameter bidirectional scattering distribution function (BSDF), which describes light scattering from rough surfaces of the optical elements. Ordinarily, the three scatter parameters of the BSDF are not known for COTS camera lenses, which makes it necessary to assess them by own experiments. Besides the experimental setup and the measurement process, we present in detail the subsequent data exploitation. From measurements on seven COTS camera lenses, we deduced a generic set of scatter parameters. For a deeper analysis, the results of our measurements have also been compared with the output of an optical engineering software. Together with our theoretical model, now stray light calculations can be accomplished even then, when specific scatter parameters are not available from elsewhere. In addition, the light scattering analyses also allow considering the glare vulnerability of optical systems in terms of laser safety.

Real-Time Detection on SPAD Value of Potato Plant Using an In-Field Spectral Imaging Sensor System.

In this study, a SPAD value detection system was developed based on a 25-wavelength spectral sensor to give a real-time indication of the nutrition distribution of potato plants in the field. Two major advantages of the detection system include the automatic segmentation of spectral images and the real-time detection of SPAD value, a recommended indicating parameter of chlorophyll content. The modified difference vegetation index (MDVI) linking the Otsu algorithm (OTSU) and the connected domain-labeling (CDL) method (MDVI-OTSU-CDL) is proposed to accurately extract the potato plant. Additionally, the segmentation accuracy under different modified coefficients of MDVI was analyzed. Then, the reflectance of potato plants was extracted by the segmented mask images. The partial least squares (PLS) regression was employed to establish the SPAD value detection model based on sensitive variables selected using the uninformative variable elimination (UVE) algorithm. Based on the segmented spectral image and the UVE-PLS model, the visualization distribution map of SPAD value was drawn by pseudo-color processing technology. Finally, the testing dataset was employed to measure the stability and practicality of the developed detection system. This study provides a powerful support for the real-time detection of SPAD value and the distribution of crops in the field.

Insulation Failure Quantification Based on the Energy of Digital Images Using Low-Cost Imaging Sensors.

Insulation faults in high-voltage applications often generate partial discharges (PDs) accompanied by corona activity, optical radiation mainly in the ultraviolet (UV) and visible bands. Recent developments in low-cost, small-size, and high-resolution visible imaging sensors, which are also partially sensitive to the UV spectral region, are gaining attention due to their many industrial applications. This paper proposes a method for early PD detection by using digital imaging sensors, which allows the severity of insulation faults to be assessed. The electrical power dissipated by the PDs is correlated to the energy of the acquired visible images, and thus, the severity of insulation faults is determined from the energy of the corona effect. A criterion to quantify the severity of insulation faults based on the energy of the corona images is proposed. To this end, the point-to-plane gap configuration is analyzed in a low-pressure chamber, where digital image photographs of the PDs are taken and evaluated under different pressure conditions ranging from 10 to 100 kPa, which cover the typical pressure range of aeronautic applications. The use of digital imaging sensors also allows an early detection, location and quantification of the PD activity, and thus assessing the severity of insulation faults to perform predictive maintenance tasks, while enabling the cost and complexity of the instrumentation to be reduced. Although the approach proposed in this paper has been applied to detect PDs in aeronautic applications, it can be applied to many other high-voltage applications susceptible of PD occurrence.

Experimental Study of Visual Corona under Aeronautic Pressure Conditions Using Low-Cost Imaging Sensors.

Visual corona tests have been broadly applied for identifying the critical corona points of diverse high-voltage devices, although other approaches based on partial discharge or radio interference voltage measurements are also widely applied to detect corona activity. Nevertheless, these two techniques must be applied in screened laboratories, which are scarce and expensive, require sophisticated instrumentation, and typically do not allow location of the discharge points. This paper describes the detection of the visual corona and location of the critical corona points of a sphere-plane gap configurations under different pressure conditions ranging from 100 to 20 kPa, covering the pressures typically found in aeronautic environments. The corona detection is made with a low-cost CMOS imaging sensor from both the visible and ultraviolet (UV) spectrum, which allows detection of the discharge points and their locations, thus significantly reducing the complexity and costs of the instrumentation required while preserving the sensitivity and accuracy of the measurements. The approach proposed in this paper can be applied in aerospace applications to prevent the arc tracking phenomenon, which can lead to catastrophic consequences since there is not a clear protection solution, due to the low levels of leakage current involved in the pre-arc phenomenon.

Terminal Sliding Mode Control with a Novel Reaching Law and Sliding Mode Disturbance Observer for Inertial Stabilization Imaging Sensor.

High-performance control of inertial stabilization imaging sensors (ISISs) is always challenging because of the complex nonlinearities induced by friction, mass imbalance, and external disturbances. To overcome this problem, a terminal sliding mode controller (TSMC) based on a novel exponential reaching law (NERL) method with a high-order terminal sliding mode observer (HOTSMO) is suggested. First, the TSMC based on NERL is adopted to improve system performance. The NERL incorporates the power term and switching gain term of the system state variables into the conventional exponential reaching law, and the convergent speed of the TSMC is accelerated. Then, an HOTSMO is designed, which considers the speed and lumped disturbances of the system as the observation object. The estimated disturbance is then provided as a compensation for the controller, which enhances the disturbance rejection ability of the system. Comparative simulation and experimental results show that the proposed method achieves the best tracking performance and the strongest robustness than PID and the traditional TSMC methods.

Modeling of the Return Current in a Light-Addressable Potentiometric Sensor.

A light-addressable potentiometric sensor (LAPS) is a chemical sensor with a field-effect structure based on semiconductor. Its response to the analyte concentration is read out in the form of a photocurrent generated by illuminating the semiconductor with a modulated light beam. As stated in its name, a LAPS is capable of spatially resolved measurement using a scanning light beam. Recently, it has been pointed out that a part of the signal current is lost by the return current due to capacitive coupling between the solution and the semiconductor, which may seriously affect the sensor performance such as the signal-to-noise ratio, the spatial resolution, and the sensitivity. In this study, a circuit model for the return current is proposed to study its dependence on various parameters such as the diameter of contact area, the modulation frequency, the specific conductivity of the solution, and the series resistance of the circuit. It is suggested that minimization of the series resistance of the circuit is of utmost importance in order to avoid the influence of the return current. The results of calculation based on this model are compared with experimental results, and its applicability and limitation are discussed.

Fast 2D Laser-Induced Fluorescence Spectroscopy Mapping of Rare Earth Elements in Rock Samples.

Due to the rapidly increasing use of energy-efficient technologies, the need for complex materials containing rare earth elements (REEs) is steadily growing. The high demand for REEs requires the exploration of new mineral deposits of these valuable elements, as recovery by recycling is still very low. Easy-to-deploy sensor technologies featuring high sensitivity to REEs are required to overcome limitations by traditional techniques, such as X-ray fluorescence. We demonstrate the ability of laser-induced fluorescence (LIF) to detect REEs rapidly in relevant geological samples. We introduce two-dimensional LIF mapping to scan rock samples from two Namibian REE deposits and cross-validate the obtained results by employing mineral liberation analysis (MLA) and hyperspectral imaging (HSI). Technique-specific parameters, such as acquisition speed, spatial resolution, and detection limits, are discussed and compared to established analysis methods. We also focus on the attribution of REE occurrences to mineralogical features, which may be helpful for the further geological interpretation of a deposit. This study sets the basis for the development of a combined mapping sensor for HSI and 2D LIF measurements, which could be used for drill-core logging in REE exploration, as well as in recovery plants.

A Visual Cortex-Inspired Imaging-Sensor Architecture and Its Application in Real-Time Processing.

For robots equipped with an advanced computer vision-based system, object recognition has stringent real-time requirements. When the environment becomes complicated and keeps changing, existing works (e.g., template-matching strategy and machine-learning strategy) are computationally expensive, compromising object recognition performance and even stability. In order to detect objects accurately, it is necessary to build an efficient imaging sensor architecture as the neural architecture. Inspired by the neural mechanism of primary visual cortex, this paper presents an efficient three-layer architecture and proposes an approach of constraint propagation examination to efficiently extract and process information (linear contour). Through applying this architecture in the preprocessing phase to extract lines, the running time of object detection is decreased dramatically because not only are all lines represented as very simple vectors, but also the number of lines is very limited. In terms of the second measure of improving efficiency, we apply a shape-based recognition method because it does not need any high-dimensional feature descriptor, long-term training, or time-expensive preprocessing. The final results perform well. It is proved that detection performance is good. The brain is the result of natural optimization, so we conclude that a visual cortex-inspired imaging sensor architecture can greatly improve the efficiency of information processing.

HoPE: Horizontal Plane Extractor for Cluttered 3D Scenes.

Extracting horizontal planes in heavily cluttered three-dimensional (3D) scenes is an essential procedure for many robotic applications. Aiming at the limitations of general plane segmentation methods on this subject, we present HoPE, a Horizontal Plane Extractor that is able to extract multiple horizontal planes in cluttered scenes with both organized and unorganized 3D point clouds. It transforms the source point cloud in the first stage to the reference coordinate frame using the sensor orientation acquired either by pre-calibration or an inertial measurement unit, thereby leveraging the inner structure of the transformed point cloud to ease the subsequent processes that use two concise thresholds for producing the results. A revised region growing algorithm named Z clustering and a principal component analysis (PCA)-based approach are presented for point clustering and refinement, respectively. Furthermore, we provide a nearest neighbor plane matching (NNPM) strategy to preserve the identities of extracted planes across successive sequences. Qualitative and quantitative evaluations of both real and synthetic scenes demonstrate that our approach outperforms several state-of-the-art methods under challenging circumstances, in terms of robustness to clutter, accuracy, and efficiency. We make our algorithm an off-the-shelf toolbox which is publicly available.

Intelligent Luminance Control of Lighting Systems Based on Imaging Sensor Feedback.

An imaging sensor-based intelligent Light Emitting Diode (LED) lighting system for desk use is proposed. In contrast to the traditional intelligent lighting system, such as the photosensitive resistance sensor-based or the infrared sensor-based system, the imaging sensor can realize a finer perception of the environmental light; thus it can guide a more precise lighting control. Before this system works, first lots of typical imaging lighting data of the desk application are accumulated. Second, a series of subjective and objective Lighting Effect Evaluation Metrics (LEEMs) are defined and assessed for these datasets above. Then the cluster benchmarks of these objective LEEMs can be obtained. Third, both a single LEEM-based control and a multiple LEEMs-based control are developed to realize a kind of optimal luminance tuning. When this system works, first it captures the lighting image using a wearable camera. Then it computes the objective LEEMs of the captured image and compares them with the cluster benchmarks of the objective LEEMs. Finally, the single LEEM-based or the multiple LEEMs-based control can be implemented to get a kind of optimal lighting effect. Many experiment results have shown the proposed system can tune the LED lamp automatically according to environment luminance changes.