Spectral Drift Research Articles

Tensorized Principal Component Alignment: A Unified Framework for Multimodal High-Resolution Images Classification

High-resolution (HR) remote sensing (RS) imaging opens the door to very accurate geometrical analysis for objects. However, it is difficult to simultaneous use massive HR RS images in practical applications, because these HR images are often collected in different multimodal conditions (multisource, multiarea, multitemporal, multiresolution, and multiangular) and learning method trained for one situation is difficult to use for others. The key problem is how to simultaneously tackle three main problems: spectral drift, spatial deformation, and band inconsistency. To deal with these problems, we propose an unsupervised tensorized principal component alignment framework in this paper. In this framework, local spatial–spectral patch data are used as basic units in order to achieve simultaneously multidimensional alignment. This framework seeks a domain-invariant tensor feature space by learning multilinear mapping functions which align the source tensor subspace with the target tensor subspace on different dimensions. In addition, an approach based on the Mahalanobis distance for dimensionality estimation of tensor subspace is proposed to determine best sizes of the aligned tensor subspace for reducing computational complexity. HR images from GF-1, GF-2, DEIMOS-2, WorldView-2, and WorldView-3 satellites are used to evaluate the performance. The experimental results show the following two points: first, the proposed alignment framework for multimodal HR images not only can align the different multimodal data more accurately than existing state-of-the-art domain adaptation methods, but also has a fast and simple procedure for large-scale data situation which is caused by HR imaging. Second, the proposed tensor dimensionality estimation method is an efficient technology for seeking the intrinsic dimensions of high-order data.

Identification and treatment approach for spectral baseline distortion in processing of gas analysis online by Fourier transform infrared spectroscopy

ABSTRACTAddressing the problem of baseline drift or distortion in Fourier transform infrared spectra when analyzing gas continuously at scene, a novel and simple method is proposed to identify and treat spectral baseline drift. Firstly, the spectral baseline was corrected with the common approach followed by standard gas analysis. Then, absorbance of analytes was reconstructed and reduced from the corrected spectrum. Finally, a Fourier transform was performed with the difference spectrum, and the first four-line strengths were used as feature variables to determine if baseline distortion occurred. In the end of this paper, this approach has been applied in the field of wellhead gas logging with Fourier transform infrared. The application results showed that all the spikes had been removed from the raw logging curves, confirming that this approach could be used to identify spectral baseline distortion accurately.

Thermal stability of volume Bragg gratings in chloride photo-thermo-refractive glass after femtosecond laser bleaching.

We demonstrate that the Joule heating of the volume Bragg grating recorded in chloride photo-thermo-refractive glass can be suppressed by bleaching the silver nanoparticles with intense ultrashort laser pulses. Measurement of the bleached grating angular selectivity showed that, at the signal wavelength at 972nm, the spectral drift is 0.5nm at the CW laser diode beam intensity as high as 145 W/cm2. Thus, the bleaching of silver nanoparticles results in the improved thermal stability of transmission gratings, allowing one to employ them to control the powerful CW laser radiation.

Sources of the Multi-Lane Type II Solar Radio Burst on 5 November 2014

We report the well-observed event of a multi-lane type II solar radio burst with a combined analysis of radio dynamic spectra and radio and extreme-ultraviolet (EUV) imaging data. The burst is associated with an EUV wave driven by a coronal mass ejection (CME) that is accompanied by a GOES X-ray M7.9 flare on 5 November 2014. This type of event is rarely observed with such a complete data set. The type II burst presents three episodes (referred to as A, B, and C), characterized by a sudden change in spectral drift, and contains more than ten branches, including both harmonic-fundamental (H–F) pairs and split bands. The sources of the three episodes present a general outward propagating trend. There exists a significant morphology change from single source (Episode A) to double source (Episode B). Episode C maintains the double-source morphology at 150 MHz (no imaging data are available at a lower frequency). The double-source centroids are separated by $\sim300 ^{\prime\prime}$ to $500^{\prime\prime}$ . The southeastern (SE) source is likely the continuation of the source of Episode A since both are at the same section of the shock (i.e. the EUV wave) and close to each other. The northwestern (NW) source is coincident with (thus, possibly originates from) the interaction of the shock with a nearby mini-streamer-like structure. Comparing the simultaneously observed sources of the F and H branches of Episode A, we find that their centroids are separated by less than $200^{\prime \prime}$ . The centroids of the split bands of Episode B are cospatial within the observational uncertainties. This study shows the source evolution of a multi-lane type II burst and the source locations of different lanes relative to each other and to the EUV wave generated by a CME. The study indicates the intrinsic complexity underlying a type II dynamic spectrum.

Structured light illumination for order sorting in Echelle spectrometers.

We report on the operation of an echelle spectrometer under structured light illumination. Each diffraction order of the spectrometer is encoded with a certain periodic structure allowing for order sorting by numerical analysis after detection. In contrast to cross-dispersed echelle spectrometers, in this approach the orders overlap at the detection plane so that the spectral calibration can be performed easily with a single reference wavelength. This operational simplification makes it possible to measure simultaneously the light source under study and the calibration wavelength giving rise to a self-calibrated echelle spectrometer. In this way the device compensates for the spectral drift due to temporal changes of environmental conditions in real time. Our proposal can be useful in a large number of applications requiring moderate, high or very high resolving power for a wide bandwidth in a non-isolated environment.

OBJECT MANIFOLD ALIGNMENT FOR MULTI-TEMPORAL HIGH RESOLUTION REMOTE SENSING IMAGES CLASSIFICATION

Abstract. Multi-temporal remote sensing images classification is very useful for monitoring the land cover changes. Traditional approaches in this field mainly face to limited labelled samples and spectral drift of image information. With spatial resolution improvement, “pepper and salt” appears and classification results will be effected when the pixelwise classification algorithms are applied to high-resolution satellite images, in which the spatial relationship among the pixels is ignored. For classifying the multi-temporal high resolution images with limited labelled samples, spectral drift and “pepper and salt” problem, an object-based manifold alignment method is proposed. Firstly, multi-temporal multispectral images are cut to superpixels by simple linear iterative clustering (SLIC) respectively. Secondly, some features obtained from superpixels are formed as vector. Thirdly, a majority voting manifold alignment method aiming at solving high resolution problem is proposed and mapping the vector data to alignment space. At last, all the data in the alignment space are classified by using KNN method. Multi-temporal images from different areas or the same area are both considered in this paper. In the experiments, 2 groups of multi-temporal HR images collected by China GF1 and GF2 satellites are used for performance evaluation. Experimental results indicate that the proposed method not only has significantly outperforms than traditional domain adaptation methods in classification accuracy, but also effectively overcome the problem of “pepper and salt”.

Minimizing residual spectral drift in laser diode bars using femtosecond-written volume Bragg gratings in fused silica.

Ultrashort laser pulses are used to inscribe volume Bragg gratings (VBGs) into fused silica. These VBGs demonstrate excellent performance for the external stabilization of laser diode bars. The stabilized system emits at a wavelength of 969 nm with a signal width (FWHM) of 100 pm and shows a spectral drift as low as 24 pm for a change in output power of 45 W for a grating surface area of 10 mm².

Visualization, Quantification, and Alignment of Spectral Drift in Population Scale Untargeted Metabolomics Data.

Untargeted liquid-chromatography-mass spectrometry (LC-MS)-based metabolomics analysis of human biospecimens has become among the most promising strategies for probing the underpinnings of human health and disease. Analysis of spectral data across population scale cohorts, however, is precluded by day-to-day nonlinear signal drifts in LC retention time or batch effects that complicate comparison of thousands of untargeted peaks. To date, there exists no efficient means of visualization and quantitative assessment of signal drift, correction of drift when present, and automated filtering of unstable spectral features, particularly across thousands of data files in population scale experiments. Herein, we report the development of a set of R-based scripts that allow for pre- and postprocessing of raw LC-MS data. These methods can be integrated with existing data analysis workflows by providing initial preprocessing bulk nonlinear retention time correction at the raw data level. Further, this approach provides postprocessing visualization and quantification of peak alignment accuracy, as well as peak-reliability-based parsing of processed data through hierarchical clustering of signal profiles. In a metabolomics data set derived from ∼3000 human plasma samples, we find that application of our alignment tools resulted in substantial improvement in peak alignment accuracy, automated data filtering, and ultimately statistical power for detection of metabolite correlates of clinical measures. These tools will enable metabolomics studies of population scale cohorts.

Low drift cw-seeded high-repetition-rate optical parametric amplifier for fingerprint coherent Raman spectroscopy.

We introduce a broadly tunable robust source for fingerprint (170 - 1620 cm-1) Raman spectroscopy. A cw thulium-doped fiber laser seeds an optical parametric amplifier, which is pumped by a 7-W, 450-fs Yb:KGW bulk mode-locked oscillator with 41 MHz repetition rate. The output radiation is frequency doubled in a MgO:PPLN crystal and generates 0.7 - 1.3-ps-long narrowband pump pulses that are tunable between 885 and 1015 nm with >80 mW average power. The Stokes beam is delivered by a part of the oscillator output, which is sent through an etalon to create pulses with 1.7 ps duration. We demonstrate a stimulated Raman gain measurement of toluene in the fingerprint spectral range. The cw seeding intrinsically ensures low spectral drift.

Synchronization-free all-solid-state laser system for stimulated Raman scattering microscopy.

We introduce an extremely simple and highly stable system for stimulated Raman scattering (SRS) microscopy. An 8-W, 450-fs Yb:KGW bulk oscillator with 41 MHz repetition rate pumps an optical parametric amplifier, which is seeded by a cw tunable external cavity diode laser. The output radiation is frequency doubled in a long PPLN crystal and generates 1.5-ps long narrowband pump pulses that are tunable between 760 and 820 nm with >50 mW average power. Part of the oscillator output is sent through an etalon and creates Stokes pulses with 100 mW average power and 1.7 ps duration. We demonstrate SRS microscopy at a 30-μs pixel dwell time with high chemical contrast, signal-to-noise ratio in excess of 45 and no need for balanced detection, thanks to the favorable noise properties of the bulk solid-state system. Cw seeding intrinsically ensures low spectral drift. We discuss its application to chemical contrast microscopy of freshly prepared plant tissue sections at different vibrational bands.

가간섭성 광섬유 센서에 대한 주파수 천이도의 영향

Effect of spectral drift in coherent fiber laser was investigated by injecting optical feedback to Fabry-Perot resonance loop. Er+3 doped fiber laser having unilateral optical feedback loop in Fabry-Perot configuration using two FBGs was fabricated. The optical feedback was found to be effective in linewidth reduction of fiber laser compared to the case without any optical feedback. The linewidth of three fiber lasers using above configuration were measured to be within 3kHz which is resolution-limited performance of self-heterodyne linewidth measurement set-up. The frequency drift measurement using Mach-Zehnder measurement set-up having 200m optical delay-line in one arm showed that the frequency drift rate of optical feedback fiber laser was measured as 300kHz/sec which was better than the case without optical feedback.

Speciation of water soluble iron in size segregated airborne particulate matter using LED based liquid waveguide with a novel dispersive absorption spectroscopic measurement technique

In this study, we present the development and evaluation of a dispersive absorption spectroscopic technique for trace level soluble ferrous detection. The technique makes use of the broadband absorption spectra of the ferrous-ferrozine complex with a novel spectral fitting algorithm to determine soluble ferrous concentrations in samples and achieves much improved measurement precision compared to conventional methods. The developed method was evaluated by both model simulations and experimental investigations. The results demonstrated the robustness of the method against the spectral fluctuation, wavelength drift and electronic noise, while achieving excellent linearity (R2 > 0.999) and low detection limit (0.06 μg L−1) for soluble ferrous detection. The developed method was also used for the speciation of soluble iron in size segregated atmospheric aerosols. The measurement was carried out during Spring and Summer in typical urban environment in Hong Kong. The measured total iron concentrations are in good agreement compared to conventional Inductively Coupled Plasma – Optical Emission Spectroscopy (ICP-OES) measurements. Investigation on ambient particulate matter samples shows the size dependent characteristic of iron speciation in the atmosphere with a more active role of fine particles in transforming between ferrous and ferric. The method demonstrated in this study provides a cost and time effective approach for the speciation of iron in ambient aerosols.

Removal of Background Drift Nonlinear Interference in 3-D Spectral Arrays for Multi-Way Calibration Using Trilinear Decomposition Methods

Fluorescent and phosphorescent techniques have been widely used in the fields of food, biology, environment, chemistry, medicine, life science and so on. However, the spectral background drift often occurs in the spectral excitation dimension, creating the need for new methods to process this phenomenon. There are many different factors which may lead to this common phenomenon, such as the changes in background noise of instrument or temperature. In the work, a new technique for removal of background drift in three-dimensional spectral arrays is proposed. The basic idea is to perform trilinear decomposition based on the alternating trilinear decomposition (ATLD) algorithm on the instrumental response data. In model building, the background drift is modeled as an additional component or factor as well as the analytes of interest and the interferents. As the optimum number of factors (N) is provided by the core consistency diagnostic (CORCONDIA), the ATLD algorithm is applied to decompose the raw data (Xraw) with the factor number of N, then three profile matrices A, B and C can be obtained. Vectors an, bn and cn that representing the signal of the background drift can be extracted from these matrices to construct a 3-D background drift data array (Xdrift ). After subtracting the Xdrift from the Xraw, the background drift is removed, leaving the new data on a flat baseline. Two simulated data sets were firstly employed to demonstrate the reasonability of the new method. The same and different levels of background drifts along the excitation dimension are added into the two simulated data sets, respectively. Then, it is successfully used to analyze two experimental data sets in which significant background drift are present. These results highlight the fact that this technique yields a good removal of background drift. In addition, the good result is obtained by secondary removal for serious background drift. The proposed method can be viewed as a good spectral pretreatment technique. Keywords chemometrics; alternating trilinear decomposition; three-dimensional spectral array; background drift; spectral pretreatment

Spectral and Spatial Proximity-Based Manifold Alignment for Multitemporal Hyperspectral Image Classification

Multitemporal hyperspectral images provide valuable information for a wide range of applications related to supervised classification, including long-term environmental monitoring and land cover change detection. However, the required ground reference data are time-consuming and expensive to acquire, motivating researchers to investigate options for reusing limited training data for classification of other temporal images. Current studies that address high dimensionality and nonstationarity inherent in temporal hyperspectral data for classification are limited for the case where significant spectral drift exists between images. In this paper, we adapt and extend two manifold alignment (MA) methods for classification of multitemporal hyperspectral images in a common manifold space, assuming that the local geometries of two temporal spectral images are similar. The first method exploits a locally based manifold configuration of a source image (considered to be the “prior” manifold), and the second approach links local manifolds of two images using bridging pairs. In addition to exploiting manifolds estimated with spectral information for MA, we also demonstrate how spatial information can be incorporated into the MA methods. When evaluated using three Hyperion data sets, the proposed methods outperform four baseline approaches and two state-of-the-art domain adaptation methods. The advantages of the proposed MA methods are more evident when significant spectral drift exists between two temporal images. In addition to the promising classification results, the proposed methods establish a domain adaptation framework for analysis of temporal hyperspectral data based on data geometry.

Fast White Light Interferometry Demodulation Algorithm for Low-Finesse Fabry–Pérot Sensors

Over decades, the signal demodulation techniques of low-finesse Fabry-Pérot interferometer (FPI) sensors have been either slow with wide dynamic range and absolute measurement capability, or fast with narrow dynamic range and relative measurement capability. The tradeoff between the speed and the measurement capability has greatly limited the application of FPI-based sensors. In this letter, a novel high-speed white light interferometry (WLI) demodulation algorithm for low-finesse FPI has been developed. By realizing high-speed absolute demodulation utilizing full spectra, the new algorithm has the advantage of spectral drift immunity, high precision, and simultaneous ac and dc signal measurement capability, such as acoustic and temperature. A 70-kHz real-time WLI demodulation experiment was conducted in lab, in which the speed was limited only by the spectrometer hardware.

Error reduction in retrievals of atmospheric species from symmetrically measured lidar sounding absorption spectra.

We report new methods for retrieving atmospheric constituents from symmetrically-measured lidar-sounding absorption spectra. The forward model accounts for laser line-center frequency noise and broadened line-shape, and is essentially linearized by linking estimated optical-depths to the mixing ratios. Errors from the spectral distortion and laser frequency drift are substantially reduced by averaging optical-depths at each pair of symmetric wavelength channels. Retrieval errors from measurement noise and model bias are analyzed parametrically and numerically for multiple atmospheric layers, to provide deeper insight. Errors from surface height and reflectance variations are reduced to tolerable levels by "averaging before log" with pulse-by-pulse ranging knowledge incorporated.

Drift reliability-based optimization method of frames subjected to stochastic earthquake ground motion

The paper describes an inter-story drift reliability-based optimization method of frames subjected to stochastic earthquake loads. First, the formulas for eigenvalue and eigenvector, drift PSD functions, drift spectral moments, drift reliabilities, their first and second derivatives are derived based on random vibration theory. The computational procedure of drift reliabilities, their first and second derivatives are given in detail. Second, optimal problem of drift reliability-based optimization design is formulated in a dimensionless way. Optimal mathematic model is converted into unconstrained mathematic model using penalty function method. Gradient and Hessian matrix of penalty function are derived using drift reliabilities and structural mass, their first and second derivatives. Third, solution step of optimal problem is constructed using conjugate gradient method. Finally, optimization designs of two planar frames are demonstrated. Sensitivity analysis of optimum design indicates the drift reliability-based optimization method can obtain local optimum design.

Development of a novel alternating quadrilinear decomposition algorithm for the kinetic analysis of four-way room-temperature phosphorescence data

Four-way room-temperature phosphorescence (RTP) data recorded by following the kinetic evolution of excitation–emission phosphorescence matrices (EEPMs) have been analyzed for the first time by third-order calibration based on parallel factor analysis (PARAFAC), alternating weighted residue constraint quadrilinear decomposition (AWRCQLD) and alternating quadrilinear decomposition (AQLD) algorithms. The AQLD constructed in pseudo-fully stretched matrix forms of quadrilinear model was a new third-order calibration algorithm, which was developed as a direct extension of alternating trilinear decomposition for quadrilinear data. These methodologies were applied to investigate the hydrolysis kinetic of carbaryl even in the presence of an uncalibrated phosphorescence background and comparisons among them were done subsequently. Spectral background drift produced in measured dataset was overcome by means of modeling the drift as an additional component as well as the analyte of interest in the mathematical model. Satisfactory results were obtained for determination of carbaryl in spiked tap water samples. The spectral and kinetic time profiles resolved by these methodologies were in good agreement with experimental observations. The present work successfully faced the difficulty in investigating the hydrolysis kinetic of analyte with the RTP techniques, opening a new approach for third-order data generation and subsequent third-order calibration.

基于Nd∶YVO4晶体四波段腔面膜的研制

The yellow laser,which is close to the sensitive area of human eyes,plays an important role in many fields,and the resonant cavity is an important component of laser.A four wavelengths cavity surface film by using a vacuum coating equipment was developed based on Nd∶YVO4 crystal(c-cut). First,in the process of film system design,the distribution of standing wave field reasonably in the film was made by analyzing the film field strength and optimizing the film system,thus to reduce the possibility of film damage theoretically.Then,through technics feedback analysis to the test result by TFCalc and several simulation experiments,it was found that in different wavelengths the film thickness ratio is different even in the same monitoring method.And the problem of spectral curve drift was solved by adjusting the tool factor.Finally,the yellow laser cavity surface film with firm film and stable chemical performance was prepared,which also met the requirements of four wavelengths spectral output.

Frequency spectrum analysis for spectrum stabilization in airborne gamma-ray spectrometer

Abnormal multi-crystal spectral drifts often can be observed when power on the airborne gamma-ray spectrometer. Currently, these spectral drifts of each crystal are generally eliminated through manual adjustment, which is time-consuming and labor-ineffective. To realize this quick automatic spectrum stabilization of multi-crystal, a frequency spectrum analysis method for natural gamma-ray background spectrum is put forward in this paper to replace traditional spectrum stabilization method used characteristic peak. Based on the polynomial fitting of high harmonics in frequency spectrum and gamma-ray spectral drift, it calculates overall spectral drift of natural gamma-ray spectrum and adjusts the gain of spectrometer by this spectral drift value, thus completing quick spectrum stabilization in the power on stage of spectrometer. This method requires no manual intervention and can obtain the overall spectral drift value automatically under no time-domain pre-processing to the natural gamma-ray spectra. The spectral drift value calculated by this method has an absolute error less than five channels (1024 resolution) and a relative error smaller than 0.80%, which can satisfy the quick automatic spectrum stabilization requirement when power on the airborne gamma-ray spectrometer instead of manual operation.