Bhattacharyya distance as a contrast parameter for statistical processing of noisy optical images

In many imaging applications, the measured optical images are perturbed by strong fluctuations or boise. This can be the case, for example, for coherent-active or low-flux imagery. In such cases, the noise is not Gaussian additive and the definition of a contrast parameter between two regions in the image is not always a straightforward task. We show that for noncorrelated noise, the Bhattacharyya distance can be an efficient candidate for contrast definition when one uses statistical algorithms for detection, location, or segmentation. We demonstrate with numerical simulations that different images with the same Bhattacharyya distance lead to equivalent values of the performance criterion for a large number of probability laws. The Bhattacharyya distance can thus be used to compare different noisy situations and to simplify the analysis and the specification of optical imaging systems.     

Hyperspectral target detection via discrete wavelet-based spectral fringe-adjusted joint transform correlation

Spectral variability remains a major challenge for target detection in hyperspectral imagery (HSI). Recently, the spectral fringe-adjusted joint transform correlation (SFJTC) technique has been used effectively for hyperspectral target detection applications. In this paper, we propose to use discrete wavelet transform (DWT) coefficients of the signatures as features for detection in order to make the SFJTC technique more insensitive to spectral variability. We devised a supervised training algorithm that uses the pure target signature and randomly selected samples from input scenery to select an optimal set of DWT coefficients for detection. We have inserted target signatures into urban and vegetative hyperspectral scenery with varying levels of spectral variability to explore the performance of our DWT-based SFJTC technique in different operating conditions. Detection results in the form of receiver-operating-characteristic (ROC) curves and area-under-the-ROC (AUROC) curves show that the proposed scheme yields the largest mean AUROC values compared to SFJTC using the original signatures and traditional hyperspectral detection algorithms.     

Spatial-scale contribution to the detection of mirror symmetry in fractal noise.

We investigated how the detection of mirror symmetry depends on the distribution of contrast energy across spatial scales. Stimuli consisted of vertically symmetric noise patterns with fractal power spectra defined by 1/f beta slopes (-2 < or = beta < or = 5). While overall rms contrast remained fixed at 25%, symmetry-detection thresholds were obtained by corrupting the signal with variable amounts of noise with identical spectral characteristics. A first experiment measured thresholds as a function of spectral slope, and performance was found to be substantially facilitated in images with power spectra that characterize natural scenes (1.2 < or = beta < or = 3.2). In a second experiment, symmetry was removed from randomly chosen octave bands and replaced by noise with the same spectral profile. Results revealed that only in images with 1/f2 spectra does performance decrease by constant amounts across all frequency bands. Together, the results imply that symmetry mechanisms extract equal amounts of information from constant-octave frequency bands but lack the ability to whiten stimuli whose spectral slopes differ from those of natural scenes. Results are qualitatively well predicted by a multichannel model that (1) relies on spatial filters with equal-volume point-spread functions and constant-octave frequency bandwidths and (2) restricts the computation of symmetry to spatial regions whose dimensions are proportional to the filters' spatial scale. These findings are also consistent with the notion that mechanisms that mediate the perception of form exploit the ability of early vision to reduce second-order redundancy in natural scenes.     

Distributed truncated ARQ protocol for cooperative diversity networks

Automatic repeat request (ARQ) retransmission in user cooperative networks, which use amplify-and-forward as a relaying strategy is dealt with. In contrast to the conventional environments, where the source responds in a retransmission demand, in cooperative systems this requirement can also be satisfied by a relay node. In addition, if a relay node can be selected according to the instantaneous channel conditions, the source is not always the optimal responder. The problem under consideration here is to find the retransmission combination which optimises the performance under a given delay quality-of-service constraint. It will be shown that the optimal combination depends on the number of available relays, the total number of retransmissions and the average signal-to-noise ratio. We provide an analytical framework for the definition of the optimal combination in function of these three system parameters. Since the practical ad hoc networks are not centralised and do not have an external control, a distributed truncated ARQ protocol is further proposed to apply the decided retransmission combination.     

Code aperture optimization for spectrally agile compressive imaging

Coded aperture snapshot spectral imaging (CASSI) provides a mechanism for capturing a 3D spectral cube with a single shot 2D measurement. In many applications selective spectral imaging is sought since relevant information often lies within a subset of spectral bands. Capturing and reconstructing all the spectral bands in the observed image cube, to then throw away a large portion of this data, is inefficient. To this end, this paper extends the concept of CASSI to a system admitting multiple shot measurements, which leads not only to higher quality of reconstruction but also to spectrally selective imaging when the sequence of code aperture patterns is optimized. The aperture code optimization problem is shown to be analogous to the optimization of a constrained multichannel filter bank. The optimal code apertures allow the decomposition of the CASSI measurement into several subsets, each having information from only a few selected spectral bands. The rich theory of compressive sensing is used to effectively reconstruct the spectral bands of interest from the measurements. A number of simulations are developed to illustrate the spectral imaging characteristics attained by optimal aperture codes.     

基于图像序列的运动目标检测识别的关键技术研究

为提高运动目标检测的识别效果,通过分析、综合比较各种运动目标检测算法的优劣性,提出了基于全局自适应帧差法和基于码本模型的背景减除法对同一运动目标进行检测。通过对运动目标检测提取运动目标的掩膜,对掩膜进行外接矩形分析,从而得到包围运动目标的矩形框;将矩形框内的图片截取出来,调整该矩形并提取图片的HOG特征,最后通过训练好的SVM进行分类。在训练过程中,针对难易情况应用自举法对训练器进行优化。实验表明,与传统HOG+SVM多尺度检测算法相比,该方法在速度和准确性上可提升20%左右,可作为运动目标检测与识别的参考方法。     

Experimental optimization of flow distributors for pressure-driven separations and reactions in flat-rectangular microchannels

We report on the results of an experimental study established to optimize the design of microfabricated flow distributors for use in pressure-driven separations and reactions in flat-rectangular channels. For this purpose, the performance of a wide variety of possible flow distributor designs etched in glass/silicon wafers was compared, using CCD camera detection to study the shape and variance of the bands eluting from them. The best performance was obtained with radially interconnected distributors with a diverging inlet section and filled with diamond-shaped pillars, oriented perpendicular to the main flow direction and with a high transversal over axial aspect ratio. It was found that the best distributor designs start with a diverging section containing some 10-12 subsequent rows of high aspect ratio pillars (with a transversal width making up 10-15% of the final channel width) and with a divergence angle selected such that the sloped side-walls run parallel with the sides of the diamond-shaped pillars. After this zone, one or more regions with pillars with a smaller aspect ratio should be provided to increase the number of exit points. To prevent the formation of dead zones in these subsequent zones, so-called distributor wedges can be used to prevent the formation of any dead zones in the wake of the large aspect ratio pillars of the preceding section.     

Thermal infrared spectral band detection limits for unidentified surface materials

Infrared emission spectra recorded by airborne or satellite spectrometers can be searched for spectral features to determine the composition of rocks on planetary surfaces. Surface materials are identified by detections of characteristic spectral bands. We show how to define whether to accept an observed spectral feature as a detection when the target material is unknown. We also use remotely sensed spectra measured by the Thermal Emission Spectrometer (TES) and the Spatially Enhanced Broadband Array Spectrograph System to illustrate the importance of instrument parameters and surface properties on band detection limits and how the variation in signal-to-noise ratio with wavelength affects the bands that are most detectable for a given instrument. The spectrometer's sampling interval, spectral resolution, signal-to-noise ratio as a function of wavelength, and the sample's surface properties influence whether the instrument can detect a spectral feature exhibited by a material. As an example, in the 6-13-mum wavelength region, massive carbonates exhibit two bands: a very strong, broad feature at ~6.5 mum and a less intense, sharper band at ~11.25 mum. Although the 6.5-mum band is stronger and broader in laboratory-measured spectra, the 11.25-mum band will cause a more detectable feature in TES spectra.     

一种目标提取新方法-光谱排序编码法

为了进一步提高光学遥感目标提取精度,拓展目标提取方法的适用范围,推进遥感智能化信息提取方法的发展,本文从谱带强度与波形特征有机结合的角度出发提出了光谱排序编码法.首先介绍了该方法的基本原理:对端元光谱与图像光谱按谱带强度大小各自进行排序的基础上,通过比较同一顺序上的波段位置的异同来计算两条光谱曲线的相似度;然后开展了与约束能量最小法(CEM)、光谱二阶导数法(SSD)的目标提取精度对比实验,结果表明该方法的目标提取精度为95%,比CEM提高了41.9%,比SSD提高了46.9%;最后利用SPOT多光谱和AVIRIS高光谱数据检验了该方法在目标提取适用范围上的能力,结果表明该方法既适用于多光谱遥感,又适用于高光谱遥感.总之,算法具有速度快、精度高、适用范围广的特点.     

Wavelength band selection method for multispectral target detection

A framework is proposed for the selection of wavelength bands for multispectral sensors by use of hyperspectral reference data. Using the results from the detection theory we derive a cost function that is minimized by a set of spectral bands optimal in terms of detection performance for discrimination between a class of small rare targets and clutter with known spectral distribution. The method may be used, e.g., in the design of multispectral infrared search and track and electro-optical missile warning sensors, where a low false-alarm rate and a high-detection probability for detection of small targets against a clutter background are of critical importance, but the required high frame rate prevents the use of hyperspectral sensors.     

Method for the estimation of the mean lorentzian bandwidth in spectra composed of an unknown number of highly overlapped bands

We introduce a method for the estimation of the mean Lorentzian bandwidth of the component bands in a spectrum. The method is computationally simple, using only the module of the Fourier transform of the spectrum, and its first derivative. Moreover, the presented method does not require knowledge of the number of bands in the spectrum, their band positions, or their band areas. Furthermore, it works on spectra containing Lorentzian bands, as well as Gaussian and Voigtian bands. Therefore, the introduced method seems especially well suited for obtaining a representative Lorentzian width for highly overlapped bands, independent of their number and Lorentzian/Gaussian character. We describe how different experimental limitations (spectral truncation, offset error, presence of noise, etc.) may affect the performance of the method, and when required we propose effective alternatives to minimize their effects. Finally, we show the application of the method to an experimental spectrum: the amide I band of a dry film of the solubilized ADP/ATP carrier. The estimation of the mean Lorentzian width can allow, for instance, for a more objective selection of the deconvolution width in Fourier self-deconvolution, allowing for a more objective and reliable analysis of the amide I band of proteins. The mean Lorentzian width can also be useful to obtain an estimation of the homogenous broadening and vibrational relaxation of the amide I vibration of proteins, without requiring complex pump-probe experiments.     

Infrared target detection with probability density functions of wavelet transform subbands

We report the development of a wavelet multiresolution texture-based algorithm that uses the probability density functions (PDFs) of the subband of the wavelet decomposition of an image. The moments of these pdfs are used in a clustering algorithm to segment the targets from their background clutter. Using the tools of experimental methodology, we evaluate the performance of this algorithm on real infrared imagery under varying algorithm parameter sets as well as scene, image, and false-alarm conditions. We estimate a set of multidimensional predictive analytic performance models that relate the detection probabilities as functions of false alarm, algorithm internal parameter, target pixel number, target-to-background interference ratio, target-interference ratio, and Fechner-Weber and local entropy metrics in the scene. These models can be used to predict performance in regions were no data are available and to optimize performance by selection of the optimum parameter and constant false-alarm values in regions with known scene and metric conditions.     

Spectral multivariate calibration with wavelength selection using variants of Tikhonov regularization

Tikhonov regularization (TR) is a general method that can be used to form a multivariate calibration model and numerous variants of it exist, including ridge regression (RR). This paper reports on the unique flexibility of TR to form a model using full wavelengths (RR), individually selected wavelengths, or multiple bands of selected wavelengths. Of these three TR variants, the one based on selection of wavelength bands is found to produce lower prediction errors. As with most wavelength selection algorithms, the model vector magnitude indicates that this error reduction comes with a potential increase in prediction uncertainty. Results are presented for near-infrared, ultraviolet-visible, and synthetic spectral data sets. While the focus of this paper is wavelength selection, the TR methods are generic and applicable to other variable-selection situations.     

Photon signal detection and evaluation in the adaptive optics scanning laser ophthalmoscope

To select a suitable photodetector for an adaptive optics scanning laser ophthalmoscope (AOSLO) and evaluate its performance, we characterized the signal and noise properties in the AOSLO photon detection and derived the signal-to-noise ratio (SNR). Using the SNR as the main criterion, we chose the best detector from a selection of four photomultiplier tubes (PMTs) and three avalanche photodiodes (APDs). We conducted a comprehensive evaluation of the performance of the selected detector on our AOSLO. The study presents a practical strategy that can be used to test the photodetector for either initial evaluation or subsequent performance in in-line inspection.     

End-to-end sensor simulation for spectral band selection and optimization with application to the Sentinel-2 mission

An end-to-end sensor simulation is a proper tool for the prediction of the sensor's performance over a range of conditions that cannot be easily measured. In this study, such a tool has been developed that enables the assessment of the optimum spectral resolution configuration of a sensor based on key applications. It employs the spectral molecular absorption and scattering properties of materials that are used for the identification and determination of the abundances of surface and atmospheric constituents and their interdependence on spatial resolution and signal-to-noise ratio as a basis for the detailed design and consolidation of spectral bands for the future Sentinel-2 sensor. The developed tools allow the computation of synthetic Sentinel-2 spectra that form the frame for the subsequent twofold analysis of bands in the atmospheric absorption and window regions. One part of the study comprises the assessment of optimal spatial and spectral resolution configurations for those bands used for atmospheric correction, optimized with regard to the retrieval of aerosols, water vapor, and the detection of cirrus clouds. The second part of the study presents the optimization of thematic bands, mainly driven by the spectral characteristics of vegetation constituents and minerals. The investigation is performed for different wavelength ranges because most remote sensing applications require the use of specific band combinations rather than single bands. The results from the important "red-edge" and the "short-wave infrared" domains are presented. The recommended optimum spectral design predominantly confirms the sensor parameters given by the European Space Agency. The system is capable of retrieving atmospheric and geobiophysical parameters with enhanced quality compared to existing multispectral sensors. Minor spectral changes of single bands are discussed in the context of typical remote sensing applications, supplemented by the recommendation of a few new bands for the next generation of optical Sentinel sensors.     

An immunosensor based on the glucose oxidase label and optical oxygen detection

A novel optical immunosensor setup is described which uses glucose oxidase enzyme as a label in conjunction with a luminescence lifetime-based oxygen sensor and phase measurements. The oxygen sensor membranes prepared on microporous filters were used as a solid phase on which the immunoassay was carried out. These sensing materials in combination with a new measurement setup provided high sensitivity for the detection of oxidase enzymes, being at nanogram per milliliter level, i.e., 10(-11)-10(-12) M, with respect to glucose oxidase and its conjugates. Experimental data on the sensitivity were validated using theoretical equations and calculations. Using the new measurement setup and IgG-anti-IgG as a model, a number of different sensing materials were studied aimed to optimize the immunosensor and evaluate its performance. This approach was then applied to a practical system for the detection of human lactate dehydrogenase isoenzymes. It provided similar sensitivity of approximately 1 ng/mL, which is comparable to that of standard ELISA. The attributes of the new immunosensor approach are discussed with respect to performance and versitility.     

Possibilistic multi-response quality design

Multi-response quality design techniques are used to identify settings of process parameters that make a product's performance close to a target value in the presence of multiple quality characteristics. In many situations, these quality characteristics, and thus their functional relations, are imprecise to some degree due to tolerances, nonspecificity, measuring errors, incomplete knowledge, vagueness of definitions, and so on. Here, possibility distributions and possibilistic regressions are used to model these imprecise natures and induced imprecise functional relationships. We first integrate and extend existing possibilistic regression methods to obtain unified measures of predictive quality characteristics or responses. The unified possibility distributions are also practical for many forecasting problems. We then propose a multiple objective programming model to obtain an appropriate combination of process parameter settings based on the unified possibility distributions of imprecise predictive responses. We not only optimize the most possible responses values, but also minimize imprecision or deviations from the most possible values. Through a die casting example, we show how to use our approach to reach an appropriate machine setting which simultaneously optimizes both porosity and temperature difference on die surfaces.     

Contrast discrimination with sinusoidal gratings of different spatial frequency

The detectability of contrast increments was measured as a function of the contrast of a masking or "pedestal" grating at a number of different spatial frequencies ranging from 2 to 16 cycles per degree of visual angle. The pedestal grating always had the same orientation, spatial frequency, and phase as the signal. The shape of the contrast-increment threshold versus pedestal contrast (TvC) functions depends on the performance level used to define the "threshold," but when both axes are normalized by the contrast corresponding to 75% correct detection at each frequency, the TvC functions at a given performance level are identical. Confidence intervals on the slope of the rising part of the TvC functions are so wide that it is not possible with our data to reject Weber's law.     

SERS detection of biomolecules using lithographed nanoparticles towards a reproducible SERS biosensor

In this paper we highlight the accurate spectral detection of bovine serum albumin and ribonuclease-A using a surface-enhanced Raman scattering (SERS) substrate based on gold nanocylinders obtained by electron-beam lithography (EBL). The nanocylinders have diameters from 100 to 180 nm with a gap of 200 nm. We demonstrate that optimizing the size and the shape of the lithographed gold nanocylinders, we can obtain SERS spectra of proteins at low concentration. This SERS study enabled us to estimate high enhancement factors (10(5) for BSA and 10(7) for RNase-A) of important bands in the protein Raman spectrum measured for 1 mM concentration. We demonstrate that, to reach the highest enhancement, it is necessary to optimize the SERS signal and that the main parameter of optimization is the LSPR position. The LSPR have to be suitably located between the laser excitation wavelength, which is 632.8 nm, and the position of the considered Raman band. Our study underlines the efficiency of gold nanocylinder arrays in the spectral detection of proteins.