Figure - available from: Applied Optics
This content is subject to copyright. Terms and conditions apply.
Source publication
A multispectral camera acquires spectral color images with high fidelity by splitting the light spectrum into more than three bands. Because of the shift of focal length with wavelength, the focus of each channel should be mechanically adjusted in order to obtain sharp images. Because progressive adjustment is quite time consuming, the clear focus...
Similar publications
There are analyzed the accuracy and the systematic errors of the known method of a flat one-sided probe for measuring the anisotropic distribution functions of electrons and ions in a plasma. It is shown that, with the exception of strong fields, ten members of the series are usually enough for an adequate description of the ion distribution functi...
L1 splines have been under development for interpolation and approximation of irregular geometric data. We investigate the advantages in terms of shape preservation and computational efficiency of calculating univariate cubic L1 spline fits using a steepest-descent algorithm to minimize a global data-fitting functional under a constraint implemente...
Problem statement: The lacunary interpolation problem, which we had investigated in this study, consisted in finding the six degree spline S (x) of deficiency four, interpolating data given on the function value and third and fifth order in the int erval (0,1). Also, an extra initial condition was prescribed on the first derivative. Other purpose o...
In this paper we propose a new algorithm for sub-sample time-delay estimation between two discrete-time signals. The new algorithm operates just over the discrete-time signals without the need for interpolation or fitting. The proposed method is compared with a method based on spline fitting, and that is referred to outperform other methods over a...
In critical design phase of aircraft, the success of stability, control and performance computations are strongly dependent upon the reliability of plant models. The aerodynamic efforts stand on the center of these models. Almost every system is fed by the aerodynamic coefficients and they should be modelled as much as accurate. It requires the inv...
Citations
... Multi-spectral images are acquired from the multi-spectral camera used in [27]. Experiments are performed on various multi-spectral images covering the range from 400−700nm. ...
Multi-spectralimages suffers from out-of-focusblur due to well focused camera at the reference imaging channel. A framework for out-of-focus images de-blurring using texture extraction and modified Fourier transform is proposed. The texture is extracted from the blurred image using region covariance. Fourier transform is modified by modification of guided image (as prior) using L0 gradient projection followed by detail amplification which is used as an input for Fourier transform. For further enhancement, the multi-spectral de-blurred images are smooth, and ringing artifacts are minimized by combining with the textures extracted to ensure maximum edge preservation. Comparison of proposed and existing schemes on different multi-spectral images explains the advantage of the proposed scheme for de-blurring in terms of edge preservation, noise and artifacts removal.
... Further, the blurring problem of image cube planes due to chromatic aberration has been addressed in [138,139]. In [138], the focus for each plane in a filter-wheel camera is set mechanically using a step-motor. ...
... Further, the blurring problem of image cube planes due to chromatic aberration has been addressed in [138,139]. In [138], the focus for each plane in a filter-wheel camera is set mechanically using a step-motor. The peak SNR and the structural similarity are used as the focus measures. ...
An alternative to the cumbersome process of traditional hyperspectral imaging is to estimate the images from the conventional snapshot multispectral cameras. Numerous multispectral design paradigms are proposed in the literature, yet little is known of their quality in terms of spectral fidelity. Similarly, spectral estimation
models are not used to their potential, despite the fact that they are mathematically well developed.
In this thesis, I address the above-mentioned issues. First, image capture is revisited from the perspective of dynamic range and signal-to-noise ratio (SNR) as a
function of wavelength and exposure. A detailed image acquisition configuration is presented and used to capture icon painting surfaces as a case study of complex objects with a high dynamic range. Second, a registration method with the focus on the outlier elimination phase is introduced for accurate spatial aligning of the multicamera image dataset. This dataset is essential in closing the loop of designing, evaluating, and refining multispectral cameras – a process that is currently performed with either simulated data or color charts to relax the registration requirement. A comparative study of the spectral fidelity of two cameras (an RGB and a seven-band beam splitter camera) is conducted using the registered dataset. Thus, this study offers immediate guidelines on improving the multispectral camera
design.
In the next section, the Gaussian copula link function is introduced and tested in combination with three kernel regression estimation models and a proposed mixture of the Gaussian model. The combination improves the estimation results of Munsell color charts with real and simulated measurements. Finally, a feature extraction method based on the concept of mathematical moments is introduced for spectral data. The proposed features are comprehensively tested in the texture analysis application of hyperspectral images. The accuracy of the classification results are improved compared to several other opponent methods.
The contributions of this thesis are to improve the pipeline of spectral imaging from the image acquisition to the estimation and processing steps, and promote its practice. The developed methods are not limited to the discussed cases, but provide the necessary tools for a myriad emerging applications of spectral big data.
... Most investigations focused on one of the captured bands and used the same focus settings for all the others [9], or used manual focusing, which is not reproducible [4,7]. Automatic focus search algorithms have been developed for different lens systems and capture devices [10][11][12], although they were not tested in multispectral imaging devices until recently [13]. In previous studies, the exposure time was usually set for the effect coated sample and the same exposure was used for the reference white used to obtain spectral reflectance. ...
Hyperspectral imaging of effect coated samples can be challenging, mainly because of the large differences in irradiance that stem from the orientation distribution of the metallic flakes contained in the coating, and from the lightness variations from one sample to another. Besides, high spatial resolution is needed to sample the details of the texture (sparkle) typical of these samples. In addition, focus search strategy and image registration are essential to achieve high quality data for further analysis. In this work, we propose and fully validate a capture framework for measuring spectral reflectance of effect-coated samples with high spatial resolution in 45/0 geometry, using an LCTF (Liquid Crystal Tunable Filter) coupled with a monochrome camera. The main features of the proposed framework are an optimized focus search method based on object movement, a very precise alignment for the images captured in different
bands (image registration), achieving sub-pixel accuracy, and a dynamic procedure that uses several white reference surfaces in exposure time estimation to cope with very dark or highly reflective samples.
The proposed capture device produces spectral reflectance values comparable to a conventional spectroradiometer using the same observation/illumination geometry, with the additional advantage of achieving a spatial resolution more than two times higher than the human visual system.
... On the other hand, the out-of-focus blur is caused by the different effective focal lengths at individual channels, which originates from the fact that the refractive indices of the lens are wavelength dependent [4]. To deal with this problem, an autofocus method [5] has been proposed to compute the focus positions of lens using a focusing device (step motor). Alternatively, when the channel images are well aligned, the blur can also be computationally removed by multispectral image deblurring [4]. ...
... We discuss the application of the proposed method in two circumstances. In the first circumstance (Subsection A), the multispectral image has a focusing device [5] and hence the acquired multispectral images are well focused. Thus the proposed registration method can be directly applied on the sharp multispectral images. ...
... Der Algorithmustyp ist als Bergsteigeralgorithmus (engl. : hill climbing algorithm) bekannt und wird in Kamerasystemen als Autofokusalgorithmus eingesetzt [155][156][157]. Es existieren zahlreiche Varianten in der Implementierung, die auf spezifische Funktionen optimiert sind, um beispielsweise den Rechenaufwand zu reduzieren oder um Zykluszeiten zu minimieren. ...
... Der größte Nachteil der diskreten Bisektion resultiert aus der starken Rauschanfälligkeit des Algorithmus, die nur teilweise durch geschickte Filterwahl kompensiert werden kann [157]. Davon ...
Diese Arbeit präsentiert die technische Nachbildung der durchstimmbaren Linse eines Säugetierauges. Die hier gezeigte technische Imitation der Augenlinse besteht vollständig aus Elastomeren und geht definierten Oberflächen aus. Dadurch unterscheiden diese Elastomerlinsen sich grundlegend von durchstimmbaren Flüssig- und Membranlinsen.
Analog zum Ziliarmuskel des Auges werden die Elastomerlinsen durch Dehnung in der Äquatorebene der Linse aktuiert. Die Deformation der Elastomerlinse ist dabei vollständig reversibel; eine 10 %ige Dehnung der Linse führt zu einer proportionalen relativen Brennweitenänderung. Aufgrund der elastischen Eigenschaften dieser Vollkörperlinsen kann das Konzept der vektoriellen Aktuierung einführt werden, wodurch sich zusätzlich bestimmte Aberrationen durchstimmen lassen.
In dieser Arbeit wurden die mechanischen und optischen Eigenschaften von Siloxanelastomeren intensiv untersucht. Eine binäre Mischung zweier Polydimethylsiloxane (PDMS) wurde ausgewählt, die sowohl den optischen Anforderungen genügt als auch eine Aktuierung durch Mikromotoren ermöglicht. Zusätzlich wurde ein flintartiges Elastomer identifiziert, welches die Fertigung eines vollständig aus Siloxanelastomeren bestehendes achromatischen Doublets erlaubt. Für dieses wurde die chromatische Längsaberration einer Einzellinse von 270 μm auf die experimentelle Nachweisgrenze des Messinstruments von 10 μm verringert.
Zur Charakterisierung der optischen Eigenschaften von durchstimmbaren Linsen wurden zwei sich ergänzende Messstände aufgebaut. Ein Messinstrument bestimmt die Wellenfrontaberration, das andere dient der Beurteilung der Optik in der Bildebene. Beide Instrumente gehen auf die besonderen Anforderungen der Optikprüfung von adaptiven Linsen ein und lassen sich leicht an verschiedene Brennweiten und Aperturdurchmesser anpassen. Wichtige optische Parameter wie die Anfangsschnittweite, die chromatische Längsaberration, die Abbildungsqualität sowie Wellenfrontaberrationen für zwei Wellenlängen können mit Hilfe einer eigens implementierten Software automatisch erfasst werden.
... When capturing BTFs, the camera arc keeps fixed, while the light arc and sample table are sequentially rotated to predefined positions. The multispectral imaging system comprises a monochrome camera, a filter wheel, and an integrating sphere [31,32]. The sample is placed in the integrating sphere. ...
Spectral bidirectional texture function (BTF) is essential for accurate reproduction of material appearance due to its nature of conveying both spatial and spectral information. A practical issue is that the acquisition of raw spectral BTFs is time-consuming. To resolve the limitation, this paper proposes a novel framework for efficient spectral BTF acquisition and reconstruction. The framework acquires red-green-blue (RGB) BTF images and just one spectral image. The full spectral BTFs are reconstructed by fusing the RGB and spectral images based on nonnegative matrix factorization (NMF). Experimental results indicate that the accuracy of spectral reflectance reconstruction is higher than that of existing algorithms. With the reconstructed spectral BTFs, the material appearance can be reproduced with high fidelity under various illumination conditions.
... When the lens is well focused at the reference channel with wavelength 560 nm, we will fail to obtain sharp images in the channels with wavelengths 420 nm and 700 nm. To deal with this problem, an autofocus method [16] has been proposed to compute the focus position of lens using a controllable step motor. A limitation of this method is that, although an efficient algorithm is adopted, the autofocus process is still time consuming. ...
... This paper proposes a multispectral image out-of-focus deblurring framework which is solely based on the captured scenarios, without using additional mechanical components [16] or calibration patterns. It is assumed that the imaging system is well focused at a reference channel. ...
... The spectral image of the 9th channel, which corresponding to wavelength 560 nm, is well focused. We also obtain the the well-focused lens positions for all other channels using an autofocus method with the assistance of a controllable step motor [16]. The lens distance, which is computed between the lens positions of target channel and reference channel, is proportional to the variance of Gaussian blur kernel [16]. ...
Out-of-focus blur occurs frequently in multispectral imaging systems when the camera is well focused at a specific (reference) imaging channel. As the effective focal lengths of the lens are wavelength dependent, the blurriness levels of the images at individual channels are different. This paper proposes a multispectral image deblurring framework to restore out-offocus spectral images based on the characteristic of interchannel correlation (ICC). The ICC is investigated based on the fact that a high-dimensional color spectrum can be linearly approximated using rather a few number of intrinsic spectra. In the method, the spectral images are classified into an out-of-focus set and a well-focused set via blurriness computation. For each out-offocus image, a guiding image is derived from the well-focused spectral images and is used as the image prior in the deblurring framework. The out-of-focus blur is modeled as a Gaussian point spread function (PSF), which is further employed as the blur kernel prior. The regularization parameters in the image deblurring framework are determined using generalized cross validation (GCV), and thus the proposed method does not need any parameter tuning. Experimental results validate that the method performs well on multispectral image deblurring and outperforms the state-of-the-arts.
... Recent advances in optics, 1-3 electronics 4 and image processing [5][6][7] have resulted in a proliferation of multispectral imaging (MSI) systems for color measurement in textile and garment. A MSI system can provide not only the spectral information but also the spatial information of a sample, i.e., multispectral images. ...
Many applications in textile require color measurement of single strands of yarns due to unavailability of a mass of consecutive yarns. While multispectral imaging systems can capture images of single strands of yarns, an unavoidable problem accompanied is how to specify their colors from images, in which pixel values vary with positions owing to three-dimensional shape of yarn. Based on the response of multispectral imaging systems to single strands of yarns, this article formulates the color specification problem as an optimization question and four methods are proposed: average of all pixels (AA), average of pixels in central area (AC), maxima of all pixels (MA), and lightness weighting method (LW). The first experiment analyzed color distribution of pixels on a single strand of yarn. Experimental results show pixels in central area have smallest color variation and largest intensity. The second experiment compared the proposed methods. Results show that CIEXYZ, L*, and C* values specified by the AA and MA methods are lowest and highest. Finally, colors of single strands of yarns specified by the proposed methods were compared with spectrophotometric colors of yarn windings using 12 pairs of single strands of yarns and yarn windings. Experimental results show the MA method yields the smallest lightness and chroma difference compared with spectrophotometric colors of yarn windings. The average color difference between spectrophotometric colors of yarn windings and multispectral imaging colors of single strands of yarns specified by the AA, AC, MA, and LW methods is 3.45, 2.72, 2.37, and 3.33 CMC(2:1) units. © 2015 Wiley Periodicals, Inc. Col Res Appl, 2015
... The passive autofocus algorithm consists of two components: focus measurement (FM) and focus search (FS) [1]. Theoretically microscopes can be simplified as a lens model and if the view position in the image space is away from the focal plane, any viewed images turn out to be blurred [6]. Passive focus takes advantage of this and regards the clarity of captured images at different Z positions as the criterion of focus level, which is called the FM [1,[7][8][9]. ...
... Sobel-Ten [3,6,11], normalized variance [12,13], and autocorrelation [3,10,[14][15][16] are the most commonly used FM algorithms, which meet all three standards. Sun et al. compared 18 algorithms including the above three and found normalized variance is the best one according to its performance in different applications such as bright field, phase contrast, or differential interference contrast (DIC) [17]. ...
... In passive focus the ideal FM is a unimodal function, of which the peak's Z position indicates the focus position, and then FS is applied to locate the peak. The most commonly used FS algorithms proposed in the previous literature are global search [18,19], hillclimbing search [3,6,[20][21][22], binary search [23,24], curve-fitting search (CFS) [25], Fibonacci search (FibS) [23,26], and golden section search (GSS) [27], among which CFS is the fastest in theory. However, frequent errors of CFS are a problem. ...
Autofocus is an important technique for high-speed image acquisition in the second-generation DNA sequencing system, and this paper studies the passive focus algorithm for the system, which consists of two parts: focus measurement (FM) and focus search (FS). Based on the properties of DNA chips’ images, we choose the normalized variance as the FM algorithm and develop a new robust FS named adaptive prediction approximation combined search (APACS). APACS utilizes golden section search (GSS) to approximate the focus position and engages the curve-fitting search (CFS) to predict the position simultaneously in every step of GSS. When the difference between consecutive predictions meets the set precision, the search finishes. Otherwise, it ends as GSS. In APACS, we also propose an estimation method, named the combination of centroid estimation and overdetermined equations estimation by least squares solution, to calculate the initial vector for the nonlinear equations in APACS prediction, which reduces the iterations and accelerates the search. The simulation and measured results demonstrate that APACS not only maintains the stability but also reduces the focus time compared with GSS and CFS, which indicates APACS is a robust and fast FS for the fluorescence microscope in a sequencing system.
A miniaturized extended short-wavelength infrared (SWIR) InGaAs microspectrometer is fabricated and characterized. An In0.83Ga0.17As detector with a cutoff wavelength of 2.6 μm is successfully grown on metamorphic InAsP layers with...
