Combination of Structured Illumination Microscopy with Hyperspectral Imaging for Cell Analysis.

Existing structured illumination microscopy (SIM) allows super-resolution live-cell imaging in few color channels that provide merely morphological information but cannot acquire the sample spectrum that is strongly relevant to the underlying physicochemical property. We develop hyperspectral SIM which enables high-speed spectral super-resolution imaging in SIM for the first time. Through optically mapping the three-dimensional (x, y, and λ) datacube of the sample to the detector plane, hyperspectral SIM allows snapshot spectral imaging of the SIM raw image, detecting the sample spectrum while retaining the high-speed and super-resolution characteristics of SIM. We demonstrate hyperspectral SIM imaging and reconstruct a datacube containing 31 super-resolution images of different wavelengths from only 9 exposures, achieving a 15 nm spectral resolution. We show time-lapse hyperspectral SIM imaging that achieves an imaging speed of 2.7 s per datacube-31-fold faster than the existing wavelength scanning strategy. To demonstrate the great prospects for further combining hyperspectral SIM with various spectral analysis methods, we also perform spectral unmixing of the hyperspectral SIM result while imaging the spectrally overlapped sample.

Snapshot spectroscopic microscopy with double spherical slicer mirrors.

Snapshot hyperspectral microscopic imaging can obtain the morphological characteristics and chemical specificity of samples simultaneously and instantaneously. We demonstrate a double-slicer spectroscopic microscopy (DSSM) that uses two spherical slicer mirrors to magnify the target image and slice it. These slits are lined up and dispersed, then mapped onto an area-array detector. An anamorphosis unit optimizes the capacity of the limited pixels. With a single shot and image recombination, a data cube can be constructed for sample analysis, and a model of DSSM is simulated. The system covers the spectral range from 500 nm to 642.5 nm with 20 spectral channels. The spatial resolution is 417 nm, and the spectral resolution is 7.5 nm.

Simultaneous multicolour imaging using quantum dot structured illumination microscopy.

Multicolour structured illumination microscopy (SIM) is a powerful tool used for the investigation of the dynamic interaction between subcellular structures. Nevertheless, most of the multicolour SIM schemes are currently limited by conventional fluorescent dyes and wavelength-dependent optical systems, and can only sequentially record images of different colour channels instead of obtaining multicolour datasets simultaneously. To address these issues, we present a novel multicolour SIM scheme referred to as quantum dot structured illumination microscopy (QD-SIM). QD-SIM enables simultaneously excitation and collection of multicolour fluorescent signals. We also propose a theoretical analysis of the image formation in two-dimensional multicolour SIM to help combine the optically sectioned and super-resolution attributes of SIM. Based on this theory, QD-SIM enables optically sectioned, super-resolution, multicolour simultaneous imaging at a single plane.

Broadband aberration-corrected snapshot spectrometer with a toroidal slicer mirror.

We demonstrate a broadband aberration-corrected snapshot spectrometer by developing a toroidal slicer mirror as the focusing mirror. A collimating slicer mirror and an integrated grating divide the entire wavelength range into several windows. The toroidal sub-mirrors of the focusing mirror compensate for the coma and correct the astigmatism at different windows to compress the spectral spots over the waveband. From 200 to 1000 nm, the RMS spot radii are less than 30 µm and the spectral resolution at 600 nm is 151.8 pm. The optics have a small volume of 11 cm×11 cm×1.5 cm by employing a folded structure.

A Multi-View Stereo Measurement System Based on a Laser Scanner for Fine Workpieces.

A new solution to the high-quality 3D reverse modeling problem of complex surfaces for fine workpieces is presented using a laser line-scanning sensor. Due to registration errors, measurement errors, deformations, etc., a fast and accurate method is important in machine vision measurement. This paper builds a convenient and economic multi-view stereo (MVS) measurement system based on a linear stage and a rotary stage to reconstruct the measured object surface completely and accurately. In the proposed technique, the linear stage is used to generate the trigger signal and synchronize the laser sensor scanning; the rotary stage is used to rotate the object and obtain multi-view point cloud data, and then the multi-view point cloud data are registered and integrated into a 3D model. The measurement results show a measurement accuracy of 0.075 mm for a 360° reconstruction in 34 s, and some evaluation experiments were carried out to demonstrate the validity and practicability of the proposed technique.

Compact high-resolution spectrometer using two plane gratings with triple dispersion.

We demonstrate a compact high-resolution spectrometer scheme using two plane gratings. In this approach, the rays are first diffracted by a fixed grating, then incident on a rotating grating at the Littrow diffraction angle, and are finally diffracted and reflected back to the fixed grating again. Thus, triple dispersion (TD) occurs during measurement, increasing the resolution. The formulae of this compact high-resolution spectrometer are rigorously derived. A design simulation with two gratings of 1050 lines/mm is performed and discussed. In addition, a prototype of this spectrometer has been built and tested. Its spectral resolution reaches a precision of 36 pm.

Detection of Micro-Defects on Metal Screw Surfaces Based on Deep Convolutional Neural Networks.

This paper proposes a deep convolutional neural network (CNN) -based technique for the detection of micro defects on metal screw surfaces. The defects we consider include surface damage, surface dirt, and stripped screws. Images of metal screws with different types of defects are collected using industrial cameras, which are then employed to train the designed deep CNN. To enable efficient detection, we first locate screw surfaces in the pictures captured by the cameras, so that the images of screw surfaces can be extracted, which are then input to the CNN-based defect detector. Experiment results show that the proposed technique can achieve a detection accuracy of 98%; the average detection time per picture is 1.2 s. Comparisons with traditional machine vision techniques, e.g., template matching-based techniques, demonstrate the superiority of the proposed deep CNN-based one.

Compact broadband high-resolution infrared spectrometer with a dihedral reflector.

We demonstrate a compact broadband high-resolution spectrometer approach. A dihedral reflector is used to reflect the dispersed light back to the grating for a second diffraction, folding the light path in a compact space, and enhancing the spectral resolution. The theoretical formulas for the system are strictly derived. In addition, a prototype of this spectrometer for fiber communication in the infrared wavelength range has been built. The optics can fit inside a volume of 12 cm × 14 cm × 5 cm and its spectral resolution is 57 pm over a wide wavelength range from 1250 nm to 1650 nm.

Coherent anti-Stokes Raman scattering imaging under ambient light.

We demonstrate an ambient light coherent anti-Stokes Raman scattering microscope that allows CARS imaging to be operated under environmental light for field use. The CARS signal is modulated at megahertz frequency and detected by a photodiode equipped with a lab-built resonant amplifier, then extracted through a lock-in amplifier. The filters in both the spectral domain and the frequency domain effectively blocked the room light contamination of the CARS image. In situ hyperspectral CARS imaging of tumor tissue under ambient light is demonstrated.

[Detection System for High-Resolution and Broadband-2D Spectrogram].

In order to meet the demand of detecting the high-resolution and broadband-2D spectrogram for echelle-prism cross-dispersion, the design method is given, and the 2D spectrogram detection system with high-resolution and low noise is designed by analyzing the relation between the detector and the optical system and proving designing method of detecting the area spectrogram. The system includes a main control unit, a detector driving unit, a signal processing unit, a data storage unit and a data transmission unit, etc. With Hamamatsu's S10141-1109S CCD as the detector, the detection system features high sensitivity, broadband spectrogram, high signal to noise ratio. Combined with the echelle-prism cross-dispersion optical path to perform experiments, the results show that, the detection system can acquire high-resolution 2D spectrogram image in the range of 200 to 600 nm, the monochromatic image at 253.652 nm of Hg lamp covers 5 pixels, and the resolvable wavelength reaches 6.3 pm.

[Optical path difference in off-plane quasi-Littrow dispersion mountings].

The present paper analyzes the relative relation between the meridian and sagittal rays in off-plane quasi-Littrow (OP-QL) dispersion mountings. It's concluded that the off-plane angle will cause the rotation of the beam and result in the mismatch between the sagittal beams on different optical elements. Therefore the total optical path difference (OPD) should be an accumulation of corresponding beams instead of the sagittal beam of each element itself. Then, a directional derivative based method is put forward to calculate the OPD for spherical mirrors in various directions. Based on the method, the numerical OPD for OP-QL mountings is solved. Finally, this methodology is validated with both echelette and echelle examples.

Approximate analytic astigmatism of unit-magnification multipass system.

We develop a way to estimate the approximate analytic astigmatism with a high accuracy for any unit-magnification multipass system (UMS). The coaxial optical transmission model for UMS is simplified based on the system's features. Furthermore, astigmatism is derived as a distinct form of vector addition and, thus, feasible analytic astigmatism can be obtained. The effectiveness of our method is verified by simulations for a Bernstein-Herzberg White cell. In our cases, the relative error of optimization for astigmatism correction by our method is smaller than 5 per thousand, which is only one-tenth of that by Kohn's method. Our method significantly improves the efficiency for astigmatism correction, and further benefits the optical design of a UMS.

Generalized method for calculating astigmatism of the unit-magnification multipass system.

A generalized method to accurately calculate astigmatism of the unit-magnification multipass system (UMS) is proposed. A practical coaxial optical transmission model is developed for the UMS. Astigmatism analysis is then made convenient by a 4 by 4 general transfer matrix. Astigmatism correction is significantly promoted, and hence further improvement in imaging quality can be expected. Good agreement between numerical simulations and Zemax ray tracing results verifies the effectiveness of this method. The resulted RMS spot size of this method is only 25% to 64% of other previous methods based on the golden section search for minimum astigmatism in real design cases. This method is helpful for the optical design of the UMS.

[Coma and resolution in wide spectral region Czerny-Turner spectrometer].

The Czerny-Turner layout, which is most frequently used in miniature spectrometers, should follow Shafer's coma-free condition and Fastie's flat-field principal to eliminate the central wave's primary coma and maximize its resolution. However, the design process does not take the comas and resolutions at non-central waves into consideration. Based on the theory of primary coma in reflection optical system, the present paper points out that in the crossed beam design, the resolutions at wide 'spectral region present a "V" shape, while in the M design, the resolutions change little over the whole region, and present an approximately straight line shape, so the latter kind of spectrometer maintains a far more consistent resolution than the former one. Accordingly, this paper designs two kinds of spectrometers with spectrum regions from 400 to 600 nm, and carries out theoretical simulation and contrast experiment. The result demonstrates that for the two designs the resolutions at the fringe wavelength are 3.8 times and 1.5 times respectively that at the central wavelength, which accords with the conclusion of the theoretical simulation.

[Algorithm for restoring spectrogram with sub-pixel resolution].

Insufficient spatial resolution of detectors is an obstacle to capturing spectra with high resolution. An effective solution is sub-pixel restoration, which can restore a high resolution spectrogram from sub-pixel shifted low resolution ones. In the present paper, an algorithm for sub-pixel restoration is suggested. It utilizes the bidirectional recursive relation between sub-pixel values and estimates each sub-pixel value from both directions of head-end and end-head. As a result, the averaged value will be accepted as the sub-pixel value. Numerical experiments on single gauss profile and superposition ones verified the effectiveness of the algorithm.

[Global modeling and analyzing of grating spectrometers].

A method for globally modeling and analyzing grating spectrometers is put forward in the present paper. Different from existing methods which are confined to parts of a spectrometer, the method takes a grating spectrometer as four functional parts, namely imaging optics, detector, reconstruction and display. Effects of each part on spectrum are considered and a global model of the spectrometer is developed, accompanied with its transfer function. With the help of the model and the transfer function, laws of each part affecting the holistic performance are summed up. It is suggested that high quality spectrogram needs enhancing baseband response and reducing spurious response, and reconstruction is an effective way.

[Comparison of dispersion parts of conventional spectrometers].

Innovation of conventional spectrometers is of actual technical and economical value. It is also an important way to accelerate the development of spectroscopic instruments. When improving a conventional spectrometer, its dispersion part is pivotal, because it is decisive to the basic performance of the spectrometer. In the present paper, the typical dispersion parts of conventional spectrometers are compared to feature them and find the evolution force among them. The basic characters of the dispersion parts, including spectral range, dispersion power, resolution and throughput, are compared separately and comprehensively by reviewing their decisive factor, formula and typical data. The results not only conclude the feature and the complementariness of the dispersion parts, but also indicate that the trade-off between resolution and throughput is ubiquitous in traditional spectrometers. Further reviewing from this point, the evolution history of traditional spectrometers shows that the conflict between resolution and throughput is an important evolution force. This is a new way to understand the evolution of traditional spectrometers. Moreover, dealing with the trade-off between resolution and throughput correctly will help to analyze and settle the core problem of spectrometers.

[Blind deconvolution algorithm for spectrogram super-resolution restoration].

Deconvolution is an important way to realize spectrogram super-resolution restoration. Blind deconvolution is superior to the traditional one in that it does not need a well prepared convolution core. Taking advantages of the features of spectrogram and the existing achievements of spectrogram deconvolution, the authors bring forward a scheme to adapt the space domain iterative blind deconvolution method to spectroscopy application. Moreover, after probing into the spectrogram degradation described by convolution, computational models for spectrum convolution and Gauss fitting are worked out to meet the requirements of blind deconvolution algorithm. Accompanying results are simulations with MATLAB7.0. They shows that for the given spectrum and point spread function of Gauss type the blind deconvolution algorithm works well and a resolution enhancement of 30% can be achieved under a signal-to-noise ratio of 50 dB.

[The effect of the photoelectric detector on the accuracy of the spectrometer].

An optimized photoelectric detector will increase the precision of a spectrometer, thus indicates an important way to develop high performance spectrometer. With an eye to this, a model describing the process that spectrogram is integrated and sampled by photoelectric detector and restored after low-pass filtering is developed. Based on the model, the influence of the characteristic parameters of the detector on the spectral line in the frequency domain is analyzed and the relation between the full width half maximum (FWHM) of the spectra line and the integral interval, sampling space and sensitivity of the detector is deduced. The conclusion indicates that both the integral interval and sampling space should be 1/6 of the FWHM for a spectral line with gaussian profile as a result of compromise between accuracy and feasibility. Moreover, the critical point deciding the right situation for scanner and array detector is given. Other guide line to optimize the photoelectric detector and increase accuracy is suggested also.

Resolution enhancement by combination of subpixel and deconvolution in miniature spectrometers.

The resolution of a miniature spectrometer with a multichannel detector is limited by its throughput. A subpixel deconvolution method is proposed to enhance resolution without physically reducing the throughput. The method introduces subpixel reconstruction to overcome undersampling during deconvolution processing. The experimental result has shown a 36.6% reduction in FWHM of spectral lines, indicating the effectiveness of the method.