Microlens array snapshot hyperspectral microscopy system for the biomedical domain

Fabrication of Large-Area Silicon Spherical Microlens Arrays by Thermal Reflow and ICP Etching

In this paper, we proposed an efficient and high-precision process for fabricating large-area microlens arrays using thermal reflow combined with ICP etching. When the temperature rises above the glass transition temperature, the polymer cylinder will reflow into a smooth hemisphere due to the surface tension effect. The dimensional differences generated after reflow can be corrected using etching selectivity in the following ICP etching process, which transfers the microstructure on the photoresist to the substrate. The volume variation before and after reflow, as well as the effect of etching selectivity using process parameters, such as RF power and gas flow, were explored. Due to the surface tension effect and the simultaneous molding of all microlens units, machining a 3.84 × 3.84 mm2 silicon microlens array required only 3 min of reflow and 15 min of ICP etching with an extremely low average surface roughness Sa of 1.2 nm.

Development of Shape Prediction Model of Microlens Fabricated via Diffuser-Assisted Photolithography

The fabrication of microlens arrays (MLAs) using diffuser-assisted photolithography (DPL) has garnered substantial recent interest owing to the exceptional capabilities of DPL in adjusting the size and shape, achieving high fill factors, enhancing productivity, and ensuring excellent reproducibility. The inherent unpredictability of light interactions within the diffuser poses challenges in accurately forecasting the final shape and dimensions of microlenses in the DPL process. Herein, we introduce a comprehensive theoretical model to forecast microlens shapes in response to varying exposure doses within a DPL framework. We establish a robust MLA fabrication method aligned with conventional DPL techniques to enable precise shape modulation. By calibrating the exposure doses meticulously, we generate diverse MLA configurations, each with a distinct shape and size. Subsequently, by utilizing the experimentally acquired data encompassing parameters such as height, radius of curvature, and angles, we develop highly precise theoretical prediction models, achieving R-squared values exceeding 95%. The subsequent validation of our model encompasses the accurate prediction of microlens shapes under specific exposure doses. The verification results exhibit average error rates of approximately 2.328%, 7.45%, and 3.16% for the height, radius of curvature, and contact angle models, respectively, all of which were well below the 10% threshold.

Dual-channel snapshot imaging spectrometer with wide spectrum and high resolution

The comprehensive analysis of dynamic targets brings about the demand for capturing spatial and spectral dimensions of visual information instantaneously, which leads to the emergence of snapshot spectral imaging technologies. While current snapshot systems face major challenges in the development of wide working band range as well as high resolution, our novel dual-channel snapshot imaging spectrometer (DSIS), to the best of our knowlledge, demonstrates the capability to achieve both wide spectrum and high resolution in a compact structure. By dint of the interaction between the working band range and field of view (FOV), reasonable limits on FOV are set to avoid spectral overlap. Further, we develop a dual-channel imaging method specifically for DSIS to separate the whole spectral range into two parts, alleviating the spectral overlap on each image surface, improving the tolerance of the system for a wider working band range, and breaking through structural constraints. In addition, an optimal FOV perpendicular to the dispersion direction is determined by the trade-off between FOV and astigmatism. DSIS enables the acquisition of 53×11 spatial elements with up to 250 spectral channels in a wide spectrum from 400 to 795 nm. The theoretical study and optimal design of DSIS are further evaluated through the simulation experiments of spectral imaging.

Spatial-spectral resolution tunable snapshot imaging spectrometer: analytical design and implementation

A snapshot imaging spectrometer is a powerful tool for dynamic target tracking and real-time recognition compared with a scanning imaging spectrometer. However, all the current snapshot spectral imaging techniques suffer from a major trade-off between the spatial and spectral resolutions. In this paper, an integral field snapshot imaging spectrometer (TIF-SIS) with a continuously tunable spatial-spectral resolution and light throughput is proposed and demonstrated. The proposed TIF-SIS is formed by a fore optics, a lenslet array, and a collimated dispersive subsystem. Theoretical analyses indicate that the spatial-spectral resolution and light throughput of the system can be continuously tuned through adjusting the F number of the fore optics, the rotation angle of the lenslet array, or the focal length of the collimating lens. Analytical relationships between the spatial and spectral resolutions and the first-order parameters of the system with different geometric arrangements of the lenslet unit are obtained. An experimental TIF-SIS consisting of a self-fabricated lenslet array with a pixelated scale of 100×100 and a fill factor of 0.716 is built. The experimental results show that the spectral resolution of the system can be steadily improved from 4.17 to 0.82 nm with a data cube (N x×N y×N λ) continuously tuned from 35×35×36 to 40×40×183 in the visible wavelength range from 500 to 650 nm, which is consistent with the theoretical prediction. The proposed method for real-time tuning of the spatial-spectral resolution and light throughput opens new possibilities for broader applications, especially for recognition of things with weak spectral signature and biomedical investigations where a high light throughput and tunable resolution are needed.

Mold-free self-assembled scalable microlens arrays with ultrasmooth surface and record-high resolution

Microlens arrays (MLAs) based on the selective wetting have opened new avenues for developing compact and miniaturized imaging and display techniques with ultrahigh resolution beyond the traditional bulky and volumetric optics. However, the selective wetting lenses explored so far have been constrained by the lack of precisely defined pattern for highly controllable wettability contrast, thus limiting the available droplet curvature and numerical aperture, which is a major challenge towards the practical high-performance MLAs. Here we report a mold-free and self-assembly approach of mass-production of scalable MLAs, which can also have ultrasmooth surface, ultrahigh resolution, and the large tuning range of the curvatures. The selective surface modification based on tunable oxygen plasma can facilitate the precise pattern with adjusted chemical contrast, thus creating large-scale microdroplets array with controlled curvature. The numerical aperture of the MLAs can be up to 0.26 and precisely tuned by adjusting the modification intensity or the droplet dose. The fabricated MLAs have high-quality surface with subnanometer roughness and allow for record-high resolution imaging up to equivalently 10,328 ppi, as we demonstrated. This study shows a cost-effective roadmap for mass-production of high-performance MLAs, which may find applications in the rapid proliferating integral imaging industry and high-resolution display.

Analytical design of a cemented-curved-prism based integral field spectrometer (CIFS) with high numerical aperture and high resolution

Snapshot hyperspectral imaging is superior to scanning spectrometers due to its advantage in dimensionality, allowing longer pixel dwell time and higher data cube acquisition efficiency. Due to the trade-off between spatial and spectral resolution in snapshot spectral imaging technologies, further improvements in the performance of snapshot imaging spectrometers are limited. Therefore, we propose a cemented-curved-prism-based integral field spectrometer (CIFS), which achieves high spatial and high spectral resolution imaging with a high numerical aperture. It consists of a hemispherical lens, a cemented-curved-prism and a concave spherical mirror. The design idea of aplanatic imaging and sharing-optical-path lays the foundation for CIFS to exhibit high-resolution imaging in a compact structure. The numerical model between the parameters of optical elements and the spectral resolution of the system is established, and we analyze the system resolution influenced by the hemispherical lens and the cemented-curved-prism. Thus, the refractive index requirements of the hemispherical lens and the cemented-curved-prism for the optimal spatial and spectral resolution imaging of the system are obtained, providing guidance for the construction of CIFS. The designed CIFS achieves pupil matching with a 1.8 f-number lenslet array, sampling 268 × 76 spatial points with 403 spectral channels in the wavelength band of 400 to 760 nm. The spectral and spatial resolution are further evaluated through a simulation experiment of spectral imaging based on Zemax. It paves the way for developing integral field spectrometers exhibiting high spatial and high spectral resolution imaging with high numerical aperture.

Backtracking Reconstruction Network for Three-Dimensional Compressed Hyperspectral Imaging

Compressed sensing (CS) has been widely used in hyperspectral (HS) imaging to obtain hyperspectral data at a sub-Nyquist sampling rate, lifting the efficiency of data acquisition. Yet, reconstructing the acquired HS data via iterative algorithms is time consuming, which hinders the real-time application of compressed HS imaging. To alleviate this problem, this paper makes the first attempt to adopt convolutional neural networks (CNNs) to reconstruct three-dimensional compressed HS data by backtracking the entire imaging process, leading to a simple yet effective network, dubbed the backtracking reconstruction network (BTR-Net). Concretely, we leverage the divide-and-conquer method to divide the imaging process based on coded aperture tunable filter (CATF) spectral imager into steps, and build a subnetwork for each step to specialize in its reverse process. Consequently, BTR-Net introduces multiple built-in networks which performs spatial initialization, spatial enhancement, spectral initialization and spatial–spectral enhancement in an independent and sequential manner. Extensive experiments show that BTR-Net can reconstruct compressed HS data quickly and accurately, which outperforms leading iterative algorithms both quantitatively and visually, while having superior resistance to noise.

Image Correction and In Situ Spectral Calibration for Low-Cost, Smartphone Hyperspectral Imaging

Developments in the portability of low-cost hyperspectral imaging instruments translate to significant benefits to agricultural industries and environmental monitoring applications. These advances can be further explicated by removing the need for complex post-processing and calibration. We propose a method for substantially increasing the utility of portable hyperspectral imaging. Vertical and horizontal spatial distortions introduced into images by ‘operator shake’ are corrected by an in-scene reference card with two spatial references. In situ light-source-independent spectral calibration is performed. This is achieved by a comparison of the ground-truth spectral reflectance of an in-scene red–green–blue target to the uncalibrated output of the hyperspectral data. Finally, bias introduced into the hyperspectral images due to the non-flat spectral output of the illumination is removed. This allows for low-skilled operation of a truly handheld, low-cost hyperspectral imager for agriculture, environmental monitoring, or other visible hyperspectral imaging applications.

Research on spectral reconstruction algorithm for snapshot microlens array micro-hyperspectral imaging system

Snapshot microlens array microscopic hyperspectral imaging systems do not require a scanning process and obtain (x,y,λ) three-dimensional data cubes in one shot. Currently, the three-dimensional spectra image data are interleaved on a charge-coupled device detector, which increases subsequent data processing difficulty. The optical design software cannot simulate actual engineering installation and adjustment results accurately and the tracking results cannot guide precise rapid online calibration of the snapshot microlens array microscopic hyperspectral imaging system. To solve these problems, we propose an accurate spectral image reconstruction model based on optical tracing, derive spatial dispersion equations for the prisms and gratings, establish an algorithm model for the correspondence between the microlens array's surface dispersion spectral distribution and its imaging position, and propose a three-dimensional spectral image reconstruction algorithm. Experimental results show that this algorithm's actual spectral calibration error is better than 0.2 nm. This meets the image processing requirements of snapshot microlens array microscopic hyperspectral systems.