Portable 4d Snapshot Hyperspectral Imager for Fastspectral and Surface Morphology Measurements

Luo, Jianxun; Lin, Zijian; Xing, Yuxin; Forsberg, Erik; Wu, Chengdong; Zhu, Xinhua; Guo, Tao; Wang, Gaoxuan; Bian, Beilei; Wu, Dun; He, Sailing

doi:10.2528/pier22021702

Cited by 16 publications

(8 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another related field is active 3D hyperspectral imaging. This method requires the use of additional equipment to project onto the target, enabling the hyperspectral camera to acquire three-dimensional information, as discussed in articles such as those by J.Luo et al [32] and others. While this method allows end-to-end acquisition of three-dimensional information, it has higher equipment requirements and usage conditions compared to fully image-based methods.…”

Section: B Hyperspectral-based 3dmentioning

confidence: 99%

Novel View Synthesis and Dataset Augmentation for Hyperspectral Data Using NeRF

Ma,

Guo

et al. 2024

IEEE Access

Self Cite

View full text Add to dashboard Cite

Hyperspectral data for the 3D domain is relatively difficult to acquire. Existing hyperspectral datasets are unsuitable for 3D research, suffer from issues of severe data scarcity, and a lack of multiperspective images of the same object, etc. To address these challenges, data augmentation with limited data is essential. In this study, we applied neural rendering method (such as Neural Radiance Field [1]) to hyperspectral images for dataset augmentation. We conducted experiments on novel view synthesis for hyperspectral images from 360-degree multi-perspectives, demonstrating that our method can generate high-quality hyperspectral images from various perspectives. Through experiments involving key points extraction and 3D reconstruction, we validated the efficacy of generating a substantial volume of high-quality hyperspectral images from a restricted set of varying perspectives. These results contribute to addressing the challenges associated with data augmentation. We also conducted experiments of neural radiance fields in the hyperspectral data domain under different network parameters and training conditions to find the appropriate settings.INDEX TERMS dataset augmentation, hyperspectral image, NeRF, novel view synthesis.

show abstract

Section: B Hyperspectral-based 3dmentioning

confidence: 99%

Novel View Synthesis and Dataset Augmentation for Hyperspectral Data Using NeRF

Ma,

Guo

et al. 2024

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

“…By scanning the whole spectral domain temporally, the input spectrum can be recovered through one‐to‐one sampling. [ 10,11 ] Compared to the snapshot spectroscopic analysis based on filter arrays, [ 12,13 ] quantum dots, [ 12,14 ] photonic crystals [ 15 ] and nanowires, [ 8 ] tunable filter‐based approaches possess the advantages of ultra‐compact volume for spectral imaging with both high spectral and spatial resolution. Relying on various active materials, plenty of narrowband tunable filters have been proposed to work as compact spectrometers or spectral imagers.…”

Section: Introductionmentioning

confidence: 99%

A High Throughput Tunable Filter Module for Multispectral Imaging

Chen,

Guo,

Lin

et al. 2024

Advanced Optical Materials

Self Cite

View full text Add to dashboard Cite

Tunable filters are vital not only for dynamic display but also in the miniaturized spectral imaging system. Yet, the present tunable devices possess low optical throughput, bulky volume, slow response, or complex configuration. In this paper, the study proposed and experimentally demonstrated a compact tunable filter module based on metagratings combined with liquid crystals (LC). By the synergetic effect between the birefringent LC and the dichroic metagratings, the module possesses distinct transmissive features by electrically tuning the LC material, resulting from the polarization interference effect. A simple nanoimprinting method, which is suitable for the mass production of nanostructures is adopted for low‐cost fabrication. These unique features endow the present tunable filter module with low cost, high optical throughput, and compact volume. As a proof of concept, a multispectral imager is built with this filter based on a commercial microscope. Experimental results show the system response is five times higher in optical throughput across the whole visible region than the traditional liquid crystal tunable filter and the signal‐to‐noise ratio can be improved by 7 dB. Considering the feasibility of nanoimprinting, this module paves the way for spectral imaging in low‐light conditions and on‐site scenarios such as tongue diagnosis and food quality inspection.

show abstract

“…Hyperspectral detection, as a non-invasive, chemical-free technique (see e.g., [11]), can provide spectral information that serves as the bacteria's fingerprint, and has been widely utilized in pathogenic bacteria identification and classification, with the advantages of rapid detection, low production cost, and convenient operation [12,13]. A few studies focused on combining hyperspectral detection with different machine learning (ML) frameworks, e.g., support vector machine (SVM) [14], fusion-net [15], and BI-Net [16], to increase the number of identified bacteria species and improve classification accuracy.…”

Section: Introductionmentioning

confidence: 99%

Detecting Multiple Mixed Bacteria Using Dual-Mode Hyperspectral Imaging and Deep Neural Networks

Zhu,

Luo,

2024

Applied Sciences

Self Cite

View full text Add to dashboard Cite

Identifying and analyzing mixed pathogenic bacteria is important for clinical diagnosis and antibiotic therapy of multiple bacterial infection. In this paper, a dual-mode hyperspectral microscopic detection technology with hybrid deep neural networks (DNNs) was proposed for simultaneous quantitative analysis of four kinds of pathogenic bacteria in mixed samples. To acquire both transmission and fluorescence spectra regarding the mixed pathogens, we developed a dual-mode hyperspectral detection system with fine spectral resolution and wide wavelength range, which can also generate spatial images that can be used to calculate the total amount of mixed bacteria. The dual-mode spectra were regarded as mixed proportion characteristics and the input of the neural network for predicting the proportion of each bacterium present in the mixture. To better analyze the dual-mode spectral data, we customized a mixed bacteria measurement network (MB-Net) with hybrid DNNs architectures based on spectral feature fusion. Using the fusion strategy, two DNNs frameworks applied for transmission/fluorescence spectral feature processing were stacked to form the MB-Net that processes these features simultaneously, and the achieved average coefficient of determination (R2) and RMSE of validation set are 0.96 and 0.03, respectively. To the best of our knowledge, it is the first time of simultaneously detecting four types of mixed pathogenic bacteria using spectral detection technology, showing excellent potential in clinical practice.

show abstract

Portable 4d Snapshot Hyperspectral Imager for Fastspectral and Surface Morphology Measurements

Cited by 16 publications

References 42 publications

Novel View Synthesis and Dataset Augmentation for Hyperspectral Data Using NeRF

Novel View Synthesis and Dataset Augmentation for Hyperspectral Data Using NeRF

A High Throughput Tunable Filter Module for Multispectral Imaging

Detecting Multiple Mixed Bacteria Using Dual-Mode Hyperspectral Imaging and Deep Neural Networks

Contact Info

Product

Resources

About