Hyperspectral Imaging Method Research Articles

Spatial-spectral feature mining in hyperspectral corn leaf venation structure and its application in nitrogen content estimation

The growth of corn plants requires a significant amount of nitrogen nutrient supply to ensure the yield. Meanwhile, overfertilization could result in adverse environmental impacts, making it crucial to monitor the nitrogen supply condition in corn plants precisely. Hyperspectral imaging (HSI) could be the solution to this challenge, as HSI plant phenotyping technology has been widely utilized to nondestructively identify plant traits, such as mineral nutrient deficiencies in corn plants. However, limitations persist where the conventional HSI nitrogen prediction models mainly relied on the overall colors of the plant tissue extracted from the spectral domain but rarely included the spatially distributed information on the leaf level. This study aimed to examine if including new information from the spatial domain could potentially improve the performance of HSI models. Because the venation structure plays a crucial role in the distribution of various nutrients across the leaf, a series of physiological responses caused by nitrogen deficiency could result in non-uniform color patterns related to the veins. The recent advancement of HSI techniques such as LeafSpec, whose imaging quality and spatial-spectral resolution have been significantly improved, made it possible to extract the venation structure together with high-resolution spatial-spectral features. In this study, the hypothesis was that utilizing high-resolution spatial-spectral features extracted based on the venation structure could yield a better-performed nitrogen content prediction model compared to the averaged spectrum of the corn leaf. First, a venation structure segmentation algorithm was developed using the Local Binary Patterns (LBP) method for hyperspectral corn leaf images scanned with LeafSpec. Second, for every hyperspectral leaf image, 131 spectral index heatmaps were generated, and 30 leaf regions were picked based on the venation structure, resulting in 3930 spatial-spectral features. Third, the features were picked using the Least Absolute Shrinkage and Selection Operator (LASSO) regression. As the results tested with a field assay involving two different corn genotypes and two nitrogen treatments, the Partial Least Square Regression (PLS-R) model built with selected spatial-spectral features outperformed the model built with the averaged leaf spectra in terms of the Root Mean Squared Error (RMSE) of predictions. The results provided encouraging proof and guidance for the potential benefit of using high-resolution spatial-spectral features derived from hyperspectral leaf images to improve the accuracy and robustness of the nitrogen content prediction models for corn plants.

Hyperspectral imaging in neurosurgery: a review of systems, computational methods, and clinical applications.

Accurate identification between pathologic (e.g., tumors) and healthy brain tissue is a critical need in neurosurgery. However, conventional surgical adjuncts have significant limitations toward achieving this goal (e.g., image guidance based on pre-operative imaging becomes inaccurate up to 3cm as surgery proceeds). Hyperspectral imaging (HSI) has emerged as a potential powerful surgical adjunct to enable surgeons to accurately distinguish pathologic from normal tissues. We review HSI techniques in neurosurgery; categorize, explain, and summarize their technical and clinical details; and present some promising directions for future work. We performed a literature search on HSI methods in neurosurgery focusing on their hardware and implementation details; classification, estimation, and band selection methods; publicly available labeled and unlabeled data; image processing and augmented reality visualization systems; and clinical study conclusions. We present a detailed review of HSI results in neurosurgery with a discussion of over 25 imaging systems, 45 clinical studies, and 60 computational methods. We first provide a short overview of HSI and the main branches of neurosurgery. Then, we describe in detail the imaging systems, computational methods, and clinical results for HSI using reflectance or fluorescence. Clinical implementations of HSI yield promising results in estimating perfusion and mapping brain function, classifying tumors and healthy tissues (e.g., in fluorescence-guided tumor surgery, detecting infiltrating margins not visible with conventional systems), and detecting epileptogenic regions. Finally, we discuss the advantages and disadvantages of HSI approaches and interesting research directions as a means to encourage future development. We describe a number of HSI applications across every major branch of neurosurgery. We believe these results demonstrate the potential of HSI as a powerful neurosurgical adjunct as more work continues to enable rapid acquisition with smaller footprints, greater spectral and spatial resolutions, and improved detection.

Development of a Hyperspectral Imaging Protocol for Painting Applications at the University of Seville

In the last decade, the hyperspectral imaging (HSI) method allowed performing non-invasive analysis in the field of cultural heritage. However, a considerable limitation was given by redundant and time-consuming features, with the necessary application of statistical algorithms and image-processing tools to extract relevant information. In this study, the Centro Nacional de Aceleradores (CNA) and the Group of Interdisciplinary Physics (GFI) of the School of Engineering (ETSI) of the University of Seville tested the application of three different hyperspectral cameras in the visible and near-infrared (VNIR) and short-wave infrared (SWIR) range for the investigation of an ancient painting. A reference-based procedure was realised to build a starting personal library and to evaluate the best working conditions for non-invasive and non-destructive characterisation with data treatment using the commercially available software Evince® and Specim IQ® to apply, respectively, the Principal Component Analysis (PCA) model functions and the classification method. The evaluation of the protocol was tested by acquiring complementary information by X-ray fluorescence (XRF), Ultraviolet Luminescence (UVL) imaging, and Infrared Reflectography (IRR). This exploration established a simplified protocol to analyse the large collection of paintings of the Archbishop’s Palace and the Cathedral of Seville.

Advancements in Hyperspectral Imaging and Computer-Aided Diagnostic Methods for the Enhanced Detection and Diagnosis of Head and Neck Cancer.

Background/Objectives: Head and neck cancer (HNC), predominantly squamous cell carcinoma (SCC), presents a significant global health burden. Conventional diagnostic approaches often face challenges in terms of achieving early detection and accurate diagnosis. This review examines recent advancements in hyperspectral imaging (HSI), integrated with computer-aided diagnostic (CAD) techniques, to enhance HNC detection and diagnosis. Methods: A systematic review of seven rigorously selected studies was performed. We focused on CAD algorithms, such as convolutional neural networks (CNNs), support vector machines (SVMs), and linear discriminant analysis (LDA). These are applicable to the hyperspectral imaging of HNC tissues. Results: The meta-analysis findings indicate that LDA surpasses other algorithms, achieving an accuracy of 92%, sensitivity of 91%, and specificity of 93%. CNNs exhibit moderate performance, with an accuracy of 82%, sensitivity of 77%, and specificity of 86%. SVMs demonstrate the lowest performance, with an accuracy of 76% and sensitivity of 48%, but maintain a high specificity level at 89%. Additionally, in vivo studies demonstrate superior performance when compared to ex vivo studies, reporting higher accuracy (81%), sensitivity (83%), and specificity (79%). Conclusion: Despite these promising findings, challenges persist, such as HSI's sensitivity to external conditions, the need for high-resolution and high-speed imaging, and the lack of comprehensive spectral databases. Future research should emphasize dimensionality reduction techniques, the integration of multiple machine learning models, and the development of extensive spectral libraries to enhance HSI's clinical utility in HNC diagnostics. This review underscores the transformative potential of HSI and CAD techniques in revolutionizing HNC diagnostics, facilitating more accurate and earlier detection, and improving patient outcomes.

Reconstructing hyperspectral images of textiles from a single RGB image utilizing the multihead self-attention mechanism

Hyperspectral images possess abundant information and play a pivotal role in enhancing the accuracy of color difference detection in textiles. However, traditional hyperspectral imaging methods necessitate costly equipment and intricate operational procedures. A novel deep learning model based on a multihead attention mechanism was proposed in this article to facilitate the extensive application of hyperspectral imaging technology in textile quality inspection. This model enabled the reconstruction of the hyperspectral information of plain weave textiles from a single RGB image. In this model, encoder-decoder architecture and pyramid pooling convolutional operations were employed to integrate multiscale features of plain weave cotton-linen textiles. This could capture details and contextual information in textile images more precisely, enhancing the accuracy of hyperspectral image reconstruction. Simultaneously, an attention mechanism was introduced to increase the model’s receptive field and improve its focus on key regions in the input image and feature maps. This resulted in a reduced weighting of redundant information during network learning, leading to an improved feature extraction capability of the network. Through these methods, successful reconstructions of plain weave textiles hyperspectral information from a single RGB image was achieved. Quantitative and qualitative tests were conducted on two datasets, namely, the NTIRE 2020 dataset and a self-made textile dataset, to evaluate the performance of the proposed method. The approach proposed in this article exhibited promising results on both datasets. Specifically, the reconstructed textile hyperspectral images achieved a root mean square error of 0.0344, a peak signal-to-noise ratio of 29.945, a spectral angle mapper of 3.753, and a structural similarity index measure of 0.955 on the textile dataset. In the reconstructed hyperspectral colorimetric experiment, the maximum value of average color difference was 2.641. These results demonstrate that the method can meet the requirements for textile color measurement applications.

Quality monitoring of glutinous rice processing from drying to extended storage using hyperspectral imaging

The quality of glutinous rice (GR) is susceptible to deterioration and losses due to biological or environmental factors during storage. Traditional quality assessment techniques are often time-consuming and challenging. In this study, a rapid and reliable hyperspectral imaging (HSI) technique is utilized to monitor GR quality during storage. Paddy samples were dried at 50 °C, 60 °C and 70 °C. Subsequently, these samples were milled and stored under three conditions: freeze storage (−10 °C), cold room (6 °C) and ambient (∼26 °C) for 6 months. The methodology involved data acquisition from both HSI and standard references methods, with data on hyperspectral reflectance, head rice yield (HRY), broken rice yield (BRY) and milled rice yield (MRY) collected every two weeks. Five machine learning (ML) models were evaluated for quality prediction using Python3, with Random Forest (RF) identified as the best performer, achieving a coefficient of determination (R2) of 0.995. Hyperparameter tuning (HPT) further improved the RF model’s R2 by 0.3 %. Parity plot analysis confirmed the accuracy of the RF model in describing GR quality during storage. The study demonstrates the significant impacts of different storage and drying temperatures on HSI data and GR quality attributes. Significant differences in reflectance were observed, with higher reflectance for samples dried at 60 °C and freeze-storage, while lower reflectance for samples dried at 70 °C and cold-room storage. These findings align with reference method results and ML predictions, revealing that drying paddy at 60 °C and storing it under freeze conditions enhances HRY and increases the commercial value of GR. Overall, this study highlights the potential of the HSI for real-time quality monitoring of GR and its applicability to other grains.

Determination of malathion content in sorghum grains using hyperspectral imaging technology combined with stacked machine learning models

The rapid and precise detection of pesticide residues remains a pressing safety issue in the food industry. In this study, a rapid method for analyzing pesticide residues in sorghum was developed, combining hyperspectral imaging (HSI) technology with stacking ensemble learning (SEL) models. The HSI spectral data were preprocessed using the multivariate scatter correction (MSC) algorithm. The gradient boosting decision tree (GBDT), extreme gradient boosting (XGBoost), light gradient boosting machine (LGBM), and categorical boosting (CatBoost) algorithms were employed to identify the feature wavelengths with high contributions to the predictive model, and the performances of SEL, GBDT, XGBoost, LGBM, and CatBoost to accurately predict the pesticide residues in sorghum samples were compared. The SEL model constructed using the characteristic wavelength selected by CatBoost has the best predictive performance, with RMSEP, RP2, and RPD values of 0.6940 mg/kg, 0.9798, and 7.029, respectively. The study demonstrated that the combination of HSI and SEL enabled the accurate analysis of pesticide residues in sorghum, providing a reference for the utilization of HSI methods to accurately measure the concentrations of pesticide residues in sorghum and other food products.

Object-plane encoding hyperspectral three-dimensional collaborative imaging method.

Different from traditional image-plane encoding spectral imaging method, this paper proposes an object-plane encoding hyperspectral three-dimensional collaborative imaging method, which is based on the idea of spectral displacement invariance and multiplexing encoding and decoding. The 4D data of the scene is scanned by coded apertures generated by the projector, and then the encoded data of the scene is obtained. During the reconstruction process, there is no need for phase unwrapping operation or data fusion. The reconstruction result shows excellent performance with a spectral resolution of 3 nm, nearly 120 spectral bands, depth accuracy of 0.25 mm. Compared with the image-plane encoding spectral imaging method, the object-plane encoding spectral imaging method exhibits better anti-interference characteristics, and the reconstruction quality of the object-plane encoding spectral imaging method is better than that of the image-plane encoding spectral imaging method in the scattering environments.

Detection and margin assessment of thyroid carcinoma with microscopic hyperspectral imaging using transformer networks.

Hyperspectral imaging (HSI) is an emerging imaging modality for oncological applications and can improve cancer detection with digital pathology. The study aims to highlight the increased accuracy and sensitivity of detecting the margin of thyroid carcinoma in hematoxylin and eosin (H&E)-stained histological slides using HSI and data augmentation methods. Using an automated microscopic imaging system, we captured 2599 hyperspectral images from 65 H&E-stained human thyroid slides. Images were then preprocessed into 153,906 image patches of dimension . We modified the TimeSformer network architecture, which used alternating spectral attention and spatial attention layers. We implemented several data augmentation methods for HSI based on the RandAugment algorithm. We compared the performances of TimeSformer on HSI against the performances of pretrained ConvNext and pretrained vision transformers (ViT) networks on red, green, and blue (RGB) images. Finally, we applied attention unrolling techniques on the trained TimeSformer network to identify the biological features to which the network paid attention. In the testing dataset, TimeSformer achieved an accuracy of 90.87%, a weighted score of 89.79%, a sensitivity of 91.50%, and an area under the receiving operator characteristic curve (AU-ROC) score of 97.04%. Additionally, TimeSformer produced thyroid carcinoma tumor margins with an average Jaccard score of 0.76 mm. Without data augmentation, TimeSformer achieved an accuracy of 88.23%, a weighted score of 86.46%, a sensitivity of 85.53%, and an AU-ROC score of 94.94%. In comparison, the ViT network achieved an 89.98% accuracy, an 88.14% weighted score, an 84.77% sensitivity, and a 96.17% AU-ROC. Our visualization results showed that the network paid attention to biological features. The TimeSformer model trained with hyperspectral histological data consistently outperformed conventional RGB-based models, highlighting the superiority of HSI in this context. Our proposed augmentation methods improved the accuracy, the score, and the sensitivity score.

Compressive strength detection of tunnel lining using hyperspectral images and machine learning

Traditional tunnel lining strength detection techniques are mostly contact-based, with relatively low detection efficiency. This study innovatively proposes hyperspectral imaging method to rapidly detect tunnel concrete lining strength from a machine vision perspective. Hyperspectral cameras were employed in indoor experiments to capture hyperspectral images of concrete specimens with different compressive strength levels. The differences of concrete strength based on hyperspectral reflectance characteristics were analysed using hyperspectral images and machine learning algorithms. Firstly, the K-Nearest Neighbors (KNN) classification algorithm was used to predict the classification of the concrete hyperspectral dataset with accuracy mostly exceeding 90 %. The results indicate distinctive differences in hyperspectral reflectance characteristics among concrete specimens. Furthermore, compressive strength prediction of different concrete specimens was carried out using Principal component regression (PCR), Partial least squares regression (PLSR), and Least Squares Support Vector Machine (LSSVM) machine learning models on both original and Savitzky-Golay(S-G) processed spectral data. LSSVM and PLSR models performed excellently in the visible light spectrums(400–1000 nm), with LSSVM excelling in the near-infrared spectrums(900–1700 nm). Finally, the feasibility of using Hyperspectral imaging(HSI) technology to detect tunnel lining strength was demonstrated at the shield tunnel model site. The predicted results were validated by combining with strength values measured by the rebound hammer, and presented promoising research path for automated non-destructive detection of concrete tunnel lining strength.

Identification and classification of buckwheat grain by microfocus radiography and hyperspectral imaging methods

Classification of buckwheat grains is important because the absence of defective grains is a guarantee of yield and quality. Buckwheat grains were randomly selected from a batch whose grains varied in quality. The identification and classification of buckwheat grains according to the degree of fulfillment was carried out by a combination of microfocus X-ray and hyperspectral image analysis and multivariate analysis techniques. Using microfocus radiography, buckwheat grains were categorized into groups according to the degree of fulfillment. Hyperspectral image of buckwheat grains in the range of 935-1720 nm was acquired using a Specim FX17 camera. Using the polygon selection function, the averaged spectra were obtained and a data matrix of grain samples was generated. The bands of the spectrum contributing most to the grading of the grain samples by degree of fulfillment were identified using the principal component analysis. The classification model of grading buckwheat grain into groups by degree of fulfillment was constructed by partial least squares discriminant analysis method. The results showed that hyperspectral image is a potential tool for rapid and accurate identification of buckwheat grains, which can be used in large-scale grain classification and grain quality determination.

Detection of different adulteration in cinnamon powder using hyperspectral imaging and artificial neural network method

Cinnamon is one of the medicinal spices that is important in point of economic and human health. The goal of the present research was to classify the levels of different adulterations in the spice using hyperspectral imaging technology. In the present research, three adulterants were investigated including sea foam powder and chickpea and wheat flour with 0, 5, 15, 30, and 50% adulteration levels. After sample preparation, the hyperspectral images of them were acquired using a line scan imaging system. The effective wavelengths were selected and image features were extracted. The effective features were selected and classified using the artificial neural network method. The classification accuracies of the classifier to identify sea foam powder and chickpea and wheat flour adulterants were equal to 98.9, 100, and 100%, respectively. The results showed the high ability of hyperspectral imaging combined with artificial neural networks method to detect adulteration in cinnamon with high reliability.

A semi-supervised segmentation method for microscopic hyperspectral pathological images based on multi-consistency learning.

Pathological images are considered the gold standard for clinical diagnosis and cancer grading. Automatic segmentation of pathological images is a fundamental and crucial step in constructing powerful computer-aided diagnostic systems. Medical microscopic hyperspectral pathological images can provide additional spectral information, further distinguishing different chemical components of biological tissues, offering new insights for accurate segmentation of pathological images. However, hyperspectral pathological images have higher resolution and larger area, and their annotation requires more time and clinical experience. The lack of precise annotations limits the progress of research in pathological image segmentation. In this paper, we propose a novel semi-supervised segmentation method for microscopic hyperspectral pathological images based on multi-consistency learning (MCL-Net), which combines consistency regularization methods with pseudo-labeling techniques. The MCL-Net architecture employs a shared encoder and multiple independent decoders. We introduce a Soft-Hard pseudo-label generation strategy in MCL-Net to generate pseudo-labels that are closer to real labels for pathological images. Furthermore, we propose a multi-consistency learning strategy, treating pseudo-labels generated by the Soft-Hard process as real labels, by promoting consistency between predictions of different decoders, enabling the model to learn more sample features. Extensive experiments in this paper demonstrate the effectiveness of the proposed method, providing new insights for the segmentation of microscopic hyperspectral tissue pathology images.

Hyperspectral Image Denoising Based on Deep and Total Variation Priors

To address the problems of noise interference and image blurring in hyperspectral imaging (HSI), this paper proposes a denoising method for HSI based on deep learning and a total variation (TV) prior. The method minimizes the first-order moment distance between the deep prior of a Fast and Flexible Denoising Convolutional Neural Network (FFDNet) and the Enhanced 3D TV (E3DTV) prior, obtaining dual priors that complement and reinforce each other’s advantages. Specifically, the original HSI is initially processed with a random binary sparse observation matrix to achieve a sparse representation. Subsequently, the plug-and-play (PnP) algorithm is employed within the framework of generalized alternating projection (GAP) to denoise the sparsely represented HSI. Experimental results demonstrate that, compared to existing methods, this method shows significant advantages in both quantitative and qualitative assessments, effectively enhancing the quality of HSIs.

Enhancing hyperspectral imaging through macro and multi-modal capabilities

Hyperspectral imaging (HSI) has emerged as an effective tool to obtain spatially resolved spectral information of artworks by combining optical imaging with spectroscopy. This technique has proven its efficacy in providing valuable information both at the large and microscopic scale. Interestingly, the macro scale has yet to be thoroughly investigated using this technology. While standard HSI methods include the use of spatial or spectral filters, alternative methods based on Fourier-transform interferometry have also been utilised. Among these, a hyperspectral camera employing a birefringent common-path interferometer, named TWINS, has been developed, showing a high robustness and versatility. In this paper, we propose the combination of TWINS with a macro imaging system for the study of cultural heritage (CH). We will show how the macro-HSI system was designed, and we will demonstrate its efficient capabilities to collect interferometric images with high visibility and good signal of both reflectance and fluorescence on the same field of view, even on non-flat samples. Our hyperspectral camera for macro studies of both reflectance and fluorescence data is a completely new asset in the CH panorama and beyond. The relevance of the macro technology is demonstrated in two case studies, aiding in the analysis of biofilms on stone samples and of the degradation of dyed textiles.

Hyperspectral image classification based on superpixel merging and broad learning system

AbstractMost spectral–spatial classification methods for hyperspectral images (HSIs) can achieve satisfactory classification results. However, the common problem faced with these approaches is the need for a long training time and sufficient training samples. To address this issue, this study proposes an effective spectral–spatial HSI classification method based on superpixel merging, superpixel smoothing and broad learning system (SMS‐BLS). The newly introduced parameter‐free superpixel merging technique based on local modularity not only enhances the role of local spatial information in classification, but also maintains class boundary information as much as possible. In addition, the spectral and spatial information of HSIs is further fused during the superpixel smoothing process. As a result, with limited training samples, using merged and smoothed superpixels instead of pixels as input to the broad learning system significantly improves its classification performance. Moreover, the merged superpixels weaken the dependence of the classification results on the superpixel segmentation scale. The effectiveness of the proposed method was validated on three HSI benchmarks, namely Indian Pines, Pavia University and Salinas. Experimental and comparative results show the superiority of the method to other state‐of‐the‐art approaches in terms of overall accuracy and running time.

Deep learning-enhanced snapshot hyperspectral confocal microscopy imaging system

Laser-scanning confocal hyperspectral microscopy is a powerful technique to identify the different sample constituents and their spatial distribution in three-dimensional (3D). However, it suffers from low imaging speed because of the mechanical scanning methods. To overcome this challenge, we propose a snapshot hyperspectral confocal microscopy imaging system (SHCMS). It combined coded illumination microscopy based on a digital micromirror device (DMD) with a snapshot hyperspectral confocal neural network (SHCNet) to realize single-shot confocal hyperspectral imaging. With SHCMS, high-contrast 160-bands confocal hyperspectral images of potato tuber autofluorescence can be collected by only single-shot, which is almost 5 times improvement in the number of spectral channels than previously reported methods. Moreover, our approach can efficiently record hyperspectral volumetric imaging due to the optical sectioning capability. This fast high-resolution hyperspectral imaging method may pave the way for real-time highly multiplexed biological imaging.

Deoxynivalenol Detection beyond the Limit in Wheat Flour Based on the Fluorescence Hyperspectral Imaging Technique.

Deoxynivalenol (DON) is a harmful fungal toxin, and its contamination in wheat flour poses a food safety concern globally. This study proposes the combination of fluorescence hyperspectral imaging (FHSI) and qualitative discrimination methods for the detection of excessive DON content in wheat flour. Wheat flour samples were prepared with varying DON concentrations through the addition of trace amounts of DON using the wet mixing method for fluorescence hyperspectral image collection. SG smoothing and normalization algorithms were applied for original spectra preprocessing. Feature band selection was carried out by applying the successive projection algorithm (SPA), uninformative variable elimination (UVE), competitive adaptive reweighted sampling (CARS), and the random frog algorithm on the fluorescence spectrum. Random forest (RF) and support vector machine (SVM) classification models were utilized to identify wheat flour samples with DON concentrations higher than 1 mg/kg. The results indicate that the SG-CARS-RF and SG-CARS-SVM models showed better performance than other models, achieving the highest recall rate of 98.95% and the highest accuracy of 97.78%, respectively. Additionally, the ROC curves demonstrated higher robustness on the RF algorithm. Deep learning algorithms were also applied to identify the samples that exceeded safety standards, and the convolutional neural network (CNN) model achieved a recognition accuracy rate of 97.78% for the test set. In conclusion, this study demonstrates the feasibility and potential of the FHSI technique in detecting DON infection in wheat flour.

CoastalWQL: An Open-Source Tool for Drone-Based Mapping of Coastal Turbidity Using Push Broom Hyperspectral Imagery

Uncrewed-Aerial Vehicles (UAVs) and hyperspectral sensors are emerging as effective alternatives for monitoring water quality on-demand. However, image mosaicking for largely featureless coastal water surfaces or open seas has shown to be challenging. Another pertinent issue observed is the systematic image misalignment between adjacent flight lines due to the time delay between the UAV-borne sensor and the GNSS system. To overcome these challenges, this study introduces a workflow that entails a GPS-based image mosaicking method for push-broom hyperspectral images, together with a correction method to address the aforementioned systematic image misalignment. An open-source toolkit, CoastalWQL, was developed to facilitate the workflow, which includes essential pre-processing procedures for improving the image mosaic’s quality, such as radiometric correction, de-striping, sun glint correction, and object masking classification. For validation, UAV-based push-broom hyperspectral imaging surveys were conducted to monitor coastal turbidity in Singapore, and the implementation of CoastalWQL’s pre-processing workflow was evaluated at each step via turbidity retrieval. Overall, the results confirm that the image mosaicking of the push-broom hyperspectral imagery over featureless water surface using CoastalWQL with time delay correction enabled better localisation of the turbidity plume. Radiometric correction and de-striping were also found to be the most important pre-processing procedures, which improved turbidity prediction by 46.5%.

Single-Pixel Infrared Hyperspectral Imaging via Physics-Guided Generative Adversarial Networks

A physics-driven generative adversarial network (GAN) was utilized to demonstrate a single-pixel hyperspectral imaging (HSI) experiment in the infrared spectrum, eliminating the need for extensive dataset training in most data-driven deep neural networks. Within the GAN framework, the physical process of single-pixel imaging (SPI) was integrated into the generator, and its estimated one-dimensional (1D) bucket signals and the actual 1D bucket signals were employed as constraints in the objective function to update the network’s parameters and optimize the generator with the assistance of the discriminator. In comparison to single-pixel infrared HSI methods based on compressive sensing and physics-driven convolution neural networks, our physics-driven GAN-based single-pixel infrared HSI exhibits superior imaging performance. It requires fewer samples and achieves higher image quality. We believe that our physics-driven network will drive practical applications in computational imaging, including various SPI-based techniques.