Hyperspectral Research Articles

Remote detection of polluted gases based on thermal infrared spectral imaging in the field

This paper reports the new progress of long wave infrared remote sensing in the field, focusing on the implementation process of window scanning spatiotemporal modulation Fourier spectroscopic imaging technology. Utilizing the self-developed CHIPED-1 device, the spatial modulation interference was achieved through the use of a cone mirror Michelson interferometer. By combining it with a cooled long-wave infrared focal plane detector component and applying data processing steps such as acquisition, recombination, calibration, etc., high-spectral resolution imaging in long-wave infrared region was accomplished. The detection sensitivity index nesr (Noise Equivalent Spectral Radiance) of the self-developed CHIPED-1 long wave infrared hyperspectral imaging principle experimental device reaches 5.6 × 10−8w/(cm−1.sr.cm2) in single pixel，Equivalent to commercial time modulation interferometric hyperspectral imagers; It reflects the progressiveness of the technology, and leaves much room for improvement.By testing the transmittance curve of polypropylene film, the spectral response range of CHIPED-1 infrared hyperspectral imaging principle experimental device reached 11.5 μm.The article also studied the hyperspectral imaging detection method for two-dimensional distributed chemical gas VOCs, taking the detection experiments of field high-rise buildings and ether gas as examples.Under complex backgrounds and low experimental concentrations, the presence of ether vapor cannot be observed from the infrared spectrum slices at the same wave number. However, after differential spectral processing, the spatial distribution of ether vapor can be clearly seen.The hyperspectral method is applied in the field of infrared detection of organic vapor VOCs, which has many advantages over wide-band thermal imaging methods, such as high sensitivity, strong anti-interference ability, and wide recognition range.

Spectral Discrimination of Common Karoo Shrub and Grass Species Using Spectroscopic Data

Rangelands represent about 25% of the Earth’s land surface but are under severe pressure. Rangeland degradation is a gradually increasing global environmental problem, resulting in temporary or permanent loss of ecosystem functions. Ecological rangeland studies aim to determine the productivity of rangelands as well as the severity of their degradation. Rigorous in situ assessments comprising visual identification of plant species are required as such assessments are perceived to be the most accurate way of monitoring rangeland degradation. However, in situ assessments are expensive and time-consuming exercises, especially when carried out over large areas. In situ assessments are also limited to areas that are accessible. This study aimed to evaluate the effectiveness of multispectral (MS) and hyperspectral (HS) remotely sensed, unmanned aerial vehicle (UAV)-based data and machine learning (random forest) methods to differentiate between 15 dominant Nama Karoo plant species to aid ecological impact surveys. The results showed that MS imagery is unsuitable, as classification accuracies were generally low (37.5%). In contrast, much higher classification accuracies (>70%) were achieved when the HS imagery was used. The narrow bands between 398 and 430 nanometres (nm) were found to be vital for discriminating between shrub and grass species. Using in situ analytical spectral device (ASD) spectroscopic data, additional important wavebands between 350 and 400 nm were identified, which are not covered by either the MS or HS remotely sensed data. Using feature selection methods, 12 key wavelengths were identified for discriminating among the plant species with accuracies exceeding 90%. Reducing the dimensionality of the ASD data set to the 12 key bands increased classification accuracies from 84.8% (all bands) to 91.7% (12 bands). The methodology developed in this study can potentially be used to carry out UAV-based ecological assessments over large and inaccessible areas typical of Karoo rangelands.

A Summative Review of Advances in Sensor Technology for Precision Agriculture

Sensor technology has become a cornerstone of precision agriculture, offering a wide array of applications that enhance farming efficiency, productivity, and sustainability. By providing real-time data on soil health, crop growth, and environmental conditions, sensors enable farmers to optimize inputs such as water, fertilizers, and pesticides, reducing waste and improving yields. The development of advanced sensors, such as multi-spectral and hyper-spectral imaging, combined with artificial intelligence (AI) and machine learning (ML) algorithms, has revolutionized crop monitoring and disease detection, allowing for timely interventions that prevent significant losses. Additionally, the integration of wireless sensor networks (WSNs) and the Internet of Things (IoT) facilitates the seamless collection and transmission of data, enabling remote and automated farm management. This technology has proven instrumental in addressing the challenges of food security by increasing agricultural productivity and reducing input costs. Sensor technologies also promote sustainable land and water management by improving irrigation efficiency and reducing chemical runoff, thus minimizing environmental impact. Economic benefits are substantial, with farmers experiencing cost savings, increased yields, and higher profitability. Moreover, sensor technology contributes to climate-smart agriculture by improving resource use efficiency and reducing greenhouse gas emissions. However, challenges remain, including the high cost of sensor installation and maintenance, limited accessibility in remote regions, data privacy concerns, and the lack of interoperability between different sensor platforms. The future of sensor technology lies in the development of low-cost, biodegradable sensors, the increased use of autonomous and robotic platforms, and enhanced AI integration for more accurate data interpretation. Expanding the use of sensors in smallholder and developing world agriculture will be crucial for global food security and environmental sustainability, offering a promising path toward a more resilient and efficient agricultural system.

Surgical Hyperspectral imaging and Indocyanine green Near-infrared Examination (SHINE) for brain arteriovenous malformation resection: a case report on how to visualize perfusion

Background and importanceArteriovenous malformations (AVMs) are complex vascular anomalies that pose significant risks, including intracranial hemorrhage and neurological deficits. Surgical resection is the preferred treatment, requiring precise intraoperative imaging to ensure complete removal while preserving critical structures. This case report presents the first combined use of hyperspectral imaging (HSI) and indocyanine green video angiography (ICG VA) to visualize perfusion during brain AVM surgery, highlighting the potential benefits of these advanced imaging techniques.Case descriptionA 66-year-old male presented with chronic headaches but no neurological deficits. MRI revealed a superficial AVM in the left frontal lobe within the superior frontal sulcus, measuring approximately 2.4 cm. The AVM was fed by feeders from the pericallosal artery, callosomarginal artery, and middle cerebral artery (MCA) branches, with drainage through a dilated cortical vein into the superior sagittal sinus. Preoperative embolization of two MCA feeding branches was performed, followed by microsurgical resection with ICG VA and HSI.ConclusionsThis case report demonstrates the successful application of HSI and ICG VA in brain AVM surgery. The combined use of these technologies provided comprehensive intraoperative assessment, enhancing surgical precision and safety. The integration of HSI offers non-invasive, contrast-agent-free imaging, potentially improving outcomes by enabling detailed perfusion mapping. Future studies should explore the broader applications of these imaging modalities in neurovascular practice.

Spectral Unmixing of Pigments on Surface of Painted Artefacts Considering Spectral Variability

Abstract. Painted artefacts, such as murals and paintings, are the treasures of human civilization. Pigment is an important component of their surfaces. It is crucial to study the composition and proportion of pigments on the surface of painted artefacts for the field of heritage conservation. In this study, hyperspectral remote sensing was used to invert the abundance of mixed pigments by spectral unmixing. Spectral variability effects are often present in hyperspectral images. Hyperspectral images of cultural artefacts also suffer from spectral variability due to factors such as particle size and impurities within the pigment, and acquisition conditions. Therefore, spectral variability was incorporated into the Linear Mixed Model of the pigment spectral unmixing. The unmixing methods are classified into two categories based on whether or not spectral libraries are used. In the experiment, computer synthesized data and laboratory-made sample blocks of mixed pigments, which mixed with different ratios of Azurite and Malachite, were selected as validation data. Five commonly used algorithms for solving the spectral variability problem, namely MESMA, Fractional Sparse SU, ELMM, RUSAL, and BCM, are used for pigment unmixing and compared with FCLS, which does not consider spectral variability. The results show that ELMM has the highest unmixing accuracy of aRMSE and xRMSE compared to other methods. At the same time, it can be seen from the metric of SAM that ELMM is able to extract reliable endmember variability spectral, which is more suitable for solving the problem of spectral variability in pigment unmixing. Finally, we apply ELMM to real artefacts Yungang Grottoes mural paintings, and obtain a lower xRMSE than FCLS, which improves the unmixing accuracy.

Estimating Soil Salinity Using HISUI Hyperspectral Data in the Western Australian Wheatbelt

Abstract. Soil salinity, caused by natural and anthropogenic factors, significantly impacts agricultural productivity, ecosystems, and global biodiversity. To mitigate severe salinity levels, it is necessary to detect early the salt-affected land for implementing solutions, such as adequate irrigation and cultivation of salt-tolerant crops. Before launching spaceborne hyperspectral sensor, Hyperspectral Imager Suite, (HISUI) developed by Japanese Ministry of Economy Trade and Industry, Kobayashi et al. (2010 and 2013) developed methods utilizing airborne hyperspectral data from HyMap to estimate soil salinity in wheatbelt region of Western Australia as a previous study. This paper aims to assess the feasibility of estimating soil salinity with HISUI in the same study area, following the past approach. Also, the goal is to estimate low soil salinity levels as same as the past objectives. As a result, past approach could not be fully applied in spaceborne hyperspectral sensor, caused by atmospheric effect in SWIR region. However, absorption of soil salinity in SWIR regions could be detected by HISUI. In the soil index, the NDSI (Normalized Difference Soil Index) showed higher accuracy than the Soil Index (SI) developed in previous studies. Then, the lowest values of RMSE were less than 100 mS/m in NDSI. As a result, HISUI data could not leach to map soil salinity at lower levels in this study. However, it showed the potential to estimate it with higher accuracy using hyperspectral data.

Rapid Detection of Single Bacteria Using Filter-Array-Based Hyperspectral Imaging Technology.

Rapid and accurate detection of bacterial pathogens is crucial for preventing widespread public health crises, particularly in the food industry. Traditional methods are often slow and require extensive labeling, which hampers timely responses to potential threats. In response, we introduce a groundbreaking approach using filter-array-based hyperspectral imaging technology, enhanced by a super-resolution demosaicking technique. This innovative technology streamlines the detection process and significantly enhances the resolution of mosaic hyperspectral imaging. By utilizing a snapshot hyperspectral camera with a 15 ms integration time, it facilitates the identification of bacteria at the single-cell level without requiring chemical labels. The integration of a 3D convolutional neural network optimizes the recognition of pathogenic bacteria, achieving an impressive accuracy of 91.7%. Our approach dramatically improves the efficiency and effectiveness of bacterial detection, providing a promising solution for critical applications in public health and the food industry.

Wheat grain classification using hyperspectral imaging: Concatenating Vis-NIR and SWIR Data for single and bulk grains

Wheat variety identification is an essential component of the official inspection of wheat grains to determine their quality and trade value. Traditionally, wheat class identification requires deep knowledge about the appearance of plants and kernels, their history and distribution. Therefore, there is an urgent need for an automatic technology to standardize the nomenclature of wheat varieties to enable both growers and producers to identify their grains practically. The present study investigated the feasibility of Visible-Near Infrared (Vis-NIR) and Short-Wave Infrared (SWIR) spectral imaging to identify wheat classes. Both sides of the wheat kernel as single kernel measurements and bulk grains were used to acquire hyperspectral data and vertical and horizontal data concatenation was implemented to explore performance differences. In addition, the classification of bulk kernels from single kernel data was investigated. Spectral data was pre-treated by standard normal variate (SNV), Savitzky-Golay first (SG-1) and second derivatives (SG-2) and linear discriminant analysis (LDA), support vector machine (SVM) and artificial neural network (ANN) were applied as the classification algorithms. Furthermore, feature selection was utilised using the minimum redundancy maximum relevance (mRMR) algorithm to investigate the performance difference between data concatenation and feature selection. The results showed that ventral (up) side data showed better classification performances than reverse (down) side data, while Vis-NIR data achieved higher classification accuracies than SWIR data. However, the best classification performance for single kernels was obtained by LDA-SNV using up and down data in the Vis-NIR and SWIR regions vertically and horizontally concatenated with an accuracy of 93.72% for 10-fold cross-validation and 94.93% for test sets. Models based on a hundred features did not achieve the accuracy of models based on concatenated data. Moreover, classification performances of bulk samples were higher than single kernels, which achieved 100% accuracy for both cross-validation and test sets. The study demonstrates that spectral imaging has a high potential to identify wheat classes non-destructively.

DMCCT: Dual-Branch Multi-Granularity Convolutional Cross-Substitution Transformer for Hyperspectral Image Classification

In the field of hyperspectral image classification, deep learning technology, especially convolutional neural networks, has achieved remarkable progress. However, convolutional neural network models encounter challenges in hyperspectral image classification due to limitations in their receptive fields. Conversely, the global modeling capability of Transformers has garnered attention in hyperspectral image classification. Nevertheless, the high computational cost and inadequate local feature extraction hinder its widespread application. In this study, we propose a novel fusion model of convolutional neural networks and Transformers to enhance performance in hyperspectral image classification, namely the dual-branch multi-granularity convolutional cross-substitution Transformer (DMCCT). The proposed model adopts a dual-branch structure to separately extract spatial and spectral features, thereby mitigating mutual interference and information loss between spectral and spatial data during feature extraction. Moreover, a multi-granularity embedding module is introduced to facilitate multi-scale and multi-level local feature extraction for spatial and spectral information. In particular, the improved convolutional cross-substitution Transformer module effectively integrates convolution and Transformer, reducing the complexity of attention operations and enhancing the accuracy of hyperspectral image classification tasks. Subsequently, the proposed method is evaluated against existing approaches using three classical datasets, namely Pavia University, Kennedy Space Center, and Indian Pines. Experimental results demonstrate the efficacy of the proposed method, achieving significant classification results on these datasets with overall classification accuracies of 98.57%, 97.96%, and 96.59%, respectively. These results establish the superiority of the proposed method in the context of hyperspectral image classification under similar experimental conditions.

Early Detection of Dendroctonus valens Infestation with UAV-Based Thermal and Hyperspectral Images

Dendroctonus valens is one of the main invasive pests in China, causing serious economic and ecological damage. Early detection and control of D. valens can help prevent further outbreaks. Based on unmanned aerial vehicle (UAV) thermal infrared and hyperspectral data, we compared the spectral characteristics of Pinus sylvestris var. mongolica in three states (healthy, early-infested, and dead), and constructed a classification model based on the random forest algorithm using four spectral datasets (reflectance, first derivative, second derivative, and spectral vegetation index) and one temperature parameter dataset. Our results indicated that the spectral differences between healthy and early-infested trees mainly occur in the near-infrared region, with dead trees showing different characteristics. While it was effective to distinguish healthy from early-infested trees using spectral data alone, the addition of a temperature parameter further improved classification accuracy across all datasets. The combination of the spectral vegetation index and temperature parameter achieved the highest accuracy at 93.75%, which is 3.13% higher than using the spectral vegetation index alone. This combination also significantly improved early detection precision by 13.89%. Our findings demonstrated the applicability of UAV-based thermal infrared and combined hyperspectral datasets in monitoring D. valens early-infested trees, providing important technical support for the scientific prevention and control of D. valens.

Machine Learning and Hyperspectral Imaging for Analysis of Human Papillomaviruses (HPV) Vaccine Self-Healing Particles.

Human papillomaviruses (HPV) are known to cause a variety of diseases, including cervical cancer and genital warts. HPV is a highly prevalent virus and is considered the most common sexually transmitted disease. Because of the risks associated with HPV, Gardasil, a quadrivalent recombinant vaccine, was developed by Merck & Co., Inc., Rahway, NJ, USA, and approved by the Food and Drug Administration (FDA) in 2006. The second generation of the vaccine, Gardasil9, was subsequently approved by the FDA in 2014, providing significant protection against HPV. The HPV vaccine may be given as 2 or 3 doses; however, vaccine administration as a single dose with a sustained release mechanism may potentially offer benefits to meet emerging health needs. To explore this, HPV vaccines were formulated within microporous self-healing particles (SHPs) to enable potential controlled release of HPV virus-like particle (VLP) antigen. Machine learning, in the form of multivariate curve resolution-alternating least-squares (MCR-ALS), with Raman hyperspectral imaging was used to determine the molecular identity and spatial distribution of all relevant species within this HPV vaccine formulation. The results indicate that machine learning with Raman hyperspectral imaging was able to spatially resolve HPV VLP antigens within SHP vaccines for the first time, providing crucial information necessary for vaccine development.

Detection of underground natural gas pipeline micro-leakage based on UAV hyperspectral remote sensing and GIS

ABSTRACT Detection of underground natural gas pipeline micro-leakage based on unmanned aerial vehicle (UAV) hyperspectral remote sensing and GIS. Hyperspectral images can indirectly detect underground natural gas pipeline micro-leakage through spectral and spatial variation characteristics of surface vegetation. However, most of existing studies were based on ground-mounted platforms, which could only perform small-range single-point detection and might occur misidentifications. UAV hyperspectral remote sensing can allow wide-range detection of surface vegetation. Moreover, underground pipeline distribution GIS data can provide prior knowledge about leakage points to exclude misidentifications. Therefore, a natural gas pipeline micro-leakage experiment was set up. UAV hyperspectral images of grasslands and pipeline distribution vector data were obtained, which led to a proposed new wide-range multi-points detection methodology of underground natural gas micro-leakage. Firstly, the vegetation identification index (WVI) R 470 + R 674 / R 555 + R 750 was designed based on WOA–VMD (whale optimization algorithm – variational modal decomposition) to segment and extract the vegetation stress zones. Then, UAV- and GIS-based natural gas micro-leakage vegetation stress identification model was constructed to obtain the leak points of natural gas micro-leakage. Finally, the recognition ability was evaluated by comparing with three stress vegetation identification indices proposed in previous studies. The result showed that there was no missing or false detection in identification results; the identification and positioning effect was better than other indices, which could meet the practical application requirements.

Lidar-derived estimates of boreal shrub biomass in Southcentral Alaska

Abstract Despite widespread observations of shrub proliferation and expansion (shrubification), few studies quantify shrub biomass at the regional scale. Here we describe and implement a two-part modeling approach to estimate and map tall shrub (diameter at root collar > 2.5 cm) expected aboveground biomass, or E[SHB], across a 16.6 million ha boreal region where shrubification occurs in wetlands and subalpine ecosystems. Using n = 384 field plots nested within m = 11 study sites across southcentral Alaska, we constructed random forest models of the probability (pSH) that shrub wood volume surpasses tree wood volume, and generalized additive models (GAMs) of aboveground biomass of tall shrubs (SHB) using rasterized aerial lidar variables collected by NASA Goddard’s Lidar, Hyperspectral, and Thermal (G-LiHT) Airborne Imager, together with gridded climate data from a Parameter-elevation Regressions on Independent Slopes Model 30-year normal climatology (PRISM). Applying those models to G-LiHT tiles, then averaging across tiles within n = 843 watersheds covering 9.2 million ha, we estimated that below 1,000 m asl, the area-weighted mean value of E[SHB] = pSH x SHB = 6.6 Mg ha-1 with sd = 4.5 Mg ha-1. This is the first study to estimate current shrub biomass density at the regional scale in southcentral Alaska and serves as a biomass baseline for measuring and modeling aboveground carbon fluxes where plant communities are undergoing climate-driven change.

Achromatic correction for birefringent interferometers that improve Fourier transform spectrometers and hyperspectral imaging

Spectroscopy and hyperspectral imaging are widely used tools for identifying compounds and materials. One optical design is a polarization interferometer that uses birefringent wedges, like a Babinet-Soleil compensator, to create the interferograms that are Fourier transformed to give the spectra. Such designs have lateral spatial offset between the no and ne optical beams, which reduces the interferogram intensity and creates a spatially dependent phase that is problematic for hyperspectral imaging. The lateral separation between the beams is wavelength dependent, created by the achromatic nature of Babinet-Soleil compensators. We introduce a birefringent wedge design for Fourier transform spectroscopy that creates collinear no and ne optical beams for optimal interference and no spatial dependent phase. Our 3-wedge design, which we call a Wisconsin interferometer, improves the signal strength of polarization spectrometers, and eliminates phase shifts in hyperspectral imaging. We anticipate that it will find use in analytical, remote sensing, and ultrafast spectroscopy applications.

A Low-Measurement-Cost-Based Multi-Strategy Hyperspectral Image Classification Scheme.

The cost of hyperspectral image (HSI) classification primarily stems from the annotation of image pixels. In real-world classification scenarios, the measurement and annotation process is both time-consuming and labor-intensive. Therefore, reducing the number of labeled pixels while maintaining classification accuracy is a key research focus in HSI classification. This paper introduces a multi-strategy triple network classifier (MSTNC) to address the issue of limited labeled data in HSI classification by improving learning strategies. First, we use the contrast learning strategy to design a lightweight triple network classifier (TNC) with low sample dependence. Due to the construction of triple sample pairs, the number of labeled samples can be increased, which is beneficial for extracting intra-class and inter-class features of pixels. Second, an active learning strategy is used to label the most valuable pixels, improving the quality of the labeled data. To address the difficulty of sampling effectively under extremely limited labeling budgets, we propose a new feature-mixed active learning (FMAL) method to query valuable samples. Fine-tuning is then used to help the MSTNC learn a more comprehensive feature distribution, reducing the model's dependence on accuracy when querying samples. Therefore, the sample quality is improved. Finally, we propose an innovative dual-threshold pseudo-active learning (DSPAL) strategy, filtering out pseudo-label samples with both high confidence and uncertainty. Extending the training set without increasing the labeling cost further improves the classification accuracy of the model. Extensive experiments are conducted on three benchmark HSI datasets. Across various labeling ratios, the MSTNC outperforms several state-of-the-art methods. In particular, under extreme small-sample conditions (five samples per class), the overall accuracy reaches 82.97% (IP), 87.94% (PU), and 86.57% (WHU).

First step toward designing effective real-time control systems in laser directed energy deposition

In-process monitoring and control are essential for quality assurance and consistency of laser-based directed energy deposition processes. Detection of irregularities during deposition in terms of defects or flaws is based on in situ monitoring of output process parameters such as temperature, melt-pool geometry, or deposition height. The real-time feedback of these output parameters allows the development of control strategies for real-time adjustment of input process parameters, such as laser power or scanning speed, to correct detected deviations from the desired output process parameters. Therefore, criteria such as sensitivity, stability, correlation, trends, and interactions of the input-output process parameters have a direct impact on controller design, establishing, for example, control limits or tolerance ranges of the output parameters. This paper focuses on the study of the characteristics of output process parameters to input process parameters. This research involves analyzing and comparing the deposition of single tracks under various input process parameters, including laser power and scanning speed. Melt-pool geometry and temperature are estimated from a visual camera and a hyperspectral line camera, whereas the final deposition geometry is obtained from a laser triangulation scanner. The results show the linearity between input and output process parameters, the steadiness of the output process parameters, the relation between melt-pool and final deposition, and offer insights to design effective in-process control systems.

An Edge and Trustworthy AI UAV System With Self-Adaptivity and Hyperspectral Imaging for Air Quality Monitoring

An Edge and Trustworthy AI UAV System With Self-Adaptivity and Hyperspectral Imaging for Air Quality Monitoring

Hyperspectral imaging of human liver allografts for prediction of initial graft function

PurposeIschemia reperfusion injury represents a significant yet difficult to assess risk factor for short- and long-term graft impairment in human liver transplantation (LT). As a non-invasive, non-ionizing tool, hyperspectral imaging (HSI) is capable of correlating optical properties with organ microperfusion. Hence, we here performed a study of human liver allografts assessed by HSI for microperfusion and prediction of initial graft function.MethodsImages of liver parenchyma of 37 human liver allografts were acquired at bench preparation, during normothermic machine perfusion (NMP), if applicable, and after reperfusion in the recipient. A specialized HSI acquisition software computed oxygen saturation (StO2), tissue hemoglobin indices (THI), near infrared perfusion indices (NIR), and tissue water indices (TWI). HSI parameters were analyzed for differences with regard to preservation technique, reperfusion sequence and presence of early allograft dysfunction (EAD).ResultsOrgan preservation was performed by means of NMP (n = 31) or static cold storage (SCS; n = 6). Patients’ demographics, donor characteristics, presence of EAD (NMP 36.7% vs. SCS 50%, p = 0.6582), and HSI parameters were comparable between both groups of preservation method. In organs developing EAD, NIR at 1, 2, and 4 h NMP and after reperfusion in the recipient was significantly lower (1 h NMP: 18.6 [8.6–27.6] vs. 28.3 [22.5–39.4], p = 0.0468; 2 h NMP: 19.4 [8.7–30.4] vs. 37.1 [27.5–44.6], p = 0.0011; 4 h NMP: 26.0 [6.8–37.1] vs. 40.3 [32.3–49.9], p = 0.0080; reperfusion: 13.0 [11.5–34.3] vs. 30.6 [19.3–44.0], p = 0.0212).ConclusionHSI assessment of human liver allografts is feasible during organ preservation and in the recipient. NIR during NMP and after reperfusion might predict the onset of EAD. Larger trials are warranted for assessment of this novel technique in human LT.

GASSF-Net: Geometric Algebra Based Spectral-Spatial Hierarchical Fusion Network for Hyperspectral and LiDAR Image Classification

The field of multi-source remote sensing observation is becoming increasingly dynamic through the integration of various remote sensing data sources. However, existing deep learning methods face challenges in differentiating between internal and external relationships and capturing fine spatial features. These models often struggle to effectively capture comprehensive information across remote sensing data bands, and they have inherent differences in the size, structure, and physical properties of different remote sensing datasets. To address these challenges, this paper proposes a novel geometric-algebra-based spectral–spatial hierarchical fusion network (GASSF-Net), which uses geometric algebra for the first time to process multi-source remote sensing images, enabling a more holistic approach to handling these images by simultaneously leveraging the real and imaginary components of geometric algebra to express structural information. This method captures the internal and external relationships between remote sensing image features and spatial information, effectively fusing the features of different remote sensing data to improve classification accuracy. GASSF-Net uses geometric algebra (GA) to represent pixels from different bands as multivectors, thus capturing the intrinsic relationships between spectral bands while preserving spatial information. The network begins by deeply mining the spectral–spatial features of a hyperspectral image (HSI) using pairwise covariance operators. These features are then extracted through two branches: a geometric-algebra-based branch and a real-valued network branch. Additionally, the geometric-algebra-based network extracts spatial information from light detection and ranging (LiDAR) to complement the elevation data lacking in the HSI. Finally, a genetic-algorithm-based cross-fusion module is introduced to fuse the HSI and LiDAR data for improved classification. Experiments conducted on three well-known datasets, Trento, MUUFL, and Houston, demonstrate that GASSF-Net significantly outperforms traditional methods in terms of classification accuracy and model efficiency.

Hyperspectral Analysis of Silicon Nanowires Manufactured Through Metal-Assisted Chemical Etching

Abstract This study used hyperspectral imaging to analyze localized near-field interactions between incident electromagnetic waves and silicon nanowire (SiNW) arrays manufactured through catalytic etching of Si wafers for different durations. The results revealed that the unetched upper surface area on Si wafers and reflection of incident light decreased with increasing etching time. A light reflection band peaking at approximately 880 nm was generated from arrays etched for more than 1 h. We used six separate hyperspectral images to analyze the wavelength-dependent spatial optical responses of the fabricated SiNW arrays. The images revealed hot spots of light reflection from unetched Si surfaces in the wavelength range of 470–750 nm and a resonant peak at 880 nm for a photonic crystal derived from a random SiNW array. Accordingly, hyperspectral imaging enables the assessment of localized optical responses of SiNW arrays, which can then be optimized to cater to various applications.