Hyperspectral Research Articles

Colour Analysis of Sausages Stuffed with Modified Casings Added with Citrus Peel Extracts Using Hyperspectral Imaging Combined with Multivariate Analysis

Recycling citrus peel waste offers several significant contributions to sustainability, transforming what would otherwise be discarded into valuable resources. In this study, the colour of sausages stored for 16 days, with varying amounts of orange extract added to the modified casing solution, was evaluated using response surface methodology (RSM) and a hyperspectral imaging system within the spectral range of 350–1100 nm for the first time. To enhance model performance, spectral pre-treatments such as normalisation, first derivative, standard normal variate (SNV), second derivative, and multiplicative scatter correction (MSC) were applied. Both raw and pre-treated spectral data, along with colour attributes, were fitted to a partial least squares regression model. The RSM results indicated that the highest R2 value, 80.61%, was achieved for the b* (yellowness) parameter using a second-order polynomial model. The interactive effects of soy oil and orange extracts on b* were found to be significant (p < 0.05), and the square effects of soy oil on b* were significant at the 1% level. The identified key wavelengths for colour parameters can simplify the model, making it more suitable for practical industrial applications.

Improving the accuracy of SIF quantified from moderate spectral resolution airborne hyperspectral imager using SCOPE: assessment with sub-nanometer imagery

Hyperspectral imaging of solar-induced chlorophyll fluorescence (SIF) is required for plant phenotyping and stress detection. However, the most accurate instruments for SIF quantification, such as sub-nanometer (≤1-nm full-width at half-maximum, FWHM) airborne hyperspectral imagers, are expensive and uncommon. Previous studies have demonstrated that standard narrow-band hyperspectral imagers (i.e., 4–6-nm FWHM) are more cost-effective and can provide far-red SIF quantified at 760 nm (SIF760), which correlates strongly with precise sub-nanometer resolution measurements. Nevertheless, narrow-band SIF760 quantifications are subject to systematic overestimation owing to the influence of the spectral resolution (SR). In this study, we propose a modelling approach based on the Soil Canopy Observation, Photochemistry and Energy Fluxes (SCOPE) model with the objective of enhancing the accuracy of absolute SIF760 levels derived from standard airborne hyperspectral imagers in practical settings. The performance of the proposed method was evaluated using airborne imagery acquired from two airborne hyperspectral imagers (FWHM ≤ 0.2-nm and 5.8-nm) flown in tandem on board an aircraft that collected data from two different wheat and maize phenotyping trials. Leaf biophysical and biochemical traits were first estimated from airborne narrow-band reflectance imagery and subsequently used as SCOPE model inputs to simulate a range of top-of-canopy (TOC) radiance and SIF spectra at 1-nm FWHM. The SCOPE simulated radiance spectra were then convolved to match the spectral configuration of the narrow-band imager to compute the 5.8-nm FWHM SIF760. A site-specific model was constructed by employing the convolved 5.8-nm SR SIF760 as the independent variable and the 1-nm SR SIF760 directly simulated by SCOPE as the dependent variable. When applied to the airborne dataset, the estimated SIF760 at 1-nm SR from the standard narrow-band hyperspectral imager matched the reference sub-nanometer quantified SIF760 with root mean square error (RMSE) less than 0.5 mW/m2/nm/sr, yielding R2 = 0.93–0.95 from the two experiments. These results suggest that the proposed modelling approach enables the interpretation of SIF760 quantified using standard hyperspectral imagers of 4–6 nm FWHM for stress detection and plant physiological condition assessment.

A graph-guided transformer based on dual-stream perception for hyperspectral image classification

ABSTRACT The excellent capabilities of Transformers and Graph Neural Networks (GNNs) in modelling long-range dependencies and handling irregular data have led to their widespread application in hyperspectral image (HSI) classification tasks. However, the Graph Transformer combining both advantages is rarely used in this field and has some limitations. Current Graph Transformers consider interactions between all nodes within the graph, adding complexity and introducing unnecessary information from noisy nodes. Moreover, the rich spectral information in HSIs is often ignored, and there is a lack of effective fusion of spatial information. In this paper, we propose a dual-stream graph-guided Transformer for HSI classification. In spatial dimension, superpixels are utilized to guide spatial graph generation, capturing global topological dependencies and local details effectively in HSIs. In terms of spectrum, we innovatively construct a spectral graph based on spectral channels and adopt a contribution score-based strategy to adaptively filter out irrelevant edges, achieving sparsity while preserving spectral context relationships. Experimental results demonstrate the significant competitive advantage of our method in HSI classification tasks on three public datasets. The code is available at https://github.com/youngboy03/GTDPNet.

Nondestructively Determining Soluble Solids Content of Blueberries Using Reflection Hyperspectral Imaging Technique

Effectively detecting the quality of blueberries is crucial for ensuring that high-quality products are supplied to the fresh market. This study developed a nondestructive method for determining the soluble solids content (SSC) of blueberry fruit by using a near-infrared hyperspectral imaging technique. The reflection hyperspectral images in the 900–1700 nm waveband range were collected from 480 fresh blueberry samples. An image analysis pipeline was developed to extract the spectrums of blueberries from the hyperspectral images. A regression model for quantifying SSC values was successfully established based on the full range of wavebands, achieving the highest RP2 of 0.8655 and the lowest RMSEP value of 0.4431 °Brix. Furthermore, three variable selection methods, namely the Successive Projections Algorithm (SPA), interval PLS (iPLS), and Genetic Algorithm (GA), were utilized to identify the feature wavebands for modeling. The models calibrated from feature wavebands generated an RMSEP of 0.4643 °Brix, 0.4791 °Brix, and 0.4764 °Brix, as well as the RP2 of 0.8507, 0.8397, and 0.8420 for SPA, iPLS, and GA, respectively. Furthermore, a pseudo-color distribution diagram of the SSC values within blueberries was successfully generated based on established models. This study demonstrated a novel approach for blueberry quality detection and inspection by jointly using hyperspectral imaging and machine learning methodologies. It can serve as a valuable reference for the development of grading equipment systems and portable testing devices for fruit quality assurance.

Hyperspectral Imaging Combined with Machine Learning Can Be Used for Rapid and Non-Destructive Monitoring of Residual Nitrite in Emulsified Pork Sausages

The non-destructive and rapid monitoring system for residual nitrite content in processed meat products is critical for ensuring food safety and regulatory compliance. This study was performed to investigate the application of hyperspectral imaging in combination with machine learning algorithms to predict and monitor residual nitrite concentrations in emulsified pork sausages. The emulsified pork sausage was formulated with 1.5% (w/w) sodium chloride, 0.3% (w/w) sodium tripolyphosphate, 0.5% (w/w) ascorbic acid, and sodium nitrite at concentrations of 0, 30, 60, 90, 120, and 150 mg/kg, based on total sample weight. Hyperspectral imaging measurements were conducted by capturing images of the cross-sections and lateral sides of sausage samples in a linescan mode, covering the spectral range of 1000–2500 nm. The analysis revealed that higher nitrite concentrations could influence the protein matrix and hydrogen-bonding capacities, which might cause increased reflectance at approximately 1080 nm and 1280 nm. Machine learning models, including XGBoost, CATboost, and LightGBM, were employed to analyze the hyperspectral data. XGBoost demonstrated the best performance, achieving an R2 of 0.999 and a root mean squared error of 0.095, highlighting its high predictive accuracy. This integration of hyperspectral imaging with advanced machine learning algorithms offers a non-destructive and real-time method for monitoring residual nitrite content in processed meat products, noticeably improving quality control processes in the meat industry. Additionally, real-time implementation in industrial settings could further streamline quality control and enhance operational efficiency. Further research should focus on validating these findings with larger sample sizes and more diverse datasets to ensure robustness.

HyperDehazing: A hyperspectral image dehazing benchmark dataset and a deep learning model for haze removal

Haze contamination severely degrades the quality and accuracy of optical remote sensing (RS) images, including hyperspectral images (HSIs). Currently, there are no paired benchmark datasets containing hazy and haze-free scenes in HSI dehazing, and few studies have analyzed the distributional properties of haze in the spatial and spectral domains. In this paper, we developed a new hazy synthesis strategy and constructed the first hyperspectral dehazing benchmark dataset (HyperDehazing), which contains 2000 pairs synthetic HSIs covering 100 scenes and another 70 real hazy HSIs. By analyzing the distribution characteristics of haze, we further proposed a deep learning model called HyperDehazeNet for haze removal from HSIs. Haze-insensitive longwave information injection, novel attention mechanisms, spectral loss function, and residual learning are used to improve dehazing and scene reconstruction capability. Comprehensive experimental results demonstrate that the HyperDehazing dataset effectively represents complex haze in real scenes with synthetic authenticity and scene diversity, establishing itself as a new benchmark for training and assessment of HSI dehazing methods. Experimental results on the HyperDehazing dataset demonstrate that our proposed HyperDehazeNet effectively removes complex haze from HSIs, with outstanding spectral reconstruction and feature differentiation capabilities. Furthermore, additional experiments conducted on real HSIs as well as the widely used Landsat-8 and Sentinel-2 datasets showcase the exceptional dehazing performance and robust generalization capabilities of HyperDehazeNet. Our method surpasses other state-of-the-art methods with high computational efficiency and a low number of parameters.

Multispectral and hyperspectral image fusion via multi-model deep prior regularization

ABSTRACT Due to the difficulty of collecting a sufficient number of photons within a narrow spectral window, hyperspectral images often have low spatial resolution. In order to solve this problem without increasing hardware costs, a large number of image super-resolution algorithms have been proposed and have shown promising results. However, the performance of these algorithms is inconsistent in different datasets. To fix this situation, we proposed a multi-model deep prior regularization method for multispectral and hyperspectral image fusion. This method leverages mechanistic knowledge to combine various super-resolution results into a single algorithm, significantly improving the quality of the hyperspectral images. Specifically, we first proposed a variable residual mixed attention network. This network takes a low-resolution hyperspectral image and a high-resolution multispectral image as inputs, producing deep prior results from three different modes by adjusting spatial and channel attention. Then, we reconstructed the prior results along the spectral dimension and used them as regularization terms in our mechanistic model. We transformed the problem of finding the optimal solution of the model into solving the Sylvester equation and applied a fast solver for this equation to create the high-resolution hyperspectral images. Additionally, we used a genetic algorithm to determine the best hyperparameters for different deep prior results. Through validation on three datasets, we have demonstrated that our method performs better than existing methods in both quantitative and qualitative comparisons.

Hyperspectral image classification with token fusion on GPU

Hyperspectral images capture material nuances with spectral data, vital for remote sensing. Transformer has become a mainstream approach for tackling the challenges posed by high-dimensional hyperspectral data with complex structures. However, a major challenge they face when processing hyperspectral images is the presence of a large number of redundant tokens, which leads to a significant increase in computational load, adding to the model’s computational burden and affecting inference speed. Therefore, we propose a token fusion algorithm tailored to the operational characteristics of the hyperspectral image and pure transformer network, aimed at enhancing the final accuracy and throughput of the model. The token fusion algorithm introduces a token merging step between the attention mechanism and the multi-layer perceptron module in each Transformer layer. Experiments on four hyperspectral image datasets demonstrate that our token fusion algorithm can significantly improve inference speed without any training, while only causing a slight decrease in the pure transformer network’s classification accuracy.

E2TNet: Efficient enhancement Transformer network for hyperspectral image classification

Recently, Convolutional Transformer-based models have become popular in hyperspectral image (HSI) classification tasks and gained competitive classification performance. However, some Convolutional Transformer-based models fail to effectively mine the global correlations of coarse-grained and fine-grained features, which is adverse to recognizing the refined scale variation information of land-cover. The combination of convolution operations and multihead self-attention mechanisms also increases the computational cost, leading to low classification efficiency. In addition, shallow spectral–spatial features are directly input into the encoder, which inevitably incurs redundant spectral information. Therefore, this paper proposes an efficient enhancement Transformer network (E2TNet) for HSI classification. Specifically, this paper first designs a spectral–spatial feature fusion module to extract spectral and spatial features from HSI cubes and fuse them. Second, considering that redundant spectral information has a negative impact on classification performance, this paper designs a spectral–spatial feature weighted module to improve the feature representation of critical spectral information. Finally, to explore the global correlations of coarse-grained and fine-grained features and improve classification efficiency, an efficient multigranularity information fusion module is embedded in the encoder of E2TNet. The experiment is conducted on four benchmark hyperspectral datasets, and the experimental results demonstrate that the proposed E2TNet is better than several Convolutional Transformer-based classification models in terms of classification accuracy and classification efficiency.

Spectral reconstruction from RGB image to hyperspectral image: Take the detection of glutamic acid index in beef as an example

The use of spectral reconstruction (SR) to recovery RGB images to full-scene hyperspectral image (HSI) is an important measure to achieve real-time and low-cost HSI applications. Taking the detection of glutamic acid index for 360 beef samples as an example, the feasibility of using 11 state-of-the-art reconstruction algorithms to achieve RGB to HSI in complex food systems was investigated. The multivariate correlation analysis was used to prove that RGB is a projection of three-channel comprehensive coverage wide-band information. The comprehensive quality attributes (PSNR-Params-FLOPS) was proposed to determine the optimal reconstruction model (MST++, MST, MIRNet, and MPRNet). Moreover, SSIM values and t-SNE were introduced to evaluate the consistency of the reconstruction results. Finally, Lightweight Transformer was used to establish the detection models of Raw-HSI, RGB and SR-HSI for the prediction of glutamic acid index for beef. The results showed that the MST++ model exhibited the best performance in SR, with RMSE, PSNR, and SSIM values of 0.015, 36.70, and 0.9253, respectively. Meanwhile, the prediction effect of MST++ (R2P = 0.8422 and RPD = 2.46) reconstructed was close to the Raw-HSI (R2P = 0.8526 and RPD = 2.69). The results provide practical application scenarios and detailed analysis ideas for RGB-to-HSI.

Electrical spectroscopy of polaritonic nanoresonators

One of the most captivating properties of polaritons is their capacity to confine light at the nanoscale. This confinement is even more extreme in two-dimensional (2D) materials. 2D polaritons have been investigated by optical measurements using an external photodetector. However, their effective spectrally resolved electrical detection via far-field excitation remains unexplored. This hinders their exploitation in crucial applications such as sensing, hyperspectral imaging, and optical spectrometry, banking on their potential for integration with silicon technologies. Herein, we present the electrical spectroscopy of polaritonic nanoresonators based on a high-quality 2D-material heterostructure, which serves at the same time as the photodetector and the polaritonic platform. Subsequently, we electrically detect these mid-infrared resonators by near-field coupling to a graphene pn-junction. The nanoresonators simultaneously exhibit extreme lateral confinement and high-quality factors. This work opens a venue for investigating this tunable and complex hybrid system and its use in compact sensing and imaging platforms.

Unsupervised Hyperspectral and Multispectral Image Blind Fusion Based on Deep Tucker Decomposition Network With Spatial-Spectral Manifold Learning.

Hyperspectral image (HSI) and multispectral image (MSI) fusion aims to generate high spectral and spatial resolution hyperspectral image (HR-HSI) by fusing high-resolution multispectral image (HR-MSI) and low-resolution hyperspectral image (LR-HSI). However, existing fusion methods encounter challenges such as unknown degradation parameters, and incomplete exploitation of the correlation between high-dimensional structures and deep image features. To overcome these issues, in this article, an unsupervised blind fusion method for LR-HSI and HR-MSI based on Tucker decomposition and spatial-spectral manifold learning (DTDNML) is proposed. We design a novel deep Tucker decomposition network that maps LR-HSI and HR-MSI into a consistent feature space, achieving reconstruction through decoders with shared parameters. To better exploit and fuse spatial-spectral features in the data, we design a core tensor fusion network (CTFN) that incorporates a spatial-spectral attention mechanism for aligning and fusing features at different scales. Furthermore, to enhance the capacity to capture global information, a Laplacian-based spatial-spectral manifold constraint is introduced in shared-decoders. Sufficient experiments have validated that this method enhances the accuracy and efficiency of hyperspectral and multispectral fusion on different remote sensing datasets. The source code is available at https://github.com/Shawn-H-Wang/DTDNML.

Hyperspectral transmittance imaging detection of early decayed oranges caused by Penicillium digitatum using NFINDR-JMSAM algorithm with spectral feature separating

Decay caused by Penicillium spp. is the main cause of postharvest citrus quality loss, moreover, the fungus can quickly infect entire batches of citrus fruit resulting in significant economic losses. However, effective detection of early decay remains a challenge due to the lack of distinct visual features. In this study, a Vis-NIR hyperspectral imaging system was developed to acquire full-transmittance images and an NFINDR-JMSAM algorithm was proposed to segment different image pixels. By extracting pure pixels and separating spectral features, the overall classification accuracy of 99.3 % was obtained for all tested samples. The proposed method can also effectively identify scars on the flavedo, citrus stem-end and navel, thereby eliminating their interference with the detection of decayed orange fruit. This study provided a new idea for accurately detecting the early decayed citrus fruit and visualizing the detection results for different tissues by combining hyperspectral transmittance imaging and NFINDR-JMSAM algorithm.

Application of nondestructive techniques for peach (Prunus persica) quality inspection: A review.

Peaches are highly valued for their rich nutritional content. Traditional fruit quality accessing methods (i.e., manual squeezing the fruit for firmness) are both subjective and destructive, which tend to diminish the integrity of fruit samples, consequently undermining their market value. Compared to traditional detection methods, nondestructive technology offers efficient and noninvasive solutions for rapidly and accurately assessing internal external quality of peaches. This can significantly enhance product classification and quality assurance while reducing the need for extensive human resources and minimizing potential physical damage to peaches. This review provided a comprehensive overview of nondestructive techniques for peach quality evaluation, including visible/near-infrared spectroscopy, machine vision technology, hyperspectral imaging, dielectric and optical properties, fluorescence spectroscopy, electronic nose/tongue, and acoustic vibration methods. It also evaluates the effectiveness of each technique in assessing internal quality, maturity, and disease detection of peaches. The advantages and limitations of each method were also summarized. This study focuses specifically on peaches and encompasses all existing nondestructive testing methods, providing valuable insights and references for future studies in the field of peach quality analysis using nondestructive testing methods.

Characterization of the in-situ degradation process of P3HT:PCBM based on hyperspectral and neural networks

In situ online observation of surface morphology during degradation processes is of paramount importance for exploring the stability of organic photovoltaic materials. In this study, we designed an in situ online characterization system based on hyperspectral and neural network technologies, and observed the degradation processes of P3HT:PCBM thin film materials. The system is capable of collecting hyperspectral image data from 101 channels within the 400–700 nm wavelength range for characterizing detailed surface features of materials. Additionally, to automate the processing of hyperspectral image data, we designed a spectral image segmentation algorithm based on neural networks and proposed a foreground attention mechanism to improve the segmentation accuracy of the algorithm. The experimental results indicate that the system can achieve high spectral characterization of P3HT:PCBM thin film materials and automate image data processing through artificial intelligence algorithms, with an image segmentation accuracy of 99.62 %. Furthermore, owing to the higher spectral resolution of this system and its computer-assisted analysis capabilities for material image data, not only are the in-situ variations in size, density, and formation rate of aggregates formed during the thermal degradation process of P3HT:PCBM thin film materials experimentally analyzed, but also the fluorescence changes at the edges of aggregates during the photodegradation process are revealed. The reliable code can be found at the following link: https://github.com/HyperSystemAndImageProc/IONFMDP-UHHNN.

APiS (AI powered intelligence spectrum analyser): hyperspectral imaging machine learning analyser app for mineral processing

ABSTRACT Hyperspectral data, encompassing hundreds of spectral wavelengths, is a powerful tool for mineral processing, as minerals display unique reflection spectra. Machine learning can identify mineral species from hyperspectral data, but its high dimensionality poses challenges for ordinary users. To address this, we developed APiS (AI Powered intelligence Spectrum Analyser), an interactive hyperspectral analysis application requiring no coding. APiS offers a user-friendly interface for data visualisation, learning data generation, advanced machine learning analysis, and model evaluation. This App surpasses previous script-based methods, enabling diverse users in geometallurgy to conduct sophisticated hyperspectral data analysis and fostering creativity in their work.

Hyperspectral imaging for pest symptom detection in bell pepper

BackgroundThe automation of pest monitoring is highly important for enhancing integrated pest management in practice. In this context, advanced technologies are becoming increasingly explored. Hyperspectral imaging (HSI) is a technique that has been used frequently in recent years in the context of natural science, and the successful detection of several fungal diseases and some pests has been reported. Various automated measures and image analysis methods offer great potential for enhancing monitoring in practice.ResultsIn this study, the use of hyperspectral imaging over a wide spectrum from 400 to 2500 nm is investigated for noninvasive identification and the distinction of healthy plants and plants infested with Myzus persicae (Sulzer) and Frankliniella occidentalis (Pergande) on bell peppers. Pest infestations were carried out in netted areas, and images of single plants and dissected leaves were used to train the decision algorithm. Additionally, a specially modified spraying robot was converted into an autonomous platform used to carry the hyperspectral imaging system to take images under greenhouse conditions. The algorithm was developed via the XGBoost framework with gradient-boosted trees. Signals from specific wavelengths were found to be associated with the damage patterns of different insects. Under confined conditions, M. persicae and F. occidentalis infestations were distinguished from each other and from the uninfested control for single leaves. Differentiation was still possible when small whole plants were used. However, application under greenhouse conditions did not result in a good fit compared to the results of manual monitoring.ConclusionHyperspectral images can be used to distinguish sucking pests on bell peppers on the basis of single leaves and intact potted bell pepper plants under controlled conditions. Wavelength reduction methods offer options for multispectral camera usage in high-grown vegetable greenhouses. The application of automated platforms similar to the one tested in this study could be possible, but for successful pest detection under greenhouse conditions, algorithms should be further developed fully considering real-world conditions.

Deep bidirectional hierarchical matrix factorization model for hyperspectral unmixing

Existing deep nonnegative matrix factorization-based approaches treat shallow and deep layers equally or with similar strategies, neglecting the heterogeneous physical structures between the shallow and progressively deeper layers, thus failing to explore the latent different sub-manifold structures in the hyperspectral image. In this paper, we propose a deep nonnegative matrix factorization model with bidirectional constraints to achieve hyperspectral unmixing. The sub-manifold structures in hyperspectral image are fully exploited by filtering and penalizing the shallow abundance layer with a denoised regularizer and a manifold regularizer. In contrast to the shallow abundance layer, the remaining layers are constrained by an extremely common regularizer to avoid over-denoising and maintain fidelity. In this way, the fine cues between different substances are exploited to a large extent. Additionally, the overall reconstruction error can be well controlled because the performance of the designed feedback mechanism can be fine-tuned by the inverse hierarchical constraints. Finally, we employ Nesterov's optimal gradient method to solve the optimization problem effectively. Experiment results are conducted on both synthetic datasets and real datasets, and all results show that the proposed method is superior to recent canonical unmixing methods.

Cnn-assisted multi-hop graph attention network for hyperspectral image classification

ABSTRACT Recently, the convolutional neural network (CNN) has gained widespread adoption in the hyperspectral image (HSI) classification owing to its remarkable feature extraction capability. However, the fixed acceptance domain of CNN restricts it to Euclidean image data only, making it difficult to capture complex information in hyperspectral data. To overcome this problem, much attention has been paid to the graph attention network (GAT), which can effectively model graph structure and capture complex dependencies between nodes. However, GAT usually acts on superpixel nodes, which may lead to the loss of pixel-level information. To better integrate the advantages of both, we propose a CNN-assisted multi-hop graph attention network (CMGAT) for HSI classification. Specifically, a parallel dual-branch architecture is first constructed to simultaneously capture spectral-spatial features from hyperspectral data at the superpixel and pixel levels using GAT and CNN, respectively. On this basis, the multi-hop and multi-scale mechanisms are further employed to construct a multi-hop GAT module and a multi-scale CNN module to capture diverse feature information. Secondly, an attention module is cascaded before the multi-scale CNN module to improve classification performance. Eventually, the output information from the two branches is weighted and fused to produce the classification result. We performed experiments on four benchmark HSI datasets, including Indian Pines (IP), University of Pavia (UP), Salinas Valley (SV) and WHU-Hi-LongKou (LK). The results demonstrate that the proposed method outperforms several deep learning methods, achieving overall accuracies of 95.67%, 99.04%, 99.55% and 99.51%, respectively, even with fewer training samples.

Local-global feature fusion network for hyperspectral image classification

ABSTRACT Hyperspectral images (HSIs) are rich in spatial and spectral information, making them crucial for accurate classification. While existing convolutional neural network (CNN) methods have shown promise in HSI classification, they primarily focus on local features and struggle with establishing long-range dependencies. To address these limitations, we propose a Local-Global Feature Fusion Network (LGFFN). LGFFN effectively learns joint local-global information from HSIs. Firstly, we employ a group-general convolution parallel module (GGCPM) to extract both local and global spectral features. Next, we utilize parallel branches: a CNN branch for capturing fine-grained spatial information and an improved transformer encoder (ITE) branch for capturing long-distance spatial-spectral features and exploiting semantic information. Finally, we elegantly integrate the diverse features obtained from these branches to obtain classification results. We validate our method on four datasets and compare it with six state-of-the-art hyperspectral classification methods. Our experimental results demonstrate that LGFFN achieves high classification accuracy, outperforming existing methods.