47 results about "Spectral space" patented technology

The invention discloses a spectrum-space depth fusion hyperspectral image classification method oriented to a small sample condition, and the method comprises the steps: firstly carrying out the multi-scale hole convolution processing of input hyperspectral data through a range attention convolution SAC module, and extracting the multi-scale context features; then, a spatial normalization attention SNA mechanism is utilized to carry out adaptive weighting adjustment of spatial dimensions on the feature map, and spatial feature representation of the key area is enhanced; the method comprises the following steps: constructing a lightweight hybrid expert model LMOE, carrying out parallel processing and gating weighting through a multi-path expert network, carrying out efficient refining and mapping on features, finally fusing processed spectral features and spatial features, and carrying out pixel-level prediction through a classifier to obtain a terrain classification result map of a hyperspectral image. The method solves the problems that in the prior art, overfitting is prone to occurring under the small sample condition, the spectrum-space collaborative modeling capacity is insufficient, the long-range dependence obtaining efficiency is low, and the recognition precision is reduced under the class imbalance scene.

The invention discloses an ultra-high-speed imaging device based on acousto-optic filtering modulation. The ultra-high-speed imaging device comprises a spectrum modulation system, a spectrum imaging system and a data processing system. The spectrum modulation system utilizes an acousto-optic tunable filter (AOTF) to dynamically modulate broadband continuous spectrum light, narrow-band illumination light which rapidly changes along with time is generated, and time information of a dynamic scene is effectively mapped to a spectrum dimension. And the spectral imaging system images the spectral coded dynamic scene to the hyperspectral camera to realize acquisition of spectral space-time information. The data processing system performs spectral crosstalk correction, channel separation and demosaicing reconstruction on the acquired original mosaic image, and recovers a time sequence image according to a spectrum-time mapping relation, thereby realizing single-shot superspeed imaging in a dynamic process. The device has the advantages of being simple and compact in structure, flexible in time window, high in imaging fidelity and the like, and has wide application prospects in the fields of rapid dynamics such as microfluidics, laser processing and ultrafast physics.

The invention relates to a geological disaster early warning algorithm based on hyperspectral and Internet of Things data fusion analysis, and relates to the technical field of geological disaster monitoring and early warning, the geological disaster early warning algorithm comprises the following steps: S1, preprocessing hyperspectral data and Internet of Things data, and respectively extracting spectral-spatial features and dynamic topology time sequence features; s2, fusing the spectrum-space features and the time sequence features to generate cross-modal joint features; s3, performing parameter optimization on the fusion features through a quantum-classical hybrid optimization algorithm, and calculating a disaster risk probability; and S4, based on the optimization result and the risk probability, executing an edge-cloud collaborative early warning decision. According to the method, through non-negative tensor ring decomposition of the hyperspectral data and dynamic topology modeling of the Internet of Things, the limitation of a traditional single data source in temporal-spatial resolution and physical relevance is solved, multi-dimensional joint extraction of spectrum-space-mechanical characteristics can be realized, and the characterization precision of a rock-soil body deformation evolution law is remarkably enhanced.

The invention relates to a wetland vegetation extraction method and system based on spatial-temporal feature fusion, and belongs to the field of wetland remote sensing intelligent interpretation. The problems that an existing wetland vegetation classification method does not make full use of remote sensing data time sequence features, and the wetland vegetation extraction precision is insufficient are solved. According to the technical scheme, a single optical multi-temporal remote sensing image is used as input, frequency domain decomposition and convolution long and short time memory network modeling long-term phenology and short-term fluctuation are carried out through a time-frequency enhancement time sequence branch, and spectral grouping of a spectral space branch and cavity convolution multi-scale feature extraction are combined; and a cross-domain cross attention fusion module is utilized to realize time sequence spectral space three-dimensional feature deep fusion, and finally a wetland vegetation community classification map is output. On the premise of not depending on multi-source data, the time-frequency enhanced double-branch space-time model is innovatively provided, the multi-temporal remote sensing data time sequence feature self-adaptive extraction capacity can be effectively improved, the classification stability and robustness are improved, and finally the remote sensing monitoring precision of a wetland ecosystem is improved.

The invention discloses a hyperspectral image target classification method based on a diffusion model, and the method achieves the high-precision classification of targets in a hyperspectral image through a deep learning diffusion model. Firstly, unsupervised diffusion training needs to be carried out on a diffusion model. The unsupervised diffusion process is based on a de-noising diffusion probability model (DDPM). Then, useful diffusion features are extracted from a pre-trained de-noising diffusion probability model (DDPM). And finally, inputting the features into a supervised classification model for classification. According to the hyperspectral image target classification method based on the diffusion model, the advantages of the diffusion model can be utilized, the relation in the spectral space can be captured from the hyperspectral image, effective and efficient deep features can be obtained, spectral space information is fully utilized, and high-precision classification of hyperspectral image targets is achieved.

The invention discloses an automatic silk damage area positioning and selecting method and system. The method comprises the following steps: S1, acquiring hyperspectral data; s2, generating a damage probability graph; s3, generating a physical parameter diagram; s4, collaborative segmentation: taking the damage probability graph and the physical parameter graph as input, performing fusion processing by using a spectrum-space collaborative segmentation network, and outputting a silk damage area segmentation mask after global optimization; according to the invention, a hyperspectral imaging system integrated with a multi-modal reference plate and active illumination compensation is designed according to unique properties of silk, a compensation graph is generated through pre-scanning, and local brightness compensation is carried out by using a DMD projector, so that high uniformity and high fidelity of spectral data are ensured from the source; and a stable and reliable precondition is created for subsequent accurate detection.

The invention relates to a hyperspectral calculation imaging technology, and aims to solve the problems that in an actual remote sensing observation environment, a single imaging mode is limited by a physical sensor, high space resolution and hyperspectral resolution are difficult to obtain at the same time, and geometry and spectrum complementarity exists among multi-source images. The invention provides a hyperspectral fusion imaging method based on topology perception and a differential homeomorphic mechanism. According to the method, differential homeomorphic transformation is introduced to construct spatial degradation priori, a dual learning framework is adopted to jointly optimize high-resolution generation and bidirectional reconstruction tasks, meanwhile, a topology sensing Transform module is utilized to explicitly model a spectrum-space double-domain manifold structure, complementary correlation among multi-source data is fully mined, and the spectrum-space double-domain manifold structure is constructed. The spatial structure consistency and the spectral fidelity of the fused image are effectively improved, the problem of fusion quality reduction caused by geometric deformation and spectral variation in a complex ground feature scene is significantly improved, and a more reliable data support is provided for applications such as remote sensing monitoring and environmental perception.

The application discloses a hyperspectral remote sensing image recognition method and device based on spectral space feature coupling, and utilizes different linear spectral filters to perform feature extraction on the whole hyperspectral remote sensing image, so that multi-class spectral features are obtained. Compared with the traditional technology, the application does not need to perform spectral dimension reduction, thereby avoiding the problem that the traditional technology weakens the subtle spectral information and high-order spectral correlation contained in the hyperspectral data. Meanwhile, the application takes the minimization of image background energy as an objective function to optimize the optimal real spectral feature vector of each linear spectral filter to the target class. Based on this, when the constructed linear spectral filter is used for feature extraction, the background energy can be minimized while the target spectral response is maintained and enhanced, so that the spectral clutter and noise in the complex environment are suppressed, and thus the detection precision of the model in the complex environment background can be improved.

The application discloses a hyperspectral target detection method based on a position-sensitive Gaussian attention Transformer, relates to the technical field of computer vision, and aims at solving the problem that in the prior art, the self-attention mechanism lacks effective spatial position prior, and thus is insufficient in suppressing complex background when modeling global context, resulting in a high false alarm rate. The application sequentially performs preprocessing on a hyperspectral image, extracts spectral-spatial features, and then performs global feature enhancement through a Transformer embedded with a position-sensitive Gaussian attention mechanism, and finally outputs results through a detection head.

The invention discloses an urban land surface coverage classification system based on a multispectral remote sensing technology, and relates to the technical field of spectral spatial data calculation. The system obtains land surface temperature data of a target area and a sub-area and calculates the change rate difference of the land surface temperature data; carrying out first-level correction on the initial contribution degree by utilizing a material thermal tuning coefficient; and performing second-level space coupling correction according to the correlation coefficient of the temperature change of the marked sub-region and the adjacent region, and finally generating an accurate urban surface heat contribution report. According to the method, the transformation of the urban thermal environment contribution from static recognition to dynamic quantitative evaluation is realized, and the accuracy, fineness and practical value of an evaluation result are remarkably improved. The technical problem that the spatial correlation dynamic evaluation adjustment accuracy of urban surface classification thermal contribution is insufficient is solved.

The application discloses a hyperspectral image classification method and system based on a quaternion deep network, mainly strengthens the learning ability of the network and improves the training efficiency of the network through quaternion algebra theory and deep learning theory, and solves the problems of few hyperspectral image data samples, spectral space variability and the like.The implementation scheme is as follows: firstly, considering the high dimensionality of data, dimensionality reduction is carried out through principal component analysis.Next, a deep attention module is constructed, important features are strengthened and noise interference is suppressed, so that the quality of data is improved.Then, a multi-branch network is constructed, the strengthened features are grouped, and a quaternion generator is designed to map the hyperspectral image data to a quaternion space for processing.Meanwhile, a multi-scale quaternion attention module is embedded to further strengthen the feature expression capability.Finally, the features of each branch are fused, and high-precision classification of the hyperspectral image is realized.

The application provides a dynamic granular ball graph neural network working method for user and item recommendation, and belongs to the technical fields of recommendation systems and graph neural networks. The method comprises the following steps: S1, constructing user-side and item-side granular balls in a low-dimensional spectral space, updating the granular balls according to the node representation of each propagation layer, calculating the membership of nodes to the granular balls by using a radius-normalized distance, and defining the collaborative purity according to the interaction distribution of the granular balls on the opposite granular balls; S2, coupling fine-grained interaction modeling and coarse-grained collaborative abstraction, and performing message propagation along two complementary paths; S3, performing bilateral degree perception gating, and adaptively balancing the contributions of the two propagation paths for each node; and S4, aggregating the representations of each layer by weighted summation to obtain the final embedding of the user side and the item side, finally adopting a Bayesian personalized ranking loss to optimize the candidate item score, so as to play a good recommendation role in ranking the user and the item.

The application discloses an intracavity lipid recognition method based on spectral space joint modeling. The method comprises the following steps: collecting intravascular optical imaging data, constructing an optical imaging data set based on the intravascular optical imaging data; constructing a spectral space joint modeling network, inputting the optical imaging data set into the spectral space joint modeling network for training, and obtaining a trained spectral space joint modeling network; the spectral space joint modeling network comprises a preprocessing module, a physical layer module, a feature extraction module and a classification module connected in sequence; inputting real-time collected intravascular optical imaging data to be measured into the trained spectral space joint modeling network for processing, and obtaining a lipid probability vector composed of all A-lines corresponding to the intravascular optical imaging data to be measured. The method has the beneficial effects of explicitly utilizing spectral information, jointly modeling spectrum and deep structure, introducing the context of intra-frame and rollback direction, and being friendly to engineering implementation.

The application discloses a hyperspectral classification method based on a reconstructed convolutional Transformer, and belongs to the field of hyperspectral remote sensing detection, and comprises the following steps: after original image data is acquired, a spectral space reconstruction network is introduced, and separation and reconstruction operations are adopted; a scaling factor in a group normalization layer is combined to evaluate the information content of different feature maps, so that a calculation feature map is acquired; a spectral feature reconstruction network is introduced by using the redundancy of features, and segmentation transformation and fusion strategies are adopted; a spatial attention module is used to overcome the limitation caused by the grid structure of a convolution kernel, and a new activation scheme is introduced; a spectral correlation module is used to establish the connection of different pixel spectra in a hyperspectral cube, and an input feature map is reconstructed; and a linearized output is obtained by using Pooling, so that a classification result is obtained. The application realizes hyperspectral image classification with higher precision and faster speed, and can maintain stable classification performance in various complex scenes.

The application discloses a kind of semi-supervised hyperspectral image intelligent classification method, accurate ground object identification is carried out using labeled sample and unlabeled sample.The application first constructs a multiscale deformable spectral spatial feature extraction module, can extract features on multiple scales and fusion, the introduction of deformable convolution can extract more in line with the distribution characteristics of sample.Then through the proposed pseudo-label generation strategy, Gaussian function is used to weight, further improve the utilization efficiency of pseudo-label and model performance.In the training process, maintain high quantity and high quality of pseudo-label to effectively use unlabeled sample.Finally, through the transformation, long-range dependencies are captured and effective feature representations of hyperspectral remote sensing images are learned.Experimental results demonstrate the effectiveness and superiority of the application under the condition of limited labeled samples.

This application discloses a method, apparatus, equipment, and medium for assessing photovoltaic power plant disaster damage, belonging to the technical field of photovoltaic power plant disaster damage assessment. The method includes: acquiring pre-disaster and post-disaster image data of the target area; mapping the pre-disaster and post-disaster image data to photovoltaic panel features based on the spectral characteristics of photovoltaic modules, obtaining pre-disaster and post-disaster feature maps with positive and negative dual-polarization distribution characteristics; performing clustering on the post-disaster feature map to obtain a multimodal spectral space corresponding to the post-disaster feature map; using a preset reference probability distribution model to assign an abnormal energy score to each pixel in the post-disaster feature map; and classifying the disaster damage according to the abnormal pixels and the multimodal spectral space to generate a disaster damage classification map, wherein abnormal pixels include pixels with abnormal energy scores greater than a preset judgment threshold. This application can achieve automated and standardized identification of damaged photovoltaic areas.

The invention provides a fire range extraction method based on a nuclear combustion index and adaptive spatial spectrum optimization, and relates to the technical field of remote sensing image processing, and the method comprises the steps: collecting remote sensing image data; mapping a spectral vector corresponding to the fire image data set to a high-dimensional space based on a Gaussian radial basis kernel function, and performing kernel principal component analysis on the spectral vector to construct a kernel difference combustion index; according to kernel density estimation, performing adaptive threshold segmentation on the kernel difference combustion index to obtain an initial fire pixel detection result; constructing a spectrum space composite kernel similarity matrix, and optimizing the initial fire pixel detection result in a conditional random field according to the spectrum space composite kernel similarity matrix to obtain a transition fire pixel detection result; and performing morphological operation and small patch area filtering on the transition fire pixel detection result to obtain a predicted fire pixel detection result. According to the invention, the overall precision of fire range extraction can be improved.

The invention discloses a hyperspectral image classification method for non-Euclidean spectral-spatial feature mining, and the method comprises the steps: obtaining input hyperspectral image data, and carrying out the preprocessing of an image, so as to obtain a pixel-level data set; the method comprises the following steps: extracting spectral-spatial features in Euclidean space by using a convolutional neural network (CNN), constructing a graph structure based on superpixel nodes, and mapping a hyperspectral image to a non-Euclidean space; based on the constructed graph structure, a graph convolutional neural network GCN with a gating mechanism is adopted to extract spectrum-space features of a non-Euclidean space; performing weighted fusion on the features extracted by the CNN branch and the GCN branch to obtain joint spectrum-space feature representation; and finally, inputting the fusion features into a classifier, and outputting a classification result of the target scene. According to the method, the misclassification risk caused by salt and pepper noise in a common classification method is remarkably reduced, and the accuracy of a classification boundary is improved.

The present application belongs to the technical field of grain informatization processing, and particularly relates to a method and system for identifying imperfect grains based on spectral spatial coupling features. The method comprises inputting visible light images and hyperspectral data of the grains into a recognition model; a visible light image branch in the recognition model processes the visible light images of the grains to obtain the output of the branch; a hyperspectral data branch in the recognition model processes the hyperspectral data of the grains to obtain the output of the branch; the outputs of the visible light image branch and the hyperspectral data branch are used to obtain the recognition result; the recognition model is trained using grain sample images; the processing of the hyperspectral data branch comprises: expanding the hyperspectral data represented by a high-order tensor into a tensor factor multiplication mode containing spectral and spatial information features, and each mode is a tensor form in which all different spectral information features are superimposed; and the tensor form is subjected to feature extraction to obtain the branch output.

The invention discloses a complex threat intelligence analysis and logical reasoning method based on an intelligent agent, and belongs to the technical field of intelligent agent intelligence processing. The objective of the invention is to solve the problem of accurate and explainable analysis of complex threat intelligence. The method comprises the following steps: performing question decomposition on a question input by a user based on a multi-strategy decomposition mechanism to obtain decomposed sub-questions; adopting self-adaptive hybrid retrieval and information refining to obtain preliminarily screened candidate documents; constructing a hybrid reordering model based on a spectrogram theory and a determinant point process, including construction of a hybrid semantic weighted graph, adaptive spectrum embedding and noise reduction and greedy k-DPP selection based on a spectrum space, and reordering the preliminarily screened candidate documents to obtain topological data of the reordered hybrid semantic weighted graph; and reasoning and generating a final answer based on the constructed global evidence graph. According to the method, logic coherence of outputting complex threat intelligence is realized, and a final answer which is accurate in reference and has a global view angle is quoted.

The application discloses a hyperspectral compression imaging method, system, terminal and storage medium in the technical field of hyperspectral image processing, and aims to solve the problem of low reconstruction accuracy caused by high-frequency detail loss, weak spectral correlation and poor noise resistance in the compression imaging process in the prior art. It comprises obtaining the compression measurement and the perception mask of the coded aperture snapshot spectral imaging system, obtaining the initial feature map according to the compression measurement and the perception mask; inputting the initial feature map into the pre-constructed U-Net model to obtain the decoding feature map; and performing convolution mapping on the decoding feature map to obtain the reconstructed hyperspectral image in combination with the initial feature map; the spectral-space state synchronizer efficiently processes long sequence data, and the wavelet variance modulation block can effectively capture the local and global features of the image through multi-scale decomposition, thereby solving the problem of low reconstruction accuracy caused by high-frequency detail loss, weak spectral correlation and poor noise resistance in the traditional compression imaging process.

The application discloses an image super-resolution reconstruction method based on space-frequency dual-domain representation learning, and relates to the technical field of image super-resolution reconstruction. The method constructs a multi-scale receptive field through four-way differentiated convolution, synchronously captures coarse-grained structure and fine-grained texture information, and combines a convolution feedforward network to complete feature refinement; the method introduces a window multi-head self-attention mechanism to capture spatial long-range dependencies to optimize low-frequency global representation, and then realizes high-frequency information refinement from global to local through an efficient distillation structure; the method modulates spectral space information adaptively by means of a learnable frequency domain filter, simultaneously generates spatial attention guided by high-frequency prior, and promotes deep interaction of dual-domain features; the method realizes deep feature mapping through six recursively connected feature representation groups, and completes high-resolution image reconstruction in combination with a pixel shuffling layer. While keeping the model lightweight, the application effectively breaks through the inherent limitations of single architecture and single-domain representation, and significantly improves the comprehensive performance of super-resolution reconstruction.

The invention discloses a small and micro land surface water body refining extraction method based on multi-scale remote sensing, which comprises the following steps: constructing a high-medium-low resolution image pyramid by using multi-source satellite remote sensing data, realizing self-adaptive wave band combination optimization based on small and micro water body spectral characteristics, and carrying out multi-scale noise suppression and radiation normalization processing; presetting a small and micro water body geometric shape quantification index, constructing a small and micro water body screening mechanism constrained by geometric features, and realizing coarse positioning of a small and micro water body candidate region; a scale weight adaptive distribution algorithm is adopted, contribution weights of data of different scales are dynamically adjusted according to the geometrical characteristics of the candidate areas, a combined discrimination model of spectrum-space-geometrical characteristics of the small and micro water body is established, and accurate extraction of the small and micro water body is achieved; and constructing a boundary optimization method based on a functional energy model, and realizing small and micro water body boundary optimization by combining small-area noise removal, connected region combination and result consistency check of time series data constraint.

This invention relates to a multispectral point cloud classification method based on spatial-spectral self-supervised pre-training, belonging to the field of multispectral lidar point cloud processing technology. The method includes: voxelizing multispectral point cloud samples and preprocessing them to obtain a self-supervised distance matrix between voxel blocks in three-dimensional Euclidean space and spectral space; extracting global and local features of the multispectral point cloud samples, performing feature pooling to obtain point-level feature representations, and then performing residual connections with the extracted voxel block spatial features; learning implicit topological relationships between voxel blocks in three-dimensional Euclidean space and spectral space through graph modeling based on a self-supervised distance matrix constraint model to obtain a pre-trained model; introducing a learnable classification head for training to obtain a multispectral point cloud classification model to achieve point cloud classification. The aim is to improve the pre-trained model's ability to capture information about common features of multispectral point cloud ground objects in complex remote sensing scenarios and effectively utilize the spatial-spectral consistency features of multispectral point cloud data.

The application provides a remote sensing panchromatic sharpening method and system based on a cross-spectrum-space fusion network, comprising: down-sampling high-resolution multispectral remote sensing images to target low-resolution multispectral remote sensing images to construct a high-low resolution sample library; taking the low-resolution multispectral remote sensing images and the panchromatic remote sensing images as inputs, extracting feature information of the low-resolution multispectral remote sensing images and the panchromatic remote sensing images through a basic residual module; using a cross-spectrum-space attention module to extract spectral information of a multispectral branch and spatial information of a panchromatic branch, then gradually enhancing spectral and spatial feature representations of the remote sensing images, so as to improve the fusion process and generate high-resolution multispectral images with detailed information. The scheme of the application is superior to other remote sensing image panchromatic sharpening methods in qualitative and quantitative evaluation, and can generate higher-precision remote sensing image high-resolution multispectral reconstruction results.

The application discloses a graph data backdoor defense method and device based on spectral domain transformation and edge weight learning, equipment, medium and program product, relates to the technical field of network security, and includes: obtaining and constructing a feature vector matrix and an eigenvalue matrix of an original graph data set; mapping each feature vector matrix from a spatial domain to a spectral space through spectral transformation to obtain a plurality of spectral domain node features, so as to construct a normal data distribution of the original graph data set, and determine whether each to-be-tested graph data is abnormal graph data; if it is determined that the graph data is abnormal, learning abnormal node features and abnormal adjacency relationships of the to-be-tested graph data through a multilayer perceptron to obtain abnormal edge weights; based on the abnormal edge weights, performing adaptive clipping on abnormal spectral components in the abnormal graph data through a Gaussian mixture model to obtain modified spectral domain node features; and performing inverse spectral transformation on each modified spectral domain node feature to reconstruct a target graph data set. The application can effectively resist multiple types of graph backdoor attacks.

The invention relates to a multispectral point cloud classification method based on space-spectrum self-supervision pre-training, and belongs to the technical field of multispectral laser radar point cloud processing. The method comprises the following steps: voxelizing a multispectral point cloud sample and preprocessing to obtain a self-supervised distance matrix between voxel blocks in a three-dimensional Euclidean space and a spectral space; extracting global features and local features of the multispectral point cloud sample, performing feature pooling to obtain point-level feature representation, and performing residual connection with extracted voxel block spatial features; an implicit topological relation between voxel blocks in a three-dimensional Euclidean space and a spectral space is learned based on a self-supervised distance matrix constraint model through graph modeling, a pre-training model is obtained, a learnable classification head is introduced for training, and a multispectral point cloud classification model is obtained to realize point cloud classification. The objective of the invention is to improve the information capturing capability of a pre-training model for the universal features of multispectral point cloud ground objects in a complex remote sensing scene, and effectively utilize the spatial-spectral consistency features of multispectral point cloud data.

The present application relates to a ground object classification method based on spectral space fusion Transformer feature extraction, and belongs to the technical field of ground object classification.The purpose of the present application is to solve the problems of the existing DL method, on the one hand, the CNN using local extraction and global parameter sharing mechanism pays more attention to spatial content information, so that the spectral sequence information in the learning feature is distorted; on the other hand, the CNN is difficult to describe the long-distance correlation between HSI pixels and bands.The process is as follows: first, a spectral space fusion multi-head double self-attention Transformer feature extraction classification network is established, and a trained network is obtained based on a training set; second, the image to be tested is input into the trained network to complete the classification of the image to be tested; the spectral space fusion multi-head double self-attention Transformer feature extraction classification network comprises SpaFormer, SpeFormer, AS 2 FM and classifier.The present application is used in the field of hyperspectral image classification.

This invention discloses a deep learning-based method and system for modeling atmospheric pollution diffusion, comprising the following steps: Step 1: Constructing a two-dimensional spatial grid; Step 2: Constructing a three-dimensional voxel space model; Step 3: Generating pollution information units; Step 4: Constructing a three-dimensional pollution diffusion state field; Step 5: Performing spatiotemporal spectral decomposition of the topological manifold to form a set of pollution diffusion trajectories; Step 6: Inputting the three-dimensional pollution diffusion state field into the state branch of an improved DeepONet model, inputting the set of pollution diffusion trajectories into the trajectory branch, constructing a pollution diffusion operator matrix through an operator projection layer, performing spectral space decomposition, and introducing a spectral manifold folding mechanism to obtain the three-dimensional pollution diffusion state at the next time step; Step 7: Obtaining the updated pollution information units; Step 8: Generating the spatial distribution of pollution concentration. This invention achieves stable modeling of atmospheric pollution diffusion through spatiotemporal spectral decomposition of the topological manifold and an improved DeepONet model.

The invention relates to the field of hyperspectral image classification, and discloses a multi-hop differential Transform and pyramid gating network for Mars hyperspectral image classification, the multi-hop differential Transform and pyramid gating network comprises a multi-hop differential Transform module, the multi-hop differential Transform module is composed of neighborhood feature mapping and a Transform encoder based on multi-hop differential attention, a PoNF is used for aggregating neighborhood spectral features, and the PoNF is used for aggregating the neighborhood spectral features; the TE with multi-hop differential attention further extracts global spectral features and captures fine local spectral changes; the spatial pyramid gating enhancement module is composed of an SPGB and an MLP, a double-threshold gating strategy is further designed, and extracted spatial features are finely selected; the spectral space adaptive fusion module is used for realizing deep interaction and fusion of spectral features and spatial features through an adaptive weighting mechanism; the network of the invention further extracts global spectral features and captures local fine spectral changes through a multi-hop differential attention mechanism, thereby enhancing the ability of the model to extract discriminative spectral features from highly similar spectrums.