47 results about "Hilbert curve" patented technology

The invention discloses a tree three-dimensional reconstruction method based on a ground laser radar point cloud, and the method comprises the following steps: carrying out the down-sampling and noise reduction preprocessing of the point cloud data of a single plant through a Hilbert curve; carrying out branch and leaf separation on the preprocessed data by adopting a LeWoS model, and extracting a clear trunk point cloud; and extracting an initial skeleton of the tree by using a semantic Laplacian contraction algorithm based on the trunk point cloud, and optimizing the initial skeleton by using a Laplacian smoothing algorithm. The method has the beneficial effects that the S-LBC algorithm has more advantages in the aspect of maintaining the integrity of the topological structure of the tree, the problems of skeleton breakage and structure simplification can be avoided, and the overall shape of the tree is more natural; according to the method, the robustness under different point cloud densities is achieved, even under the condition of data missing or point cloud sparseness, the overall structure of the tree can still be well reconstructed, and reliable technical support is provided for high-precision tree modeling.

The invention discloses a serialization-based point cloud over-segmentation method, device and equipment and a medium, and relates to the technical field of point cloud over-segmentation. The point cloud over-segmentation method comprises the following steps: obtaining point cloud data set preprocessing; serializing the preprocessed point cloud data into a Hilbert curve; dividing the serialized point cloud data into a plurality of initial segments based on spatial continuity, and obtaining a multi-scale segment set; the feature similarity between adjacent segments is calculated, and a similarity matrix is obtained and established to provide a basis for subsequent super-point clustering; carrying out super-point clustering on the initial segments based on a self-adaptive updating algorithm; updating super-point features through a cross attention mechanism, establishing dynamic association between points and super-points, and obtaining a new super-point structure; and according to the updated super-point features, constructing a super-point graph, performing enhancement processing through a graph convolutional network, fusing features from the backbone point cloud network and corresponding multi-level super-point features, then transmitting the fused features to a segmentation head for segmentation, and obtaining semantic segmentation output.

The invention provides a gas detection method and a gas detection device.The method comprises the steps that gas spectral data are obtained, the gas spectral data are obtained by detecting mixed gas in a gas cell through a quantum cascade laser, and the mixed gas comprises gas to be detected; according to global index information of the gas spectrum data, obtaining a plurality of local index information, the global index information representing coordinate curves of a plurality of signals in different coordinate system quadrants in the gas spectrum data; on the basis of a Hilbert space mapping method, quadrant overturning and rotation processing is carried out on a coordinate system quadrant of the at least one piece of local index information, and initial global coordinates are obtained; calculating a new initial global coordinate by taking the initial global coordinate as new global index information, and determining the new initial global coordinate corresponding to the target order of the Hilbert curve as target spectral data; and inputting the target spectral data into the gas concentration detection model, and outputting gas concentration information of the to-be-detected gas.

The invention provides an optical high-dimensional data time domain compression calculation method and system based on Hilbert mapping, and the method comprises the steps: employing a Hilbert mapping method, enabling original high-dimensional data to be expanded into a one-dimensional vector from a memory according to the dimension and size Size of the original high-dimensional data and a Hilbert curve with the order n of the same dimension, and carrying out the time domain compression calculation of the original high-dimensional data, and a digital-to-analog converter drives a high-speed optical modulator to load the vector to the time-varying intensity of the optical carrier so as to form signal light. Then, the signal light enters a basic time domain optical operation unit to be further processed, including weighted secondary intensity modulation, or optical difference (or differential) operation used for extracting a neighborhood information change relation and the like. And finally, vector information carried by the processed one-dimensional time domain signal light is extracted and stored, Hilbert inverse mapping of a corresponding dimension is carried out on the vector information, and the vector information is restored into high-dimensional feature information for subsequent operations such as judgment, segmentation and the like by an artificial intelligence algorithm.

The invention discloses a point cloud registration method based on an attention mechanism and a state space model. The method comprises the following steps: obtaining a standard surface shape precision point cloud and a deformed surface shape precision point cloud of an astronomical instrument; carrying out different levels of downsampling processing on the point cloud; sequentially carrying out serialization of a Hilbert curve, a transposed Hilbert curve and a zero-order curve on the processed point clouds of different levels; inputting the serialized point cloud into a coarse registration network fusing an attention mechanism and a state space model to extract features, and obtaining a corresponding relation of point cloud features; and estimating initial point cloud rotation and translation transformation through a dense-sparse fine registration network based on the corresponding relationship, and finally optimizing initial transformation through a weighted singular value decomposition algorithm to obtain a final registration result. According to the point cloud registration method, the registration effect of the point cloud registration method in a large-scale precise point cloud scene is improved, good registration performance is also ensured, and the effect of point cloud registration in surface type detection of high-precision equipment such as astronomical instruments is greatly improved.

This invention provides a method, apparatus, device, and medium for UAV route planning based on Hilbert curves. The method includes: collecting real-time position, attitude, and environmental information of the UAV; constructing an initial UAV route using a Hilbert curve based on the UAV's mission requirements and the collected information; smoothing the initial UAV route and removing redundant paths using a Bézier curve to obtain an optimized UAV route; correcting the optimized UAV route based on real-time environmental changes to generate a safe route; and converting the safe route into executable control commands for the UAV to drive its operation, thereby solving the problems of insufficient space filling efficiency, poor adaptability to dynamic environments, and low path smoothness in dynamic path planning.

The application discloses a non-full-field data synchronization method based on a Hilbert curve and related equipment thereof, and belongs to the technical field of big data. The application obtains the primary key information of non-full-field data, performs a hash operation on the primary key information to obtain a hash value corresponding to the primary key information, maps the hash value to a two-dimensional space where a Hilbert curve of full-amount data is located, obtains coordinate data of the hash value on the Hilbert curve, inputs the coordinate data into a preset Hilbert space curve algorithm model to obtain position information of the non-full-field data, obtains a synchronization mode corresponding to the non-full-field data, and synchronizes the non-full-field data based on the synchronization mode and the position information. In addition, the application also relates to the technical field of blockchains, and the non-full-field data can be stored on nodes of a blockchain network. The application can realize fast query and positioning of non-full-field data, quickly complete synchronization of non-full-field data, and improve the efficiency of data synchronization.

The application belongs to the field of computer application, and discloses a directed road network error limited Hilbert tree learning index construction method, which projects a space-time sample into a position-time grid, sorts cells of the grid based on a Hilbert curve, then divides the sorted sample into logical blocks and stores the logical blocks into a disk, and simultaneously establishes a binary tree for dividing nodes according to Hilbert curve serial numbers, representing a mapping relationship between the Hilbert serial numbers and the logical blocks, and ensuring that the mapping error of leaf nodes is less than a specified error. The application divides a two-dimensional space according to the Hilbert curve, ensures that trajectory data in the position-time grid is uniformly distributed, ensures full recall of query data through bounded error, and thus improves the query performance on inclined data.

This invention discloses a liquid-cooled plate flow channel structure based on Hilbert curves, belonging to the field of power battery thermal management technology. It mainly involves a coolant flow channel network composed of multiple flow channel units. The coolant flow channel network includes multiple integrated main channels, each consisting of multiple flow channel units connected in series. Each flow channel unit includes a unit inlet end, a Hilbert curve flow channel, and a unit outlet end. The liquid-cooled plate has inlets on both sides of its long side and outlets on both sides of its short side, forming intersecting transverse flow channels. This invention utilizes the self-similarity and high spatial coverage of the Hilbert curve to achieve near-full coverage flow of the coolant, expanding the heat dissipation area. Furthermore, it enhances heat transfer through continuous turns and flow channel contraction-expansion, significantly improving temperature uniformity and heat dissipation efficiency, and solving problems such as localized overheating in traditional straight channels or serpentine flow channels.

The invention discloses an event flow action identification method based on a selection state space. The method comprises the following steps: step 1, carrying out noise reduction processing on original event stream data through a DBSCAN noise reduction algorithm; 2, scanning and serializing the event stream data after noise reduction through a Hilbert curve and a Zigzag curve, and converting an unstructured event stream into a serialized event stream; 3, encoding the serialized event stream data through a PointNet network to generate a spatial feature Token, extracting a time feature Token through a time aggregation module, and splicing the spatial feature Token and the time feature Token to obtain a complete spatial-temporal feature; and 4, inputting the spatial-temporal characteristics into a Bi-SSM module, and capturing bidirectional time dependence in an event stream through a bidirectional selection state space model, thereby realizing accurate recognition of the action. The method effectively solves the problems that in the prior art, calculation complexity is high, time information of event streams cannot be fully utilized, and model efficiency is low, and has the advantages of being low in calculation complexity and long in time sequence dependence modeling.

This application provides a method and apparatus for classifying encrypted traffic based on path signatures, relating to the field of artificial intelligence and also applicable to the financial field. The method includes: segmenting acquired encrypted traffic at the session granularity to obtain traffic paths; performing traffic path signatures on the traffic paths to obtain corresponding traffic path signature feature vectors; and inputting the traffic path signature feature vectors, after applying a Hilbert curve transform, into a pre-trained encrypted traffic classification model to obtain the encrypted traffic classification corresponding to the encrypted traffic. This application can generate signature information for encrypted traffic, thereby reflecting the geometric and statistical characteristics of encrypted traffic and exhibiting strong generalization ability.

The invention discloses a CNN-RNN-based nuclide identification method. The method comprises the steps of S1, gamma energy spectrum data acquisition and expansion; s2, gamma energy spectrum data preprocessing: performing smoothing and normalization processing on the energy spectrum data obtained in the step S1, mapping the processed energy spectrum data to 0-255, then transforming the processed energy spectrum data to a two-dimensional space through a Hilbert curve, generating a nuclide energy spectrum, and dividing the obtained nuclide energy spectrum data into a nuclide training set and a nuclide test set; s3, a nuclide identification model training and testing process: building a CNN-RNN model, decoding the nuclide training set obtained in the S2, and completing model training of the CNN-RNN model based on decoded data to obtain a trained CNN-RNN model; s2, testing the trained CNN-RNN model based on the nuclide test set obtained in S2, and when corresponding test indexes are met, exporting the corresponding CNN-RNN model as a nuclide identification model; and S4, completing the identification of the nuclide to be detected based on the obtained nuclide identification model. Therefore, the problem of the accuracy of nuclide identification is solved.

The invention relates to the technical field of location service privacy protection, in particular to a location privacy protection method based on regional confusion and differential privacy, which comprises the following steps: S1, dividing regions, processing a location set uploaded by a user, constructing a probability distribution matrix, and generating a user pseudo location; and S2, request forwarding: the relay server receives a user query request and generates a hidden area, forwards the query request to the LBS server based on the hidden area, and finally returns a query result to the user. According to the method, the positions are mapped and sorted through the Hilbert curve, so that the spatial proximity of the positions is accurately reflected; and the optimal clustering number k is adaptively determined in combination with a QK-means clustering algorithm, so that PLS division is more reasonable, and the accuracy and matching efficiency of region division are improved.

The present invention relates to the field of mechanical engineering and heat transfer. In particular, it provides a plate heat exchanger based on a surface constructed from the Hilbert curve, characterized in that it comprises a three-dimensional structure of substantially quadrangular shape, formed by a plurality of stacked metal plates; a metal shell operatively covering said substantially quadrangular three-dimensional structure; a set of double channels for the circulation of a fluid 1 and of a fluid 2, which completely covers each of the metal plates forming said plurality of metal plates, following a Hilbert curve of arbitrary order; and a first main collector double duct for the fluid 1 that has passed through the channels in the plates, and a second main collector double duct for the fluid 2 that has passed through the complementary channels of the plates.

The application discloses a land ownership space information construction method based on cadastral survey data, relates to the technical field of data processing, and comprises the following steps: adopting cadastral elements and ownership identification in cadastral survey results to perform coordinate unification, geometry repair, reference relationship calibration and ownership relationship calibration, and generating a cadastral space object sequence; performing Hilbert curve coding sorting on parcel space objects in the cadastral space object sequence, and dividing data sections according to cadastral sub-areas and continuous coding intervals, to generate a cadastral segmented directory; constructing a Hilbert global directory, an R-star tree local space index and a B-plus tree ownership attribute index through the cadastral segmented directory, and generating an ownership space construction index table. Through vectorized R-tree space connection, the application can perform batch judgment on the spatial relationship among parcel surfaces, boundary lines and boundary points, and can generate segmented ownership space results in combination with ownership attribute matching, so that the spatial relationship calculation efficiency and the ownership corresponding accuracy in a big data processing scene can be improved.

The application provides a kind of laser radar point cloud fast query matching method based on local sensitive hash, first through laser radar scanning collection environment data, point cloud segmentation is carried out to the point cloud data collected using K-means clustering segmentation algorithm;Then the feature of data object is extracted, and the data set S after preprocessing is obtained;The point cloud data in voxel is generated one-dimensional index sequence using pseudo-hilbert curve generation algorithm;The pseudo-hilbert curve sequence obtained is encoded using LSH, and the index is constructed, the Euclidean distance between query point cloud and candidate point cloud is confirmed, and the nearest point cloud in voxel is confirmed.The method of the application improves the speed of point cloud feature extraction query based on local sensitive hash index method, shortens the query matching time and improves the query matching precision;LSH index gathers similar point clouds together by hashing coding pseudo-hilbert sequence, reduces the number of point clouds that need to calculate distance, thereby improves the retrieval efficiency.

The invention discloses a sea wave disaster early warning information precision verification method, and relates to the technical field of intelligent management, and the method comprises the following steps: dividing a three-dimensional space-time grid, calculating sea wave energy characteristics, reconstructing the monitoring data of a problem node, generating an analysis link, and carrying out the first verification of alarm information based on the time characteristics of a sea wave disaster; and updating the alarm information according to the first verification result. According to the method, the global sea area is divided into the three-dimensional grids with the unified temporal-spatial resolution, spatial-temporal alignment of multi-source monitoring data is achieved through three-dimensional Hilbert curve coding, the precision and consistency of sea wave energy feature extraction are improved, a comparison mechanism of theoretical energy and actual energy is constructed based on an inter-node energy propagation model, and the real energy feature extraction accuracy is improved. The reliability and robustness of monitoring data are improved, the accuracy and authority of early warning information are improved by constructing an analysis link, and the node with the highest error risk is preferentially processed by adopting a strategy of firstly verifying problem nodes and then rechecking reliable nodes.

The present invention discloses a data distributed clustering method and device based on local direction centrality. The method includes the following steps: S1. Submit the parameters required for the algorithm task in a distributed cluster environment and read the data to be clustered; S2. Construct a priority search K-means tree global index based on the complete data and share the index variables to each working node in the cluster; S3. Divide the complete data by combining data sampling and Hilbert curve partitioning method; S4. Parallelly execute CDC local clustering on each working node; S5. Perform inter-region cluster merging according to the maximum reachable distance of the local clusters to generate complete clusters; S6. Output the clustering result to the distributed file system. The method of the present invention performs distributed optimization and acceleration on the CDC clustering algorithm from two perspectives of algorithm process optimization and parallel processing optimization, aiming to improve the computing efficiency of the CDC algorithm and provide a feasible optimization scheme for the application of this algorithm in massive data mining and machine learning tasks.

The invention relates to the technical field of geographic information processing, in particular to a real estate surveying and mapping resource sharing system and method, and the system comprises a heterogeneous primitive analysis module which constructs a spatial topological incidence matrix and generates a standardized semantic feature set, and a distributed index module which generates spatial position hash codes by using a Hilbert curve to perform fragmentation storage, the vision field verification module uses a ray crossover algorithm to calculate a space inclusion state and generate a dynamic access authorization token, and the increment synchronization module calculates a geometric element difference package to complete instant redrawing. According to the method, topological dislocation caused by source data format difference is eliminated by constructing a heterogeneous semantic mapping matrix, bandwidth occupation and loading delay are reduced based on a differential transmission mechanism of viewpoint coordinates, and self-adaptive adjustment of access granularity is realized by utilizing a dynamic geofencing algorithm; the problem of I / O blocking under high-concurrency requests is solved by adopting distributed hash fragment index logic, and accurate distribution and real-time synchronization of resources are ensured.

The present invention relates to a color image encryption method based on a four-dimensional memristive hyperchaotic system. The present invention includes the following steps: S1: Construct a four-dimensional memristive hyperchaotic system and discretize the system. S2: Read the color plaintext image information and split the color image to be encrypted into three primary color component matrices. S3: Obtain the initial value containing the plaintext information; combine the initial value with the discrete memristive hyperchaotic system for iteration to obtain a hyperchaotic sequence. S4: Preprocess the hyperchaotic sequence so that the chaotic sequence is applicable to image data to obtain a key sequence. S5: Use the key sequence to perform encryption operations on the rows and columns of the R, G, and B channels of the color plaintext image respectively. The initial key of the present invention contains the Hash value of the original image, and completes the permutation and diffusion operations of the image under the introduction of the Hash table structure and the Hilbert curve, making the relationship between the plaintext, ciphertext, and key more complex.

The invention discloses a wind power generation power estimation method based on data fusion processing, and relates to the technical field of wind power generation power prediction, and the method comprises the steps: collecting real-time multi-modal data, carrying out the preprocessing, generating a real-time feature set, extracting a power sequence, calculating the number of IMFs through a cyclic autocorrelation function and a similarity coefficient, and carrying out the prediction of the IMFs. The method comprises the following steps: performing EEMD decomposition in combination with noise amplitude to obtain an IMF sequence, dividing a wind power plant into equal-proportion grids, mapping geographic coordinates of a fan into a two-dimensional grid index, converting the two-dimensional grid index into one-dimensional codes through a Hilbert curve algorithm, constructing space-time grids in combination with time binary codes, mapping multi-dimensional feature vectors to generate an aggregation space-time feature cube, and performing power prediction; according to the method, the input quality and precision of the prediction model are improved through time alignment and feature extraction of the multi-modal data, and the prediction robustness and the system adaptability are improved by using EEMD decomposition and space-time grid modeling.

The invention discloses a wood detection method and system based on a quantum image enhancement algorithm, and relates to the technical field of image processing, and the method comprises the steps: placing a to-be-detected wood sample in a terahertz imaging device, collecting a multi-view reflection signal, obtaining a frequency domain signal through FFT transformation, and mapping a frequency spectrum into a two-dimensional image to form an initial view image set; performing quantum DNA coding on the pixel value of each image, performing mapping by adopting four bases to generate a DNA image matrix, rearranging pixel addresses by using a Hilbert curve traversal method, and outputting a disturbed image; converting the disturbance image into a one-dimensional time sequence signal, and establishing an AR model to obtain a preliminary enhanced image; according to the method, the separability and identifiability of the defect edge are effectively improved, and three-dimensional expression and highlight display of the geometric morphology of the defect are realized through three-channel pseudo-color mapping and three-dimensional voxel grid modeling.

The invention discloses a large-scale spiking neural network efficient partitioning method based on a connection mode index, and the method comprises the following steps: aiming at the characteristics of brain-like computing hardware, carrying out the topological sorting, carrying out the locality-keeping index sorting of neurons through employing a Hilbert curve, and carrying out the resource-limited segmentation algorithm based on binary search, neurons with common pre-synaptic inputs are clustered into the same computing core. According to the method, a pulse sharing mechanism of hardware can be utilized to the maximum extent, the pulse broadcast flow of the network-on-chip is remarkably reduced, and core resource waste caused by axon memory limitation is reduced. Compared with the prior art, the compiling speed is increased by several orders of magnitude when the network with the gigons of neuron scale is processed, and the communication delay and the energy consumption during operation are greatly optimized at the same time.

The invention discloses a metro shield tunnel leakage detection method based on Hilbert curve point cloud serialization. The metro shield tunnel leakage detection method comprises the following steps: acquiring point cloud data of a metro shield tunnel by using a mobile laser radar system; preprocessing the point cloud data, including tunnel ancillary facility removal and point cloud denoising; leakage instance labels are manually labeled, point cloud data are divided into a leakage class and a background class, and a leakage detection data set is constructed; performing data enhancement on the data set; serializing the point cloud data in the leakage detection data set based on a Hilbert space curve so as to optimize the spatial continuity of the point cloud and improve the input efficiency of a subsequent deep learning model; and inputting the serialized point cloud data and the corresponding point cloud intensity value into a Point Transformer deep learning network for model training, constructing a Lovasz Hinge loss function, and obtaining an optimal model by optimizing the loss function. The method is suitable for leakage monitoring and maintenance of an actual subway tunnel, and the overall detection efficiency and detection precision are greatly improved.

The invention provides an anti-counterfeiting image texture feature extraction method and system, and the method comprises the steps: carrying out the YCbCr color space conversion of an anti-counterfeiting image, carrying out the two-dimensional dual-tree complex wavelet transformation of Y, Cb and Cr components, and obtaining a multi-scale multi-direction high-frequency sub-band coefficient matrix; energy of each high-frequency sub-band is calculated, global texture complexity parameters are determined based on the energy, a dynamic selection threshold value is generated, and the high-frequency sub-bands with the energy exceeding the threshold value are effective texture sub-bands; for the effective texture sub-band, calculating information entropy, determining the size of a sub-block, dividing the sub-block, and reserving the sub-block of which the standard deviation is greater than the mean value of the standard deviation of the sub-block in the sub-band; the effective texture sub-band reserved sub-block coefficients are scanned and spliced into a one-dimensional feature vector according to a Hilbert curve, sliding windows with odd lengths are used for traversing, a bit 1 is generated when window center coefficients are larger than a left neighborhood coefficient mean value and a right neighborhood coefficient mean value at the same time, otherwise, a bit 0 is generated, and an anti-fake image texture feature hash sequence is generated through splicing.

The present application provides a gas detection method and a gas detection device, the method comprising acquiring gas spectral data, wherein the gas spectral data is obtained by detecting a mixed gas in a gas pool using a quantum cascade laser, and the mixed gas includes a gas to be detected; obtaining a plurality of local index information based on global index information of the gas spectral data, the global index information representing coordinate curves of a plurality of signals in the gas spectral data in different coordinate system quadrants; performing quadrant flipping and rotation processing on a coordinate system quadrant of at least one local index information based on a Hilbert space mapping method to obtain an initial global coordinate; using the initial global coordinate as new global index information to calculate a new initial global coordinate, and determining the new initial global coordinate corresponding to a target order of the Hilbert curve as target spectral data; inputting the target spectral data into a gas concentration detection model, and outputting gas concentration information of the gas to be detected.

The invention provides a three-dimensional geological block bidirectional mapping method based on a W-Hilbert curve, and relates to the field of geological data coding, and the method comprises the steps: obtaining a to-be-coded three-dimensional geological block, and carrying out the W-Hilbert coding, and obtaining a to-be-converted code; constructing a positive mapping algorithm; mapping the to-be-converted code to a storage position through a positive mapping algorithm; constructing an inverse mapping algorithm; and the storage position is converted into the W-Hilbert code through an inverse mapping algorithm, so that seamless conversion between the W-Hilbert code and the storage position is realized. And the conversion between the W-Hilbert coding and the storage index is realized by respectively calculating the coding position relationship at each level.

A file is received that includes locus information. The received file is preprocessed, including by selecting a size for a Hilbert curve. File contents are transformed using the Hilbert curve. The transformed file contents are down-sampled. A hashing operation is performed on the down-sampled result.

The invention provides a flow channel structure and a liquid-cooling cold plate, the flow channel structure comprises a fractal curve flow channel, the fractal curve flow channel is a Hilbert curve flow channel or a Petronol curve flow channel, and the fractal curve flow channel is used for increasing the heat dissipation area and strengthening the convection heat exchange efficiency in a limited plane space; wherein the Hilbert curve flow channel is constructed on the basis of a Hilbert curve, and the Peano curve flow channel is constructed on the basis of a Peano curve. According to the flow channel structure and the liquid-cooling cold plate provided by the embodiment of the invention, the heat dissipation performance is improved.

This invention discloses a data completion method based on spatiotemporal privacy protection, which relates to the field of data completion technology. The method comprises the following steps: obtaining spatiotemporal data to be completed; obfuscating the spatiotemporal data to be completed; mapping the obfuscated spatiotemporal data onto a three-dimensional Hilbert curve, assigning a unique Hilbert value to each position, and constructing a location privacy set; generating a confusion probability matrix by permuting and flipping the location privacy set; and inputting the confusion probability matrix, the obfuscated spatiotemporal data, and the spatiotemporal data to be completed into a completion model to obtain the completed spatiotemporal data. The completion model includes a diffusion process and an inverse process. The present invention improves completion accuracy in privacy protection scenarios.