23results about How to "Improve matching speed" patented technology

The invention discloses a target object pose estimation method, system and device and a medium, and relates to the technical field of pose estimation, and the method comprises the steps of offline data set establishment and online real-time multi-target pose estimation. In the off-line process, continuous RGBD streams surrounding a single object are collected, the relative poses of front and back frames are calculated frame by frame, and meanwhile, the three-dimensional surface of a target object is reconstructed to obtain a key frame off-line data set; in the online process, a visible target set in a data stream is firstly detected, VOS segmentation is performed on each target to obtain real-time effective texture and point cloud, offline key frame images of to-be-detected targets are arranged to realize single matching of all feature points and find out a closest reference frame at the same time, and RANSAC is adopted to calculate relative poses of a current frame target and the reference frame; according to the method, an expensive data labeling process is avoided, rapid integrated deployment of detection of an object without a new target is realized, the matching speed is improved, and meanwhile, online real-time multi-target pose estimation can be realized.

The invention relates to the technical field of smart home, and particularly provides a method for obtaining a clothes display image based on image recognition, an application, a control device and a medium. The method for obtaining the clothes display image based on image recognition comprises the following steps: establishing a clothes database; wherein the clothes database comprises a plurality of clothes images; acquiring a user's shot image of the to-be-displayed clothes; and determining a display image according to similarity analysis of the shot image and the clothes image. The invention further provides application of the method in clothes management. According to the invention, the quality of the display image for clothes management is improved, and the aesthetic degree is increased.

The invention discloses an attitude immune pedestrian double-magnetic speed measurement method and device, and belongs to the technical field of indoor pedestrian positioning. According to the method, a static attitude projection method is adopted, and the problems of pedestrian attitude change, magnetometer zero offset and magnetometer installation inconsistency are solved; a double-window weighted dynamic time warping algorithm is adopted, the capacity of capturing the change trend of the magnetic field is improved, the capacity of dynamically adjusting the magnetic field intensity and the contribution degree of the change trend in a cost function is improved, meanwhile, the influence of transient disappearance of the magnetic field gradient on a matching result is reduced, and the matching speed and the matching precision are improved; therefore, the pedestrian movement speed can be accurately estimated.

The invention discloses an SA-CNN-BiLSTM-based intelligent substation virtual loop automatic generation method, which comprises the following steps: in an SCD file, extracting IED equipment information of historical IED equipment and virtual terminal information contained in the historical IED equipment, and storing the IED equipment information and the virtual terminal information in a historical SCD database; forming a semantic similarity model; and based on the semantic similarity model, finding a final historical IED equipment association pair corresponding to the to-be-designed IED equipment association pair, and instantiating the to-be-designed IED equipment association pair based on the final historical IED equipment association pair. According to the method, the virtual terminator information contained in the historical SCD file is rapidly extracted in a DOM mode and stored in the database in a classified mode, and the query and matching speed is remarkably increased. The semantic similarity model can effectively extract semantic features and endow the semantic features with higher weights, the semantic recognition success rate is high, and the method has a good effect in intelligent substation virtual connection automatic design.

The present invention belongs to the field of project data management, and discloses a cloud-native supply chain service platform, including a demand order management module and an order matching module; the demand order management module is used to obtain the customer's demand order; the order matching module includes a database unit, a pre-matching unit, a display unit and a confirmation unit; the database unit is used to store the supplier's supply information and a set A of keywords of the supply information; the pre-matching unit is used to obtain a set B of keywords in the demand order, and calculate the matching value between the set B and the set A corresponding to each supplier, and transmit the list of suppliers with matching values ​​greater than a set threshold and the supply information to the display unit; the display unit is used to display the list of suppliers and the supply information sent by the pre-matching unit; the confirmation unit is used for the staff to select the final confirmed supplier from the list of suppliers. The present invention effectively improves the matching speed of demand orders and improves the distribution speed of the supply chain.

The invention provides a shape-based template matching method, and belongs to the technical field of machine vision, and the method comprises the steps: obtaining a plurality of template images with different scales and angles; respectively calculating the gradient, the feature vector and the feature point at each pixel point on each template image; for a test image, calculating feature vectors and feature points of the test image, performing gradient direction expansion on each feature point to obtain a gradient direction quantized value, and enabling the gradient direction quantized value to correspond to a unique binary character string; a plurality of lookup tables are created, the index number of each lookup table corresponds to each binary character string, and the index value corresponds to the cosine value of the feature point and the feature vector corresponding to the binary character string; extracting template images and feature information of the template images, performing sliding window matching on the template images on the test images in the horizontal and vertical directions, acquiring overall similarity, and judging whether the test images are successfully matched with the corresponding template images or not; according to the invention, the matching speed and accuracy can be effectively improved.

The application discloses an automatic matching embroidery thread color embroidery method for embroidery plate making, and belongs to the technical field of embroidery embroidery, and has the technical scheme as follows: S1, a numbering database is established, corresponding embroidery thread numbers are set according to the RGB values of embroidery thread colors, the numbering database is adapted to plate making software, and the numbering database is stored on a tablet computer of an embroidery machine; S2, plate making software generates embroidery plate making with embroidery thread numbers through the numbering database, and the embroidery plate making is input to the tablet computer of the embroidery machine; and S3, the embroidery plate making is read on the tablet computer of the embroidery machine, the embroidery thread numbers of the embroidery threads placed on a thread drum placing rack of the embroidery machine are identified, the embroidery threads with colors corresponding to the embroidery thread numbers in the embroidery plate making are selected, and the embroidery threads are controlled to perform embroidery. The application can improve the matching accuracy of the embroidery threads in embroidery machine embroidery processing, and ensure the matching speed of the embroidery threads.

The application provides a power grid frequency modulation method and system, the method comprising: a power grid node generates frequency modulation demand information and synchronizes the frequency modulation demand information to a block chain; each EV aggregator node generates first frequency modulation resource information corresponding to the node, and if first target demand information is matched in each frequency modulation demand information according to the first frequency modulation resource information, first frequency modulation supply and demand information is generated and synchronized to the block chain; the power grid node performs device line safety checking on the first frequency modulation supply and demand information according to pre-established power grid device line information, and when the checking is passed, a first smart contract signed between the power grid node and the EV aggregator node corresponding to the first frequency modulation supply and demand information is generated, and the first smart contract is synchronized to the block chain; each EV aggregator node pushes charging and discharging task information to each electric vehicle corresponding to the node according to the first smart contract signed by the node, so that each electric vehicle accesses the power grid.

The present application provides a method for processing a sludge transfer form, which belongs to the field of data processing technology. The method includes: obtaining multiple initial images, including multiple sludge transfer form images; determining multiple target images based on an initial image, wherein the target image is a partial image in the initial image, wherein each target image includes at least one of a signature image, a date image, and a seal image; performing text recognition on the target image, and determining multiple feature data based on the recognition result, wherein the feature data includes signature data, date data, and seal data; classifying and marking multiple initial images based on the multiple feature data, and classifying the sludge transfer form images corresponding to the initial images according to the classification marks. The present application also provides a sludge transfer form processing system and an electronic device.

The application provides a multi-image array splicing method for large cement prefabricated components, and the splicing method comprises the following steps: step one: collecting original images of the cement prefabricated components, dividing the original images into regions, and adjusting the gray scale of each region to obtain preprocessed images; step two: searching for feature points between every two adjacent images in each row in the images, identifying and extracting key feature points in the images through a SIFT algorithm, and then matching the key feature points of the adjacent images through a KNN algorithm to determine the relative position relationship between the key feature points; step three: searching for feature points between every two adjacent images in each column and matching the feature points; and step four: calculating the boundary positions of each image in the splicing area, and then splicing. The application realizes high-precision splicing of the cement prefabricated components, can effectively improve the shape detection precision of subsequent cement components, and improves the product detection quality.

The invention relates to a real-time revision method and system for an interlocking relay drive and acquisition simulation rule, and the method comprises the following steps: S1, loading a BOOL operation rule definition file and a relay drive and acquisition simulation rule configuration file into an operation display subsystem, respectively determining an available relay driving code bit set and an available relay acquisition code bit set according to the BOOL operation rule definition file, and determining all initial driving code bit mapping items according to the relay driving simulation rule configuration file, the available relay driving code bit set and the available relay acquisition code bit set; s2, the operation display subsystem starts a BOOL operation simulation module and a relay simulation module, and the BOOL operation simulation module is operated based on a BOOL operation rule definition file; and the relay simulation module operates based on the drive and acquisition code bit mapping item, and provides a GUI interface for viewing and revising the current drive and acquisition code bit mapping item. According to the method, the defect of high coupling of an existing method is overcome, and the operation efficiency of the relay simulation module is improved.

The invention provides a medical safety adverse event early warning method and computer equipment, and relates to the technical field of medical information, and the method comprises the steps: obtaining early warning rules of a plurality of preset medical safety adverse events under each clinical medical business system of a target medical institution; matching a plurality of clinical medical record files of each clinical medical service system in a preset historical time period by adopting the pre-warning rule of each preset medical safety adverse event to obtain a target clinical medical record file matched with the target pre-warning rule of the target medical safety adverse event; and according to the target early warning rule and the target clinical medical record file, generating an early warning message of the target medical safety adverse event. According to the method and the device, the early warning message of the medical adverse safety event is automatically generated, and the department inpatient area with missing report is effectively supervised, so that the medical quality is improved.

The present invention discloses a multi-level flow table matching method and device based on Bloom filter. The method comprises the following steps: obtaining a data frame of a computer network data flow to be matched, extracting key fields from the data frame to obtain the value of the key field; inputting the values ​​of matching keywords corresponding to each level of the flow table into a preset counting Bloom filter in order of priority from high to low for detection, until the detection is passed; matching the values ​​of matching keywords corresponding to the flow table that passes the detection with a preset multi-bucket hash table based on DDR, and outputting the matching result; the device comprises a frame processing module, a serial matching module, a counting Bloom filter and a DDR memory; the present invention adopts a DDR memory to store the multi-bucket hash table, thereby increasing the flow table scale supported by the matching circuit; adopting a Bloom filter when matching flow table items, thereby reducing the number of accesses to the off-chip DDR storage and improving the matching speed of the data frame.

The invention belongs to the technical field of scene matching, and particularly relates to a rapid scene matching method for a large-range search area, which greatly improves the scene matching speed during large-range search and solves the problem of high time consumption in initial positioning of current scene matching. According to the method, a four-dimensional descriptor is firstly designed, so that the robustness of a back-end matching strategy can be greatly improved; furthermore, on the basis of the multi-dimensional reference descriptor and the multi-dimensional real-time descriptor, a convolution coarse matching and fine matching strategy is designed, firstly, the multi-dimensional reference descriptor is segmented to perform software acceleration, coarse matching is performed on the obtained nine regions by using convolution, and fine matching is performed on the obtained nine regions; cutting an image with the same size as the real-time image on the original reference image according to the matching position for accurate matching, and finally designing a matching judgment rule to judge a matching result. According to the method, a mode of combining segmentation rough matching and fine matching is adopted for a large-range search region, so that the matching speed of the large-range region is greatly improved.

The present application discloses a biometric identification method and device, which belongs to the field of computer technology. The method includes: after acquiring a biological image under TEE, extracting biometric data in the biological image under the TEE, and then matching all or part of the biometric data with a biometric template in SE to obtain a biometric identification result. In the present application, since the security level of SE is higher than that of TEE, the biometric identification function reaches a high security level. In addition, since the calculation complexity of the template matching of biometric data is not high and the processing pressure of SE is relatively small, the template matching speed can be guaranteed to be high, thereby ensuring a high biometric identification speed. In this way, the present application enables biometric identification to be protected at a high security level while only slightly increasing the latency and without affecting the biometric identification performance, thereby improving the security of biometric identification.

The invention relates to the technical field of artificial intelligence big data processing, in particular to a user intelligent matching method and system based on a data collision algorithm. The establishment of the user tag is the basis of intelligent matching, the basic features of the feature information are correspondingly associated with the user tag, the behavior features are specific behaviors of the user, the achievement features are results generated by the specific behaviors, and the behavior features and the achievement features are associated with the user tag through the basic features. By using a data collision algorithm, matching points can be quickly found in large-scale and multi-dimensional data. Compared with a traditional matching algorithm, the data collision algorithm can process a complex data set more efficiently, and the matching precision and efficiency are remarkably improved.

The invention discloses a natural gas supply and demand matching method based on a smart contract, and particularly relates to the technical field of block chains, and the method specifically comprises the steps: building a prediction model through a machine learning algorithm, carrying out the optimization of the smart contract according to the prediction model, transmitting external data to the smart contract through an integrated oracle machine, and carrying out the optimization of the smart contract. Performing primary evaluation according to the external data to verify an applicability condition matched with the transaction, and performing secondary evaluation on the smart contract to verify the real-time performance of the smart contract when the applicability condition does not meet a preset applicability limit; when the real-time performance of the intelligent contract is insufficient, a state channel is introduced to carry out under-chain transaction, a transaction result is transferred to the chain to carry out verification settlement through the intelligent contract, natural gas supply and demand matching is carried out according to the introduced state channel, and payment settlement is carried out on the transaction through the chain after matching is completed, so that the flexibility of the intelligent contract is improved; and the transaction security is improved while the transaction cost is reduced.

The invention discloses a method for constructing a three-dimensional model on the surface of a hidden area. The method comprises the following steps: performing low-altitude oblique photography on a hidden area at a low-altitude position to obtain a low-altitude photography data set carrying first POS data; hidden area textures which cannot be obtained by low-altitude unmanned aerial vehicle oblique photography are collected at the ground position in a ground supplementary photography mode, and a ground photography data set carrying second POS data is obtained; matching RTK positioning coordinates, namely second pos data, corresponding to each photo through the shooting time of terrestrial photography; and based on the terrestrial photography data set and the low-altitude photography data set, combined aerial triangulation adjustment is carried out under the assistance of POS data, and a hidden area surface model is constructed. The invention aims to provide the method for constructing the hidden area surface three-dimensional model, the full-scene dead-corner-free surface three-dimensional model can be duplicated, digital twinning in the real world can be truly realized, the modeling quality and the mapping efficiency can be greatly improved for the surveying and mapping industry, and the investment cost of equipment personnel is reduced.

The application provides a kind of based on DTW's multivariate time sequence feature prediction method and system, comprising: satellite telemetry data is segmented and handled, and original time sequence is obtained;Respectively extract features, combine to generate multivariate feature vector as initial sample dataset for preprocessing;Time scale alignment equal interval interpolation processing is carried out to the processed feature sequence, and sample time sequence feature set is generated;The initial sample dataset and corresponding classification label set are used as the input of LSTM network, and model training is carried out;According to the feature sequence of sample data and corresponding classification label, the corresponding type feature template is generated;Input trained LSTM model, and obtain classification label;The similarity of each segment feature template and mobile feature sequence is calculated in turn;The biggest template sequence of similarity is selected to obtain the predicted value of specified time and the final predicted value is output by inverse normalization processing.The application can simultaneously parallelly calculate the similarity value between each segment sequence, thereby significantly improving the matching speed.

The application discloses a virtual power plant user sensitive data identification method and device based on graph learning, and the method comprises the following steps: obtaining to-be-identified user data; generating a to-be-identified graph network according to the relationship of keywords in the to-be-identified user data; inputting the to-be-identified graph network into a pre-trained graph convolutional neural network model to obtain a minimum connected dominating set corresponding to the to-be-identified graph network; generating a simplified graph of the to-be-identified graph network based on the minimum connected dominating set corresponding to the to-be-identified graph network; and performing matching degree calculation on the simplified graph of the to-be-identified graph network and a simplified graph corresponding to standard sensitive level data to obtain a sensitive identification result of the to-be-identified user data. Through the implementation of the application, the minimum connected dominating set is generated by using the graph learning method, the complexity of subsequent graph matching is reduced, the matching speed is improved, and the application has higher efficiency compared with a traditional sensitive identification method.

The application relates to the field of FPGA debugging technology and discloses an FPGA signal observation method, system, device, equipment and storage medium, wherein the method comprises the following steps: setting a unique identifier for a to-be-observed FPGA function module where a target observation signal is located, and setting the same identifier in an FPGA top-level module; constructing a signal transmission link between the function module and the top-level module based on the identifier, so that the target observation signal is transmitted to the top-level module; and transmitting the signal to a debugging interface through the top-level module to complete online observation. The method solves the problems that the signals need to be repeatedly and manually connected and the observation target cannot be dynamically switched in the prior art through identifier matching and link construction, and the debugging efficiency is significantly improved.

The invention discloses a stripe splicing robust processing method and system for cluster multiple beams, and the method comprises the steps: carrying out the curvature and perpendicularity calculation of the multi-beam stripe data of an unmanned ship, and constructing a homonymy feature point set; performing weight estimation on the homonymous feature point set through a Huber function to obtain a homonymous feature point set weight value; constructing a homonymy feature point set seven-parameter transformation model, and performing iterative solution through a nonlinear least square adjustment method to obtain a solved homonymy feature point set seven-parameter transformation model; and splicing the multi-beam stripe data of the unmanned ship based on the solved homonymy feature point set seven-parameter transformation model. According to the method, the strong matching relationship between homonymy points can be fully utilized, and the weak matching relationship is limited, so that the multi-beam strip splicing precision is improved. The stripe splicing robust processing method and system oriented to cluster multiple beams can be widely applied to the technical field of multi-beam stripe data splicing.

This invention discloses a robot relocalization method, apparatus, and robot. In the method, laser point cloud data and image data acquired by the robot at its current position are received; keyframe information corresponding to the image data is determined according to a visual dictionary; physical location information corresponding to the keyframe information is obtained from a visual map; feature point information of the image data is matched with feature point information corresponding to the keyframe information; and calculation is performed by combining the physical location information of the feature points in the visual map and the pixel location information of the feature points in the image data. After obtaining the robot's first pose data, a predetermined area is set based on the first pose data, and iterative matching calculations are performed within the predetermined area using the laser map and laser point cloud data to obtain the robot's second pose data. By using the above scheme, the iteration range of laser matching can be reduced by utilizing the initial pose data, thus accelerating the matching speed and quickly achieving robot relocalization.