53results about How to "Guaranteed data volume" patented technology

The embodiment of the invention provides a man-machine interaction data processing method and device. The man-machine interaction data processing method comprises the steps of receiving data collection information sent by a user terminal, wherein the data collection information comprises identification information of the user terminal, sensor data and data collection time information; obtaining application service content information corresponding to the identification information and the data collection time information, and extracting user activity behavior characteristics from the application service content information; carrying out label processing on the sensor data according to the user activity behavior characteristics of the user terminal. The sensor data of all users can be collected through the method, active cooperation of the users is not needed, a large sensor data volume is guaranteed, and therefore subsequent processing is simpler and more accurate.

The invention relates to a stream media data request transmission method for multimedia telephone peer-to-peer (P2P) video-on-demand. A combination of one or more of a node marking mechanism, a safety zone compensation strategy and a data request scheduling strategy is adopted, wherein the node marking mechanism marks corresponding nodes according to a response of P2P data nodes to a data request and acquires data from a high-quality P2P data node; the safety zone compensation strategy pre-detects a data insufficient situation in a buffer zone and supplies sufficient buffer data to data reduction playback by quickly downloading deficient data blocks to a content delivery network (CDN) server; and the data request scheduling strategy schedules a data transmission rate and a data block requesting sequence among data transmission channels established between a user node and the various P2P data nodes so as to acquire a minimum data transmission delay. By adopting the stream media data request transmission method, even a stream media data quantity is larger, continuous and smooth playback of stream media can be realized; and the stream media data request transmission method is economic and convenient and has a wide application prospect.

The embodiment of the invention provides a method and device for processing data. The method comprises the steps that when dirty data are detected, the first bitmap of a first storage block which is built in advance and used for indexing at least two first reserved capacities is inquired, and the first storage block is determined by dividing data storage space; whether the corresponding position of the first storage block which the dirty data belong to in the first bitmap is provided with dirty data identification or not is judged; if not, the corresponding position of the first storage block which the dirty data belong to in the first bitmap is provided with the dirty data identification, the first storage block is divided into the second storage blocks of at least two second reserved capacities, and the second bitmap used for indexing the second storage blocks is built; if yes, the second bitmap built for the first storage block which the dirty data belong to is determined, and the corresponding positions of the second storage blocks which the dirty data belong to in the second bitmap are provided with the dirty data identification. Through the method and device for processing the data, resources occupied by dirty data markers are saved.

The invention belongs to the technical field of bridge detection, and discloses a bridge load testing method. The bridge load testing method comprises the following steps that 1, a dynamic weighing system is arranged at the bridge head of a bridge, and the measuring content comprises the axle distances and the axle weights of passing vehicles; 2, a plurality of shooting devices are arranged on the bridge for full-bridge-face covering shooting, and position distribution of the vehicles on the bridge face is measured; 3, sensors are arranged on the bridge, and static load parameters and dynamic parameters of the bridge structure are measured; 4, the data acquisition time of the full-bridge-face shooting devices in the step 2 and the data acquisition time of the sensors in the step 3 are synchronized; 5, the load bearing capacity of the bridge is evaluated, wherein the step 1 to the step 5 are conducted under the normal traffic condition of the bridge. According to the method, the test cost is reduced; due to the fact that traffic does not need to be interrupted or blocked, continuous multi-time testing can be conducted, the real-time performance, the continuity and the data volume of testing are guaranteed, and then the load test data integrality and the analysis accuracy are improved.

The invention provides a data cleaning method and device and a server, and relates to the technical field of wind power plant data processing, and the method comprises the steps: obtaining an actual measurement data sample; establishing joint probability distribution of wind speed and power in the Copula space; uniformly dividing the value range of the joint probability distribution into a plurality of probability intervals; and performing data cleaning on the actually measured data sample in each probability interval along the dimensions of the wind speed and the power. According to the datacleaning method and device and the server provided by the invention, the actually measured sample data can be cleaned based on the probability space to obtain a finer data cleaning result, meanwhile,the effectiveness and the data size of the remaining data can be ensured, and a solid foundation is laid for subsequent data analysis.

The invention provides a high-throughput sequencing-based animal-derived ingredient identification method and application thereof. The method comprises the following steps: taking a deoxyribonucleic acid (DNA) sequence, obtained by high-throughput sequencing, as an analysis object, taking a public DNA sequence database as a reference, and investigating the pairing number of the analysis object and the reference sequence data, so as to reflect the existence or number of specific categories. The method is simple in operation steps and high in accuracy. The detection limit of specific types of animal-derived ingredients reaches 1%, and the species identification types of the method theoretically cover all species. Therefore, when being used for identifying the animal-derived ingredients in a sample, the method can analyze the animal-derived ingredients in the sample with unknown composition, and simplifies the operation process, so that the method has wide application prospect in the aspect of animal-derived ingredient identification.

The invention provides a product information recommendation method and device. The method comprises the steps of receiving customer basic information of a target customer; determining preference scores of the target customer corresponding to different product types by applying the customer basic information and a plurality of preset product preference prediction models corresponding to the different product types, and taking the product type corresponding to the maximum preference score as a target recommended product type; sending product purchase information corresponding to the target recommended product type to the target customer, wherein each preset product preference prediction model is an iterative decision tree model obtained by pre-training according to the historical customer basic information of a plurality of historical customers and the respective preference score label of each historical customer for the same product type. According to the invention, the timeliness and intelligence degree of product recommendation can be improved, and the success rate of product recommendation can be further improved.

The invention discloses an Internet spatial data search system and method. The system comprises a detection module, a grasp module, a cleaning module, a storage module, a search module and middleware,wherein the detection module is used for detecting surviving equipment and host in the Internet, and the grasp module is used for acquiring data information of the surviving equipment and host; the cleaning module is used for cleaning the data information; the storage module is used for saving the cleaned data information to a database; the search module is used for searching the data informationof the equipment or host from the database; the middleware connects the storage module and the search module, and the middleware is used for synchronizing the data information in the storage module into the search module. The provided Internet spatial data search system and method can improve the storage and search reading efficiency of the search system.

The invention discloses a village and town development state measurement method based on a remote sensing noctilucent dark target enhancement technology. The method comprises the following steps: 1, carrying out dark target enhancement processing on monthly-scale remote sensing light data of a measurement area; 2, forming a data cube according to the monthly-scale remote sensing light data subjected to dark target enhancement processing in a first area in the measurement area in combination with monthly-scale surface temperature data and monthly-scale vegetation index data corresponding to the first area; 3, training and verifying the constructed long-short-term memory network, and when it is determined that the long-short-term memory network converges, determining the town boundary of the measurement area according to the convergent long-short-term memory network and the data cube; and 4, according to the boundaries of the villages and the towns, based on a spatial pattern and a trend analysis method, performing measurement analysis on the development states of the villages and the towns. According to the technical scheme of the invention, the method effectively improves the recognition capability of the first-level dark targets of villages and towns, and guarantees that the enhancement of the dark targets is not affected by obvious noise.

The invention relates to a wireless multifunctional data collector for energy-saving monitoring, and the collector comprises a signal collection module which is used for receiving various collected signals, the signal collection module comprising a one-way collection unit and a two-way collection processing unit, the one-way collection unit being used for collecting current, voltage, active power and power factors, two-way collection processing unit comprising a serial port level conversion processing unit, an Ethernet transmit-receive management unit and a hand-held terminal acquisition unit; acontrol module, used for processing the signal received by the signal acquisition module; a transmission module, used for transmitting the signal processed by the control module; and an upper computer, used for receiving the transmitted signal and adjusting the acquisition frequency of the bidirectional acquisition processing unit in the signal acquisition module according to the data volume of the transmitted signal. By effectively adjusting the data volume, the matching between the data processing speed and the accuracy of the evaluation result is realized, so that the energy-saving monitoring process is more efficient actually.

The invention discloses an online car-hailing metering method and a computer readable storage medium, and the method comprises the steps: notifying a driver terminal and a corresponding passenger terminal to start metering after a central server receives a start instruction transmitted by the driver terminal; receiving a first set uploaded by the driver terminal and a second set uploaded by the passenger terminal; respectively replacing the position points in the first set and the second set the positioning mark is no with corresponding path points in the road data; after arriving at the destination, dividing the position points in the first set and the second set where the positioning mark is yes into credible points and uncredible points, and replacing the uncredible points with the corresponding path points in the road data; obtaining a third set according to the replaced first set and second set; and performing calculating to obtain the total mileage according to the third set. According to the invention, the accuracy of online taxi-hailing distance measurement can be improved, so that the accuracy of taxi-hailing cost is improved.

The invention relates to an infrared small target image labeling method and a system. The method comprises the following steps: converting an original image in video data acquired by infrared acquisition equipment into an 8-bit single-channel image; splicing the three frames of single-channel images with similar time into a three-channel image; marking each three-channel image by using a marking frame; and determining the centroid of the mark box according to the gray value of each point in the mark box, and readjusting the size and position of the mark box through the centroid. Three frames of single-channel images with similar time are spliced into a three-channel image, and time sequence information can be reserved in the single image; then, a target in the three-channel image is marked; and finally, the mass center of the mark box is calculated. The size and the position of the mark box are adjusted according to the mass center, the target and the background within the set range around the target are contained in the mark box, time domain information and richer space domain information are fused into image marks according to a small target neural network model, and the accuracyand the practical value of the model in infrared investigation monitoring equipment are improved.

The invention provides an enterprise operation risk monitoring method and a device, and the method comprises the steps: generating a plurality of beneficiary network maps taking entities as nodes according to a pre-obtained enterprise operation physical map, enabling an edge between two nodes to represent the association relation between the entities, wherein the nodes comprise a unique candidateenterprise node and a plurality of beneficiary nodes; obtaining a preset enterprise operation virtual graph, judging whether an edge in the preset enterprise operation virtual graph is in the beneficiary network graph or not, and if not and two nodes corresponding to the edge are in the beneficiary network graph, updating the beneficiary network graph based on the edge; and monitoring whether theupdated beneficiary network map has a closed-loop link containing candidate enterprise nodes, and if so, determining that the candidate enterprises corresponding to the candidate enterprise nodes haveoperation risks. According to the invention, the efficiency and the accuracy of obtaining the enterprise operation related data and the association relationship between the data can be improved, andthe accuracy, the efficiency and the automation degree of enterprise operation risk monitoring can be further improved.

The invention discloses a high-precision map crowdsourcing system based on an optimal time-stop rule and a data collection method of the system. The system comprises a task issuing module, a member optimization module and an optimal time stop module. The task issuing module is used for issuing a high-precision map crowdsourcing task; the member optimization module is used for selecting superior participants from the candidate participant set, receiving data of the superior participants and providing rewards for the superior participants; and the optimal time stop module controls the stop opportunity of the crowdsourcing task of the high-precision map to maximize the crowdsourcing utility. Based on the system, the invention provides the data collection method of the high-precision map crowdsourcing system based on the optimal time-stop rule. According to the method, sufficient data sources are provided for updating of the high-precision map, the cost of data collection is reduced, and updating of the high-precision map is more real-time and efficient.

The invention discloses a rear axle system nonlinear strength analysis method. The method comprises the steps that 1, a rear axle material curve is processed; 2, an error judgment value e is set; 3, an initial point counter i=1 and an ending point counter j=3 are generated; 4, initial points and ending points are connected to generate judgment lines; 5, errors e between original data, between the initial points and the ending points, and the judgment lines are calculated; 6, the calculated errors are judged, if all the errors are smaller than e, the ending point counter j is equal to j+1, an array storing the errors is reset, and the step 4 is returned to, and if at least one of the errors is greater than e, the initial point counter i is equal to j-1, the ending point counter j is equal to j+1, the array storing the errors is reset, and the step 4 is returned to; and 7, the initial point of the judgment line generated each time and the last point of material curve data processed in the step 1 are used as reserved points, other points are deleted, and optimized material data is obtained. Through the method, a material stress-strain curve and a bushing stiffness curve are reasonably shrunk, and therefore calculation efficiency is improved.

The invention discloses an abnormal signal perception based abnormal step size acquisition method for vehicle electrical performance data. According to vehicle electrical performance data acquisitionrequirements, acquisition point pretreatment and an acquisition time prediction mechanism are added, and vehicle electrical performance data acquisition is converted into abnormal signal perception based abnormal step size acquisition; the pretreatment mechanism of acquisition points is defined, signal deviation estimation [rho]c and abnormal estimation which is shown in the description are introduced, and an abnormal signal in the acquisition points is corrected through the sampling of the acquisition points; and a signal margin which is shown in the description is defined, a signal bias coefficient a, zero drift b and a signal skipping coefficient which is shown in the description are introduced, and a real-time acquisition time prediction mechanism is designed. Through the pretreatmenton data, the acquisition time can be predicted to prepare the next acquisition, so that the vehicle electrical performance data abnormal step size acquisition can be realized; and the method can correct abnormal points during acquisition and dynamically change sampling step sizes, so that situation of high hardware requirements for high frequency, fixed cycle and longtime acquisition can be avoided, equipment costs can be reduced, and the efficiency and reliability of vehicle electrical performance data acquisition can be enhanced.

The method discloses a method for extracting an elevation of a height measuring foot point through CryoSat-2 SARIn mode L1b level waveform data. The method comprises the following steps of 1) extracting echo coherence and differential phase information from the CryoSat-2 SARIn mode L1b level data; 2) carrying out smoothing processing on echo coherence and differential phase data; 3) carrying out threshold segmentation according to echo coherence, and taking an area where the coherence is greater than a threshold as an alternative area; 4) merging the alternative areas to obtain an effective differential phase to be unwrapped; 5) carrying out relative phase unwrapping in a time delay direction; 6) calculating an three-dimensional coordinate of an elevation point through a differential phase;7) converting a reference coordinate system; and 8) carrying out absolute phase unwrapping based on an external DEM. In the invention, a high-precision height measuring foot point elevation can be obtained from L1b level data; compared with L2 level data released by ESA, there is a larger foot point coverage density, and the digital elevation model with a higher resolution can be acquired to obtain a finer terrain.

The invention relates to the field of business security of computer security, and discloses a risk model training method and device, computer equipment and a readable storage medium, and the method comprises the steps: obtaining long-term historical data through a long-term extraction rule, and obtaining recent historical data through a recent extraction rule; summarizing the long-term historical data and the recent historical data to obtain comprehensive training data; adjusting data volumes of long-term historical data and short-term historical data in the comprehensive training data, and adjusting a positive-negative proportion between a sum of a long-term positive sample and a short-term positive sample in the comprehensive training data and a sum of a long-term negative sample and a short-term negative sample in the comprehensive training data to obtain comprehensive adjustment data; and training the initial intelligent model through comprehensive adjustment data to obtain a mature model. The invention further relates to a block chain technology, and information can be stored in a block chain node. According to the invention, the risk identification efficiency is high, the accuracy is high, and the problems that experience and rules become vulnerabilities and normal users are identified as risk users are avoided.

The invention discloses a popup window processing method and device, computer equipment and a storage medium. The pop-up window processing method comprises the steps of monitoring a calling event of afunction module triggered by an online user, and obtaining triggering data of the function module; when the function module is a new function module, judging whether the trigger data meets a triggercondition of an initial popup event or not; popping up an evaluation popup window corresponding to the function module when the trigger data meets the trigger condition of the initial popup window event; when the triggering condition of the initial popup window event is not met, judging whether the triggering data meets the triggering condition of the popup window probability event or not; and when the triggering condition of the popup probability event is met, popping up the evaluation popup corresponding to the function module. The invention also relates to a blockchain technology, and the evaluation data can be stored in the blockchain. The method does not affect the use of the user on the basis of ensuring the data volume of the evaluation data.

The embodiment of the invention provides a model training method and device, computer equipment and a readable storage medium, and the method comprises the steps: obtaining an original sample set, training a sample recognition model according to the original sample set, and obtaining an original recognition model, wherein the original sample set comprises a plurality of original samples; calling the original identification model to identify each original sample to obtain an original identification result of each original sample; dividing the original sample set into a target sample set and a first activation sample set according to the original identification result of each original sample; calling an activation model to activate each target sample to obtain a second activation sample set;and training a recognition processing model according to the first activation sample set and the second activation sample set to obtain a target recognition model, so that the translation accuracy ofthe trained translation model is effectively improved while the data volume of the trained translation model can be ensured.

The invention discloses an image processing method and device, equipment and a medium, and the method comprises the steps: obtaining sample data, calling an image processing model to convert an annotated real face image Ai into a reconstructed cartoon face image B'i, and converting a non-annotated real face image Ci into a reconstructed cartoon face image B''i; inversely converting the reconstructed cartoon face image B''i into a reconstructed real face image C''i; inversely converting the marked cartoon face image Bi into a reconstructed real face image C'i; obtaining a first loss according to the difference between B'i and Bi; obtaining a first loss according to B'i, B''i and Bi, obtaining a second loss according to B'i, B''i and Bi, obtaining a third loss according to Ci and C'i, obtaining a fourth loss according to the difference between Ci and C''i; and training an image processing model based on the first loss, the second loss, the third loss and the fourth loss. By means of the method, the model training effect can be improved, and the universality and stability of the trained image processing model are improved.

The invention discloses a data processing method, computer equipment and a storage medium. The method comprises the steps of when a first picture frame of an application client is obtained through coding, enabling an application server to detect the data volume of the first picture frame; if the data volume of the first picture frame is greater than or equal to a data volume threshold value, performing picture reduction adjustment on the first picture frame to obtain a first adjusted picture frame; encoding the first adjusted picture frame to obtain a picture frame code stream of the first adjusted picture frame; sending the picture frame code stream to an application client, so that the application client displays a second picture frame according to the picture frame code stream; and the second picture frame being obtained after the application client decodes the picture frame code stream to obtain the first adjusted picture frame and adjusts the resolution of the first adjusted picture frame back to the first resolution. By adopting the method, the code stream transmission speed for the first picture frame can be improved, and the picture frame display fluency is further improved.

An embodiment of the invention provides a flight equipment communication method and equipment, a server and a storage medium. The method comprises the steps that: target flight equipment sends an access request message to a plurality of adjacent nodes, and receives channel state information returned by the plurality of adjacent nodes based on the access request message, wherein the adjacent nodes comprise other flight equipment adjacent to the target flight equipment and / or a base station; according to the channel state information, link communication information between the target flight equipment and each adjacent node is determined; according to the link communication information, at least one candidate access node is determined from the plurality of adjacent nodes, and the initial transmitting power and the initial data transmission rate of the target flight equipment corresponding to each candidate access node are respectively determined; and a target access node is determined from the at least one candidate access node based on the initial transmitting power and the initial data transmission rate, and the target flight equipment is controlled to access the target access node. Therefore, the data transmission quality and data volume of the flight equipment are ensured.

The invention provides a pipeline leakage detection method based on secondary screening of acoustic real-time data, and relates to the technical field of pipeline leakage detection. It is implemented using a device including an acoustic sensor module, DSP processing module, ARM control module, A / D conversion module, data storage module, GPRS wireless communication module, GPS timing and positioning module, signal conditioning amplification module and solar battery module. The acoustic signal is processed After collection and conversion, it is an analog quantity, and then converted into a digital signal and sent to the DSP processing module and data storage module. After processing, analysis, and screening, the data packets to be sent are sent to the ARM control module, and the corresponding data packets are sent from the data storage module. Sent to the host computer via GPRS wireless communication module. Based on the real-time acoustic data information, typical signals are screened twice, and pipeline leakage signals are sent wirelessly in the wild environment. The amount of data transmitted is effectively reduced, which greatly saves the storage space of the host computer and improves efficiency.

The invention discloses a method, system and device for improving the quality of a classified learning data set and a storage medium, and belongs to the technical field of image classification. The method comprises the following steps: updating a data set by utilizing a pre-designed updating method; outputting the data set in response to the detection that the proportion of the clean labels in the data set is not increased; in response to the detection that the proportion of the clean labels in the data set is increased, updating the data set by using the pre-designed updating method again, wherein the pre-designed updating method comprises the following steps: acquiring an error transition probability matrix of a label through network output of an anchor point sample; according to the error transition probability matrix of the label, obtaining an error rate and a weight of the label, and according to the error rate and the weight of the label, obtaining a weighted average error rate of the data set; sorting the data samples according to the probability of label labeling errors, screening out error label samples in combination with the weighted average error rate of the data set, correcting the labels of the error label samples by using the error transition probability matrix of the labels, and updating the data set.

The invention belongs to the field of image processing, and discloses a TB-level 3D cranial nerve data reduction method and system based on deep learning, wherein the method includes the following steps: S1: establish a test data set; S2: establish and train a segmentation model; S3: Use the segmentation model to process each test image in the test data set to obtain the nerve fiber distribution information corresponding to each test image; S4: Use morphological operations to optimize the nerve fiber distribution information to make the three-dimensional space The maximum connected domain corresponding to the region where the nerve fibers are located in the nerve fiber contour map below is optimized, and finally the data reduction results of TB-level sparse whole-brain nerve fibers are obtained. The present invention can quickly, accurately and effectively reduce cranial nerve data sets of TB level and above, greatly reduce the amount of data for subsequent neuron reconstruction, and improve reconstruction efficiency.