418results about How to "Improve statistical efficiency" patented technology

The invention relates to a blockchain network, an object transaction method, a device and node equipment, wherein the object transaction method comprises the steps of transmitting a transaction request of a preset object to authorized nodes in a blockchain to which a preset object transaction is related; respectively receiving transaction verification information which is obtained by each authorized node according to the received transaction request; acquiring a verification result according to the transaction verification information; and transmitting the verification result to each authorized node, wherein the verification result is used for confirming or canceling transaction of the preset object by each authorized node. According to the blockchain network, different objects correspond with different blockchains; and in a transaction confirmation process, through co-verification of the nodes in multiple blockchains, the verification result is obtained, thereby ensuring high transaction safety, reducing the length of a single blockchain, reducing data volume of each node, improving searching efficiency of the blockchain, improving gathering efficiency and statistics efficiency of single kind of objects and improving concurrence.

The network element management system includes database server and application server. The performance reporting table is set in the database server for use in saving the reported performance data. The method includes following steps: 1) the table for storing different time grain size data is set in database server, and the grain size presents the data reporting time, and there is inclusion relation between grain sizes; 2) The application server periodically merges the reported performance data according to time grain size, and then saves it into grain size data storing table; 3) when the performance statistics is made for mass performance data in the network element management system, the application server retrieves the data storing table relating to grain size according to the grain size required by statistics.

The invention provides an automobile remote monitoring system and a monitoring method thereof. The system comprises automobile ECUs, a wireless communication network and a monitoring center communicated with the automobile ECUs by the wireless communication network, wherein the wireless communication network comprises an automobile ECU-inside wireless communication module, an antenna, a mobile base station, an SGSN server, a GGSN server, a GPRS / GSM / 3G backbone network and an Internet network; the mobile base station is communicated with the automobile ECUs and communicated with the GPRS / GSM / 3G backbone network by the SGSN server, and the GPRS / GSM / 3G backbone network is communicated with the Internet network by the GGSN server; and the monitoring center comprises a data server and a control terminal, and the control terminal and the server are connected with the Internet network respectively. By the monitoring system, the monitoring center can remotely monitor the running state of a batch of automobiles by a wireless communication technique, and the maintenance and management cost of sold automobiles is greatly lowered.

The invention provides a front-end user behavior statistical method and device. The front-end user behavior statistical method comprises the steps of acquiring point burying rules preset for a page, monitoring a page clicking event for the page in the running process of the page, and performing statistics on behavioral data of a front-end user of the page based on the point burying rules when the page clicking event is monitored. In the invention, the point burying rules are firstly set for the page in order to realize automatic point burying for the page, a buried point is triggered to open by the page clicking event in running of the page, and thus a problem that the existing manual point burying makes errors and is tedious and poor in convenience. In addition, global statistics can be performed on behaviors of the front-end user of the page, and the statistical efficiency is high.

The invention provides a log management method based on a micro-service framework. The method includes the steps of outputting a request ID generated by an interceptor configured in each application and collected calling information of the applications and user information into a log, collecting log information of each application on each application server, caching the log information of all theapplications in a name with a Key of logstash in a unified mode, pulling the log information, conducting conversion from the original data format to the data format required by an ElasticSearch searchserver, capturing keywords from the log information, and storing the corresponding log information in the ElasticSearch search server based on the keywords. Dynamic elastic shrinkage of the whole logprocessing process can be conducted according to actual use conditions, physical resources are effectively used, large-scale log conditions can be responded to, and inquiry and statistics efficiencyis improved. The invention further discloses a log management system based on the micro-service framework.

A method for counting data stream includes presetting charge configuration table, generating counting flow table according to charging data and user IP address , finding counting flow table according to source IP and destination IP of data packet and combining with configuration data in charging configuration table when data packet is arrived , recording relevant flow data of the data packet if counting flow table is inquired .

The invention provides a cinema information acquisition system. The cinema information acquisition system comprises a passenger flow acquisition module used for acquiring cinema audience passenger flow information, an audio acquisition module used for acquiring audio information in a video hall of a cinema, a facial expression acquisition module used for acquiring facial expression information of audience in the video hall of the cinema, a comment acquisition module used for acquiring comment information on a film from the audience in the cinema and an analytical and statistical module which is connected with the passenger flow acquisition module, the audio acquisition module, the facial expression acquisition module and the comment acquisition module and is used for identifying the acquired passenger flow information, audio information, facial expression information and comment information. The cinema information acquisition system is high in statistical efficiency, the passenger flow counting accuracy is greatly improved, and real feeling of the audience can be analyzed more accurately.

The invention relates to a cardinal number estimating method and a cardinal number estimating device under a multi-section query condition of big data. The cardinal number estimating method comprises the following steps of dividing the big data into a plurality of sections in advance according to numerical attributes; establishing a tree-shaped index structure, and utilizing each section as a node of the tree-shaped index structure; acquiring a data source to be written in the tree-shaped index structure, and reversely indexing the data source which supports the section query condition; writing the data source which is reversely indexed into the nodes of the tree-shaped index structure, and respectively writing corresponding portions of the data source into data files and a cardinal number estimator; and querying the nodes meeting the section query condition in the tree-shaped index structure according to the section query condition to obtain the cardinal number estimator in the nodes, and logically processing the cardinal number estimator to obtain cardinal number estimated values. The cardinal number counting efficiency is improved by reducing data calculating precision; the query efficiency is high under the optional multi-section query condition; and the online index data updating efficiency is improved by using a big data incremental updating technology.

The invention discloses a configuration method for a distributed statistical analysis system. The distributed statistical analysis system comprises a ZooKeeper cluster, a service node and a computational node cluster. The method comprises the following steps of electing a leader node in the computational node cluster, fragmenting data in each computational node according to a data fragmentation principle and electing a leader fragment in a copy of the data fragment; after the service node receives a statistical analysis request, applying to the leader node for the computational node and making the leader node feed the computational node with the minimum task load back to the service node; after the service node obtains the feedback computational node, sending the statistical request to the computational node; and making the computational node search the leader fragment, apply to the leader fragment for an idle copy of the data fragment and distributing a statistical task to the copy of the data fragment to execute the statistical task. The invention also provides the distributed statistical analysis system based on the configuration method.

The invention discloses a flow statistic method and device. The method comprises the steps that connection is newly established according to quintuple information of a received message; if the message is matched with a strategy of the pre-configured flow statistic type, an ID value matched with the matched strategy of the flow statistic type is recorded on the established connection; a corresponding statistic element is found on a hash table in a preset flow statistic type management table according to the ID value and by the adoption of the hash algorithm for corresponding flow statistics. According to the flow statistic method and device, when the message passes through the device, first the strategies of various statistic types are matched, the matched ID value will be recorded to the connection, then the corresponding statistic element can be found according to the hash algorithm, a parameter of flow is directly recorded, a subsequent message is matched to the same connection, and the corresponding statistic element can be found for flow statistics. Strategy matching is not needed to be conducted on every message, statistic efficiency is improved, and the statistic types are further conveniently managed, so that efficient and flexible flow statistics are achieved.

The invention provides a vehicle flow statistical method based on convolutional neural network vehicle model recognition in a cloud environment. The vehicle flow statistical method comprises the steps of: constructing a vehicle model database; designing a convolutional neural network used for vehicle model recognition; training the convolutional neural network; carrying out vehicle model recognition by utilizing a real-time cloud platform; and tracking vehicles by utilizing a Kalman filter, and counting vehicle flow of corresponding vehicle models. According to the vehicle flow statistical method based on convolutional neural network vehicle model recognition in the cloud environment, an image recognition technology based on the convolutional neural network is combined with a cloud computing technology, vehicle flow statistics based on vehicle model recognition is subjected to parallelization by utilizing the cloud computing technology, and vehicle flow statistical rate is increased; moreover, on the premise of fully utilizing computing resources, an optimal network is found, the vehicle models are recognized, the vehicle flow is counted, and a real-time purpose is achieved.

The invention belongs to the technical field of machine vision, and provides a pedestrian counting method based on head detection. The method comprises that S1) multiple training sample sets are trained by an SVM classifier to obtain multiple head classifiers; S2) a video stream image of a monitoring area is obtained and used as an input image; S3) a frame difference method is used to calculate the frame difference of the input image and further to obtain a moving pedestrian area; S4) the head classifiers are used to detect the moving pedestrian area to obtain a head area; and S5) a Kalman nearest neighbor coupling tracking method is used to track and count central points of the head area. According to the invention, the SVM classifier is used for training to obtain the multiple head classifiers, the head classifiers are used to detect the moving pedestrian area to obtain the head area, and the alman nearest neighbor coupling tracking method is used to track and count the central points of the head area; and thus, the calculation efficiency and accuracy of the number of pedestrians in a public area is improved.

The invention discloses a system and a method for identifying spam message sources by combining message contents and transmission frequency. The system comprises a parameter configuration interface module, a real-time acquisition and interception module, a service module and a management module, wherein the real-time acquisition and interception module acquires all messages, transmits the messages to the service module and receives interception information transmitted by the management module; the service module receives the parameter configuration transmitted by the management module, counts the messages of a user, identifies the message sources reaching a preset threshold and transmits all information of the message sources exceeding the threshold to the management module; and the management module receives and records the threshold exceeding information transmitted by the service module and transmits the interception information to the real-time acquisition and interception module. In the method, content characteristic values are calculated according to the contents, the messages having the same contents of the message sources are counted according to the characteristic values, only messages which are sent by the message sources and of which the contents meet characteristic value discovering conditions are intercepted, the characteristic values intercepted at the same time are counted and message sources meeting a blacklist condition are recorded in the blacklist. The system and the method can improve the spam message identification rate and lower spam message interception error rate.

A test use case display method comprises the steps of reading XML data of a mind map of test use cases, information carried by the path from a root node to each leaf node in the mind map forming one test use case; generating a mind map picture with an appointed height according to the XML data, wherein paths in the mind map picture diffuse towards a single side from the root node and the leaf nodes are arranged uniformly in the longitudinal direction; opening a specified EXCEL table which contains a use case column used for recording test use cases and a result column used for recording test results; adjusting the width of the use case column to adapt to the width of the mind map picture and inserting the mind map picture into the use case column and adjusting the row height of the EXCEL table to align each row of the lateral side of the mind map picture with one leaf node of the mind map picture horizontally; displaying the EXCEL table. The method can increase the overall efficiency in a test process. The invention also provides a test use case display device.

The invention discloses a flow calculating method, a flow calculating device and a flow calculating system based on a private network VPN. On the basis of private VPN information, the flow is calculated and analyzed on a device of a service provider of an MPLS LSVPN network, so the invention solves the original problem of relying on administrator to manually deduct and calculate the flow, and enables the flow analysis to automatically complete by matching a flow analyzing server with the device, thereby enhancing the efficiency of the flow calculation, and lowering the labor cost.

A classroom behavior analysis method and system. The system is provided with a classroom audio data module, a speech recognition module, a first analysis module, a second analysis module, and an information statistics module. The method includes: S1, the classroom audio data module combining multiple audio and video data The information is sent to the speech recognition module; S2, the speech recognition module recognizes the speech information generated by the classroom audio data module into words and a certain person's voice, and then sends the recognized content of these two parts to the second analysis module and the first analysis module respectively. Analysis module; S3, the first analysis module records the voice data information sent by the speech recognition module; the second analysis module matches and analyzes the text content sent by the speech recognition module; S4, the information statistics module combines the first analysis module and The analysis results of the second analysis module are summarized and a statistical table is generated. The classroom behavior analysis method and system proposed by the present invention can significantly improve the statistical efficiency and statistical accuracy of helping students.

The invention relates to a microscopic image-based cell analysis method and device, equipment, and a storage medium, and relates to the technical field of image processing. The method comprises the following steps: acquiring color microscopic images for respectively displaying and marking cytoplasm of the cell nucleus and the positive cell; identifying cell nucleuses in the color microscopic imageto obtain cell nucleus coordinates of each cell; performing region division on the color microscopic image based on a pixel value threshold value; obtaining at least two image areas respectively corresponding to various cell attributes, combining cell nucleus coordinates of each cell; and at least two image regions, obtaining percentages of cells indicating various cell attributes in each cell, according to the scheme, the cell percentage is automatically counted through artificial intelligence AI, manual classified statistics is not needed, the statistical time of the cell percentage of eachattribute is greatly shortened, and the statistical efficiency of positive / negative cells in pathological diagnosis based on immunohistochemistry is improved.

The invention discloses a production data management information system. The system comprises a production plan scheduling system, a production data acquisition system, a production data storage center, an analytical and statistical system platform and a result presentation module. The production data management information system utilizes the production plan scheduling system to schedule production plans, utilizes the production data acquisition system to product according to the production plans, records actual production information and stores actual production information in the production data storage center, and then, utilizes the analytical and statistical system platform to process the actual production information according to a preset strategy to obtain a production statement, thus, the statistical efficiency of production efficiency can be improved, production efficiency analysis is relatively thorough, a large amount of working time of plan management personnel used in paper file statistics of production efficiency is released, the efficiency of generating the production statement is high, the work efficiency of plan management personnel can also be improved, in addition, the production statement is presented through the result presentation module in the form of characters or graphs, and the readability is high.

The invention discloses a distributed statistics based method for frequently passing vehicles. The method comprises the following steps: generating a request for statistics on frequently passing vehicles according to a preset statistical condition for frequently passing vehicles; sending the request for statistics on frequently passing vehicles to all nodes of a hadoop; receiving data of frequently passing vehicles through passing gates, found by the nodes in response to the request and in compliance to requirements of preset gate information, vehicle information and vehicle passing time, and storing the data according to the gate information and vehicle identification; searching all stored data of passing vehicles through the gates, having them combined and finding out the how many times each vehicle passes through the gates and recording the times; search how many times one vehicle passes the gates; according to a preset threshold value on the passing times, deleting the data of passing vehicles that are not in compliance with the threshold value; and obtaining and storing the gate information, vehicle information, gate passing time and the passing times of frequently passing vehicles. The method and the system raise statistical efficiency on frequently passing vehicle.

The invention discloses a microsatellite loci marker combination which comprises 17 microsatellite loci markers, and simultaneously provides an efficient detection method based on the 17 microsatellite loci markers; the markers have high polymorphism, are not linked with each other and have proper fragments interval, and are easy to be detected simultaneously. The invention is improved in a detection technology, optimizes an experimental system by utilizing a multi-marker amplification technology and a fluorescent semi-automatic microsatellite typing method, thus establishing high flux, accurate and effective 17 cattle microsatellite marker compound amplification and a gene typing method.

The invention discloses a vehicle counting method based on movement track, and relates to the field of traffic detection. The method of the invention comprises the following steps: stable feature points are selected and tracked; clustering is carried out according to the change of relationship between the same-time relative distances of the track of the stable feature points in a real space; and the count value of a vehicle counter is not updated when the variance of the same-time relative distances is smaller than a preset threshold, and 1 is added to the count value of the vehicle counter when the variance of the same-time relative distances is greater than the preset threshold. According to the invention, the vehicle track in a shot road video image is extracted and processed to obtain the relative distances of the track at the same time, and the number of vehicles in the video image is determined according to the change relationship between the same-time relative distances. Therefore, the influence of the environment in the vehicle counting process is avoided, the limitation that only vehicles of a single model can be counted is eliminated, and the precision and efficiency of vehicle counting are improved.

The invention provides a data statistic method and system. The method comprises the following steps: pre-creating a data extraction record chart, presetting a target condition and monitoring at least one relational database connected with a big data platform; when the target condition is satisfied, extracting incremental data in the relational database with the incremental data into the big data platform; updating the data extraction record chart; and when a statistic instruction is obtained, responding the statistic instruction according to the data extraction record chart. According to the scheme, the data extraction record chart is updated after the data in the relational database is extracted into the big data platform; and when statistic is carried out on the data in the big data platform, response can be carried out by utilizing the data extraction record chart without traversing all the data in the big data platform, so that the statistic efficiency can be improved and a mass of computation and network resources are saved.

The invention discloses an adaptive quick focusing device for an optical camera. The adaptive quick focusing device comprises a focus detector (1), an image collector (2), a data arithmetic unit (3), a focusing drive component (9) and an encoder (7), wherein the encoder (7) is used for acquiring the position of an image surface (10) of the optical camera; the image collector (2) is used for acquiring a quantized image; the data arithmetic unit (3) is used for calculating a configuration entropy of the image; and the focus detector (1) is used for calculates configuration entropies of images at two different positions and sending focusing information to the focusing drive component (9) according to a result to drive the image surface (10) of the optical camera to reach the position of the maximum configuration entropy. An adaptive quick focusing method for the optical camera comprises the following steps of comparing the configuration entropies of images formed by the image surface (10) of the optical camera at two different positions, adjusting the moving direction according to the comparison result, and taking the position of the maximum configuration entropy of the image as the focusing position of the image surface (10) of the optical camera. The focusing speed is high, the focusing accuracy is high, and the requirement of weak contrast objects on space remote sensing is met.

The invention discloses a farm management method based on a neural network, and the method comprises the steps: collecting various images of a farm, and processing the images to form an image data set; Forming a target detector by adopting a YOLO network, wherein the target detector adopts three scales; Inputting the image data set into a YOLO network to complete training; enabling The trained target detector to be used for detecting and counting targets bred in a breeding farm, and distinguishing other types of species; And positioning the detected human head, controlling a camera to automatically focus the human head to extract a clear head portrait, and storing and recording the clear head portrait. According to the method provided by the invention, the types of outbound harmful speciescan be accurately judged, and farmers can conveniently judge whether to take defense measures or not; The method provided by the invention has the advantages that experimental verification proves that the accuracy is higher when small targets such as chickens are detected; People entering and exiting the breeding pen are positioned, the camera is controlled to automatically focus, clear images are extracted and recorded and stored, and theft prevention, damage prevention and poisoning prevention are effectively achieved.

The invention relates to a GIS (Geographic Information System)-based vehicle-mounted recorder for monitoring a vehicle, which consists of a core plate, a GPS (Global Positioning System) module, a power supply circuit, a vehicle-mounted coder and a 3G (the 3rd Generation) network card. The power supply circuit is input by 12V, and the core plate is respectively connected with the GPS module, the power supply circuit, the vehicle-mounted coder and the 3G network card. Scientific means are provided for analysis and decision of online real-time monitoring and management of the vehicle by combining real-time vehicle tracks and work state data information obtained by a monitoring network consisting of a central processing unit and a vehicle-mounted wireless communication detector by means of strong spatial inquiring analyzing and pattern visualizing function of the GIS. The invention not only can position the vehicle precisely, playbacks and displays the information by the map and coverage loading function of the GIS, but also can inquire and display full-time dynamic data of the vehicle, and analyzes the path by means of the spatial inquiring analyzing and thematic map preparing function of the GIS, thereby improving the statistical efficiency of data of environmental sanitary vehicles in various forms.

The invention discloses a data sending method and device, and relates to the field of computers. The method includes acquiring distribution data to be distributed and a first time node and a second time node which are to be distributed; acquiring a total number N of the distribution data in the distribution data, and randomly determining N different time nodes i between the first time node and the second time node; corresponding each time node i to the distribution data, and placing the distribution data into data queues in sequence; after a data acquiring request from a client terminal is received, acquiring the initiate time of the request and the time node i according to the data, and determining whether or not the client terminal has the permission to acquire the distribution data according to the data queues; if so, acquiring and sending the distribution data to the client terminal according to the data queues. The method is convenient to maintain, occupation of system resources can be reduced, data performances can be further improved, and statistical efficiency can be improved in process of distribution data statistics.

The invention discloses a curtain wall production and processing method based on a BIM technology, and belongs to the technical field of curtain walls. The curtain wall production and processing method comprises the following steps of S1, designing a drawing; S2, performing statistics on shared parameters; S3, performing node deepening arrangement; S4, performing location optimization; S5, performing software programming; S6, performing information statistics; S7, providing a drawing; and S8, processing and installing. In the construction preparation stage, the BIM technology is adopted, a curtain wall three-dimensional model is established first, and an external wall in the design is divided and subjected to typesetting; before factory production and processing, early mounting simulationcontrol is carried out, and relevant content of a curtain wall discharging list is checked, so that the occasions that material models, sizes and quantities do not fit real on-site conditions during later mounting, so that manpower, materials, the construction period and other costs are increased invisibly are reduced. Meanwhile, the appearance and cost risks of the project are integrally controlled in advance.

The invention discloses a data processing method, device and system. The data processing method comprises the steps of, aiming at any client, obtaining online behavior information collected by a client; matching an online behavior of the online behavior with a preset online behavior corresponding to each task in a preset task data table; if the online behavior is succeeded in matching with at least a task, accumulating continuous data of the online behavior information with the preset behavior data table and an accumulated value of the preset online behaviors corresponding to the online behavior information collected by the client; and determining tasks which the online behavior information collected by the client is met based on a judgment threshold in the task data table and the accumulation value in the behavior data table. Compared with the prior art, the data processing method reduces workload during new task development by building a new task in the task data table rather than modifying any program at the front end, the rear end and the client when a new task is added.

The present invention discloses a method for constructing a school district pressure model through adoption of a map technology. The method comprises the following steps: drawing a school district area on a map, performing statistics of school-age children's household registration information in the household registration information and converting the school-age children's household registration information to longitudes and latitudes, performing statistics of latitude and longitude points falling into each school district, wherein the number of persons corresponding to the longitudes and latitudes are the number of the school-age children in the corresponding area, and finally displaying the school district information and the statistics data. Compared to the prior art, the method for constructing the school district pressure model through adoption of the map technology, the method can assist schools in effectively perform pre-prediction of the number of the school-age children so as to greatly reduce the work cost of artificial statistics and effectively improve the efficiency of calculation and statistics of the number of the school-age children.

The invention discloses a method for calculating the score of a test paper, and relates to the field of education and examination. The method comprises the steps of acquiring an image of the test paper, and performing preprocessing on the image of the test paper, wherein the preprocessing is used for improving the contrast ratio of the image of the test paper and removing noises in the image of the test paper; performing regional positioning on the preprocessed image of the test paper, and acquiring a target image only containing handwritten score information; performing score character segmentation on the target image to acquire a character image of each test question; recognizing the character image of each test question to acquire the score of each test question; and calculating and inputting the total score of the test paper. According to the method for calculating the score of a test paper, the total score of the test paper can be calculated on the basis of the handwritten score of each test question, so that the workload of teachers is reduced while artificial marking habits of the teachers are reserved, and the application range is wide.