528 results about "Time overhead" patented technology

The invention relates to a method and a system for storing mass data of the Internet of Things (IoT). The system comprises a plurality of data receiving nodes, a master node server and a database cluster. The method comprises the following steps of: (1) carrying out preprocessing on IoT data, and putting the preprocessed data in the database cluster consisting of a master node, slave nodes and the data receiving nodes; (2) creating sample records, which take sample elements as storage units, on the master node according to the static and dynamic information of the data in the database cluster; (3) after the sample records are encapsulated, sending the encapsulated sample records to all the slave nodes by the master node so as to be subjected to fragmentation processing and/or separated storage; and (4) after the slave nodes complete storage, uploading results to the master node, and updating the data in the database cluster by the master node, thereby completing storage. According to the method and the system, the cluster is extended by fully utilizing the existing storage technologies so as to store the mass data, the IoT data are divided into lightweight data and multimedia data, and particularly, a fragmentation strategy is adopted aiming at a large-scale amount of data, so that the time expenditure caused by the extension of a storage space is avoided.

The invention relates to a video dynamic super-resolution processing method and system. The processing method comprises the steps of obtaining a motion vector of any pixel in a current to-be-processed frame according to motion estimation and calculating reliability degree of the motion vector; extracting the corresponding pixel of the pixel in a former high-resolution frame according to the motion vector of the pixel and carrying out Kalman filtering processing according to the corresponding pixel, thereby obtaining a first processing result; carrying out interpolation processing on the to-be-processed frame to obtain a second processing result; and carrying out weight fusion on the first processing result and the second processing result through utilization of the reliability degree, thereby obtaining a super-resolution frame after fusion. According to the method and the system, the problems of protecting edge information well and improving image edge sawteeth are realized; more image details are restored effectively on the condition of not increasing excessive extra time cost; moreover, relatively high robustness is achieved; and disorder of a super-resolution result or motion blurring resulting from inaccurate motion estimation is avoided.

The invention discloses a topic detecting device and a topic detecting method based on distributed multistage cluster. The topic detecting device mainly comprises a news acquiring module, a news classifying module, a topic detecting module, a topic integrating module and a topic displaying module. The topic detecting method is characterized by comprising the steps of A, acquiring news; B, classifying the newly acquired news; C, performing multistage cluster for various channels concurrently; and D, computing hotness of all topics, and screening hot topics of a total system and hot topics in each channel. By the topic detecting device and the topic detecting method, a sharp contradiction between a detecting effect and time cost in topic detection under the condition of quick updating of a large number of files in an internet environment can be solved.

The invention relates to a method for identifying multi-class facial expressions at a high precision based on Haar-like features, which belongs to the technical field of computer science and graphic image process. Firstly, the high-accuracy face detection is achieved by using the Haar-like features and a series-wound face detection classifier; further, the feature selection is carried out on the high-dimension Haar-like feature by using the Ada Boost. MH algorithm; and finally, the expression classifier training is carried out by using the Random Forest algorithm to complete the expression identification. Compared with the prior art, the method can reduce training and identifying time while increasing the multi-class expression identification rate, and can implement the parallelization conveniently to increase the identification rate and meet the requirement of real-time processing and mobile computing. The method can identify the static image and the dynamic image at a high precision, is not only applicable to the desktop computer but also to the mobile computing platforms, such as cellphone, tablet personal computer and the like.

A system and method for reducing the overhead associated with direct data placement is provided. Processing time overhead is reduced by implementing packet-processing logic in hardware. Storage space overhead is reduced by combining results of hardware-based packet-processing logic with ULP software support; parameters relevant to direct data placement are extracted during packet-processing and provided to a control structure instantiation. Subsequently, payload data received at a network adapter is directly placed in memory in accordance with parameters previously stored in a control structure. Additionally, packet-processing in hardware reduces interrupt overhead by issuing system interrupts in conjunction with packet boundaries. In this manner, wire-speed direct data placement is approached, zero copy is achieved, and per byte overhead is reduced with respect to the amount of data transferred over an individual network connection. Movement of ULP data between application-layer program memories is thereby accelerated without a fully offloaded TCP protocol stack implementation.

The invention discloses a fast high-resolution SAR (synthetic aperture radar) image ship detection method based on feature fusion and clustering. The fast high-resolution SAR image ship detection method comprises the following steps: on the basis of the back scattering characteristics of each ground object and the prior information of a ship target in an SAR image, positioning a target potential position index map by an Otsu algorithm and range constraint; on the index map, pre-screening to obtain a detection binary segmentation map by a CFAR (constant false alarm rate) algorithm based on a local contrast; carrying out morphological processing to a detection result, and extracting a potential target slice from the SAR image and a detected binary segmentation map according to a processing result; and carrying out K-means clustering to the extracted slice by a designed identification feature to obtain a final identification result. According to the fast high-resolution SAR image ship detection method based on feature fusion and clustering, the data volume of a detection stage is effectively reduced by pre-processing, and point-to-point detection is not needed/the time of point-to-point detection is saved. Meanwhile, a target identification problem under the condition of insufficient training samples at present can be solved by the designed characteristic and a non-supervision clustering method, the target can be effectively positioned, and the size of the target can be estimated.

The present invention discloses an alignment device for alignment between a mask and a working table of a lithography machine, and can determine a basic alignment position by one time alignment with at least four marks. Compared with the prior art, the motion process required by the alignment of the mark is further reduced, thus further reducing the time cost of the alignment and improving the working efficiency of the system. Meanwhile, the present invention has a plurality of ways for processing alignment signals, can obtain different alignment accuracy, and can combine the further optimization of two aspects of the alignment accuracy and alignment efficiency by the selection of different processing strategies for the alignment signals based on the needs of users, and the detection of image quality can be realized.

This invention relates to a method for controlling location distribution of sensing nodes, selection of sensing nodes, control of sensing implementation and process of performing spectrum sensing in cognitive radio networks of open wireless architecture (OWA) systems. Specifically, the invention relates to an efficient and reliable method minimizing time overhead consumed during spectrum sensing with open wireless architecture (OWA).

The invention provides a time advance compensation satellite mobile communication random access method and system and a medium, and the method comprises the steps: a downlink broadcast sending step: obtaining a cell formed by a wave beam of a satellite on the ground, obtaining the minimum TA in the cell, recording the minimum TA as TAmin, and enabling the satellite to send satellite position information and TAmin in the downlink broadcast; And a transmission delay calculation step: enabling the UE to receive the downlink broadcast, obtaining the satellite position information and the TAmin according to the received downlink broadcast, and calculating the difference value TAdiff of the TA estimated by the UE per se relative to the TAmin. In the random access process of the satellite, the transmission time delay is estimated in advance at the UE side and compensation is performed when the PRACH is sent, so that the problem of high time cost caused by long time delay in the random accessof the satellite is solved, and the consumption of uplink resources by the PRACH is greatly reduced.

The invention relates to a lung nodule detection method based on a three-dimensional convolution neural network, which adopts a network structure of a feature pyramid and an attention mechanism,. Themethod integrates features of low-level and high-resolution and high-level abstract features, and enables the network to concentrate attention on a target area. The method has the effects of using three-dimensional convolution neural network detection method, end-to-end detection of pulmonary nodules, and reducing the time overhead. Compared with the traditional methods, the method can improve therecall rate and average accuracy of nodule detection.

The invention discloses a video service quality-based hybrid selective repeat method, which mainly solves the problems of reduced data transmission quantity, large transmission delay and poor video information continuity caused by unnecessary cost in the prior art. The method is implemented by the following steps of: first, calculating a video subjective quality objectified value by utilizing a video subjective quality objectifying model, and calculating a data transmission redundancy range which cannot cause network congestion by using a network broadband which feeds back from a receiving end to a transmitting end; then, according to the data transmission redundancy range, allocating proportions to forward error correction redundancy and selective repeat redundancy to optimize the video subjective quality objectified value; and finally, selecting a code pattern for coding according to the forward error correction redundancy, and determining the number of packets needing to be repeated according to the selective repeat redundancy. In the method, forward error correction control and selective repeat error control are taken into consideration at the same time, so that the time cost and the network cost are reduced; therefore, the method can be used in a video transmission system.

The present invention belongs to the field of three-dimensional reconstruction technology and relates to a fast three-dimensional reconstruction method based on Kinect camera. Compared with the conventional technique, the method of the present invention uses the method of main optical axis constraints to remove the imaging error points in a single-view point cloud and improves the quality of the point cloud to a certain extent, making the three-dimensional reconstruction higher accuracy. In addition to that, considering that the point cloud has poor accuracy and less coincidence, the existing point cloud registration algorithm based on RGB-D data is improved. Although some computational time overhead is increased, the registration accuracy of the algorithm is greatly improved. This means the great possibility of the algorithm converging to a local minimum.

The invention discloses a multiple-mode-string matching method and device. The method includes that a plurality of mode strings are sequenced according to their respective characters, each character is written into a node along a root node of a tree structure downwards to generate a decision tree structure, and main strings to be matched are matched downwards along the decision tree. By means of the method and device, accurate matching of multiple mode strings can be achieved, meanwhile sub nodes are searched according to the Hash values that the sub nodes correspond to, the width change of the decision tree cannot affect time expense of a central processing unit (CPU) matched with the strings, and the time expense of the algorithm depends on the average depth of the decision tree and is unrelated to the number of the mode strings. For string matching with a large number of mode strings, the algorithm can greatly reduce the time expense of the CPU and improve application response speed.

System and method are described for register optimization during code translation utilizes a technique that removes the time overhead for analyzing register usage and eliminates fixed restraints on the compiler register usage. The present invention for register optimization utilizes a compiler to produce a register usage bit vector in a NOP instruction within each basic block (i.e., subroutine, function, and/or procedure). Each bit in the bit vector represents a particular caller-saved register. A bit is set if, at the location of NOP instruction, the compiler uses the corresponding register within that basic block containing the NOP instruction to hold information to be used at a later time. During the translation, the translator examines the register usage bit vector to very quickly determine which registers are free and therefore can be used during the register optimization without the need to save and restore the register values.

The invention discloses a violent behavior detection system based on human body posture estimation, and the system comprises a video obtaining unit which obtains a monitoring video and synchronously transmits the monitoring video to a monitoring center and a cloud end; T. The monitoring center is used for displaying and storing the monitoring video in real time and supporting playback; T. The cloud is used for converting the monitoring video into multiple frames of images, estimating the human body posture in each frame of image by adopting OpenPose, judging whether a violent behavior exists in the frame of image or not according to four limb directions of the human body posture, judging the type of the violent behavior, and transmitting an analysis result to the alarm unit; and the alarmunit is used for sending alarm information in real time if the cloud identifies that a violent behavior exists in a certain frame of image. The invention further discloses a violent behavior detectionmethod based on human body posture estimation. According to the method, a bottom-up human body posture estimation method is adopted, the robustness is higher, the time cost is stable, and the real-time requirement of the method can be met.

The invention discloses a method for remapping virtual network resources, and relates to the technical field of virtual networks. The method comprises the following steps of: acquiring a key virtual node periodically, wherein the key virtual node is a virtual node in a virtual node set which is born by a physical node having a resource utilization rate greater than a threshold value; mapping resources of the key virtual node to a physical node having the resource utilization rate smaller than the threshold value, so that a resource load of a physical network is balanced; and remapping resources of a virtual link which is connected with the key virtual node. According to the method, a whole physical resource load is analyzed periodically, whether virtual resource is required to be remapped is judged, and part of key virtual nodes which cause unbalance of the physical resource load are selected to be remapped during remapping, so that the conventional physical resources can be integrated, the physical resource load can be balanced, the receiving rate of virtual network requests is improved, time consumption and computation cost can be saved, and the influence caused by the interruption of virtual network normal service can be maximally reduced.

The invention relates to a prediction method of operational performance based on a historical data modeling in grid, belonging to an operation completing time modeling and prediction method in high-performance grid. The prediction method is characterized by comprising the steps of establishing a historical operational information bank based on a CGDP grid software and a CGSV grid software in grid nodes, wherein the historical operational information bank contains N historical operational information, relating to four aspects of resource allocation, resources loading, operation request and operation actual performance; and simultaneously establishing a set comprising one or more candidate regressive functions, wherein when predicting, the N+1th operation submitted by a user is acquired according to a regressive model of the Nth operation, and the regressive model of the Nth operation is acquired by selecting a candidate regressive model with the smallest difference value from differences of predicted value results of operation actual performance of the Nth operation and the actual performance of the candidate regressive models of the N-1th operation. According to a simulation experiment, the invention can solve the problem of surging operation time and operation expense caused by excessive resource load.

The invention provides a parallel TCP (Transmission Control Protocol) technology based wide area network (WAN) communication acceleration method. The method comprises the following steps of: when a proxy server monitors that a client in a local area network (LAN) sends a session connection request to a server in a target LAN, establishing a session connection between the proxy server and a proxy client, i.e. a first session connection; based on a WAN environment, establishing a parallel TCP tunnel between the proxy server and the proxy client; connecting the proxy client and the server through a session connection based on session pool distribution, i.e. a second session connection; and carrying out data communication by the client in the LAN with the server in the target LAN through a communication channel formed by the first session connection, the parallel TCP tunnel and the second session connection. The invention can effectively reduce the time expense of the data transmission of the WAN and accelerate the speed of file and data transmission.

The present invention relates to a system and method of data management for controlling a plurality of peripheral devices. In an example embodiment, a data management device for managing data communications between a data control device and the plurality of peripheral devices is added between them, and the data management device controls the plurality of peripheral devices by grouping them according to their characteristics, whereby various protocols can be integrated into minimized optimal protocols by the grouping, and the data control device can reduce technical and temporal overheads due to detection and control of the plurality of peripheral devices.

The invention relates to the technical field of brain function network research and more particularly relates to a brain function network construction method based on a micro-state. Starting from theconstruction of an EEG brain function network, according to the theory of EEG micro-state related aspects, namely that the EEG micro-state is the most basic state which reflects the brain activity, proposing that the EEG micro-state can also be applied to construction of the brain function network. The method provided by the invention solves the problem that the time consumed for construction of the traditional brain function network is too long and reduces a large quantity of redundant information in an original EEG time sequence by using the time sequence after micro-state reconstruction. Byapplying the method to an EEG data set of mental disease working memory collected in a hospital, the EEG brain function network for schizophrenia patients and normal persons is constructed; and compared with the brain function network constructed by using an original EEG signal, the brain function network constructed by the time sequence after micro-state reconstruction has the advantage that thecharacteristics of the human brain network can be more accurately reflected.

To solve the technical problem that the existing particle image classification methods have too much time overhead and low accuracy, the present invention provides a deep clustering method facing single particle cryo-electron microscope image. The deep clustering method comprises the following steps: a first step, performing data preprocessing, and sending data to an autoencoder to perform pre-training; a second step, training the autoencoder: clustering output vector features of the encoder; calculating a loss function by using the clustering result; and optimizing the autoencoder weight by using the stochastic gradient descent method; and a third step, inputting all the particle image data into the autoencoder, obtaining the clustering result and analyzing the clustering accuracy rate, determining whether the loss function and the accuracy rate change are less than a threshold, outputting the clustering result if so and performing ending, otherwise, returning to the second step. Theinvention can perform pre-training under various noise data to improve the noise reduction capability of a network, and adaptively train the weight of the loss function by using the stochastic gradient descent method to further improve the classification accuracy rate.