92results about How to "Reduce storage capacity requirements" patented technology

The invention discloses a VFTO (very fast transient overvoltage) measurement method and system, relating to the technical field of overvoltage protection of a power system. The high-frequency component of VFTO is measured by high-resolution continuous sampling, the power-frequency component in the low-frequency component of VFTO is measured by low-resolution single sampling, and the measured data of the high-frequency component and low-frequency component of VFTO is synthesized into a VFTO full waveform. Through the VFTO measurement method and system disclosed by the invention, the high-frequency component and low-frequency component of VFTO are separately measured, and the measured results are synthesized into a full-waveform VFTO, thus the difficulty of VFTO measurement and the requirement for the storage capacity for storing the measured results are reduced, and the accuracy of VFTO measurement is improved.

The invention discloses a traffic accident recorder, which comprises an embedded processor, a system interface module, a sensor, an internal storage, a memory controller, an external storage, an ATA bus, a PCI bus and a video recorder. The video recorder transmits video data recorded to the internal storage through the PCI bus; the internal storage is connected with the embedded processor, the ATA bus, the PCI bus through the memory controller respectively; the sensor is connected with the embedded processor through the system interface module to output environmental information collected to the embedded processor through the system interface module so as to judge unexpected accidents; when an accident happens, the corresponding video data in the internal storage is stored in the external storage by the embedded processor through the ATA bus. The invention can record the occurrence process of the accident timely and promptly when the traffic accident happens.

The invention discloses a method for recognizing Chinese characters in a natural scene. The method recognizes characters by a modeling training stage and a character recognition stage. The modeling training stage comprises successively establishing a tree structure expression of template characters, synthesizing a training set, training a convolutional nerve network, extracting the deep template characteristic of a node, and establishing and training a Markov random field. The character recognition stage comprises preprocessing a picture to be recognized, extracting the deep characteristic of an input picture, minimizing a Markov random field energy function, and finally recognizing characters. The method takes account of the local characteristic and the global structure of the characters while recognizing the characters, overcomes an influence on a recognition effect due to fuzzy characters and large deformation in the natural scene by combining the Markov random field technique with the deep characteristic of the node in the tree structure, thereby increasing recognition efficiency.

The invention belongs to the technical field of block chain network communication, and discloses a high-efficiency asynchronous Byzantine consensus method with low communication complexity and a network communication platform. A node in a network calculates transaction data blocks, redundant data blocks, a data block hash and a fingerprint cross checksum, and sends related data to each node in thenetwork by using an asynchronous verifiable information transmission protocol. After the nodes confirm that the transactions broadcast and sent by some nodes are correctly stored in the network, a binary vector consistency protocol is executed, and consensus is carried out on the transactions; and the node requests a specific data block from the network and reconstructs to obtain original transaction data. According to the invention, the problem of high communication complexity based on an erasure code protocol in the prior art is solved, and the throughput of the algorithm is improved; a binary vector consistency protocol is adopted, the defect that an original protocol can only consensus transactions broadcasted by a single node each time is overcome, the complexity of algorithm messages is reduced, the worst time complexity of an algorithm is optimized, and the high efficiency of the algorithm is guaranteed.

The invention discloses a server log viewing method, device and system used for reducing the processing load of a log server and improving the processing speed of server log storage and query. The server log viewing method comprises the following steps: receiving a log data query request sent by a requester, wherein the log data query request carries a service identifier; determining, according toa corresponding relationship between the pre-created service identifier and a service server identifier, at least one service server identifier corresponding to the service identifier carried in thelog data query request; sending, according to the determined at least one service server identifier, the log data acquisition request to a corresponding service server, wherein the service server stores log data of own running services; receiving log data returned by the service server for a log data acquisition request; and sending the received log data to the requester.

The invention discloses a satellite borne shared storage exchange device which comprises a wire in processing unit, a plurality of memorizers and a wire out processing unit. The wire in processing unit comprises a plurality of interface processing units, a plurality of packet processing units, an address list management module, a parallel writing-in control module and a multiplexing module. The wire out processing unit comprises a demultiplexing module, a parallel reading control module, a plurality of output caching modules and a plurality of sending modules. Each input interface corresponds to an interface processing module, a packet processing module and a memorizer. Each output interface corresponds to an output caching module and a sending module. The number of the input interfaces and the output interfaces is identical to the number of the interface memorizers. The satellite borne shared storage exchange device can effectively reduce the requirements on the access rates and the capacity of the memorizers, and meet the development requirements of a high-speed large-capacity satellite borne routing switch.

The invention discloses a multi-agent system consistency problem optimization method based on sub-graph processing, and the method is characterized in that the method comprises the following steps: 1)constructing a graph signal model of a multi-agent system; the method comprises the following steps of (1) solving a problem to be solved, (2) adding an auxiliary constraint condition to the problemto be solved, (3) carrying out local inversion in sub-graphs, (4) carrying out fusion averaging between the sub-graphs, and (5) iteratively eliminating errors, and the method is capable of reducing operation complexity, achieving consistent averaging, high in convergence speed and low in requirements for computing capacity and storage capacity of a single node of a system.

The invention discloses an extension visual encryption and decryption method. In the extension visual encryption process, a cover image is chunked first, and pixels in each block are divided into secret information pixels, auxiliary black pixels and common pixels; then (2,2) lines of a visual cryptography fundamental matrix are filled at the position of secret information, share images are divided into four areas, and secret images and watermark images are sequentially embedded into the four areas; at last, limited error diffusion is used for incorporating the characteristics of the cover image. In the extension visual decryption process, the two share images are overlaid in a staggered mode, and watermarks are recovered; then the two images align to each other to recover secret images. The extension visual encryption and decryption method has the advantages that image sharing is meaningful, the dimension is not extended, decryption does not need a computer, a tampering authentication function is achieved, and the recovered secrete images and the watermarks can be free of interference of the cover images at the same time.

The invention discloses a digital predistortion processing device which includes a pre-distorter, a self-adapting updating module, a control module and a calibration sequence generation model used for generating the source calibration sequence and providing the source calibration sequence for the pre-distorter; the self-adapting updating module is used for acquiring the feedback calibration sequence from the power amplification output; the latest pre-distorter is obtained according to the source calibration sequence and the feedback calibration sequence; the pre-distorter is used for storing the predistortion function, processing the source calibration sequence and the source signal through the predistortion function and sending the processed source calibration sequence and source signal to the power amplifier; meanwhile, the latest predistortion function of the self-adapting updating module can be adopted to update the predistortion function stored in the self-adapting updating module; the control module is used for controlling the pre-distorter, the self-adapting updating module and the calibration sequence generation module for coordination operation. The digital predistortion processing device adopts the newly-constructed calibration sequence to control all the amplitudes from the minimum value to the peak value; namely, the pre-distorter can be updated so that the updating speed is improved and the algorithm difficulty is simplified. The digital predistortion processing device also provides an updating method of the digital pre-distorter.

The invention relates to a method and device for providing complex web applications in a resource-constrained environment. The method includes the following steps (1) the operating environment of all the interface logic and part of service logic of the web applications is transferred to a browser client from a traditional server to lower the requirements of the server when CPU processing capacity and/or a RAM in operation are/is constrained; (2) on the premise that executing effect is not affected, the web applications and codes of the server of the web applications are decompressed to lower the requirements of the server when ROM storage capacity is constrained. The method and device for providing the complex web applications in the resource-constrained environment has the advantages that the requirements for the CPU processing capacity and the RAM in operation are substantially lowered; complex interface effect and interaction of a large amount of structural data can be achieved; the requirements of the server for the ROM storage capacity are lowered; more than 99% browser occupancy is covered, and the constraint of only supporting a specific browser (IE in general) caused by the mode that the applications are encapsulated in browser plug-ins is gotten rid of.

The invention relates to a high speed synchronization technique of character and figure sequence used for detecting bit error, comprising the following steps: writing the set character and figure data into a programmable figure and character storage unit, namely a static random access memory (SRAM), by a CPU; then switching to an address generator to control the SRAM to read the character data in a circulation manner; and carrying out buffer and displacement processing on the read-out characters and sending to a synchronous judging circuit to perform synchronous judgment with the received character data, wherein if the received characters are different from the local characters, then sending out a stop signal, prohibiting the address generator to generate new addresses and maintaining to compare the local character read by the previous address with the next received character coming continuously; and if the received characters are identical to the local characters, then allowing the address generator to generate new addresses and carrying out displacement processing on the read-out data and comparing in sequence; and when the same characters are greater than the synchronous threshold, then carrying out synchronization on the received character and figure and the local character and figure. The invention provides a high speed synchronization technique of character and figure sequence which is reliable and economical and is used for detecting bit error.

The invention discloses a method for rapidly dispatching short messages in the public phone network short message system comprising, first the short message platform establishing the connection with the terminal according to the destination contained in the short message, the storage transfer module in the short message platform sending a short message to the dispatching module, the dispatching module sending a short message to the terminal, the terminal sending the dispatching result message to the dispatching module after receiving the short message which keeps the connection with the terminal and transfers the dispatching result message to the storage transferring module, the storage transferring module checks the storage region after receiving the dispatching result messages. The method by the invention can improve sending efficiency of short messages.

A lightweight file system management method in a communication device comprises a file storage process, a file retrieval process and a file operation process. In the file storage process, the file header information and the file data are separately stored in a system storage device, and the file header information content comprises file mapping information, a file size, a file offset pointer, a space use identifier and a file system containing a file mapping information table; the file retrieval process comprises the following steps: the file header information is retrieved in segments according to the file name, and whether the file header information is used or not is judged by the space use identification. The file operation process comprises obtaining a file name of a terminal, performing file retrieval on the file name, matching the file header information corresponding to the file name, and directly operating the physical storage address of the file data. The invention can directly read and write the physical address of the file data only by searching the file header information, thus reducing the file searching speed and improving the file operating efficiency.

The invention relates to a method for predicting the post forwarding amount in a social network in the field of Internet, and particularly designs a method for estimating the event popularity by predicting the post forwarding amount in the social network based on a Kalman filter. The method is aimed at solving the problem of predicting the total post forwarding amount when the time after a post isissued tends to be infinite. A post influence predicting model is established for different states in the post cascaded forwarding process, the theory that the post state at the current moment is only influenced by the post state at the last moment is put forward, and branch factors are introduced; by establishing an information cascaded tree, the predicting model of the total future post forwarding amount is converted into geometric progression summation, and therefore the final post forwarding amount is predicted. The method is widely applied to application scenes of public sentiment control, new product popularization and the like.

The invention discloses a family cloud center file sharing system based on mobile cross-platform architecture to solve the problems that in a plurality of mobile terminals existing in modern families, data files are various and hard to manage. A PhoneGap cross-platform framework is established under different operating system platforms, a family file sharing system is compiled, a PhoneGap encoder is used for online compiling the family file sharing system into different application programs of operating systems, and the application programs are allocated to client terminals. Appropriate installation file packages compiled through the PhoneGap are selected according to the operating systems of the client terminals to be downloaded, installed and used. The network is connected, all the client terminals set terminal labels and generate and upload metadata XML files of all files of the client terminals to a server. The server comprehensively obtains a metadata set of the files of the client terminals, and broadcasts and distributes the metadata set to all the client terminals. After that, all the file systems can be seen from any client terminal.

In order to solve the problem of high storage overhead in image data processing of the conventional image sensor, the invention provides an image data processing method and an image data processing device for an image sensor. The image data processing method comprises the following steps of: a, acquiring image information in a row scanning and sampling mode, and caching the image information by using a row as a unit; b, when the predetermined conditions are met, suspending the acquisition of the image information, and compressing the cached image information; and c, after the compression is finished, outputting the compressed image information, recovering the acquisition of the image information, and repeating the steps a, b and c. In the invention, the image data processing device only needs to cache the image information sampled and acquired before compression, and does not need to cache new image information during compression. Therefore, the requirement for the storage capacity of a cache module is greatly reduced, and the image data processing cost can be effectively reduced.

The present invention relates to a method for error correction in position measuring devices with material measures, which are scanned by at least one scanning unit, wherein for a defined number of correction points on the scale embodiment respectively correction values are kept, which recovered from a place before the measuring mode calibration and in measuring mode correction the detected position values are used. The correction values obtained during the calibration are compressed for the measurement operation.

The invention relates to a block chain distributed storage method and system, and the method comprises the following steps: dividing the time into a plurality of continuous time periods according to aset time length, and dividing a plurality of storage intervals in the storage space of a database according to the sequence of the time periods; dividing the data file into storage intervals corresponding to time periods containing the time points according to the standard that the time points of the data file stored in the database are taken as the standard; dividing each data file in the storedstorage interval into a plurality of data packets, and randomly combining all the obtained data packets to obtain a plurality of integrated data packets; Storing the plurality of integrated data packets in different blocks on the block chain, deleting the integrated data packets while the storage is completed, and storing the returned private keys in the original storage intervals of the data packets; Not only can high-efficiency and safe distributed storage of data files with strong time continuity be ensured, but also the requirement on the storage capacity of the memory is greatly reduced.

The invention provides a video monitoring method and device based on salient object detection, relates to the technical field of artificial intelligence and can also be used in the technical field of finance, and the method comprises the steps: obtaining a plurality of frames of images of a target monitoring region within a preset time period; applying a preset salient object detection model, and selecting an image with the highest salient degree from all the frame images as a key frame image; and obtaining a monitoring result of the target monitoring area according to a preset image recognition model and the key frame image, and outputting the monitoring result. According to the method and the device, the video monitoring efficiency can be improved on the basis of ensuring the accuracy of video monitoring, so that the security of a bank is ensured.