44results about How to "Avoid repeated reading" patented technology

The invention relates to a three-dimensional GIS (Geographic Information System) technology based power grid visualization system and method. The three-dimensional GIS technology based power grid visualization method comprises the steps of: dividing models to be uploaded into different model upload sub-tasks based on different types of objects included in a scene to be uploaded, and calling model documents of the multiple sub-tasks in a parallel manner through a multi-thread way to improve an uploading speed; and meanwhile, on the basis of dividing the uploading tasks into different model uploading sub-tasks based on the type of the uploaded objects, only reading the model document of the objects of each type for once, wherein the repeated use of each model document is not limited to the uploading task of one client; and different clients can multiplex the read model documents. Therefore, the three-dimensional GIS (Geographic Information System) technology based power grid visualization system and method, disclosed by the invention, have the advantages of giving a full consideration to the characteristics of limited type and relative consistent specification of power devices, avoiding repeated reading of the model documents of the same type, obviously improving the uploading efficiency, guaranteeing rapid uploading and guaranteeing real-time browsing of the clients.

The invention discloses a file fragment sorting method and system applied to a file system. The method comprises the steps that offset used for identifying the reading progress of each aggregate file is set in the same aggregate file in advance; read-write request information for a target aggregate file by a first client is acquired; whether other clients perform read-write operation on the target aggregate file currently is judged; if not, a read-write exclusive permission is given to the first client, the offset of the target aggregate file is acquired through the first client, and fragment sorting is performed on the target aggregate file from the position of the offset; and when the first client finishes fragment sorting operation on the target aggregate file, a file length read by the first client is acquired, and the sum of the offset of the target aggregate file and the read file length is calculated to serve as new offset of the target aggregate file to perform offset updating. Through the file fragment sorting method and system, repeated reading of the same source file by multiple clients is avoided, and the fragment sorting efficiency of the aggregate files is improved.

The invention discloses an image data reading method for a video coder, a computer device and a computer readable memory medium. The method comprises the steps that a cache receives an instruction forreading video data, reads compressed image data from an external memory and stores the compressed image data to a compressed image data storage area; the compressed image data is decompressed and then is stored to a decompressed image data storage area of the cache; coordinates of image data are determined according to the instruction for reading the video data, to-be-searched image data is searched from the decompressed image data storage area according to the coordinates of the image data, and after the to-be-searched image data is stored to the decompressed image data storage area, the searched image data is sent to a subordinate data utilization module. According to the method, the computer device and the computer readable memory medium, the bandwidth cost of the memory in an image coding process can be reduced.

The invention provides a scanning specifying unit, a first transportation path, a second transportation path, a path switching unit, and a document detection unit. The scanning specifying unit specifies any of a double sided scanning and a single sided scanning for a scanning method of a document. The first transportation path is a transportation path for double sided scanning for scanning a first surface and a second surface of the document in order. The second transportation path is a transportation path for single sided scanning, shorter than the first transportation path, for scanning the second surface of the document. The path switching unit selects any of the first transportation path and the second transportation path based on the specified scanning method. The document detection unit detects the document with the first surface scanned and the second surface not scanned when the double sided scanning is interrupted. The path switching unit selects the second transportation path for the transportation path of the first document in the resumed double sided scanning based on a detection result of the document detection unit.

The invention relates to a method for importing data into multiple Hadoop components simultaneously, and is characterized in that the method comprises the following steps of (1) extending an import tool of the Sqoop, and adding service for importing data to the Kafka; (2) importing configuration parameters of the components according to a database, and writing parameter verification program; and (3) extending the import tool of the Sqoop, and adding service for simultaneously exporting the data to the HDFS, Hive, Hbase and Kafka. On basis of original functions of connecting the database and reading the data of the Sqoop, the function of simultaneously exporting the data to the multiple components is added, the database data are read for one time, multiple user-specified export modules are started simultaneously, and efficient and convenient data import is achieved. On one hand, multiple export tasks are prevented from being written for the same data, on the other hand, the same data are prevented from being repeatedly read, and therefore efficiency is improved.

The invention provides a convolution hardware acceleration method and a convolution hardware acceleration circuit. The convolution hardware acceleration method comprises the following steps: segmenting a feature map into a plurality of map blocks; during segmentation, the column width of the segmented image blocks is matched with the number of the MAC operation units, and the row height of the segmented image blocks is the maximum row number which can be accommodated by the internal storage module; reading data required by a block and a corresponding convolution kernel for convolution operation in sequence; extracting the extraction data of the image block and the corresponding convolution kernel in sequence; and performing convolution operation and outputting an operation result to an external storage module. By adopting the convolution hardware acceleration method provided by the invention, a smaller internal cache area can be used for supporting feature maps with any resolution to carry out convolution operation; by adopting the block segmentation mode, the utilization efficiency of the MAC operation unit is relatively high, the reading of boundary data in the vertical direction can be reduced as much as possible, and the convolution operation speed is further improved.

The invention provides a rectangular window variable two-dimensional OS-CFAR detection device and method using center point sorting, and the method comprises the steps: S1, setting a detection window on the basis of a two-dimensional distance Doppler power matrix, and obtaining a column vector in the detection window; s2, performing edge filling on the edge data; s3, generating an ascending order center point sequence by using the current center to-be-detected unit data and the adjacent center to-be-detected unit data detected for multiple times; s4, performing dichotomy insertion comparison on the column vector and the ordered center point sequence to generate a column binary accumulation partial sum of the current column; s5, obtaining the binary accumulation sum of the current window, and judging whether a target exists in a corresponding point of the detection window in the two-dimensional distance Doppler power matrix; and S6, obtaining all detection results of the two-dimensional distance Doppler power matrix, and solving the radial speed and distance of the target. According to the invention, the dynamic configuration of the detection window can be realized, the detection of RDM matrixes of various sizes can be realized, and the method can adapt to wide application scenes and requirements.

The invention provides a data processing method based on a photonic neural network chip and a related device or equipment, and the method comprises the steps: segmenting a to-be-processed image into aplurality of sub-images with different sizes, carrying out the multilayer convolution processing of each sub-image one by one, and determining the size of a buffer corresponding to each sub-image based on the convolution kernel of each layer and an intermediate result; and determining a target buffer size from the plurality of buffer sizes. According to the scheme, the size of the buffer is set to be larger than or equal to the sum of the occupied sizes of the buffers required by the convolution kernel of each layer and the intermediate result, that is, it is ensured that the convolution kernel and the intermediate data of one sub-image can be stored in the buffer at a time during multi-layer convolution, and repeated data reading is avoided. Since the size of the sub-image is smaller than the size of the image before cutting, the smaller the size of the input multi-layer convolution image is, the higher the convolution calculation efficiency is. According to the scheme, the calculation efficiency is improved while the memory reading bandwidth is reduced by adopting a relatively small buffer size.

A tag data processing method is disclosed. The method includes: receiving tag information sent by an adapter, updating the stored tag data according to the tag information, determining a tag state after the tag information has been received, and sending the tag of which the tag state is active to an upper layer for processing. A tag data processing device is also disclosed. By using the method and the device, the burden of data processing is reduced for a system, and the generation of redundant data is avoided.

The invention discloses a method for improving the readability of modeling codes of a register model and a storage medium, and the method for improving the readability of the modeling codes of the register model comprises the following steps: creating a UVM register model, deriving a field base class of the UVM register model, and generating subclasses; a series interface method is expanded or newly added in the subclass and used for supporting method calling of enumeration type values; creating an enumeration type variable corresponding to a register field of the preset UVM register model; replacing a field base class with the subclass to reconstruct an example register; and performing read-write access on the register field by using a subclass interface method. According to the method, the readable coding mode of the register field can be improved, repeated browsing of the design document is avoided, the time of design and verification developers is saved, and the working efficiency of the developers is improved.

The invention discloses a report form processing system based on a dynamic domain and a using method of the report form processing system based on the dynamic domain. The system comprises a data storage device, a data processing device, a data generating device, a data managing device, an input and output device and a report form generating device. The data storage device is used for storing data source files of report forms and report form configuration files. The data processing device is used for acquiring the data source files and the report form configuration files from the data storage device and performing field reorganization on the data source files. The data generating device is used for acquiring the report form configuration files and the data source files from the data processing device and performing special field processing and data fractionation on the data source files. The data managing device is used for acquiring the report form configuration files and the data source files from the data generating device. The input and output device is used for receiving a report form inquiring command of a user, sending the command to the report form generating device, receiving band number form samples returned by the report form generating device and showing the band number form samples to the user. The report form generating device is used for acquiring the data source files and the report form configuration files from the data managing device according to the report form inquiring command, generating the band number form samples and transmitting the band number form samples to the input and output device. Therefore, showing efficiency of the report forms is improved.

The present disclosure relates to a method and device for displaying electronic documents, and belongs to the technical field of computers. The method includes: receiving an object relationship query instruction in a state of displaying an electronic document, wherein the object relationship query instruction carries a target object, and the target object is any one of all objects contained in the electronic document; Generate an object relationship display model corresponding to the target object according to the pre-stored object relationship data, wherein the object relationship data is used to record all the relationship entries contained in the electronic document, and all the relationship entries in the Each association relationship entry includes any two objects in all the objects and at least one kind of association relationship information between the two objects; and an object relationship display model corresponding to the target object is displayed. By adopting the present disclosure, the efficiency of reading electronic documents can be improved.

The invention discloses an image data reading method for a video coder, a computer device and a computer readable memory medium. The method comprises the steps that a cache receives an instruction forreading video data, reads compressed image data from an external memory and stores the compressed image data to a compressed image data storage area; the compressed image data is decompressed and then is stored to a decompressed image data storage area of the cache; coordinates of image data are determined according to the instruction for reading the video data, to-be-searched image data is searched from the decompressed image data storage area according to the coordinates of the image data, and after the to-be-searched image data is stored to the decompressed image data storage area, the searched image data is sent to a subordinate data utilization module. According to the method, the computer device and the computer readable memory medium, the bandwidth cost of the memory in an image coding process can be reduced.

The invention relates to a method for importing data into multiple Hadoop components simultaneously, and is characterized in that the method comprises the following steps of (1) extending an import tool of the Sqoop, and adding service for importing data to the Kafka; (2) importing configuration parameters of the components according to a database, and writing parameter verification program; and (3) extending the import tool of the Sqoop, and adding service for simultaneously exporting the data to the HDFS, Hive, Hbase and Kafka. On basis of original functions of connecting the database and reading the data of the Sqoop, the function of simultaneously exporting the data to the multiple components is added, the database data are read for one time, multiple user-specified export modules are started simultaneously, and efficient and convenient data import is achieved. On one hand, multiple export tasks are prevented from being written for the same data, on the other hand, the same data are prevented from being repeatedly read, and therefore efficiency is improved.

The application discloses an information output method and device. A specific embodiment of the method includes: obtaining a pending order set pre-stored in the cache; reading the inventory information set from the inventory database for storing inventory information into the cache; for each order in the order set, according to the cache Each inventory information stored in determines the storage location of the item matching the order and the inventory quantity of the item matching the order; for each order in the order set, if the storage location of the item matching the order stored in the inventory database corresponds to If the inventory of the item is not less than the inventory of the item matching the order, modify the inventory information in the inventory database according to the storage location of the item matching the order and the inventory of the item matching the order, and output the storage of the item matching the order The location and stock availability of items matching this order. This embodiment reduces the times of reading the inventory database and improves the positioning efficiency.