98results about How to "Reduce the amount of stored data" patented technology

The invention provides a data processing method based on a distributed storage system. The method comprises the following steps: defining a compressed configuration item that comprises at least a compression mode, off-line compression start time and off-line compression termination time; when to-be-processed data is written in, using a client or a server, or the client and the server to successively perform chain type compression processing on the to-be-processed data, and determining the time of the server intervening to perform the chain type compression processing only in the off-line compression mode according to the off-line compression start time and the off-line compression termination time; when the to-be-processed data is read, at least enabling the client to perform chain type decompression processing; after the chain type compression processing or the chain type decompression processing is performed, responding to an opposite end device through a network. After the data processing method based on the distributed storage system is used, the bandwidth utilization rate is increased, the memory space is saved, the data reconstruction speed is accelerated, and the data handling capacity of the distributed storage system is improved when a write operation or read operation is performed.

The invention discloses an SS Table file data processing method and an SS Table file data processing system, wherein an SS Table file is provided with a framework table, a column sequence and column attribute information of row data of the SS Table file are defined in the framework table. The method comprises the following steps of reading row data to be written into the SS Table file; writing a row major key into the SS Table file, and writing each column data of the read row data corresponding to the row major key according to the column sequence and the column attribute information of the row data which is defined in the framework table of the SS Table file. When the method and the systemare adopted to carry out data processing on the SS Table file, data reading and writing processing can be carried out according to the framework table, so that storage of row data only is realized by only storing the row major key and column values, the column values are stored according to the prescribed sequence in the framework table, and information such as column names or line IDsis needed to be stored, so that data storage volume can be reduced.

The invention discloses an identity authentication method based on biological characteristics and relates to the technical field of biological characteristic data processing. The method comprises the steps of obtaining and reconstructing identity information of a registered user; collecting biological characteristic information of the registered user and generating biological characteristic data; through adoption of a preset key generation method, generating an encryption key through utilization of the biological characteristic data of the registered user; encrypting the reconstructed identity information through utilization of the encryption key, thereby generating a ciphertext; collecting the biological characteristic information of a to-be-authenticated user and generating the biological characteristic data; generating a decryption key through utilization of the biological characteristic data of the to-be-authenticated user; decrypting the ciphertext through utilization of the decryption key and generating a plurality of pieces of decrypted data; and screening the decrypted data consistent with a reconstruction rule, carrying out reverse reconstruction to generate to-be-identified data, screening the to-be-identified data with the highest occurrence frequency as the identity information of the to-be-authenticated user, and judging the user as the registered user. The invention also discloses another identity authentication method based on the biological characteristics.

The invention discloses a portable overhead line geometric parameters detection method and a portable overhead line geometric parameters detection system based on a binocular linear array CCD camera. The method includes the steps of: 1, disposing two imaging devices on two tracks, disposing a line light source at the center between the two imaging devices; 2, transmitting linear illuminating light towards the overhead line, wherein generated reflected light respectively enter the two imaging devices for imaging; 3, calculating the geometric parameters of the overhead line. The system is used for implementing the above method. The portable overhead line geometric parameters detection system has the advantages of being simple in structure, convenient to operate and carry and capable of working all the time.

The invention discloses a video data storage method, device and system and a cloud storage-based video access method and device, and relates to the technical field of video data processing and storage. The video data storage method is applied to a cloud storage system, and the cloud storage system comprises a management node and a storage node. The method comprises the steps that the management node detects the available residual storage capacity; the management node compares the available residual storage capacity of the storage node with a first threshold, a second threshold and / or a third threshold; when it is judged that the available residual storage capacity is lower than the first threshold, the storage node receives and stores double code rate video data generated by front-end equipment; and according to the condition of the available residual storage capacity, the storage node deletes a stored part or first code rate video data and / or second code rate video data in all the double code rate video data. According to the method, video data with the higher definition can be saved when the storage space is enough, and longer-time video data can be saved when the storage space is not enough.

The invention provides a delay parameter computation method for a multi-channel scanning, which unifies delay computational formulas of three probes such as a linear array, a convex array and a phased array, and the formulas are applicable to non-deflection scanning and deflection scanning; and scanning lines with different densities can be calculated only by changing location variables of the scanning lines. In the invention, only an array element position, a scanning starting point position and relative focus parameters need to be stored, thus having small quantity of stored data; real-timecomputation can realize accurate delay without a large quantity of the stored data; delay in various situations for the probes, the density and deflection is calculated using a unified circuit to greatly simplify a circuit structure; and delay of each scanning line is independently controlled to conveniently realize arbitrary density scanning, for if no integral multiple relationship exists between spacing among the scanning lines and spacing among array elements of the probes, different scanning lines have different delay data. In the invention, the scanning lines are independently controlled, thus achieving arbitrary density scanning.

The invention provides a DVR monitoring system supporting single-channel dual-stream video recording and a method thereof. A video and audio acquisition channel acquires video and audio data, and the video and audio acquisition channel comprises analog video cameras or IPCs; a DSP encoder encodes analog signal input of the video and audio data acquired by the same path of analog video cameras into mainstream data with higher resolution and sub-stream data with lower resolution, or repackages the mainstream and sub-stream data acquired by the same path of IPCs into the mainstream data with higher resolution and the sub-stream data with lower resolution; the mainstream data with higher resolution are cached in a mainstream video recording buffer area, and the sub-stream data with lower resolution are cached in a sub-stream video recording buffer area; and a mainstream and sub-stream video recording scheduling module schedules the mainstream data with higher resolution and the sub-stream data with lower resolution to be stored to video recording storage equipment. Hard disk data storage overhead can be reduced so that network bandwidth can be effectively reduced, and overhead of a client-terminal decoder and the equipment performance requirement can be reduced.

The invention provides a drive method for reducing the response time of twisted and super-twisted LCDs. According to the method, the data of different dynamic frames pass through a dynamic frame determination mechanism behind an input tricolor data bus of red, green and blue. Overdrive compensation potentials higher or lower than prior output potential are given flexibly N times within the time for each data bus to be updated. In this way, liquid crystal is rotated to reach the expected object brightness within a shorter time, thus significantly improving the problem of blurred images caused by too slow response of passive matrix-type LCDs of twisted nematic type (TN), super-twisted nematic type (STN) and so on.

An image processing device includes a statistical information acquiring unit, an optical information acquiring unit, a filter information calculating unit and a filter coefficient calculating unit. The filter information calculating unit obtains filter information of a restoration filter for point image restoration processing according to at least one of statistical information and optical information. The filter information includes information related to number of taps of the restoration filter and information indicating a kind of symmetry of the restoration filter. The filter coefficient calculating unit calculates a filter coefficient of the restoration filter according to the statistical information and the optical information with at least the information related to the number of taps of the restoration filter and the information indicating the kind of the symmetry of the restoration filter of the filter information as a constraint condition.

The invention provides an advancing method and device of a movable robot. The method comprises the following steps that at least one walked path which the movable robot passes through is determined; information of a pair of end points of the two ends of the path corresponding to each walked path is determined; the current position where the movable robot is located is determined; and according tothe current position and the determined information of the end points, the to-be-advanced path of the movable robot is determined. Through the scheme, the data storing quantity and calculating quantity of the movable robot in the process of to-be-advanced path determination can be decreased.

The invention discloses a cartoon manufacture method and device. In the invention, digital image processing technology and fractal geometry principle are used for parting off cartoon components from an image; dynamic and logic cartoon characteristics and the like are extracted according to dynamic geometry principle and combinational logic principle; moreover, a dynamic logic relationship is usedfor representing cartoon characteristics to describe the cartoons through formal languages; and then cartoon files with different styles can be manufactured and generated according to the cartoon components, the cartoon characteristics and the logic relationship. Compared with a traditional cartoon manufacture method, the invention greatly reduces the requirements on storage space, and is suitable for being realized on a wireless data transmission system with limited bandwidth and an embedded terminal with limited calculation capacity and store resources.

The invention provides a convolutional neural network system and a convolutional neural network quantization method, and the system comprises: a quantization module which is used for carrying out thequantization of the input data of an ith convolution layer of the system and the weight and bias of the ith convolution layer, and wherein i is a positive integer; and a convolution module that is used for carrying out convolution calculation on the quantized input data of the ith convolution layer, the quantized weight and the quantized bias to obtain a convolution result of the ith convolution layer. According to the convolutional neural network system provided by the invention, the weight and bias of the convolutional layer needing to be quantized in the convolutional layer and the input data input into the convolutional layer are quantized, and convolution calculation is performed by using the quantized input data, the quantized weight and the quantized bias so as to obtain the calculation result of each convolutional layer. The calculation amount of the convolutional neural network is reduced, and the quantization precision of the convolutional neural network is improved.

The invention discloses a writing note data storage method and device, a terminal and a storage medium. The method comprises the following steps of receiving a writing operation of a user on a canvas;displaying a corresponding writing handwriting on the canvas according to the writing operation; and when a storage operation used for storing the writing handwriting on the canvas is detected, generating a writing handwriting file according to the canvas and the writing handwriting, so that when the writing handwriting file is opened, the corresponding writing handwriting is displayed and the writing handwriting file is transmitted, wherein the writing handwriting file comprises at least two levels, different data blocks in each level are separated by a special character, and special characters of different levels are different. According to the method and the device, the data amount of the generated writing handwriting file can be reduced on the premise of keeping the edibility of handwriting data, so that the required storage space is reduced.

The invention discloses a Gaussian filter with the self-adaptive data rate. The Gaussian filter with the self-adaptive data rate is a digital Gaussian filter based on an application look-up table method and formed by an N-system counter, an M-N decoder, a Gaussian ROM and a parallel recombiner. Particularly, according to the Gaussian filter with the self-adaptive data rate, the input end of the N-system counter is connected with output of a self-adaptive frequency divider, input of the self-adaptive frequency divider is a bit rate of a clock signal and an input signal, frequency division is performed on the clock signal to make the clock signal be N times of the bit rate, output of the self-adaptive frequency divider and a base band signal are jointly connected into synchronization frequency, a phase and the starting and stopping time of a synchronizer, and output of the synchronizer and output of the N-system counter are jointly connected into a caching device; output of three continuous code elements of the caching device is connected into the Gaussian ROM, and meanwhile the middle code element output of the caching device is connected into the parallel recombiner. When the digital Gaussian filter is applied for performing modulation, the number of sampling points, the counter, the decoder and the like can be kept unchanged according to the input signals of different bit rates, and the digital Gaussian filter has universality and is small in amount of stored data.

The invention provides an intelligent lamp, a brightness and color temperature adjusting method thereof and a computer readable storage medium, and the method comprises the steps: obtaining a set target brightness value and a set target color temperature value, and obtaining a first actual brightness value and a first actual color temperature value of ambient light; obtaining a brightness adjustment change curve of the intelligent lamp according to the target brightness value, the target color temperature value, the first actual brightness value and the first actual color temperature value, and obtaining a color temperature adjustment change curve of the intelligent lamp; and adjusting the brightness change of the intelligent lamp according to the brightness adjustment change curve, and adjusting the color temperature change of the intelligent lamp according to the color temperature adjustment change curve. The invention further provides the intelligent lamp for realizing the method and the computer readable storage medium. Intelligent adjustment of the brightness and the color temperature of the intelligent lamp can be achieved, adjustment of the brightness and the color temperature is more accurate, sudden change of the brightness and the color temperature is avoided, and discomfort of a user is reduced.

The invention discloses a machine tool manufacturing BOM storage query and tree structure construction method based on an MES system. The method comprises the steps: constructing a manufacturing BOM storage structure; setting a cache data structure to search part material numbers; setting a BOM reverse query process data structure, and searching part numbers through a coding rule for subsequent manufacturing BOM search; respectively searching for a first-level component manufacturing BOM and a multi-level component manufacturing BOM, removing same-level repeated data and non-same-level repeated data, and searching for the components; searching a machine tool manufacturing BOM directly containing a specific part and a machine tool manufacturing BOM containing a group part, removing the same-level repeated data, removing the non-same-level repeated data, and searching to obtain machine tool manufacturing BOM data; according to the method, building a tree structure according to machine tool manufacturing BOM data, providing a basis for subsequent material updating and correcting on a production site, and achieving BOM reverse checking on the production site; and performing result display on the tree structure through an MES system. According to the invention, the workload is reduced, the query efficiency is improved, fast storage, fast taking and fast updating are achieved, and the query speed is increased.

It is judged whether or not impositioning is applied to an image, based on layout information. If the image is the imposed one, layout information (storage layout information) for each page constituting one impositioning and output joint information for impositioning based on the storage layout information are generated. A page raster for each page constituting an imposed image is saved without duplication.

The method provides a method for reducing a real-time data transmission amount of a color image, which is used for a color CCD camera to acquire and transmit image data to a computer. The method comprises the following steps: by use of a camera development instruction, setting the color CCD camera to be at a monochromatic original raw data mode (i.e., a gray scale mode); operating the camera to acquire in real time original raw data whose images, transmission and storage are all single channeled; and finishing acquisition and transmission, the computer performing color interpolation calculating on stored monochromatic original raw data to obtain a three-channel color image after each pixel is interpolated. The operation and calculating are simple and feasible, and during a transmission process, a three-channel data amount is reduced to a single channel, i.e., the amount is reduced to be one third of an original amount, so that real-time transmission and storage data amount can be effectively reduced, hardware pressure is relieved accordingly, and the problem of frame loss is solved. Therefore, the method provided by the invention is a simple and feasible method which exerts no effects on data, and can be applied to the field of color image real-time acquisition of a large data mount.

A method for extending EC ability of a cycle code includes following steps: a, traversing error patterns with the code weight less than 1+1 and computing out the adjoint formula corresponding to one error pattern, b, storing the patterns and adjoint formulas in filel, c, traversing error patterns with 1+1 code weight and computing out adjoint formulas corresponding to the error patterns, d, comparing computed adjoint formulas with those stored in filel, e, operating the same way on error patterns with 1+1 code weight till traversing all error patterns with 1+1 code weight, f, comparing each adjoint formula in file 2 with the rest in file2, g, processing the next in the same way till traversing all formulas finally to array and store the error patterns from small to big in terms of the adjoint formulas.

The invention discloses an intelligent checkerboard with the track recording function. The intelligent checkerboard includes a checkerboard body and several chess pieces. Each chess piece is internally provided with a corresponding unique RFID label, 90 chess positions arranged in an arrayed mode are arranged on the checkerboard body, the positions, corresponding to the chess, in the checkerboardbody are provided with antennas used for emitting radio-frequency signals, the checkerboard body is internally provided with a processor and nine card readers, each card reader scans ten antennas, scanning of the chess positions of the whole checkerboard is completed within 500 ms, the card reader is electrically connected with the signal end of the processor, collected data is pre-processed, andthe processor is provided with a memory module; when the chess positions are changed, the chess positions and the current time are stored in a non-volatile memory, the radio frequency identification technology is used for recording the change of the position of each chess, the storage data volume is small, response is rapid, real-time display can be conducted through a displayer, data can be uploaded through wireless communication for storage, the storage cost is effectively reduced, and the technology exchange and analysis about chess are facilitated.

The embodiment of the invention discloses a data storage method, device and equipment for a solid state drive and a computer readable memory medium. The method comprises the steps of storing change quantity of management data to first storage areas when the management data of LBAes is changed in a solid state drive operation process; copying the management data corresponding to the LBAes from DDR(Double Data Rate) to second storage areas corresponding to the LBAes according to preset data copy quantity when residual space of the first storage areas of the LBAes satisfy a preset condition, andsending a notice for flashing data in the first storage areas and the second storage areas of the LBAes to an NAND management module; and flashing the data in the first storage areas and the second storage areas of each LBA to the NAND when the solid state drive is powered off, and flashing L2P tables of the updated management data to a control manager, wherein the data in each LBA is obtained bypre-dividing the management data of the solid state drive according to preset data storage quantity. According to the method, the device, the equipment and the computer memory medium, the data quantity needing to be stored when the SSD is powered off is reduced, and the data integrity in the SSD is ensured.

The embodiment of the invention provides a data query method and device, and electronic equipment. The method comprises the following steps: presetting an index of a list of which each chain length isgreater than a preset threshold in a database; receiving a query request input by a user; wherein the query request carries a query condition and a target identifier; determining a query list in thedatabase according to the target identifier; wherein the query list is a list corresponding to a user identifier matched with the target identifier; and when the chain length of the query list is greater than a preset threshold and the query condition conforms to a first preset condition, querying in the query list in an index manner to obtain a query result. According to the method of the invention, an index does not need to be set for each list in the database, so that the storage data volume of the indexes is reduced, and the storage space is saved. And the query is performed in the list ofwhich the chain length is greater than the preset threshold in an index manner, so that the query time can be greatly shortened, and the query performance is improved.

The invention provides an image compression method. The image compression method comprises the following steps: raw data of Ni frame of newly acquired images and raw data of Ni-1 and Ni-2 frames of images acquired before are compared one by one according to corresponding pixels; an identification code for representing the comparison result is output; raw data of pixel values in Ni frame of the new images are modified according to the identification code and saved. The image compression method adopts the continuous 3-frame images for comparison to eliminate influences caused by tiny data fluctuation, reduce the data storage amount, improve the data compression proportion and guarantee complete data storage so as to meet the technical requirement that mainstream hard discs on the market complete data storage at the present write-in speed. The image compression method is simple to use, and can be widely applied to various configured PCs and embedded or FPGA systems.

The invention provides a moving track information storage method, a server and a computer storage medium, and mainly aims to reduce the storage data volume of the moving track information on the server. The method comprises the following steps that a server receives indication information, wherein the indication information is used for indicating a first UE and a second UE to form a group to perform a mobile activity; the server receives first movement track information reported by the first UE and second movement track information reported by the second UE; when the server determines that thefirst moving track information and the second moving track information are equal moving tracks, the server stores third moving track information, and the third moving track information is obtained bystoring one of the first moving track information and the second moving track information; when the server receives a calling request sent by the third UE, the server sends third movement track information to the third UE; and the server sends the movement track information of the first UE and the second UE except the third movement track information in the movement activity to the third UE.

The invention discloses a vector map feature expression-based large-scene unmanned aerial vehicle image rapid positioning method, and belongs to the field of unmanned aerial vehicle positioning and navigation. According to the method, vector patches of four typical targets including a football field, a basketball field, a building monomer and a water body area are used as an unmanned aerial vehicle image positioning reference base map, and feature vector expression is performed on each vector patch; the method comprises the following steps of: quickly identifying four types of targets, namely a football field, a basketball field, a building monomer and a water body area, of a high-resolution remote sensing image acquired by an unmanned aerial vehicle, and performing feature vector calculation on a central patch of the image; and finally, feature vector matching is carried out, and rapid matching of the unmanned aerial vehicle image and the base map data is realized. According to the method, the data volume carried by the unmanned aerial vehicle can be greatly compressed, the image matching efficiency of the unmanned aerial vehicle can be greatly improved, and rapid positioning of the image of the unmanned aerial vehicle under the condition of large-scene GNSS signal rejection is realized.

It is judged whether or not impositioning is applied to an image, based on layout information. If the image is the imposed one, layout information (storage layout information) for each page constituting one impositioning and output joint information for impositioning based on the storage layout information are generated. A page raster for each page constituting an imposed image is saved without duplication.

The invention relates to the technical field of virtual reality, in particular to a three-dimensional fitting simulation system. In order to solve the problems that in the prior art, the clothes wearing effect can only be observed from several fixed angles such as the front face and the back face, clothes deformation is caused due to lack of figure changes, and consequently size selection is inaccurate, the technical key points are that the system comprises a client used for inputting registration information of a user and generating a corresponding unique identification code; the data acquisition system which is used for reading the unique identification code of the client and acquiring human body data information of a user; and the server side which is in communication connection with the client side and the data acquisition system, and is used for receiving the registration information sent by the client side and the human body data information sent by the data acquisition system, establishing a three-dimensional human body model, and feeding back an effect picture obtained after the three-dimensional garment model selected by the user is combined with the three-dimensional human body model to the client side. According to the invention, the effect that the clothes are worn on the human body can be visually presented so as to meet the ever-increasing demand of people for continuous improvement of online shopping activities.