45results about How to "Make full use of computing power" patented technology

The invention discloses a computer multiscreen expanded display method and aims to provide a display method for implementing support for a plurality of screens in a more flexible mode and under the low cost. The computer multiscreen expanded display method comprises the following steps of: simulating a video processing or display operating function of a display card in an operating unit of a CPU (Central Processing Unit) or an operating unit of a GPU (Graphics Processing Unit) of a computer, decoding an image displayed in each expanded display, acquiring display encoded data of each frame of image to be displayed on each display and sending out the display encoded data; by a display output card, receiving the display encoded data, splitting the display encoded data and then respectively packaging and outputting the display encoded data to the displays corresponding to the split display encoded data; and by each display, receiving and displaying the correspondingly split display encoded data sent by the display output card. The invention also discloses a computer multiscreen expanded display system. The computer multiscreen expanded display method and the computer multiscreen expanded display system can be used for multiscreen display.

The invention discloses a method for directly analyzing the Android package (APK) file information of an Android installation file on a computer. The method comprises the following steps of: analyzing the software information of an APK file; and analyzing the identification information of a manufacturer. In the step of analyzing the software information of the APK file, the APK file is decompressed in a Zip mode, an Android Manifest.xml file which is acquired by decompression is subjected to inverse operation, and an original extensible markup language (XML) file which is obtained by inverse operation is analyzed. In the step of analyzing the identification information of the manufacturer, the APK file is read in a JarFile mode, all elements are traversed from META-INF, and each element is operated in a JarEntry mode to acquire a digital certificate of the element. By the method, the APK file is directly analyzed on the computer, so the method is easy to operate, the operation capacity of the computer is fully utilized, batched and concurrent analysis of the APK file information can be realized, and the analysis computation efficiency of data information is greatly improved.

The invention relates to a distributed type-B ultrasonic inspection system and a processing method for ultrasonic detection signals thereof, wherein the system comprises a type-B ultrasonic image processing center for communication through a communication network, and a plurality of type-B ultrasonic detecting terminals; the type-B ultrasonic image processing center comprises a communication module, a user and task management server, a task input/output buffer area and an image processing server; the communication module is connected with the user and task management server, and the user and task management server, the task input/output buffer area and the image processing server are mutually connected; the type-B ultrasonic detecting terminals comprise probes, probe control modules, a cache, communication modules, detection terminal control modules, display modules and control panels; the probes, the probe control modules, the cache and the communication modules are connected in turn, and the detection terminal control modules are connected with the probe control modules, the cache and the communication modules, the display modules and the control panels respectively; and the system and the method have the advantages of a host with shared computing power, high resource utilization rate, flexible use, low cost, strong expanding capability, and the like.

The invention relates to the field of model construction, and a filter and a feature map are segmented according to a preset rule when a convolution step length is not 1, and the convolution is enabled to be equivalent to a convolution with a step length of 1 so as to be adaptive to a pulsation array. The invention particularly discloses a neural network acceleration method and device based on a pulse array, computer equipment and a storage medium. The method comprises the steps of acquiring convolution parameters of a convolution filter; if the convolution step length is not 1 and the size ofthe convolution filter is greater than 1 * 1, segmenting a plurality of sub-filters from the convolution filter according to a preset filter segmentation rule; obtaining a to-be-convolutional featuremap and segmenting a plurality of feature sub-maps from the to-be-convolutional feature map according to a preset feature map segmentation rule; based on the pulsation array, carrying out convolutioncalculation on the characteristic sub-graphs corresponding to the sub-filters according to the sub-filters, wherein the step length of convolution calculation is 1; and superposing convolution calculation results corresponding to the sub-filters, and outputting the result of superposition as a result of convolution calculation of the to-be-convolutional feature map by using the convolution filter.

The invention discloses an intelligent home system and an application method thereof. The system comprises a television mainboard, a communication device, intelligent home terminals, an artificial intelligence computing power strip, a modem and a television display. A network television becomes a novel home entertainment and intelligent home system center. Edge computing by taking a home as a unitis realized. The computing power of the network television is utilized fully. According to the intelligent home system and the application method, a control system of the existing network televisionis specially designed and computing tasks of a special system and communication contact of an internal special intelligent home can be received at the same time, so the network television becomes thenovel home entertainment and intelligent home system center. According to the control system, the internal artificial intelligence computing power strip can be controlled and the intelligent home terminals are uniformly formed into Internet of things through the communication device, so the integrated artificial intelligence control of the home is realized.

The invention discloses a simulated mouse pointer drawing method and device, equipment and a medium. The method comprises the following steps: creating a game process based on a game starting request sent by a streaming client; adding the process ID of the game process to a monitoring list in a kernel driving module preset on the target cloud host; obtaining main window information of the game process, and creating a transparent top-layer window on the upper layer of a main window display area of the game process based on the main window information, so that the transparent top-layer window serves as a moving area and a drawing area of a simulated mouse pointer corresponding to the game process; creating a simulated mouse pointer drawing thread corresponding to the game process, and initializing a Direct2D drawing environment; and calling the simulated mouse pointer drawing thread to draw the simulated mouse pointer on the transparent top-layer window according to the simulated mouse pointer information. In this way, each game process can have an independent user-visible simulation mouse pointer, and the simulation mouse pointers do not interfere with each other.

The invention relates to the field of audio mixing, and specifically relates to a video conference audio mixing method and device. The method comprises the steps that a master audio mixing resource receives audio code streams sent by slave audio mixing resources, wherein there is one or more slave audio mixing resources, and each slave audio mixing resource is used for carrying out audio mixing on audio code streams of at least two conference places; judging whether the master audio mixing resource is fully loaded or not; searching an idle slave audio mixing resource when it is judged that the master audio mixing resource is fully loaded; distributing partial audio code streams received by the master audio mixing resource to the idle slave audio mixing resource when the idle slave audio mixing resource is searched; and after audio mixing is carried out on the partial audio code streams by the idle slave audio mixing resource, sending the processed audio code streams to the master audio mixing resource. Through adoption of the method and the device, the problem that in the prior art, the number of the conference places capable of accessing audio mixing is liable to be limited by the audio mixing processing capability is solved, and the limitlessly extendable intelligent audio mixing of the conference places with extra large number of squares can be realized.

The invention provides a synchronous method, a host computer, a mobile terminal and a synchronous system. The method comprises the following steps: receiving synchronous data in the mobile terminal; performing synchronous operation for the synchronous data in the host computer and the synchronous data in the mobile terminal in the host computer to generate server data and client data; and deleting the server data. The embodiment of the invention can increase the synchronous speed.

The invention provides a task backup management method and device. The method comprises the following steps that a main node collects average load data of all backup nodes in a preset time period and selects the backup nodes corresponding to the average load data which is lower than a preset value as backup execution nodes; the main node conducts backup task splitting according to the average load data of the backup execution nodes and allocates backup tasks obtained after splitting to the corresponding backup execution nodes; and the backup execution nodes execute the allocated backup tasks obtained after splitting. Load balance among the nodes is achieved, the computing capacity of the nodes is fully utilized, and the backup efficiency is improved; and in addition, the scheme is automatically executed by a system, backup automation is achieved, and the backup process is simplified.

The dynamic chord number regulating method is used for an electronic device. The electronic device has S sounders, the first operation mode and the second operation mode with maximum number A and B separately, and it's A or B sounders set in on state, with S, A and B being positive integrals. The detected electronic device is first operation mode switched from the first one to the second one and the maximum chord number changed from A to B. The number of sounders with on state is then regulated from A to B.

The invention discloses a GNSS-based wireless antenna. The GNSS-based wireless antenna comprises an antenna body, a power dividing unit, a filtering and low-noise amplifying unit, a filtering and adjusting unit and a radio frequency front end processing unit, and further comprises a related processing module and a wireless transmission module. The antenna body is connected with the power dividing unit. The power dividing unit is connected with the filtering and low-noise amplifying unit which is connected with the filtering and adjusting unit. The filtering and adjusting unit is connected with the radio frequency front end processing unit which is connected with the related processing module. The related processing module is connected with the wireless transmission module. By means of the technical scheme, the GNSS-based wireless antenna which is convenient to use and can be operated remotely is manufactured, and receiving of wireless radio frequency signals and resolving of digital signals are separated. The antenna can be wirelessly connected with a base band receiving and processing unit through digital signals, flexibility is provided for outdoor workers or workers working under the dangerous environment, safety is improved, and the bodily injury to the workers is lowered.

The invention provides a dual-spectrum military vehicle CAN bus optical detector capable of accurately detecting fire alarms, which is suitable for military vehicles in harsh environments, including a power filter conversion circuit, an infrared light sensing circuit, an ultraviolet light sensing circuit, a temperature Compensation circuit, CPU interface circuit, CAN bus interface circuit, indicator light circuit and other structural designs. The invention integrates the CAN bus interface, can be directly connected to the vehicle electric information network, and transmits fire alarm, working status and other information to the upper computer in time. Due to the selection of photosensitive elements suitable for the two strong spectral ranges of the flame, the accuracy of fire detection is high and the ability to prevent false alarms is strong; the temperature compensation is carried out through CPU sampling and calculation, so that the detection sensitivity remains consistent in a wide temperature environment, The degree of intelligence has been significantly improved; when the vehicle is severely bumped or hit by a fire, causing a power supply or bus failure, it can still work normally.

The invention provides a grid point data processing method which comprises the following steps: creating a database table in a database according to a preset lattice point data structure, wherein the database table comprises a grid point data field and at least one attribute field; reading multiple groups of to-be-processed grid point data and at least one piece of attribute information corresponding to each group of grid point data from the data set; for each group of grid point data, packaging the grid point data into a grid point data object according to a preset grid point data object structure; and for each group of grid point data, writing the grid point data object into a grid point data field of a data record of the database table, and respectively writing the at least one piece of attribute information into an attribute field corresponding to the at least one piece of attribute information in the data record. Corresponding to the grid point data processing method, the invention further provides a grid point data processing device, electronic equipment and a computer readable storage medium.

The invention provides a task scheduling method and device, a household appliance and a computer readable storage medium. The task scheduling method comprises the steps of determining an execution right of a target task preempting a processor; and controlling the processor to read the data stored in the information storage space of the target task so as to execute the target task. The target taskhas the information storage space and belongs to the static array structure, so that task scheduling can be achieved on the basis that the processor does not support the dynamic memory allocation function, and particularly, when the target task preempts the execution right of the processor, namely, the target task can enter the execution state, the processor is controlled to read the data stored in the information storage space of the target task, the data of the information storage space of the target task is recovered to the processor to execute the target task, and the chip computing poweris fully utilized based on the control process.

The invention provides a treatment method for enterprise atmospheric pollutant emission reduction grading evaluation, which comprises the following steps: according to the type of a to-be-treated enterprise, obtaining at least one atmospheric pollutant in waste gas discharged by the to-be-treated enterprise; from a data storage device, acquiring data of the generation amount and the emission amount of each atmospheric pollutant in the to-be-treated waste gas emitted by the enterprise; determining the removal efficiency corresponding to each atmospheric pollutant according to the data of the generation amount and the emission amount of each atmospheric pollutant; according to the removal efficiency corresponding to each atmospheric pollutant, determining a score value of the emission reduction capability of the to-be-treated enterprise; according to the score value of the emission reduction capability of the to-be-processed enterprise, determining the grade of the comprehensive emission reduction capability of the to-be-processed enterprise; and outputting an atmospheric pollutant treatment strategy matched with the to-be-treated enterprise according to the grade of the comprehensive emission reduction capability of the to-be-treated enterprise.

The invention discloses a distributed image analysis method, which comprises the following steps that a first server acquires a result set through image collision analysis, wherein the result set comprises an index image recording a result object, the index image corresponds to an object frequency, and the object frequency is used for indicating the shooting times of the result object; a pluralityof first servers respectively send result sets to a second server; the second server performs feature extraction on the index image in each result set to obtain a feature value of each index image; the second server performs image collision analysis on the extracted feature values of all the index images to obtain the confidence coefficient of each index image; when the second server determines that the confidence among the index images in the result set sent by the plurality of first servers is greater than or equal to a preset value, the sum of object frequencies corresponding to the indeximages is obtained.The embodiment of the invention further provides a distributed image analysis system and a storage medium. Calculation power of all levels of servers is fully utilized, and analysisefficiency is improved.

The embodiment of the invention discloses a data processing method and device, equipment and a storage medium, and the method comprises the steps: enabling a main control program of test equipment tomap a transmission control protocol port of the test equipment into a socket of each terminal after the test equipment builds hardware connection with the terminal, establishing a transmission controlprotocol connection with a background program of the at least one terminal; enabling the main control program to use the transmission control protocol connection to receive terminal information reported to the test equipment by a background program of the at least one terminal; and enabling the main control program to realize data processing of the terminal according to the terminal information.

The invention discloses an operation card driver implementation method based on an application layer. The operation card driver implementation method comprises the following steps of 1, finishing initialization operation in a kernel space; and 2, completing all business processes in the application layer space. Compared with the prior art, the method has the positive effects that a hardware DMA memory and a control register of the operation card are mapped to a user space, an interface is provided to allow an application program to directly put data into a DMA area, and operation card hardwareis allowed to be controlled. After operation is completed, a completion function provided by an application program in advance is called asynchronously, and copying operation of a large amount of data is avoided. And an asynchronous completion mode is used, so that a plurality of complex operations are allowed to be realized in one thread, and CPU consumption caused by excessive CPU task switching is avoided. According to the method, the operational capability of the CPU is fully utilized, and the overall performance of complex operation is improved.

The invention discloses an edge artificial intelligence computing task scheduling method based on a peer-to-peer network, which comprises the following steps: broadcasting a request packet to other Internet of Things nodes of the peer-to-peer network according to an artificial neural network model, receiving reply packets which are sent by the other Internet of Things nodes and respond to the request packet, establishing a routing table between the Internet of Things nodes according to the reply packet, establishing associated information of the Internet of Things nodes and the calculation task of the artificial neural network model according to the routing table, and broadcasting a calculation task allocation packet according to the routing table, thereby realizing distributed calculationof the calculation task of the artificial intelligence calculation model at the edge end. The computing power of the Internet of Things node is fully utilized, the processing speed of the artificialintelligence computing task is improved, and various defects of completing the computing task of the artificial intelligence computing model at the cloud end are avoided.

The invention provides an arithmetic device, comprising: an arithmetic control module for determining block information; The operation module is used for dividing, transposing and combining the operation matrix according to the block information to obtain the transposing matrix of the operation matrix. The present disclosure also provides an arithmetic method. The invention discloses an arithmeticdevice and a arithmetic method, which reduces the overhead of storage resources and computing resources and improves the arithmetic speed.