173results about How to "Guaranteed running speed" patented technology

The invention discloses a human face detection method and a detection device based on multi-task cascade-connection convolution neural network, wherein the method comprises: establishing a cascade-connection multi-level convolution neural network; using the human face front samples, the human face back samples, some parts of the human face and the human face's key point samples as the training samples to train the multi-level convolution neural network to learn the tasks of human face categorizing, human face area position regression and human face's key point positioning; and utilizing the well trained multi-level convolution neural network to make human face detection from the to-be-detected image wherein in the training stage, both the online manner and the offline manner are combined to extract the human face back samples as the training samples. According to the invention, based on the cascade-connection multi-level convolution neural network, it is possible to learn the characteristics with stronger robustness, and at the same time, through the combination of the online manner and the offline manner to extract the back samples, the categorizing capability of the network is enhanced, so that the detection capability and the accuracy of the network are increased, and that the running speed of the method in an actual product is ensured.

The invention provides a method for constructing a dense small target detection model, the model and the detection method. Based on the information fusion of the target midpoint context, through cutting the picture with high resolution, the method avoids the picture sampled under the input network from losing too much image information, and affects the network feature extraction. The residual pyramid feature extraction network is used to fuse the features of different scales, which improves the detection accuracy of the network for different size targets, especially for small targets. A RoIAlign layer instead of RoIPooling layer is used to solve the position deviation of candidate frames caused by feature mismatch of candidate regions. Because the small target features are easily lost in the network transmission, the center point context features are fused with the original RoI features, so that the network can make full use of the target context information, ensure the network runningspeed, more accurately locate and identify the dense small target, and improve the network performance.

The invention discloses an application process starting control method and device. The application process starting control method comprises the steps of obtaining process information of an application process to be started, matching a process analysis list according to process information to obtain a matching result, and forbidding automatic starting of the application process when the matching result indicates that starting of the application process is not allowed. Due to the fact that automatic starting of application processes which are not needed in a terminal is avoided by matching the process analysis list, internal storage resources of the terminal can be saved, the operation speed of a terminal operating system is ensured, and accordingly the terminal usage experience of a user is improved.

The invention provides a banknote management method. The method comprises that a banknote information processing device is used to collect, identify and process banknote features and further to obtain banknote feature information; the banknote feature information, business information as well as information of the banknote processing device is transmitted to a master control server; and the master control server processes the received information and classifies banknotes. The invention also provides a corresponding banknote management system. According to the method, the operation speed is ensured, the identification robustness is improved, and accuracy and practicality in practical application are ensured.

The invention provides a prefix number identification method and system. The method comprises: extracting the gray level image of the area where a prefix number is located, and performing edge detection on the gray level image; rotating the image; performing fine mapping on a single number in the image, tightening characteristics included in the image of a positioned number, and performing normalization processing on each number image, normalization including dimension normalization and light and shade normalization; and employing a neural network to identify a normalized number image to obtain the prefix number. The method can guarantee an operation speed, meanwhile improve identification robustness, and guarantee accuracy and practicality in practical application.

The invention provides a method of realizing a medical data external query in digital telemedicine, comprising a general medical data exchange interface, an encryption and decryption module, a database and a classified query module. The invention realizes that in the digital telemedicine, external searches for the medical data of a hospital safely and conveniently via cooperation of each module, and at the same time also reduces load of the database of the hospital information system.

The invention discloses a universal debugging interface-based SoC (System on Chip) hardware debugger, which comprises a universal debugging interface, a TAP controller, a scan chain, a debugging control register module, a synchronizing module, a JTAG-AHB (Joint Test Action Group-Advanced High-performance Bus) protocol conversion module, an AHB-DMA (Advanced High-performance Bus-Direct Memory Access) module, an AHB bus controller, a debug support unit (DSU), an SoC peripheral module and a processor for supporting debugging. According to the universal debugging interface-based SoC hardware debugger, a JTAG port is applied to the DSU of a GAISLER laboratory open source for the first time, the scan chain is prevented from being plugged into a key path of the processor, and the structure and the performance of the SoC hardware debugger are superior to those of the normal hardware debugging modules at present. The SoC hardware debugger has beenis subjected to tape-out verification to prove that the function is correct and reliable.

The invention belongs to the technical field of remote sensing image ground object change detection, and particularly relates to a remote sensing image ground object change detection method based on multi-task learning, which is used for solving the problem of error accumulation caused by the fact that the ground object change detection precision depends on the ground object classification precision in the prior art. A multi-task learning ground object change detection model is adopted and comprises two semantic segmentation model branches and a change detection model branch. A semantic segmentation model is constructed through a segmentation network, a feature extraction module of the model can effectively extract features of a remote sensing image, then a twin network is constructed to train a ground object change detection model, and a multi-task learning mechanism is constructed. In conclusion, the ground object change detection area can be determined, change detection results of different ground objects and ground object types before and after area change can be obtained, meanwhile, the problem of error accumulation is avoided, and the change detection precision is improved.

The invention discloses a hydraulic machine and a double hydraulic system. The double hydraulic system comprises a first route of hydraulic system body and a second route of hydraulic system body. The double hydraulic system is of a logic linkage structure, and therefore the first route of hydraulic system body is used when the hydraulic machine is in the idle stroke (fast advancing) and the return stroke (fast retreating), and the large-flow rapid movement is achieved; high-pressure hydraulic oil is directly driven by the second route of hydraulic system body when the hydraulic machine is in the feeding stroke, and therefore the high-pressure system pressure can be directly obtained by the hydraulic machine, the higher working force can be obtained, and the overall weight of the hydraulic machine is reduced at a large proportion. By means of the double hydraulic system, the hydraulic machine can be applied under the power of 40 Mpa or higher, for instance, 40 Mpa, 55 Mpa, 70 Mpa, 100 Mpa, 120 Mpa, 150 Mpa and 200 Mpa, the overall weight of the hydraulic machine can be greatly reduced, all types of use efficiency, such as the running speed, of the hydraulic machine can be kept, and the production efficiency can be ensured.

The embodiment of the invention discloses a voice recognition method and device. The voice recognition method comprises the following steps: receiving an input voice segment; calculating multiple acoustical scores of each frame in the voice segment according to the spatial distribution clustering (SDC); comparing the obtained multiple acoustical scores with vocabularies in a vocabulary library in a mobile terminal, and accumulating the compared acoustical scores of each frame, wherein one has the highest aggregate score is named as the optimal acoustical aggregate score; taking the sum of the highest acoustical score in each frame of the voice segment as a background acoustical aggregate score; comparing whether the optimal acoustical aggregate score and the background acoustical aggregate score can meet a reserved threshold or not; if no, rejecting to recognize the voice segment. The method can show that larger difference values generated by the comparison of the sum of the highest acoustical score in each frame of the voice segment and the input voice segment as well as the sum of the highest acoustical score in each frame of the voice segment and the vocabularies in the vocabulary library in the mobile terminal, so that the frequency of the false operations of the mobile terminal according to the voice input is greatly reduced, and the user experience is improved.

The invention discloses a failure alarm method and system for a Lustre parallel file system. The alarm method comprises the steps that (1) a monitor module periodically scans an OSS (Open Source Software) log; (2) the log information is analyzed to judge whether failure information is included, and then an alarm report for the failure information is created; (3) whether failures in the alarm report are serious is judged, and a separate report for serious failures is created for further analysis by a program; and finally the serious failure information and common alarm information are collected to be written in a txt document which is transmitted to an administrator through a mail hub. The failure alarm system of the invention comprises a LAToolkit service terminal, a storage client cluster, a main server cluster and a LAToolkit client. Intelligent failure analysis is achieved through the control of the system disclosed by the invention, a concise report is created, the approximate situations of the failures can be known remotely through a mobile phone, the detection time is saved, the cost of the method and the system is low, and just using the original equipment instead of adding new equipment can achieve the purpose.

The invention provides a railway passenger train body. The railway passenger train body comprises a passenger train body bottom plate and passenger train body side walls arranged at two sides of the passenger train body; a sliding door is arranged at a through port on one of the passenger train body side wall; the passenger train body bottom plate is covered with an interior floor board; and a first gap is formed between the internal floor board and the top surface of the passenger train body bottom plate; a pedal is arranged in the first gap; a gear mechanism is fixedly arranged between the pedal and the sliding door; the gear mechanism comprises a transmission shaft capable of being self rotated and a first gear and a second gear which are respectively fixedly arranged on the transmission shaft; a first rack extending along the longitudinal direction is fixedly arranged on the sliding door; a second rack extending along the horizontal direction is fixedly arranged on the pedal; the first gear is engaged with the first rack, the second gear is engaged with the second rack, so that the second gear which rotates along with the transmission shaft can drive the pedal to move towards the external side of the passenger train body bottom plate. According to the railway passenger trail body provided by the invention, the pedal can be stretched out when the sliding door is opened, the gap between the station and the passenger train body is blocked, and thus, the safety on a railway passenger train and the speed of the railway passenger train are improved.

The invention discloses an intelligent acquisition method and device for vehicular photographing equipment, and a car digital video recorder. The intelligent acquisition method comprises the following steps of: performing photographing according to a photographing frequency so as to generate video data, and sampling the video data according to a sampling frequency so as to obtain sample images; calculating the brightness average value of the sample images, wherein the brightness average value of the sample images is an average value of brightness values of continuous X-frame sample images, and X is a positive integer; and determining a pre-set photographing frequency corresponding to the brightness average value according to the brightness average value, and performing photographing according to the pre-set photographing frequency corresponding to the brightness average value so as to generate the video data. According to the intelligent acquisition method disclosed by the invention, the brightness values of the continuous X-frame sample images in the video data are acquired, so that the brightness average value is obtained and used for adjusting the photographing frequency in real time; therefore, the better video data is obtained; and recorded data with relatively high referential meanings is provided for users.

The invention provides a method and a system for searching for a vehicle by a graph. The method comprises the following steps: a model for vehicle filtering is trained in advance, wherein the model comprises a vehicle type model, a color model, a sub-brand model, a license plate information model and a feature region classification model. A front overhead picture of a target vehicle to be detectedis obtained, and license plate information of the target vehicle is detected. The retrieval result is: a licensed vehicle or an unlicensed vehicle; according to the different conditions of the targetvehicle license plate information, the retrieval of the licensed vehicle or the retrieval of the unlicensed vehicle is started. A variety of models perform filtering first to reduce the impact of interference factors, and the feature region classification model is used for feature alignment ranking. Moreover, different feature comparison models are used according to the usage scenario. At the same time, according to the characteristics of vehicle samples, a feature comparison model is specially designed to ensure the balance among accuracy, memory occupation and running speed in the application process.

The invention discloses a method and system for on-demand loading of a TR069 parameter node in home gateway equipment, and relates to the field of home gateway service configuration management. The method includes the following steps: TR069 parameter configuration information in home gateway equipment that needs to be used is used to form a TR069 parameter configuration file, and load the TR069 parameter configuration file to a memory; a parameter cutting file is configured, and target node cutting processing flow is performed according to the parameter cutting file; a parameter loading file is configured, and target node loading flow is performed according to the parameter loading file; the TR069 parameter configuration file after cutting and modification is stored to a Flash file; and a memory object tree is initialized according to the TR069 parameter configuration file, finished. The method for on-demand loading of the TR069 parameter node in the home gateway equipment is relatively strong in flexibility, has few configuration parameters, does not exist a great deal of information redundancy, and is not only very convenient to maintain, but can also ensure normal starting speed and operation speed of the home gateway equipment, thereby facilitating use of people.

The invention discloses a multi-layer configurable mobile control system for a foot type climbing robot. The multi-layer configurable mobile control system comprises a user view layer for being interacted with a user, a hardware relation layer for being interacted with climbing rotor hardware, a path and gait layer for conducting path planning and gait management, a data center layer for completing data configuration structure storage and analysis, and a main control layer for dispatching operation of all the layers. The main control layer is in communication with the user view layer, the path and gait layer, the hardware relation layer and the data center layer, and the data center layer is connected with the user view layer, the hardware relation layer, the path and gait layer and the main control layer. The multi-layer configurable mobile control system is applied to the foot type climbing robot, is complete in function, can be freely configured according to different robots and is high in universality.

The invention belongs to the technical field of networks and discloses an application running method and an application running device. The method includes: receiving a starting instruction for a first plug-in application in a platform application; extracting application feature information of plug-in applications installed in the platform application; matching the application feature information of the plug-in applications in a malicious plug-in feature library to recognize malicious plug-in applications; searching and killing the recognized malicious plug-in applications; judging whether the plug-in applications installed in the platform application are all matched or not, and clearing redundant progresses in a terminal system after matching is finished; running the first plug-in application in the platform application after clearing is finished. The application running method and the application running device have advantages that in each running process of the first plug-in application, potential safety hazards caused by malicious plug-in applications such as malicious deduction, privacy stealing and information fraud can be effectively avoided to improve safety of the terminal system while electric energy can be saved, and internal storage occupied by the redundant progresses can be released, so that running speed of the first plug-in application is guaranteed.

The invention discloses a plant disease long-distance diagnostic device which comprises a main processor, a data collecting module, a wireless transmitting module and a data display module, wherein the main processor is provided with a video processing front end for converting and preprocessing format of images and a video processing rear end for digital output, the data collecting module is connected with the video processing front end for collecting image data of ill plants, the wireless transmitting module is connected with the data collecting module for transmitting collected image data of ill plants to a server to diagnose disease causes and obtain diagnostic results, and the data display module is connected with the video processing rear end for receiving and displaying the diagnostic results from the server. The plant disease long-distance diagnostic device has excellent versatility, and can be applied to the collection and transmission of plant image information and environment parameter in various occasions.

The invention relates to the technical field of multimedia information processing, and provides a video reorganization system capable of automatically partitioning a shot and a video reorganization method thereof. The video reorganization system is as follows: a partition processor is respectively connected to an original video library and a reorganizer; and a generation and storage module is connected with the reorganizer. The video reorganization method comprises the following steps that: a user opens the original video library and extracts a video in need of being processed from the original video library; the partition processor performs automatic shot partition of the video, such that an independent video processing unit is obtained; a single-video effect library, a replacement background module and a multi-video operation module of the reorganizer are used for correspondingly processing a video processing unit respectively; and the processed video is transmitted to the generation and storage module and then stored and played. The video reorganization system disclosed by the invention is convenient to install and low in system hardware requirement and is capable of realizing operations, such as automatic shot partition, single-video effect processing, background replacement and multi-video mixing, on the same platform; therefore, the video processing efficiency of the video reorganization system is ensured; and thus, the video reorganization system has perfect functionality.

The invention relates to a rail vehicle brake system (100) comprising a conditioning device (10), having at least one disc brake per axle (4), the conditioning device (10) having a control device (11). The conditioning device (10) is formed having at least one wagon control unit (14) and at least one coupling unit (16) for selective actuation of the at least one disc brake per axle (4). The invention further relates to a conditioning device (10) and to a method for operating a conditioning device (10), and to a method for deicing and / or for preventing icing of a brake unit (5) of a rail vehicle brake system (100).

The invention discloses a region growth detection method based on multiple judgments and belongs to the technical field of image processing. The region growth detection method is used for processing medical MRI (Magnetic Resonance Imaging) and CT (Computed Tomography) images and accurately extracting out the required tissue images. The region growth detection method provided by the invention comprises the steps: firstly, manually choosing an initial point of a target image to be detected, wherein the choosing range is a target range; then, gradually and forcefully expanding the growth range to all image data of the choosing range based on an original region growth scheme, optimizing matching of discrete points and detecting whether the discrete points belong to the growth region based on edge detection; continuously optimizing the growth results in a growth process through the initial point optimization standard gray to guarantee that the required large region images are matched and avoiding matching interference points which can meet the judgment conditions; finally, performing corrosion and expansion treatment on the growth result, removing the interference points, truly optimizing the whole matching process on the whole and obtaining the more accurate results.

The present invention provides a method for quickly generating high dynamic images based on local characteristics. The method includes the following steps: an image segmentation step, which divides images with different exposures in the same scene into blocks of equal size; a pixel-level weight generation step, which obtains The block-level weight of each block is smoothed by mean filtering to filter out the edges between the weight blocks; the image fusion step is to fuse images with different exposures according to the image fusion model through the generated pixel-level weights together. This method enables the fused high dynamic video image to contain information of all luminance ranges, and can ensure fast algorithm running speed.

The invention provides a multi-station ceramic tile transferring and stacking method and system based on pallets, and belongs to the field of ceramic tile production equipment. The multi-station ceramic tile transferring and stacking system based on the pallets is characterized in that the system comprises a stacking work area which is defined by fences (5), a pallet storage station (9) and ceramic tile stacking stations (4) are arranged in the stacking work area, a conveying device (1) is arranged at the rear end of the stacking work area, a manipulator (3) is arranged above the stacking work area, a control cabinet is arranged outside the stacking work area, and a control unit is arranged in the control cabinet. Through the multi-station ceramic tile transferring and stacking system based on the pallets, automatic clamping and separated transferring and stacking of ceramic tiles of different grades are achieved, great waste of manpower resources and material resources is avoided, the probability of collision of products during manual transferring is avoided while the labor efficiency is greatly improved, and the appearances of the products are guaranteed.

A method and a system for retrieving a fake-licensed vehicle are disclosed. The method comprises: pre-training a classification model for vehicle filtering, obtaining the type of target vehicle, color, sub-brand, license plate number and feature area information, removing the license plate number information of the target vehicle to obtain a second image, removing a license plate number information of an original test set vehicle from that vehicle passing record to obtain a second test set, filtering a second test set is associated with the target vehicle type, colors, sub-brands and differentvehicle images to get a third test set, selecting a vehicle image set according with the second image in similarity demands from the third test set, restoring the license plate number information ofthe vehicle in the image set, and matching and verifying the license plate information of the vehicle in the image set and the license plate information of the first image, so as to achieve the purpose of matching and verifying. The vehicle image set in the third test set and the vehicle image in the second image meet the similarity requirement. The technical scheme of the invention solves the technical problems of precision retrieval and verification that the prior art can not realize for the brand-set car.

The invention discloses equipment for accounting teaching, and relates to the technical field of accounting teaching. The equipment comprises a machine body, wherein a host is installed in the machinebody, a lifting mechanism is arranged in the machine body, the lifting mechanism comprises a threaded rod and a servo motor, a driving bevel gear is installed on the servo motor, a driven bevel gearis installed at the bottom part of the threaded rod, the driven bevel gear is meshed with the driving bevel gear, a lifting plate is installed on the threaded rod in a threaded mode, and a display mechanism is installed on the lifting plate. According to the equipment for accounting teaching, the liftable storage type display mechanism is adopted, the display mechanism comprises a display for a teacher to check and three displays for students to watch, thus the defect that a teacher shields a display device in teaching can be effectively avoided, and the teaching quality and effect are improved; meanwhile, the heat generated by the computer host can be rapidly discharged by arranging a heat dissipation fan in the machine body, the running speed of the computer host is guaranteed, and accountant teaching can be conveniently and smoothly carried out.

The invention relates to a computer unified device architecture (CUDA)-based multi-system signal frequency rapid acquisition implementation method. The method is characterized by being based on a high-performance private computer (PC), using the CUDA, and realizing rapid acquisition of the multi-system signal frequency through a parallel way. The method comprises the implementation steps of setting Doppler and Doppler changing rate segments; determining the number of blocks of each graphics processing unit (GPU) kernel function and the number of threads of each block; reading signal data into a GPU card overall storage space; starting a down-conversion and integrator module; performing parallel fast Fourier transform (FFT) operation; searching the maximum spectral line and corresponding Doppler changing rate segments and Doppler segments through a parallel reduction manner; and computing estimated values of carrier frequency offset and the Doppler changing rate to complete acquisition. The signal frequency rapid acquisition algorithm is in accordance with the CUDA and working characteristics of the GPU, so that the method can take full advantage of the parallel computing power of the GPU, is suitable for multiple systems of signals, effectively improves the acquisition speed and is easy to realize.

The invention provides an unmanned surface vessel path planning method, and the method comprises the steps of: S1, global path planning: generating an improved global path based on an improved RRT algorithm, optimizing the improved global path through a beetle antennae search algorithm, thus obtaining an optimal global path; the global path planning algorithm is simple, the calculated amount is small, the searching speed is high, and searching of the global optimal path can be rapidly achieved. S2, local path planning: carrying out local path planning by adopting a dynamic window method, and avoiding a situation that a target is unreachable through parameter adjustment; local path planning guarantees the running speed through dynamic parameters, and the situation that a target is unreachable is avoided. and S3, based on different dynamic obstacle meeting conditions, different collision avoidance strategies are adopted, so that the unmanned surface vessel can end the meeting situation in a short time, and then a best dynamic obstacle avoidance effect can be achieved. And a strategy of temporary obstacles and temporary target points is adopted, so that a safer and more efficient dynamic obstacle collision avoidance effect is achieved.

The invention discloses an intelligent switch control device control method, which comprises the following steps: when a home appliance is detected to be accessed to a network, the accessed home appliance is used as a current home appliance, and the identity of the current home appliance and the home appliance communication address are acquired, wherein the current home appliance is in the same network as the intelligent switch and the current home appliance is not associated with the intelligent switch; the type of the current home appliance is determined according to the identity, and according to the type of the home appliance, a home appliance number is allocated for the current home appliance; and the type of the home appliance, the home appliance number and the home appliance communication address are associated and saved to a target memory. The invention also discloses an intelligent switch control device and a readable storage medium. The running speed of a control chip in theintelligent switch is ensured, and the problem of delayed response of the home appliance is avoided.