Product
Patsnap Eureka
For R&D, Patsnap Eureka makes reading and utilizing patents & technical documents easy.
Patsnap Eureka AIR
Designed for self-driven R&D workflows. Generate viable solutions, solve complex R&D challenges, empower your innovation with AI.
Patsnap Eureka Materials
Designed for material experts only. Revolutionize your material R&D, from search, analyze, to developing new materials.
Feature
TechResearch
Generate reliable direction feasibility study reports for your R&D in just a few steps.
TechSeek
Discover and master advanced knowledge NOW. Basics, ideas, possibilities, all at once.
TechMind
As an expert in R&D Theories, TechMind can generates customized viable solutions instantly.
TechRisk
Analyze your overall solution with one click, know your potential R&D risks in advance.
TechMonitor
Get weekly tech updates, stay abreast of the latest tech innovations and key insights.
Patents
Literature
Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
106results about How to "Does not affect real-time performance" patented technology
Efficacy Topic
Property
Owner
Technical Advancement
Application Domain
Technology Topic
Technology Field Word
Patent Country/Region
Patent Type
Patent Status
Application Year
Inventor
Rapid generation method for facial animation
InactiveCN101826217AImprove descriptive powerQuick changeImage analysis3D-image renderingPattern recognitionImaging processing
The invention relates to a rapid generation method for facial animation, belonging to the image processing technical field. The method comprises the following steps: firstly detecting coordinates of a plurality of feature points matched with a grid model face in the original face pictures by virtue of an improved active shape model algorithm; completing fast matching between a grid model and a facial photo according to information of the feature points; performing refinement treatment on the mouth area of the matched grid model; and describing a basic facial mouth shape and expression change via facial animation parameters and driving the grid model by a parameter flow, and deforming the refined grid model by a grid deformation method based on a thin plate spline interpolation so as to generate animation. The method can quickly realize replacement of animated characters to generate vivid and natural facial animation.
Owner:SHANGHAI JIAO TONG UNIV
Real-time speech endpoint detection method and device
ActiveCN109545188AAvoid misjudgmentPrevent strong noise misjudgmentSpeech recognitionZero-crossing rateSpectral entropy
The invention relates to the technical field of speech, in particular to a real-time speech endpoint detection method and device. The method comprises the following steps that signal framing and emphasis are carried out; pulse removal processing is carried out; direct current components are removed; the short-time energy and zero-crossing rate of each frame of signal are calculated; windowing processing is carried out; spectrum reduction processing is carried out; spectral entropy is calculated; transformation smooth spectral entropy is calculated; a speech frame and a noise frame are preliminarily judged; the transformation smooth spectral entropy and a threshold are processed; a start frame and end frame in a speech segment are judged. The real-time speech endpoint detection method and device have the advantages that according to the conditions under which a signal is judged and a judged result, thresholds of parameters, such as a spectrum reduction threshold, the transformation smooth spectral entropy, the corresponding short-time energy, corresponding short-time average energy and a spectrum reduction power spectrum are weighted and updated, so that the thresholds are more andmore accurate, and finally the judged speech start frame and end frame are also more and more accurate; the method can efficiently and accurately detect speech in real time.
Owner:深圳市友杰智新科技有限公司
Method and system for resolving attitude of unmanned aerial vehicle with double controller
InactiveCN109506646ADoes not affect the real-time performance of the attitude calculationDoes not affect real-time performanceNavigation by terrestrial meansNavigation by speed/acceleration measurementsAttitude controlComplementary filter
The invention relates to a method and a system for resolving an attitude of an unmanned aerial vehicle with double controller. An unmanned aerial vehicle attitude resolving system which includes a main engine and a slave engine is arranged; in the main engine, an attitude angle resolved by an accelerometer and an electronic compass is adopted for performing drift compensation on the attitude angleresolved by a gyroscope, a complementary filtering algorithm is adopted for primarily fusing the attitude angle resolved by the accelerometer and the electronic compass with the attitude angle resolved by the gyroscope, noise is filtered and attitude angle drifting is restrained; the attitude angle after complementary filtering is transmitted to the slave engine, an EKF is adopted for promoting the resolving accuracy of attitude angle and the resolved high-precision attitude angle is transmitted by the slave engine to the main engine for attitude compensation and control in the manner of taking 50HZ as a period; the EKF is an expanded kalman filter algorithm; when no attitude angle information is received by the main engine from the slave engine, the attitude angle resolved by the complementary filtering algorithm in the main engine is adopted for controlling attitude. The system provided by the invention can promote the accuracy of resolved attitude and has high practicability.
Owner:SHIJIAZHUANG TIEDAO UNIV
Method for identifying place image on the basis of improved probabilistic topic model
ActiveCN102609719AGood effectReduce complexityCharacter and pattern recognitionFrequency vectorScale-invariant feature transform
The invention discloses a method for identifying a place image on the basis of an improved probabilistic topic model, belonging to the technical field of image identification. The method provided by the invention can be used for well solving the problems that the image identification is uncertain due to different angles, illumination, and height dynamic changes of figures and objects. The method comprises the following steps: an image acquiring step, an image preprocessing step, a feature extraction step, a feature clustering step, a feature distribution step and a potential topic modeling step, wherein in the image acquiring step, the features of the image are extracted by adopting a SIFI (scale invariant feature transform) algorithm; in the feature clustering step, all the features are clustered so as to obtain a plurality of clustering centers; in the feature distribution step, the feature of each image is voted in the clustering center so as to obtain a frequency vector corresponding to each clustering center; in the potential topic modeling step, the potential topic distribution of the image is learned by adopting the improved probabilistic topic model; and a classifier is adopted to identify the images at unknown places. According to the invention, a quantization function is added in an LDA (latent dirichlet allocation) model, and the potential topic of the image is learned through the improved probabilistic topic model, so that the identification performance is effectively improved on the premise of guaranteeing instantaneity.
Owner:BEIJING UNIV OF TECH
Automatic sound identifying treating method for embedded sound identifying system
InactiveCN1588535AStrong representativeImprove recognition rateSpeech recognitionSound input/outputIdentification rateSupport vector machine algorithm
The invention is a voice automatic-identification processing method of embedded voice identifying system, composed of a front-end processing part, a real-time identifying part, a back-end process part and a template training part, adopting self-adapting end-point detecting technique to draw voiced segments, adopting synchronous mode to identify input voice, applying vector-supporting algorithm to realize fast rejection of non-command voice, thus improving identification reliability and practicality, and adopting multistage vector quantization method to train the voice template and assorted with McE / GPD distinctive training to optimize the voice template so as to improve identifying property. The used acoustic model has a small memory space, thus effectively increasing the identification ratio of the system to above 95%, its algorithm load is small, its memory space is small and its identification rejection ratio is higher than 80%.
Owner:SHANGHAI JIAO TONG UNIV
Distributed optical fiber Raman temperature measurement system
ActiveCN103115693ARealize wavelet denoisingProcessing speedThermometers using physical/chemical changesWavelet denoisingPulsed laser device
The invention relates to the field of distributed optical fiber temperature measurement systems, and discloses a distributed optical fiber Raman temperature measurement system. The system comprises a pulse laser device, a wavelength division multiplexer, a sensor optical fiber, a double-channel avalanche photodiode, and a digital signal processor (DSP). The whole optical fiber laser device emits pulse lasers, the pulse lasers enter into a to-be-tested sensor optical fiber after passing through the wavelength division multiplexer, the pulse lasers continuously generate backscattering inside the optical fiber in the spreading process, and then backscattering light returns back to the wavelength division multiplexer, after the backscattering light passes through the wavelength division multiplexer, Stokes scattering light and anti-Stokes Raman scattering light are respectively filtered out and enter the double-channel avalanche photodiode to be conducted photovoltaic conversion, and after an electrical signal of the double-channel avalanche photodiode is processes by the DSP, a temperature signal is obtained. The high-speed DSP is used for achieving wavelet denoising, and therefore processing speed is quick, and real time performance of temperature measurement is not influenced on the premise that precision is guaranteed.
Owner:YANGTZE OPTICAL FIBRE & CABLE CO LTD
Method and device for realizing persistence in flow calculation application
ActiveCN104424186AExtended time intervalImprove real-time computing efficiencySpecial data processing applicationsOrder of magnitudeThroughput
The invention discloses a method and a device for realizing persistence in flow calculation application. The method comprises the following steps that: when the current batch message is successfully consumed, whether the persistence operation needs to be performed or not is judged according to the first initial offset and the preset persistence interval; when the persistence operation needs to be performed, the persistence processing is carried out according to the message position indicated by the second initial offset; and after the persistence succeeds, the first initial offset and the second initial offset are updated into the initial offset of a next batch message. The persistence operation is performed after the persistence interval, and the disk persistence time interval is prolonged, so that the real-time calculation efficiency is greatly improved. During fault recovery, at most the batch message in the persistence interval needs to be consumed again; the performance bottleneck caused by frequent disk writing in the existing synchronous persistence process is avoided; the real-time calculation message throughput performance is improved by an order of magnitude; and meanwhile, the delay caused by the fault recovery is reduced to the second stage, and the real-time performance cannot be influenced.
Owner:阿里巴巴华南技术有限公司
Multitask unloading method and system in time-sensitive network
ActiveCN112039965ADoes not affect real-time performanceImprove real-time performanceTransmissionEnergy efficient computingOptimal decisionEdge computing
The invention belongs to the technical field of edge computing and time sensitive networks, and provides a multitask unloading method and system in a time sensitive network, and the method comprises the steps: building a task model for a task to be unloaded according to the attribute of the task, and the attribute at least comprises the code quantity of the task, the data quantity input by the task and the code computing density; in the task model, solving the total time consumed by the task under each decision according to an unloading decision basis, and constructing an optimal decision function of each task by taking minimization of the consumed total time as a target; according to a game algorithm, in a multi-task scene, performing finite iterative computation on the optimal decision function of each task, and calculating a decision result that multiple tasks reach Nash equilibrium. According to the technical scheme, insufficient computing resources can be relieved, meanwhile, thereal-time performance of tasks is not affected, and even the real-time performance of task delivery is improved to a certain extent, so that loads between computing nodes are balanced, and the methodhas good practical significance and application value.
Owner:致瓴探(上海)科技有限公司
Intrusion detection system specific to programmable logic controller (PLC) control system
ActiveCN110320890AGuaranteed stabilityGuaranteed reliabilityProgramme controlElectric testing/monitoringNetwork packetProgrammable logic controller
The invention provides an intrusion detection system specific to a programmable logic controller (PLC) control system. The intrusion detection system comprises an operation data acquisition module, adata-driven intrusion detection module, a network data acquisition module, a network communication protocol intrusion detection module, an intrusion response output module and a monitoring server. A normal network communication protocol model is built by capturing and parsing a network data packet, analyzing a network data feature influence factor and extracting a network data feature value, so that network communication protocol intrusion detection is realized. Meanwhile, field operation data are acquired to build a steady operation prediction model, and residual assessment is performed on the practical output of the control system and the output of model prediction, thereby realizing data-driven intrusion detection. When intrusion is found, braking processing is performed on a controller, and an alarm is made. The intrusion detection system is suitable for the technical field of industrial control system information safety, can effectively realize the intrusion detection of the PLC control system under the condition of not intruding an industrial control system, and improves the safety guarantee capability of the industrial control system.
Owner:UNIV OF SCI & TECH BEIJING
Real-time operating system and real-time method thereof
InactiveCN101324852AReduce overheadImprove real-time performanceMultiprogramming arrangementsNon real timeOperational system
The invention discloses a real-time operating system and a real-time operating method. The processor of the system comprises a fast interrupt controller and an interrupt controller, wherein the fast interrupt controller and the interrupt controller can respectively perform the real-time interrupt processing and the non-real-time interrupt processing to a real-time task and a non-real-time task, and respectively perform controlling to the processes of the real-time interrupt processing and the non-real-time interrupt processing through a software, and the priority level of the real-time interrupt processing is higher than that of the non-real-time interrupt processing. The real-time interrupt allocating and processing program is separated form the non-real-time interrupt processing program, so that the caching is locked to ensure that the method for determinacy of the interrupt response is easy to be realized, and thereby the real-time interrupt response time of the system can be further improved. The real-time operating system and the real-time method have the advantages that not only better immediacy can be realized, but also the extra cost of the processor cannot be increased, and the real-time property of the system cannot be affected.
Owner:BEIHANG UNIV
Method and apparatus for producing an image through operation of plotting commands on image data
InactiveUS6011564AImprove process capabilityDoes not affect real-time performanceImage codingMultiple digital computer combinationsCoprocessorImaging processing
An image processing system for producing an image from three-dimensional image data. The system uses a command generator to generate one or more plotting commands for producing the image, and features a coordinate calculator for transforming the three-dimensional data into two-dimensional data. The coordinate calculator operates as a coprocessor of the command generator, and as such does not require use of the system bus to perform its functions.
Owner:SONY COMPUTER ENTERTAINMENT INC
Configuration method for control channel, and receiving method, device and system
InactiveCN105188140ASolve the problem of high number of busy checksReduce energy consumptionHigh level techniquesWireless communicationInformation typeControl channel
The invention discloses a configuration method for a control channel, and a receiving method, device and system. The method comprises the steps that a base station carries out the grouping of serving cells for terminal detection, and enables the serving cells to be divided into at least two CGs (cell groups), wherein one cell group is an MCG (main cell group), and one or more other cell groups are SCGx (x is integer greater than zero); and the base station predefines or configures a control channel information type and / or a downlink public control information format of the control channel in a terminal detection public searching space, wherein the control channel information type and / or the downlink public control information format are / is needed by downlink public control information. The method solves a problem that the introduction of double links causes the larger number of times of blind detection of the terminal on the basis of CA in related technologies, reduces the energy consumption of the terminal, and also can guarantee that the maximum number of detection times of the control channel of the terminal does not exceed the predefined number of detection times of the terminal under the condition of not affecting the average throughput capacity and real-time business dispatching of the terminal.
Owner:ZTE CORP
Method for quickly processing "kidnapping" problem of mobile robot positioning
PendingCN109556607AEasy to handleFast processing of mobile robot localizationNavigational calculation instrumentsNetwork modelConvolution
The invention discloses a method for quickly processing a "kidnapping" problem of mobile robot positioning, and belongs to the technical field of mobile robots. The method comprises the following steps of: training a convolution neural network model based on deep learning through training data collected in advance to obtain a rough pose of a camera at the time of taking a picture by inputting thepicture only; when the robot judges that there is a "kidnapping" problem in itself, sending a picture taken by a camera on the robot at the moment to the convolution neural network model to obtain a rough pose; and then adopting a particle filter positioning algorithm in a specific format to reduce the number of iterative convergence in the algorithm and achieve a goal of quickly solving the "kidnapping" problem. According to the method for quickly processing the "kidnapping" problem of the mobile robot positioning, the mobile robot can be quickly and accurately repositioned after the robot ismanually moved to any position.
Owner:SHANGHAI UNIV
Method for dynamically calculating and correcting course angle of navigation system of magnetometer
ActiveCN103954303AStrong real-timeImprove real-time performanceNavigational calculation instrumentsDistribution characteristicWorking environment
The invention provides a method for dynamically calculating and correcting a course angle of a navigation system of a magnetometer. The method can be used for dynamically calculating, correcting and compensating the course angle of the navigation system in a working environment with a changing external magnetic field. The method comprises the steps of firstly, establishing a dynamic error model of the magnetometer, secondly, comparing the corresponding relations of three axis output at two adjacent moments of the magnetometer and analyzing distribution characteristics of a geomagnetic field so as to acquire a group of equation including course angle information, and finally, dynamically calculating and correcting the course angle by solving the equation set. The method has the advantages of high applicability, strong instantaneity, high reliability, low cost, high precision, simplicity in processing of a calculation process and high stability and is capable of effectively improving the course precision of the navigation system. The method can be used for correcting and compensating the error of the course angle of an inertial integrated navigation and positioning system comprising an inertial measurement unit (IMU), the magnetometer (or an electronic compass).
Owner:SOUTHEAST UNIV
Electric vehicle CAN network anomaly detection method and device
InactiveCN106878130ALow computing power requirementsDoes not affect real-time performanceBus networksData terminalAnomaly detection
Owner:CHENGDU RAJA NEW ENERGY AUTOMOTIVE TECH CO LTD
Rearview mirror simulation method for vehicle driving simulation system
ActiveCN104408989ARealization of driving simulationRealization of Rearview Mirror Simulation for Simulated Driving SystemCosmonautic condition simulationsSpecial data processing applicationsVehicle drivingMirror effect
The invention discloses a rearview mirror simulation method for a vehicle driving simulation system. The method comprises the following steps: firstly, obtaining specific position and coordinate of a vehicle according to vehicle motion, capturing a rearview mirror scene by a rearview mirror camera to obtain a rearview mirror top point coordinate required to be drawn, connecting top points in sequence to form a rearview mirror surface which is rearview mirror texture, and finally pasting the rearview mirror texture onto a rearview mirror model; during driving simulation, pre-binding initial texture, replacing old texture with new texture in real time through a created pointer function, and finally finishing simulation of a whole rearview mirror. According to the method, the rearview mirror effect is creatively displayed on a vehicle model in a three-dimension manner, so that the trouble caused by occupation of main view angle can be avoided, the refresh rate of the system can be reduced, and the vividness of whole driving simulation can be improved; the method has good commercial values.
Owner:NANJING ZHONGCHE PUZHEN URBAN RAIL VEHICLE CO LTD
Target tracking optimization method based on tracking learning detection
InactiveCN107967692AImprove reliability and robustnessImprove robustnessImage analysisFrame sequence
The invention discloses a target tracking optimization method based on tracking learning detection. The target tracking optimization method comprises the steps of: determining L frame grayscale imagevideo frame sequences, wherein a tracking target position in the first frame grayscale image video frame sequence is known, and tracked target positions in the rest L-1 frame grayscale image video frame sequences are unknown; performing initialization, wherein t belongs to a set {1, 2, ..., L}, an initial value of t is 1, and L is a positive integer greater than 1; determining a main body part ofa first frame tracking target and an edge part of the first frame tracking target; acquiring Nt tracking points from an edge part of a t-th frame tracking target and a main body part of the t-th frametracking target, and further acquiring a limiting frame tb<t+1> of a (t+1)-th frame tracking target in the (t+1)-th frame grayscale image video frame sequence; acquiring N<t+1><db> detection target limiting frames in the (t+1)-th frame grayscale image video frame sequence from the (t+1)-th frame grayscale image video frame sequence; further acquiring a determined position of the (t+1)-th frame tracking target; and adding 1 to the value of t until determined positions from the second frame tracking target to the L-th frame tracking target are obtained, and recording the determined positions astarget tracking optimization results based on tracking learning detection.
Owner:XIDIAN UNIV
Image retrieval method and system
InactiveCN108255858AAvoid detectionAvoid retrievingCharacter and pattern recognitionSpecial data processing applicationsGreedy algorithmSample image
The embodiment of the invention discloses an image retrieval method and an image retrieval system and belongs to the technical field of image retrieval. The method comprises the steps of performing redundancy elimination processing on a to-be-retrieved image based on a global contrast saliency detection algorithm; segmenting the to-be-retrieved image after the redundancy elimination processing into a plurality of regions based on a greedy algorithm; performing extraction of image characteristic points on each region based on an image characteristic extraction algorithm to obtain a plurality ofimage characteristic points; generating a bag-of-visterms model of the to-be-retrieved image based on the plurality of image characteristic points; performing matching in a prestored sample bag-of-visterms model library to obtain a sample bag-of-visterms model matched with the bag-of-visterms model of the to-be-retrieved image; and outputting a sample image, as a retrieval result, corresponding to the bag-of-visterms model obtained through matching. The image retrieval method disclosed by the embodiment of the invention has relatively high accuracy and retrieval efficiency and has a certain robustness during processing of various complex changing images.
Owner:BEIJING UNION VOOLE TECH
Debugging-free composite communication device and method for electric vehicle charging station
ActiveCN102752299ASolve the problem of protocol conversionEliminate the trouble of configuring the protocol separatelyTransmission noise suppressionElectric carsBus interface
The invention discloses a debugging-free composite communication device for an electric vehicle charging station, and the debugging-free composite communication device comprises a central processing unit, a data conversion unit, an entrance processing module, an exit processing module, a measurement processing module, a data bus interface and an optical-fiber Ethernet module, wherein the central processing unit is connected with the data conversion unit, the entrance processing module, the exist processing module, the measurement processing module, the data bus interface and the optical-fiber Ethernet module through a system bus and is used for collecting data of the entrance processing module, the measurement processing module and the data bus interface. The debugging-free composite communication device has the beneficial effects that the problems such as equipment configuration, regulation conversion and communication interface conversion when charging devices of difference manufacturers are interconnected inside the electric car charging station can be solved, different charging devices of different manufacturers can conform to the national uniform standard after being processed by the composite communication device, the equipment interconnection and system access process can be simplified, convenience in modifying, maintaining, expanding and upgrading the charging station can be realized, and the automation degree and the management level of the charging station can be improved.
Owner:STATE GRID CORP OF CHINA +2
Message header compression method, compressor and transmission system
ActiveCN1968211ADoes not affect real-time performanceImprove transmission efficiencyNetwork traffic/resource managementData switching networksComputer hardwareTransport system
The invention relates to a method for compressing report head, which comprises that: based on data report, generating initial frame, comparing the first data report head and the frame head of initial frame, finding out and removing the abundant information of said data report head, to generate the first compressed frame; comparing the frame heads of second data frame and last frame, finding out and removing the abundant information of data report to generate the second compressed frame. The invention also discloses a compressor with comparer and generator; the generator can generate initial frame and compressed frame via data report; the comparer can find out abundant information. The invention also discloses a transmitter with compressor and decomrpessor; the compressor can generate and send initial frame and compressed frame; the decompressor is used to receive initial and compressed frames, reduce data report via compressed frame, and store the head information. The invention can effectively compress report head, and improve transmission efficiency.
Owner:HUAWEI TECH CO LTD
Railway wagon small part bearing stop key nut looseness and loss fault detection method
ActiveCN111080650AUnified operating standardsLow costImage enhancementImage analysisImage extractionFreight trains
The invention discloses a railway wagon small part bearing stop key nut looseness and loss fault detection method, and belongs to the technical field of freight train detection. The objective of the invention is to solve the problems of high cost, low efficiency and the like of a current detection mode depending on manual image viewing and the problem of low accuracy of detection by an existing image automatic processing technology. The method comprises the following steps: acquiring an image, extracting a bearing blocking key region image, constructing a sample data set, and training a deep learning network by using the sample data set to obtain a trained deep learning network; extracting a bearing blocking key area image for a real vehicle passing image, and obtaining a three-value segmentation image by using the trained deep learning network; carrying out fault detection by utilizing the three-value segmentation image, and if no nut exists in a bolt area, indicating that the stop key nut is lost; if partial area of the bolt is above the nut, indicating that the stop key nut is loosened; and if the situation is detected, carrying out fault alarming. The method is mainly used forbearing stop key nut looseness and loss fault detection.
Owner:HARBIN KEJIA GENERAL MECHANICAL & ELECTRICAL CO LTD
Adaptive anti-interference system and method
ActiveCN102571139AAutomatic identification of interference frequency bandsDoes not affect real-time performanceTransmissionTime domainSelf adaptive
The invention relates to an adaptive anti-interference system and an adaptive anti-interference method. The adaptive anti-interference system comprises a signal acquisition device, an A / D (analog-to-digital) conversion device, a digital automatic gain control module, a multi-channel signal reproducer, a time-domain filtering processing module and a transform-domain filter generation module, wherein the digital automatic gain control module is respectively connected with the time-domain filtering processing module and the transform-domain filter generation module by the multi-channel signal reproducer; the digital automatic gain control module is used for adjusting the amplitude of a digital signal; the multi-channel signal reproducer is used for reproducing the digital signal outputted by the digital automatic gain control module into at least two channels of digital signals, transmitting one channel of digital signal to the time-domain filtering processing module, and transmitting the other digital signals to the transform-domain filter generation module; the transform-domain filter generation module is used for generating filtering processing coefficients according to interference in the digital signals which are detected in real time, and transmitting the filtering processing coefficients to the time-domain filtering processing module; and the time-domain filtering processing module is used for carrying out real-time filtering on the interference in the digital signals according to the filtering processing coefficients.
Owner:北京华信创通科技有限公司
Speech recognition method and device and terminal
ActiveCN105679321ADoes not affect real-time performanceImprove noise immunitySpeech recognitionSpeech identificationComputer terminal
The invention discloses a speech recognition method and device and a terminal, and is to improve anti-noise performance of an existing speech recognition mode. The method comprises the following steps: obtaining a frame of speech signal, and extracting d-dimension MFCC parameters from the speech signal; carrying out cepstrum calculation on the d-dimension MFCC parameters to obtain d-dimension cepstrum MFCC parameters; carrying out iteration processing on each dimension of cepstrum MFCC parameters according to preset iteration times to obtain d-dimension iteration cepstrum MFCC parameters; and carrying out speech signal recognition based on the d-dimension iteration cepstrum MFCC parameters.
Owner:シェンジェンインムーテクノロジーシーオーエルティーディー
Key data acquisition method for bus type numerical control system
ActiveCN103163827ATroubleshooting Cycle Mismatch IssuesImprove real-time performanceNumerical controlData acquisitionMotion controller
The invention relates to the field of numerical control systems, in particular to key data real-time acquisition method for a bus type numerical control system. A numerical control system debugging agent module applies for a sharing buffer queue and is initialized. A numerical control system motion controller thread shares the sharing buffer queue of the debugging agent module. The key data acquisition operation of the numerical control system is achieved in the applied sharing buffer queue through write operation of the motion controller and read operation and sending operation of a data output device. A serial port for outputting is directly used in the output process of the data output device. A display terminal receives and displays debugging information output by the numerical control system in real time through the serial port. The problem that periods of the motion controller and a user interface are mismatched can be solved, the real-time performance is effectively improved, and the key data acquisition method has the advantages of being small in occupied space and good in generality.
Owner:SHENYANG GOLDING NC & INTELLIGENCE TECH CO LTD
Target tracking method and device, computer device and storage medium
InactiveCN109887001AReduce Tracking LossImprove tracking accuracyImage analysisImaging FeatureComputer vision
The invention relates to a target tracking method. The method comprises the following steps of obtaining a current video frame image containing the target object, extracting image features in the current video frame image; taking the image features as the input of a position model; obtaining an output value of the position model, determining whether the target object in the current video frame image is shielded or not according to the output value; when it is determined that shielding exists, taking the position model corresponding to the previous video frame image as a position model corresponding to the current video frame image, when it is determined that shielding does not exist, determining the target position of the target object in the current video frame image according to the output value, then updating the position model, and obtaining a position model corresponding to the current video frame. According to the method, the tracking accuracy is improved, and the target trackingefficiency is improved. In addition, the invention further provides a target tracking device, a computer device and a storage medium.
Owner:SHENZHEN INST OF ADVANCED TECH CHINESE ACAD OF SCI
Video-based flame detection method and system and readable storage medium
PendingCN111489342AImprove accuracyImprove recallImage enhancementImage analysisFlame detectionNetwork model
The invention belongs to the field of flame detection, and discloses a video-based flame detection method and system, and a readable storage medium, and the method comprises the following steps: obtaining YUV image data of a target region in real time, and cloning a current image frame and a former image frame, and obtaining F1 and F2; carrying out target detection and identification on F1 througha neural network model, and cutting F1 to obtain a plurality of first rectangular images; performing flame color feature inspection on the plurality of first rectangular images; cutting the F2 to obtain a plurality of second rectangular images, and performing flame dynamic feature detection on the first rectangular images and the second rectangular images; when both flame color feature detectionand flame dynamic feature detection are met, wherein the current target area contains flame and is calibrated in the YUV image data, generating alarm information; otherwise, carrying out S1. The advantage that the convolutional neural network extracts object features is fully utilized, meanwhile, the dynamic features of flames are effectively combined with the convolutional neural network, and theaccuracy, recall rate and detection speed of flame detection are effectively improved.
Owner:西安星舟天启智能装备有限责任公司
Real time three-dimensional image smoothing process method based on pattern processor
The invention provides a real-time three-dimensional image smooth processing method based on graphics processing units. The method orderly comprises the following steps that: (1) the data of a three-dimensional image to be processed is transmitted to a graphics processing units; (2) coordinates of related points around a pixel point to be processed are figured out; (3) the related points are sampled, so as to obtain the data values of the related points; (4) the average value of the data values of the related data points is calculated; (5) the squares of the difference of the data values of the related data points and the average value are respectively calculated, and the numerical values calculated are saved in an array; (6) the numerical values in the array are compared, and a minimum value is figured out; (7) the prior data value of the pixel point to be processed is replaced by the data value of the related data point corresponding to the minimum value; and the steps (2) to (7) are carried out in the graphics processing units. The method can eliminate the speckle noise of the three-dimensional image, beautify the image and well preserve the prior boundary information of the image without affecting the real-time performance of three-dimensional image reconstruction.
Owner:SHANTOU INST OF UITRASONIC INSTR CO LTD
High-frequency ultrasonic superficial organ imaging method capable of lowering random noise
InactiveCN103300888ARandom noise reductionDoes not affect real-time performanceUltrasonic/sonic/infrasonic diagnosticsInfrasonic diagnosticsSonificationSuperimposition
The invention discloses a high-frequency ultrasonic superficial organ imaging method capable of lowering random noise. The method is characterized in that an ultrasonic transducer continuously transmits N times of ultrasonic pulses at intervals within a dwell time of 200 microseconds on each scanning position in a scanning area, after the superimposition average operation processing is carried out by adopting a linear average method for the ultrasonic echo information, and the processing result is used as a scanning line of the scanning position; and then the scanning position is moved to repeat the process until all scanning lines are completed, and the two-dimensional high-frequency ultrasonic imaging can be realized. The echo information acquired through N times of detection is superimposed and averaged by adopting the linear average method, the random noise in the echo information of the scanning line is lowered, the signal-to-noise can be theoretically improved by N times, no frame frequency is reduced, and the imaging real-time property is not influenced. Since the N-time sampling time anticipating in the average calculation is about 200 microseconds, a probe and a tissue can be easily maintained relatively still, and the ghosting phenomenon caused by the frame average technology can be avoided.
Owner:INST OF BIOMEDICAL ENG CHINESE ACAD OF MEDICAL SCI
KCF tracking target loss detection method and system based on foreground detection
ActiveCN110689555AAccurate judgmentUnable to detectImage enhancementImage analysisBackground informationForeground detection
The embodiment of the invention relates to the technical field of visual target tracking, and discloses a KCF tracking target loss detection method and system based on foreground detection. The KCF tracking target loss detection method based on foreground detection comprises: extracting a foreground image by adopting a foreground detection algorithm; extracting a target detection frame by using aKCF tracking algorithm; obtaining a foreground region corresponding to the target detection frame, calculating the ratio of the target contour area in the foreground area to the total area of the foreground area; judging whether a KCF tracking target is lost or not by setting a ratio threshold. According to the KCF target tracking method and device, the problems that the KCF algorithm cannot be perceived after the target is lost and the KCF tracker takes the background information as the target to continue tracking are solved, and accurate judgment when the tracking target is shielded in a large area or leaves a lens is realized on the basis of not influencing the real-time performance of KCF target tracking.
Owner:四川航天神坤科技有限公司
Multi-MCU data interaction intelligent acquisition device and method
PendingCN107132799AGuaranteed correctnessGuaranteed real-timeProgramme controlComputer controlBroadcast packetData pack
The invention provides a multi-MCU data interaction intelligent acquisition device and method. The device comprises a main MCU processing module, a slave MCU acquisition module and a bottom layer module, wherein the main MCU processing module is used for sending broadcast packets and conducting data interaction with the slave MCU acquisition module and an upper computer; the slave MCU acquisition module is used for acquiring data of intelligent equipment and feeding back a data packet to the main MCU processing module after caching the data; the bottom layer module comprises various kinds of intelligent acquisition equipment. According to the multi-MCU data interaction intelligent acquisition device, the acquisition module is responsible for collecting, primarily processing and packing the data of a sensor, the processing module is responsible for integrating the data of the acquisition module and further processing the data, and finally the face-to-face interaction function is achieved through the modules and the upper computer together. An IIC bus time-sharing transmission mode is used among the modules for interaction, serial ports and network interface resources of the collected data are not occupied, the situations that self-spinning waiting is caused by access of critical resources and the real-time performance of a system is affected are avoided, and the correctness of the data can be ensured.
Owner:HUNAN LINENG TECH
Who we serve
- R&D Engineer
- R&D Manager
- IP Professional
Why Patsnap Eureka
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com