220results about How to "Real-time computing" patented technology

The present invention discloses a camera positioning correction calibration method and a system for realizing the method, belonging to the technical field of image virtual production. Through the intrinsic relationship among a lens parameter, an imaging surface and an optical tracking device, by using the world coordinate and image point coordinate of N mark points on a background screen and the internal parameter and lens distortion parameter of a camera lens, the rotation matrix between a camera coordinate system and a world coordinate system and the translation vector of a camera perspective center in the world coordinate system are obtained, combined with the current position information given by a camera posture external tracking device in a current state, camera correction calibration information and viewing angle are obtained, the lookup table with a focusing distance and a focus length relationship is established, thus when a camera position, a lens focal distance and a focusing distance change, the position of the virtual camera of the virtual production system is fully automatically positioned and corrected, and thus a real video frame picture and a virtual frame image generated by a computer can be matched perfectly.

Provided in the invention is a multi-mode neural morphological network core comprising a mode register, an axon input unit, a synaptic weight storage unit, a dendrite unit and a neuron computing unit. According to the multi-mode neural morphological network core, both artificial neural network computing and impulsive neural network computing can be carried out; and switching between an artificial neural network operation mode and an impulsive neural network operation mode can be realized according to demands.

The invention discloses an unmanned aerial vehicle autonomous flight control method and system based on deep learning. According to the invention, a computer vision target detection and orientation estimation algorithm of an artificial neural network based on deep learning is put forward, images are collected by using a multi-rotor unmanned aerial vehicle carrying a cradle head camera, target detection and target orientation estimation are performed by using an operation unit for operating the neural network, and finally a control instruction is fed back to an unmanned aerial vehicle flight controller by combining a control algorithm. A distributed data collection, operation and control loop comprising a data collection unit, a deep learning operation unit and a flight control unit is constructed. Experimental results show that the unmanned aerial vehicle autonomous flight control method and system can achieve the effects that real-time neural network operations are performed, and thusthe unmanned aerial vehicle discovers a target quickly and autonomously, gets close to the target and tracks the target with a specific attitude by combining the control algorithm.

The invention discloses a video monitoring system and a moving target detecting and tracking method thereof. The method is applied to the video monitoring system on the condition that a camera is in a moving state. The method comprises the following steps: a moving target detection step, in which combination of a Harris corner and an HOG descriptor is employed as an image feature as well as processing including image feature extraction, image projection conversion, and image subtraction operation and the like is carried out to obtain a plurality of moving targets; a moving target model establishment step, in which moving target models are established for all the moving targets as well as a combination mode of a Harris corner, an extreme point and an HOG descriptor is employed to express features of the moving targets; and a target tracking step, in which target tracking is carried out according to the moving target models. According to the invention, corner feature matching is employed to reduce the calculated amount and precision of moving target detection is ensured; a corner and an extreme point are employed as target features, so that target identification has high robustness and tracking sustainability is strong; and a target motion estimation method is employed to reduce the calculated amount.

The invention provides a biped robot walking plan and control method. The method comprises the steps of establishing an equivalent model of a robot; determining a phase change graph meeting the requirement of the equivalent model; determining a phase change graph of a single-step horizontal movement track according to a human gait walking formula, namely, calculating a relationship between the speed of a single step and the step length/period of the single step; calculating a foothold according to the horizontal movement track of the single step; and controllably converting the position of the calculated foothold into an angle required for finishing motion of each joint. According to the method, the human gait formula is associated with the relationship between the walking speed and the step length/period of the single step, so that the calculation of the step length/period of the single step can be simplified; the period of the single step serves as a trigger event of single-step starting and ending, so that the end speed of the single step can be well predicted and the real-time calculation and adjustment of the foothold can be facilitated; and based on the motion description of the phase change graph, the analysis of a motion law of the biped robot can be more visual.

A mobile body surrounding surveillance 1A apparatus according to the present invention includes: an image capturing section 2; a particular region setting section 31 for setting a particular region for image data captured by the image capturing section; an own mobile body movement direction detection section 11 for detecting a movement direction of the own mobile body and outputting a movement direction information indicating the detected movement direction; a movement vector information extraction section for taking the image data of the particular region for each of one or a plurality of frame images captured by the image capturing section in time series and adjusting a detection sensitivity based on the movement direction information so as to extract a movement vector information on another mobile body; and a change detection section 35 for detecting a moving state of the other mobile body which is present in the surrounding of the own mobile body based on the movement vector information.

The invention discloses a cloud computing platform system fault prediction method based on hidden semi-Markov models. The cloud computing platform system fault prediction method includes the following steps that error events are extracted from mass log files, and endowed types and time information of the error events are obtained; repeated reports of the same error events are combined into one event; multiple continuous events are combined into event sequences, and the event sequences are divided into fault related event sequences and fault unrelated event sequences; according to a sequence likelihood value, all the event sequences are clustered; noise elimination is carried out in each cluster; the sequence likelihood value between the event sequences extracted in real time and the fault related hidden semi-Markov model (HSMM) and the sequence likelihood value between the event sequences extracted in real time and the fault unrelated HSMM are each calculated, and a Bayes classifier is used for calculation to judge whether a system fails or not. Based on the machine learning theory, the function of predicting system faults in real time by a cloud computing platform is achieved, accurate prediction results can be provided, and overall performance is higher.

The invention provides a distributed stream-oriented computation based spatial data processing method and system. A message queue of Kafka is established and used for real-time monitor, so that when data are stored in the message queue, Kafka can timely draw the monitored data, further, the data are processed in real time, and the real-time performance of data processing is realized. Besides, data are stored in a database and are further backed up, and the processed data can be updated to the database after the data are processed, so that the data in the database are updated. Real-time computation of geospatial big data is realized through distributed stream-oriented computation, and the method and system can support development and application of geospatial data with high real-time performance.

The invention relates to a method and a system for calculating theoretical line loss of a power grid. The method comprises steps: device parameters, operation parameters and topological data are acquired; a to-be-measured object is acquired, and a corresponding calculation method is determined according to the type of the to-be-measured object; and a corresponding calculation method is called, data needed by the corresponding calculation method are acquired from the device parameters, the operation parameters and the topological data, and theoretical line loss of the to-be-measured object is calculated and outputted. According to the power grid theoretical line loss calculation method of the invention, a corresponding calculation method is matched for a different to-be-measured object, data needed for calculation are automatically acquired from an existing power grid service management platform, the theoretical line loss of the power grid can be calculated in real time, and as a different to-be-measured object adopts a different calculation method, features of each to-be-measured object and features of each calculation method are fully considered, and each to-be-measured object is matched with a corresponding calculation method so as to improve the theoretical line loss calculation accuracy.

The invention provides a power cable real-time state monitoring and operation management system and a measuring method of the temperature of a cable. The power cable real-time state monitoring and operation management system comprises a distributed Raman temperature measurement sensing unit, a data acquisition subunit, an optical magnetic field sensing unit, a current data demodulator, an FPGA core processing subunit, an ARM core processing subunit, a magnetic field sensing head and an operation management server, wherein a temperature demodulator of the distributed Raman temperature measurement sensing unit and the current data demodulator of the optical magnetic field sensing unit are connected with the operation management server through Ethernet interfaces, an optical filter carries out band-pass filtering processing on Stokes light and Anti-Stokes light, light intensity signals of filtered back Raman scattering light are converted into electric signals through a photoelectric detection unit, the electric signals are amplified to obtain voltage signals corresponding to the temperature, the data acquisition subunit carries out A/D conversion on the voltage signals, an actual temperature value is obtained through calculation, the hidden danger of faults, with unobvious surface temperature rise, of the cable is eliminated, real-time fault recording is carried out on important cable taps, rapid contact locating of the hidden danger of the faults and accident analysis are facilitated, and real-time calculation of the current-carrying capacity is achieved.

The invention discloses a high-quality method for rapidly fusing a multi-spectrum image and a full-color image. The method comprises the following steps of: keeping detailed information of the full-color image through a multidirectional gradient operator and utilizing a lower sampling operator to keep multispectral color information, so as to obtain a multispectral image with a high resolution ratio. Furthermore, a complicated space domain is calculated and converted into a frequency domain through equivalent conversion; not only can the occupation of an internal memory of a computer be reduced, but also the operation complexity is reduced, so that real-time calculation can be realized. The high-quality method for rapidly fusing the multi-spectrum image and the full-color image, disclosed by the invention, solves the difficult problem that the existing image fusing method cannot simultaneously keep both detail information of the full-color image and detail information of the multi-spectrum image; and with the adoption of the method disclosed by the invention, the space resolution ratio and the spectral resolution ratio of the fused image are simultaneously optimized.

The invention discloses a method and a device for balancing network load and a cluster service system. The method comprises the following steps of acquiring health degree information of available nodes in the cluster service system; calculating issue probability of the nodes according to the health degree information of the nodes and selecting corresponding target nodes for sending task requests according to the issue probability; sending the task requests according to the selected target nodes. The method can be used for reasonably selecting the target nodes for sending the task requests according to the health degree information of the nodes, thereby preventing from selecting abnormal nodes with the health degrees which cannot meet the index, enabling the abnormal nodes to have chance to quickly restore the functions, and improving the stability of the system.

This invention discloses one method to realize abnormal judgment and dynamic alarming, which comprises the following steps: customer end alarming parameter setting step; setting control pattern judgment rules and trend top and down limit and selecting alarming type; abnormal judging and processing real time analyzing and abnormal judging on collection spot real data; alarming step to receive abnormal information and distributing alarm information.

The invention discloses a calculating method for outlet temperature of a combustion chamber of a combustion gas turbine. The calculating method comprises the following steps: processing a gas expansion formula linearly to establish a function relationship formula between a measurable value and the outlet temperature of the combustion chamber; correcting the function relationship formula between the measurable value and the outlet temperature of the combustion chamber according to change of an adjustable guide vane at the inlet of a gas compressor; and obtaining a function relationship formulabetween the measurable value and the outlet temperature of the combustion chamber at different atmospheric temperatures in different atmospheric temperature conditions and correcting the formula to obtain a calculating relationship of the outlet temperature of the combustion chamber. The outlet temperature of the combustion chamber is calculated by adopting the directly measurable performance parameters of the combustion gas turbine, the calculating method is small in calculating quantity, high in precision and high in speed, and the outlet temperature of the combustion chamber can be calculated in real time.

The invention discloses a calculation method of interior highest temperature of lining concrete for a door-opening-shaped fracture surface during a construction period. The method comprises the steps of taking lining thickness, concrete strength grade, pouring temperature, water cooling and water temperature, an air temperature and others to a calculation formula in order to obtain corresponding lining structures and interior highest temperature of concrete during the construction period for the scheme of temperature control measures.The calculation formula is capable of optimizing and designing a temperature control scheme by calculating multiple groups of simulated measurement schemes for interior highest temperature with the calculation value less than the highest temperature in economically feasible manner. The calculation method of interior highest temperature of lining concrete for the door-opening-shaped fracture surface during the construction period can reflect impact of lining thickness, concrete strength grade, pouring temperature, water cooling and water temperature on interior highest temperature. Errors of the calculation value, on-site measurement and calculation results of a finite element method are small. The error rate of a real example is only 2.3%. The method is high in precision and easy in calculation. The calculation method of interior highest temperature of lining concrete for the door-opening-shaped fracture surface during the construction period is capable of calculating interior highest temperature of concrete and can be used for optimizing and designing a temperature control anti-cracking scheme. The method can be prompted and applied and especially used for real-time rapid design and calculations of pre-design and construction periods.

The invention discloses a blue light detecting method for a medical display device. The method comprises the following steps: an image generating module provides a test image with the image resolutionup to a pixel level for a medical display device; and a spectral radiation luminance meter module is used for measuring red-green-blue three-primary-color spectrum data of the medical display deviceand upper computer software analyzes the degree of blue light radiated by the medical display device according to the spectrum data. Meanwhile, the invention provides a comprehensive blue light hazardevaluation method for a medical display device with three physical indicators of a blue light hazard weighted radiance value LB, a blue light hazard efficiency value KB, V and a rhythm factor KC so as to overcome defects of blue light measurement and objective evaluation of the medical display device. Besides, the invention also provides a blue light detecting apparatus for a medical display device pixel level, thereby realizing the high-precision measurement and accurate evaluation of blue components of a pixel-level test screen of a medical display device.

The invention discloses a current frequency converter test method based on a virtual instrument. The method comprises the steps as follows: a programmable current source is subjected to initial setting, then the programmable current source is started to output electric currents, and a digital multi-meter is used for measuring the size of an input current after a setting time is delayed; a counter is started for measuring the frequency, and an impulse at one end of an output signal FO- and one end of an output signal FO+ is counted; after the frequency is measured, a current/frequency point is obtained, the size of the input current is changed, a next current/frequency point is measured, and so on; after all the current/frequency points required to be measured are measured, a current value is taken as a horizontal variable Xi, a frequency value is taken as a longitudinal variable Yi, the linearity is calculated with a least square method, and a current frequency converter with a linearity test result smaller than 0.03 is qualified. According to the method, the measurement speed is high, complex analysis and calculation can be performed in real time, and a test conclusion is obtained rapidly; and the test accuracy is high, and the test process is simple.

The embodiment of the invention discloses a road network matching method, device and equipment and a storage medium. When to-be-processed track data is collected for road network matching, a spatial index can be generated according to multiple track points included in the to-be-processed track data, and the spatial index reflects position information of line segments formed by the multiple track points; therefore, the candidate road network set corresponding to the to-be-processed trajectory data can be obtained according to the spatial index. The candidate road network set comprises a plurality of road networks, and the plurality of road networks have high possibility to be matched with the to-be-processed track data. And the matching probability of the to-be-processed trajectory data andthe road networks in the candidate road network set is is determined, and further a target road network matched with the to-be-processed trajectory data is determined according to the matching probability. Visibly, complete road network data does not need to be loaded in advance, the candidate road network set is selected by utilizing the spatial index, a road network matched with newly collectedand stored track data can be calculated and solved in real time in a light-weight and rapid manner, the used calculation resources are very small, and the calculation efficiency is improved.

The invention relates to a measuring and positioning system of an underwater evener, which comprises an evening boat measuring and positioning system, an evener measuring and positioning system and a microprocessing unit. The invention adopts a positioning mode that an evening boat and an evener are respectively independent, accurately determines the plane position, elevation and horizontal attitude when the underwater evener works and the actual elevation of a foundation bed surface in acceptance through the equipment of a GPS (global positioning system), a tilted sensor, a position sensor, a depth sounder and the like, simultaneously realizes the position and attitude positioning of the evening boat, and realizes the functions of design and management of a working zone, data acquisition of the equipment, real-time calculation, real-time display, early alarm and alarm for limit overrun, data storage, data playback and the like through the microprocessing unit.

The invention discloses an embedded system on programmable chip (SOPC) having an image coprocessor. The system mainly comprises the image coprocessor and an embedded system (including a central processing unit (CPU), a bus, a memory controller, an interrupt controller and the like). The CPU is mainly used for managing system resources and processing complicated transactions, and the image coprocessor is mainly used for carrying out calculation of a plurality of specific image processing algorithms. The system has the advantages that: the self-developed image coprocessor has concurrent processing capacity and pipelining processing capacity; the system can be applied to the plurality of specific image processing algorithms; and calculation efficiency is improved greatly, so that the CPU can get rid of heavy image processing to process some more complicated transactions attentively, so as to improve the processing capacity of a whole SOPC signal processing platform.

The invention discloses a universal floating point full-pipeline FFT (Fast Fourier Transform) operation IP (Internet Protocol) core, which comprises an in-situ operation unit, a base-2 pipeline type butterfly operation unit and a result processing unit, wherein the in-situ operation unit is used for segmenting and sorting input digital signals, completing in-situ operation by means of address order reversal, sequentially caching the input digital signals in a way of keeping 256 points in each group by using a ping-pang cache method, and generating data which are convenient for butterfly operation input via screening of odd and even sequences; the base-2 pipeline type butterfly operation unit comprises multiple stages of butterfly pipeline operation units; a first-stage butterfly pipeline operation unit is used for receiving data generated by the in-situ operation unit, and then each stage of butterfly pipeline operation unit is connected with a previous stage butterfly pipeline operation unit respectively to calculate FFT operation output corresponding to the input of each point; the result processing unit is connected with the base-2 pipeline type butterfly operation unit, and is used for sorting the FFT operation output calculated by using the base-2 pipeline type butterfly operation unit.

The invention provides a human body shape measuring method and device based on deep learning key point detection. The method can detect and identify key points of a human body shape through a computervision technology based on deep learning, and can achieve quick and real-time calculation of main human body shape indicators (for example, height, shoulder width, sleeve length, pant length, chest circumference, waist circumference, hip circumference and the like) and visual display of measurement effect. The method has strong anti-noise ability, high robustness and accurate measurement result for various complex environments, and effectively meets the needs of tooling batch measurement and clothing shopping guide, the measuring device can achieve automatic measuring and calculating of key indicators of the human body shape through deep learning training, the measurement is accurate and the operation is simple.