34results about How to "Meet real-time processing needs" patented technology

The invention belongs to the field of radar signal processing, and discloses a missile-borne SAR sub-aperture forward squint high-order nonlinear chirp scaling imaging method. According to the method, missile-borne SAR forward squint high-order nonlinear chirp scaling imaging is achieved through sub-aperture data, and the method can be used for SAR imaging of airborne platforms or missile-borne platforms. The method mainly includes the steps of firstly, conducting distance pulse pressure and time domain walking correction on an echo signal; secondly, converting the signal into the two-dimensional frequency domain, conducting frequency domain migration correction and secondary compression, and compensating for the high order phase in the orientation direction; thirdly, leading high-order nonlinear chirp scaling disturbance factors into the orientation frequency domain, and correcting the space variability of the Doppler frequency modulation and the high-order terms in the orientation direction; fourthly, focusing an image on the orientation frequency domain through spectrum analysis. The method solves the problem of decoupling of distance orientation and the orientation focus depth problem caused by time domain walking correction, and the method can meet the requirements for different scenes and high resolution and can be used for the field of ground mapping and other fields.

The invention discloses a method, a device and a system for scheduling in a multiprocessor nuclear system and relates to the field of the multiprocessor nuclear system. The method, the device and the system can meet real-time network input/output (I/O) processing requirements so as to promote efficiency of the whole multiprocessor nuclear system. The method for scheduling in the multiprocessor nuclear system comprises obtaining a first control parameter, a second control parameter, a third control parameter and a fourth control parameter in an operation period of the multiprocessor nuclear system; transmitting data packages of data streams entering the multiprocessor nuclear system to an idle processor nucleus for processing according to the first control parameter, the second control parameter and the third control parameter; and switching the processor nucleus in the multiprocessor nuclear system between a cutting-off mode and a polling mode according to the fourth control parameter. The method, the device and the system are mainly used for processor nucleus scheduling.

The invention discloses a fast Realization method for simulation 3D scene SAR radar echo based on forward method. The main idea of the fast Realization method comprises the following steps: acquiring digital elevation model (DEM) data within SAR radar beam range, wherein the data comprises K point targets; randomly distributing the K point targets on a 2D ground in the geographic coordinate system, and calculating DEM data within the SAR radar beam range after refinement processing; acquiring back scattering property distribution of the DEM data within the SAR radar beam range after refinement processing; utilizing an improved downward angle comparison method to determine shaded areas of Q small surface elements, calculating final actual back scattering coefficients of the Q small surface elements after shaded area determination, and then calculating SAR radar echo of the Q small surface elements. At this time, the whole 3D scene echo simulation is completed in a graphic processing unit (GPU), and finally an echo signal in the GPU is sent to a CPU processor to be output.

The invention relates to a rapid target detection method based on adaptive convolution, and the method comprises the steps: obtaining image data with a label, and forming a training set; constructinga target detection network based on an adaptive convolution module; training the proposed target detection network on the formed training set until the target detection network converges to obtain a trained target detection network; and detecting the image by using the trained target detection network, removing repeated results, and outputting a final result. By dynamically adjusting the parameters of the convolution filter, the expression capability of the detection network for extracting the features is improved, the irrelevant interference features are inhibited, the feature response of theinterested target is enhanced, the extracted features are more suitable for the detection requirement of the current scene, and the detection performance of the extracted features on the small-scaletarget in the complex scene is improved.

The invention discloses an FPGA-based solid-state power controller inverse time over-current protection device, mainly solving the problems that the existing time over-current protection device is poor in reliability, high in power consumption, low in accuracy and poor in strong electromagnetic interference resistance. In the invention, a clock management module, a multiple data register group module, an FSM control module, an inverse time over-current protection module and an output control module are arranged in the interior of FPGA; wherein the multiple data register group module receives an FSM control module address signal and outputs data to the FSM control module which judges the data, the judged data enters the inverse time over-current protection module, result calculated by the inverse time over-current protection module is returned to the FSM control module which compares the returned data, the comparison result is sent to the output control module which outputs a control signal to an outside circuit, and the clock management module provides a system clock for all modules. The invention has the advantages of high accuracy, strong reliability, strong electromagnetic interference resistance, low power consumption and the like.

The invention discloses a rapid two-dimensional CFAR (Constant False Alarm Rate) target detection system and detection method. The detection system comprises a phase reference accumulation result reading module, a distance dimension CFAR detection module, and a velocity dimension CFAR detection module. The detection method comprises steps that the phase reference accumulation result reading modulereads and records an echo phase accumulation result; the distance dimension CFAR detection module performs CFAR detection at the distance dimension on the phase reference accumulation result; and thevelocity dimension CFAR detection module performs CFAR detection at the velocity dimension through the point of the distance dimension CFAR. According to the rapid two-dimensional CFAR target detection system and detection method, rapid two-dimensional CFAR detection is realized by decomposing the two-dimensional phase reference accumulation result into distance dimension and velocity dimension separately, and performing CFAR detection; and the detection time can be greatly reduced, and the real-time processing demands of radars can be meet through engineering practice verification.

The invention discloses a pyramid-based scatter diagram sampling method and system. The method comprises the following steps: converting a scatter diagram of acquired sample data into a density map; constructing a data density pyramid and a visual density pyramid based on the density map; taking the top layer value of the visual density pyramid as the initial sample number allocated to the corresponding area in the sample allocation graph, calculating the sample number allocated to each area in the sample allocation graph layer by layer according to the data density pyramid and the visual density pyramid, and obtaining a final sample allocation graph; and selecting sample points based on the final sample distribution diagram to obtain a sampling result. A pyramid structure commonly used for image processing is used for scatter diagram sampling so as to solve the problem of over-drawing, and effect sampling points capable of keeping high relative density and abnormal values at the same time are selected by fusing a data density map and a visual density map.

The invention discloses a collaborative computing method and system. The method is applied to a system comprising a transfer module, at least one pre-computation module and at least one post-computation module, and the each pre-computation module and the each post-computation module are configured in different devices. The method comprises the following steps: after any one pre-computation moduleobtains input data of any one computation task, executing pre-processing to obtain an intermediate computation result; determining a theme type to which the computing task belongs; constructing a theme message containing the intermediate calculation result and a theme type identifier, and transmitting the theme message to a transfer module; enabling the transfer module to receives the theme message sent by each front computing module, and forwards the theme message identified by the specific theme type to the corresponding rear computing module according to the corresponding relationship between the theme type and the rear computing module; and after receiving the theme message, enabling any postposition calculation module to obtain s an intermediate calculation result through analysis, and executes postposition processing for the intermediate calculation result to complete a calculation task.

The invention provides a medical cloud platform for medical big data processing. The medical cloud platform comprises a plurality of edge computing terminal chains. The edge computing terminal chainsinclude a plurality of edge computing terminals. The medical cloud platform further comprises display control devices which are distributed in different branch consulting rooms of different floor areas of a hospital and used for displaying visual information of the edge computing terminal chains corresponding to the types of the branch consulting rooms. After a user enters the selected consultingroom for a preset time period, the edge computing terminal used for inputting the basic information of the user by the user is deleted from the added edge computing terminal chain or the newly established edge computing terminal chain. If the number of the edge computing terminals in the edge computing terminal chain exceeding the first preset value is larger than a second standard value, warninginformation is sent to a far-end cloud data processing center; and the far-end cloud data processing center sends an adjustment signal to a plurality of different types of edge computing terminals ineach floor area of the hospital. According to the technical scheme, different types of medical treatment information data can be efficiently processed in real time.