43 results about "Probability mapping" patented technology

A method for assessing breast density executed at least in part by a computer system, identifies breast tissue from the electronic image data for at least one mammographic image, then performs an initial segmentation of fibroglandular tissue within the breast tissue according to at least one of gradient and uniformity data that is derived from the image data. The initial segmentation is refined using a pixel clustering process. A localized segmentation is obtained from the refined segmentation by generating and combining a density probability mapping and a homogeneity mapping from the image data. A percent density value for the at least one image is calculated and stored in a memory.

The invention provides a method for identifying a trip mode based on mobile phone signal data. The method is characterized by comprising the following steps of: establishing a virtual sensor network; then configuring reasonable value ranges of three properties of five trip modes according to prior knowledge, wherein the value ranges are represented by using segmented linear functions and are called as probability mapping functions; acquiring the mobile phone signal data; computing a probability numerical value based on a line by using path matching based on the virtual sensor network; computing probability numerical values based on three properties such as distance, time consumption and speed; and judging which trip mode a mobile phone user belongs to by using the probability numerical values. The method has the advantages that: the conventional communication network resource is fully relied on, a method which realizes flexible structure, easiness of expansion and simple computation is adopted, the mobile phone signal data of a single complete trip process of a traveler are taken as inputs, automatic identification of five conventional trip modes is realized, and trip mode-relatedinformation with high timeliness is further acquired.

The invention relates to a skin identification method based on a skin color model. The skin identification method comprises the steps of (1) carrying out face identification on an image, and obtaining a face area, (2) carrying out mean value computing on the obtained face area obtained in the step (1), and obtaining a mean skin color, (3) computing a skin color probability mapping table of the current image according to the mean skin color obtained in the step (2), and (4) carrying out skin color identification on the current image according to the skin color probability mapping table obtained in the step (3), and obtaining a result graph of the skin color probability of the current image. According to the skin identification method, the skin color mean value after face detection is used for building the skin color model, the skin identification method can adapt to images shot under various conditions, and wide covering of the application range is achieved. In addition, according to the skin identification method, training data are not required to be accumulated, implementation can be easily achieved, comparison with huge training data is not required, and the operation efficiency is improved. Identification is carried out through the skin color model, the identification accuracy is high, and the defect that under excessively bright or excessively dark conditions, skin identification is not accurate is overcome.

A T1WI-fMRI image tumor collaborative segmentation method based on 3D-Unet and graph theory segmentation is provided, the method comprises the following steps: step 1, obtaining a T1WI and fMRI tumorsegmentation data set, and carrying out preprocessing; Step 2, generating a training sample; Step 3, training two independent 3D-Unet network; carrying out tumor and tumor sub-region segmentation onthe T1WI image and the fMRI image respectively, and helping high-quality voxel-level tumor and tumor sub-region / non-tumor masks and probability graphs to be generated through the strong description capability of the network; and step 4, carrying out fine segmentation based on a graph theory method, and simultaneously generating final tumor and tumor subregion segmentation results on the T1WI and fMRI images by using the two obtained probability mapping graphs and the coarse segmentation mask in a continuous graph segmentation model with label consistency constraint. According to the invention,automatic and accurate description of tumors and tumor subregions (including tumor necrosis, active tumors and tumor peripheral edema) is realized.

The invention relates to a system for achieving transient signal capture and spectral analysis. The system comprises a superheterodyne receiver and a digital intermediate frequency capture and analysis module. The digital intermediate frequency capture and analysis module comprises an analog-digital conversion unit, an IQ modulation unit, a rapid Fourier transform unit and a mapping and recording unit. The mapping and recording unit is used for calculating the signal amplitude of each frequency component output by the rapid Fourier transform unit, conducting probability mapping on the signal amplitudes and recording the appearing frequency of frequency points. According to the system for achieving transient signal capture and spectral analysis, by the adoption of the structure, the information of the frequency, amplitude and occurrence probability and time of signals can be analyzed and recorded at the same time, the method for comprehensively analyzing suddenly-burst signals is provided, the occurrence probability and the specific time of the signals are analyzed from the perspective of the frequency domain, four-dimensional analysis can be carried out on the signals, and the system is widely applied in the actual signal analysis, has good flexibility and expandability and is applicable to large-scale popularization and application.

The embodiment of the invention provides a modulation method and device based on dual-path multi-directional probability mapping. The method comprises the following steps: performing serial-to-parallel conversion on a to-be-modulated signal to generate six sub-signals, performing probability mapping on four sub-signals to obtain a mapping signal set; determining an amplitude mapping result corresponding to a signal value of first two mapping signals in the mapping signal set at each moment according to a first constellation diagram, and determining a location mapping result corresponding to the signal value of last two mapping signals in the mapping signal set and two sub-signals in the six sub-signals except four sub-signals at each moment according to the second constellation diagram, and acquiring an APPM signal according to the amplitude mapping result and the location mapping result. Through the modulation method provided by the embodiment of the invention, the to-be-modulated signal is mapped as the mapping signals with different probabilities, and then the APPM signal is acquired according to the first constellation diagram and the second constellation diagram, a complicatedconstellation diagram is unnecessary to design, and the complexity of the signal modulation can be lowered.

The embodiment of the invention discloses an image semantic segmentation optimization method and device, a storage medium and a terminal. The method comprises the steps of obtaining superpixels in a to-be-segmented image; obtaining a probability mapping graph of the to-be-segmented image, and determining probability distribution information of a label class of the superpixels according to the probability mapping graph; determining a target superpixel meeting a set condition according to the probability distribution information, and adjusting a label class of a pixel point in the target superpixel; constructing a conditional random field model based on the adjusted target superpixel and the remaining superpixels in the to-be-segmented image, and determining an image semantic segmentation result of the to-be-segmented image according to the conditional random field model, thereby realizing optimization of the segmentation result of the to-be-segmented image according to the conditional random field, and improving the segmentation effect of the image semantic segmentation result at the boundary.

The invention relates to a visible light communication system and method based on probabilistic shaping coding, and application equipment, and relates to the field of light communication systems. Bitsof input data are grouped at a sending end, and mapped into new bits according to the set probability; then, forward error correction coding is carried out; the coded data forms an OFDM frame; then,a data flow is formed and sent; and a receiving end receives the data flow from the sending end, recovers the OFDM frame, performs inverse mapping on the decoded data after performing OFDM frame decoding and forward error correction decoding, and recovers the original data bits by using a probability mapping rule corresponding to the sending end. Due to addition of probability mapping, the net information rate of the system can be adjusted; the communication capability of the system can be increased; adjustment is carried out flexibly; and furthermore, the complexity is relatively low.

The invention discloses a skin recognition method based on saliency detection. The skin recognition method comprises performing face recognition and saliency detection on an image to be detected to obtain a face skin region, calculating the face skin region to obtain an average skin color and a skin color probability mapping table, and finally, performing skin color recognition on the image to be detected according to the skin color probability mapping table and obtaining the result map of the skin color probability of the image to be detected; the skin recognition method has the advantages that the collection of huge training data is avoided, the recognition result is not affected by the deviation of the training data, and as saliency detection is performed, non-skin regions in the face region can be excluded, and therefore, the accuracy of recognition is higher and the problem of misrecognition due to that the image to be detected is too bright or too dark is overcome.

The advantage of the present invention is to appropriately detect the object. The object detection apparatus in the present invention has a plurality of cameras to determine the distance to the objects, a distance determination unit to determine the distance therein, a histogram generation unit to specify the frequency of the pixels against the distances to the pixels, an object distance determination unit that determines the most likely distance, a probability mapping unit that provides the probabilities of the pixels based on the difference of the distance, a kernel detection unit that determines a kernel region as a group of the pixels, a periphery detection unit that determines a peripheral region as a group of the pixels, selected from the pixels being close to the kernel region and an object specifying unit that specifies the object region where the object is present with a predetermined probability.

The invention provides a high-resolution remote sensing image vegetation extraction method based on sensitive feature focusing perception, which is carried out on the basis of a weighted fusion probability mapping graph and comprises the following steps of: processing a satellite image by adopting a machine learning algorithm, and manufacturing a vegetation training label image; setting the size of an input image according to the semantic segmentation network, and making a training image and a label data set through cutting and segmentation; coupling an adaptive affinity field and an attention mechanism into the semantic segmentation network, constructing an adaptive context reasoning and vegetation sensitive feature focusing sensing module, and training the semantic segmentation network based on the training sample set to obtain a trained vegetation element extraction model; predicting the test set based on the trained network model to obtain a predicted plaque image; and based on the obtained predicted plaque image, carrying out splicing by using a weighted average strategy, recovering the resolution of the area before cutting, and eliminating the splicing seam effect. According to the method, end-to-end automatic extraction of different types of vegetation regions of the high-resolution image can be realized.

The invention discloses an edge computing network task unloading method equipped with a cache auxiliary device. The method comprises steps of establishing execution mode of a user task into differentunloading decision variables to compute energy consumption of the user task in different execution modes such as local CPU processing, edge server execution and content cache processing; constructingan energy consumption value function for minimizing the worst case of a user during task execution, converting the energy consumption value function into a standard QCQP model, and obtaining a cache-assisted pre-selected unloading set by using a semi-definite relaxation and random probability mapping method; and finally, optimizing the transmission power and edge computing resource allocation in the preselected unloading set by adopting a Lagrange dual decomposition method and a dichotomy, and further obtaining an ideal unloading decision by comparing the equipment energy consumption in the set.

A method for multi-direction search of dynamic system state probability mapping matrix include the following: establishing the model of system simulation and fault injection, performing the discretization of system state space, randomly sampling a system originate grid element point, configuring system component status, setting simulation time step length and simulation test times, importing system originating grid element point information into simulation model through function assignment, executing and completing simulation test, counting system trajectory migration distribution in grid element, calculating system state conditional migration probability, generating system state transition probability matrix, and establishing system dynamic probability mapping network; initializing the search parameters, carrying out matrix dynamic search analysis, realize the path identification and search of the important risk sequence, calculate and predict the probability distribution of the system reliability with time. The invention can be used for generating system dynamic trajectory and state transition probability matrix, supporting matrix multi-direction searching calculation and obtaining dynamic reliability evaluation index.

This disclosure provides systems, methods and apparatus, including computer programs encoded on computer storage media, for sensing touch and / or gestures in the near-field area overlying a light-guiding layer within a device or other optical sensing system. In one aspect, modulated infrared light is emitted into the overlying area, and the light reflected by objects within the overlying area is redirected through the light-guiding layer to infrared sensors. In one aspect, a masking structure can be located between the light-guiding layer and the infrared sensors. In some aspects, probability mapping or backtracing can be used to estimate the locations of objects within the overlying area.

The invention relates to a novel photon probability mapping method based on terrace forming, which belongs to the technical field of mapping method. The method comprises the following steps: matchingaccording to channel conditions and generating an optimal Gaussian beam by a laser; performing probability matching by a probability matcher, and changing the shape of the Gaussian beam by a spatial phase array to obtain a novel beam formed in terrace; performing the 64QAM projection mapping on the data by the novel beam; concentrating and converging the 64 constellation point optical signals; performing pre-emphasis; amplifying the optical signals after passing through a channel, detecting by a photodetector, performing low-pass filter, orthogonalizing and normalizing, and performing data convergence and equalization; matching probabilistic solutions, removing step noise by wavelet transform analysis, performing constellation demapping; estimating channel conditions according to the received data, feeding back channel state information to the probability matcher, and performing real-time dynamic regulation on the probability matcher. The novel photon probability mapping method based on the terrace forming can effectively reduce the signal transmission power, reduces the bit error rate, improves the frequency band utilization, and lengthens the transmission distance.

An evaluating method for internal network performance by simulation of a flow network model with an acceptable threshold for the required flow of the internal network is applied to a computer system. The network model comprises plural nodes and document flow paths, wherein each said node corresponds to one employee of the internal network with a capacity, which is tested to get a testing result, and each said document flow path has a flow, which maps into a probability distribution data by a probability mapping function. According to the probability mapping function and operation relationship between the capacity and the flow together with the acceptable threshold and the probability distribution data, a system reliability value is calculated out and shown in a display screen.

The invention discloses a time frequency dynamically changing multi-frequency and multi-probability optical millimeter wave generation method comprising the following steps: an optical millimeter wavegeneration system is prepared; optical multi-carriers are generated, a crystal oscillator is used as a clock source, and the adjustable local oscillator signal is outputted through radio frequency tobe modulated with the pulse optical signal emitted by a pulse light source in a cyclic frequency shifter; the optical multi-carriers are grouped; and different probability mapping is performed on thegrouped optical multi-carriers. The better spectral efficiency and the better anti-noise performance under the condition of constant emission power can be obtained.

The invention belongs to the technical field of wind power generation, and particularly relates to a wind speed power probability mapping method in wind power forecast. The wind speed power probability mapping method comprises the following steps of acquiring historical running wind speeds, power data and information of a wind power generator; equally dividing a normal wind speed interval (0-25m / s) into a plurality of wind speed subintervals, and solving a probability density function and an accumulation distribution function of the interval power distribution with regard to each wind speed subinterval; and solving a mapping power value P according to a reverse function of the accumulation distribution function of the interval with regard to a given wind speed v, and obtaining a final result by calculation for many times to perform average value taking operation, wherein the result is the power mapping value P of the wind power generator under the wind speed v. By employing the method,a conventional wind power curve mapping method in wind power forecast can be substituted, the probability and the diversity which are considered during the forecast process are improved, and the actual running state of a fan can be truly reflected.

The invention provides a processing method of case report information. The method specifically includes: acquiring geographic keywords in case report data of each case of a historical case set; acquiring information of a grid center to which each case of the historical case set is delegated; generating a geographic location-based grid-center delegation probability table on the basis of the geographic keywords of each piece of case report data and the grid center to which each case is delegated, wherein the grid-center delegation probability table includes a delegation probability mapping relationship of each keyword and each grid center; acquiring geographic keywords in case report data of a new case; and executing grid-center recommendation on the basis of the geographic keywords in the case report data of the new case and the grid-center delegation probability table to determine a grid center to which the new case is delegated. The invention also provides a processing system of the case report information, computer equipment and a computer-readable storage medium.

The invention discloses a structure early warning analysis method based on a Bayesian model. Firstly, the invention provides a high-rise building structure parameter identification method based on Bayesian inference. The method comprises the following steps: constructing a Bayesian model likelihood function by utilizing instantaneous frequency and instantaneous energy of a structural response signal intrinsic mode function of a high-rise building; obtaining the instantaneous frequency and instantaneous energy of the intrinsic mode function by conducting empirical mode decomposition and Hilberttransform on the structural response signals of the high-rise building. Compared with a traditional method, the method has the advantages that the calculation complexity is greatly reduced, and the calculation efficiency is improved; a probability mapping relation model between multiple structural parameters and the environment temperature, the relative humidity and the average wind speed based on the Bayesian model is established; therefore, the analyzed and studied change mechanism of different high-rise building structure parameters along with a plurality of environmental factors and the proposed high-rise building structure health condition comprehensive early warning analysis method based on the plurality of structure parameters are more accurate and reliable.

The invention relates to a Chinese cross-language abstract method fusing word granularity probability mapping information, and belongs to the technical field of natural language processing. The method comprises the following steps: collecting corpora; performing word segmentation preprocessing on the collected corpus; using a fast-align tool and a statistical thought to obtain a Chinese and Chinese probability mapping pair; adopting a coding and decoding attention mechanism to obtain keywords based on the abstract of the Chinese article; constructing a probability mapping mechanism; fusing word-level probability mapping information; According to the method, Chinese-cross word granularity information and chapter-level texts thereof are represented; secondly, based on an attention mechanism, carrying out joint characterization on information of word granularity and a text-level text; and finally, by fusing the alignment information of the word granularity to the abstract of the target language, the accuracy of the abstract is improved. Experiments are carried out on a Chinese-cross language abstract data set, and the experiments prove the effectiveness and superiority of the method.

The invention discloses an ultrasonic image recognition method, terminal equipment and a storage medium. The method comprises the steps of obtaining a probability mapping feature map corresponding toa to-be-recognized ultrasonic image; determining and obtaining a plurality of candidate feature points corresponding to the ultrasonic image based on the probability mapping feature map; and updatingthe probability mapping feature map based on the plurality of candidate feature points, and determining a target area corresponding to the ultrasonic image based on the updated probability mapping feature map. According to the method, the probability mapping feature map of the ultrasonic image is acquired, a plurality of candidate feature points are corrected based on the probability mapping feature map to update the probability mapping feature map, and finally ovarian follicles in the ultrasonic image are segmented. In this way, the accuracy of the target area determined based on the probability mapping feature map can be improved, and therefore doctors can better observe and analyze the development situation of ovarian follicles.

The invention provides a hardware SAT solver for multi-thread parallel execution. The hardware SAT solver comprises a first register module, a first evaluation module group, a buffer module group, a counter module, a variable element overturning module, a probability mapping table, an FIFO tree group, a second register module group, a second evaluation module and a third register module. All the data is stored in an on-chip memory of an FPGA; time expenditure caused by reading off-chip data is avoided. Meanwhile, a multi-thread strategy is utilized; different threads interact with the same address mapping table and clause mapping table. Compared with the prior art, the invention has the advantages that the resource overhead of the on-chip memory is greatly reduced, the total flipping amount per second is multiplied due to the use of a multi-thread strategy, the solution of the problem is more likely to be found in the specified time, and the relatively large-scale SAT problem can be solved more efficiently and conveniently.

The embodiment of the invention discloses an object recognition method and device, a network training method and device, equipment and a medium. The object recognition method comprises the following steps: acquiring a to-be-recognized video frame of which a picture comprises a target object; the to-be-identified video frame is any video frame in a video stream of the target object; determining an initial attitude sequence of the target object based on a to-be-recognized video frame and a historical video frame of the to-be-recognized video frame in the video stream; performing probability mapping on the initial attitude sequence to obtain a target attitude sequence of the target object in the to-be-identified video frame; performing feature conversion on the target attitude sequence in space and time to obtain an attitude feature track of the target object in the to-be-identified video frame; and determining a behavior state of the target object in the to-be-identified video frame based on the attitude feature trajectory.

The invention discloses a fiber laying process parameter model construction method and system based on imitation learning. The system comprises a manual laying demonstration module and the like, and is characterized in that a demonstrator holds a small fiber laying device to carry out the fiber laying operation on the surface of a die, an action capturing module records the real-time position of alaying point on the die, a video recording module records the operation process in real time, a process parameter sensing module records the fiber laying process parameters in real time, the data istransmitted to an industrial personal computer for data storage and processing, the learning process of manual laying teaching operation is completed by simulating a learning algorithm module, a probability mapping model of the process parameter-surface curvature is output, According to the automatic fiber laying system based on the cooperative mechanical arm, the learned model result is combinedwith an intelligent control algorithm, and automatic laying work of fibers on a curved surface is achieved. According to the invention, the control precision of the laying path and process parametersin the composite material laying process is improved, so that the laying defects are reduced.

The invention provides a data sending method, a data receiving method, a terminal, a system, equipment and a readable storage medium, and the data sending method comprises the steps: obtaining to-be-sent source data; probability shaping processing is conducted on the source data, probability mapping data are obtained, the probability mapping data comprise shaped data and unshaped data, the shaped data are stored in a shaped data array, and the unshaped data are stored in an unshaped data array; performing pre-interleaving processing on the probability mapping data to obtain interleaved data; performing forward error correction coding processing on the interleaved data to obtain verification data, and performing post-interleaving processing on the interleaved data and the coded data to obtain data to be modulated; and performing modulation processing on the data to be modulated, and obtaining modulation symbol data. According to the embodiment of the invention, the signals can be arranged according to the preset mode in the signal interleaving process, so that the signals can be modulated and mapped, and shaping gain is realized.

The embodiment of the invention provides an input method and device based on sample probability quantization and electronic equipment, and the method comprises the steps of obtaining user input information, and carrying out the calculation to obtain candidate words; performing probability prediction calculation on the candidate words to obtain probability values of the candidate words; inputting the probability values of the candidate words into a mapping function to obtain probability mapping values corresponding to the candidate words, wherein the mapping function is used for mapping the probability value into a specified probability mapping value domain range and adjusting the dispersion degree of the probability mapping value to be an expected dispersion degree in the specified probability mapping value domain range, and the probability value and the probability mapping value are in a one-to-one mapping relationship; rounding the probability mapping value to obtain a probability mapping quantized value; and determining a sorting order of the candidate words according to the probability mapping quantized value so as to output a candidate word list according to the sorting order. According to the embodiment of the invention, the probability value distortion degree after quantization is reduced, so that the candidate word list sequence determined based on the probability value after quantization is consistent with the candidate word list sequence before quantization as far as possible.

The invention provides a maritime target trajectory prediction method and device based on multi-source data. The prediction method comprises the following steps: acquiring multi-source sensing data of a maritime target; on the basis of the multi-source sensing data, track grid mapping coding is carried out on the marine target; performing data fusion on the multi-source sensing data in the track grid within a preset time period; obtaining an input sequence based on the sensing data after data fusion, and inputting the input sequence into a prediction model; and calculating through a preset model to obtain an output sequence of marine target trajectory prediction. Aiming at the problem that a ship target cannot be continuously tracked under incomplete observation data information expansion, complementarity of multi-source data in ship target monitoring is fully utilized, confidence of multi-source means is considered, and through key technologies such as track grid mapping coding, multi-source data fusion and a prediction model, the ship target can be continuously tracked. The existence probability mapping of the marine target trajectory in a long-time sequence and a large-scale space under global grid subdivision is realized, and the prediction analysis of the marine target trajectory is realized.

An evaluating method for internal network performance by simulation of a flow network model with an acceptable threshold for the required flow of the internal network is applied to a computer system. The network model comprises plural nodes and document flow paths, wherein each said node corresponds to one employee of the internal network with a capacity, which is tested to get a testing result, and each said document flow path has a flow, which maps into a probability distribution data by a probability mapping function. According to the probability mapping function and operation relationship between the capacity and the flow together with the acceptable threshold and the probability distribution data, a system reliability value is calculated out and shown in a display screen.

The invention discloses a SAR image water body probability mapping method combined with prior probability estimation, comprising the following steps: step 1, establishing a statistical model assumption for the distribution of backscatter coefficients of SAR image pixels; step 2, estimating the prior probability of water body distribution; Step 3, according to the backscattering coefficient σ of the image in the study area 0 Estimate the distribution parameters; step 4, calculate the conditional probability that the pixel belongs to the water body. The present invention uses Bayesian inference to make a Gauss distribution assumption on the backscattering coefficient of the image pixel in the research area, and then combines the k-means clustering algorithm to classify the pixel into water body and non-water body, and calculates the proportion of water body pixels in the research area as the water body Finally, combined with the estimated value of the prior probability, the theoretical probability density function of the backscatter coefficient is superimposed on the statistical distribution histogram, and the nonlinear least squares fitting is used to complete the model parameter estimation, and the research area is obtained The probability that each pixel of the image belongs to a water body.