308 results about "Data synthesis" patented technology

This invention provides a process and system for a comprehensive surgical assist system, called a Therapy Imaging and Model Management System (TIMMS), which combines and integrates all of the necessary information and communication technology; workflow analysis, data processing and data synthesis; interactive interfaces between surgeon and mechatronic devices; and, cognitive agents; to provide comprehensive assistance and guidance throughout complex medical and surgical therapies, such as image guided surgery. The components of this invention, which are modular, scalable and may be distributed in location, act synergistically to provide functionality and utility that exceeds the sum of its individual parts.A method of performing surgery on a patient comprising the step of comparing a chosen patient's data to statistical data in a repository of patient data to develop a patient specific model, wherein the data comprises information from two or more sub databases selected from the group consisting of workflow data, electronic medical records, diagnostic data, biological data, measurement data, anatomical data, physiological data, genetic data, molecular data, imaging data, chemical data, clinical laboratory data, simulated data, coordinate data and surgical result and wherein the patient specific model aids in the preoperative, operative or post operative phase of surgery performed in real time on the patient.

The invention provides a medical image synthesis method based on double generative adversarial networks. The method comprises the main content of performing data management, generating the generativeadversarial networks, training a U-NET, establishing assessment indexes and obtaining processed pictures, wherein first, a DRIVE database is used to manage a first-stage GAN, then the first-stage GANgenerates a partitioning mask representing variable geometry of a dataset, a second-stage GAN converts the mask produced at the first stage into an image with a sense of reality, an generator minimizes a loss function of the true data in classification through a descriminator, then the U-NET is trained to assess the reliability of synthetic data, and finally the assessment indexes are establishedto measure a generated model. According to the method, by use of a pair of generative adversarial networks to create a new image generation path, the problem that the synthetic image contains a fake shadow and noise is solved, the stability and the sense of reality of the image are improved, and meanwhile image details are clearer.

The invention achieves stabilization of the verification performance and augmentation of the verification performance. The fingerprint data synthesis apparatus includes a fingerprint data extraction section for extracting minutiae from each of a plurality of fingerprint images to produce fingerprint data, a fingerprint data storage section for storing the fingerprint data, a common minutia searching section for investigating a correspondence of the minutiae between the plurality of fingerprint data to search for the same minutiae included commonly in two or more of the plurality of fingerprint data as common minutiae, a fingerprint data synthesis section for selecting one of the common minutiae as a minutia representative of the common minutiae to produce one synthetic fingerprint data, and a validation section for validating the synthetic fingerprint data. The apparatus is applied to production of fingerprint data used in a system which identifies an individual.

A method of integrating physiologic data including receiving physiologic treatment data from at least two bedside machines and converting the physiologic treatment data into a machine independent format. The physiologic treatment data can be presented in perceptual real-time. For example, physiologic treatment data from multiple sources can be simultaneously presented within a single graphical user interface. The graphical user interface can be interactive so that displayed graphs can be replotted and timelines revised responsive to user input, such as a point-and-drag motion.

The invention discloses a wearable personnel gait-detection indoor-positioning system and a method therefor. The wearable personnel gait-detection indoor-positioning system comprises attitude sensors, a data acquisition device, a data transmission unit and a data processing unit. The attitude sensors are installed respectively at leg parts and the waist part of a user. A first input end of the data acquisition device is connected to the attitude sensors. A second input end of the data acquisition device is connected to a correction module. The data transmission unit and the data processing unit are orderly connected to an output end of the data acquisition device. The data acquisition device acquires data measured by the attitude sensors. The data transmission unit transmits the acquired data to the data processing unit. The data processing unit carries out data synthesis processing to obtain attitude angles of the corresponding parts, carries out gait detection according to the attitude angles, acquires walk step length and a walk direction according to leg and waist length parameters of the user, and carries out vector superposition of the walk step length and the walk direction to obtain user positions in the room. The wearable personnel gait-detection indoor-positioning system and the method can simultaneously realize gait detection and positioning and has high positioning precision.

An autonomous control system generates synthetic data that reflect simulated environments. Specifically, the synthetic data is a representation of sensor data of the simulated environment from the perspective of one or more sensors. The system generates synthetic data by introducing one or more simulated modifications to sensor data captured by the sensors or by simulating the sensor data for a virtual environment. The autonomous control system uses the synthetic data to train computer models for various detection and control algorithms. In general, this allows autonomous control systems to augment training data to improve performance of computer models, simulate scenarios that are not included in existing training data, and / or train computer models that remove unwanted effects or occlusions from sensor data of the environment.

The invention discloses a method, a device and a system for compressing and decompressing transparent images and belongs to the field of digital images. The method comprises the following steps of: acquiring color channel information and transparent color channel information of original transparent images; performing loss compression on the color channel information to obtain color channel compression information, and performing lossless compression on the transparent color channel information to obtain transparent color channel compression information; and performing compression data synthesis on the color channel compression information and the transparent color channel compression information to obtain compressed transparent images. By the scheme provided by the invention, the transparent color channel information of the transparent images is subjected to lossless compression, while the color channel information is subjected to loss compression, the compressed transparent color channel information and color channel information are organized according to a specific format, redundancy contained in the image information is removed to the greatest extent, the compression efficiency of the transparent images is improved and the subjective quality of the images is not obviously reduced.

The invention provides a fully-automatic face seamless synthesis-based video synthesis method. With the fully-automatic face seamless synthesis-based video synthesis method adopted, insufficient real-time property in high-definition video processing of a face detection algorithm in the prior art can be solved. The fully-automatic face seamless synthesis-based video synthesis method of the invention comprises the following steps that: a video communication application provided by an intelligent television terminal is utilized to perform video connection; an image or video file which is locally arranged or arranged in a cloud server is adopted as a background (BG) to be synthesized; face detection is respectively performed on data foreground (FG) and background (BG) data which are acquired by a camera through using the face detection algorithm, and geometric transformation coefficients are calculated through face internal key point positioning and facial contour lines or a face minimum bonding rectangle frame; and accurate registration from the foreground (FG) to the background (BG), and face region data synthesis can be accomplished. With the fully-automatic face seamless synthesis-based video synthesis method of the invention adopted, the face image of a user can be conveniently synthesized into any existing images or videos in the process of video communication, and therefore, a sense of science and technology and interestingness can be added in the video communication, and fully-automatic seamless face synthesis of non-specific people can be realized.

A personal identification apparatus is disclosed by which personal identification can be performed making use of an impulse response obtained from a pulse signal transmitted by a living organism. A pulse generation section generates a pulse, and a speaker applies the pulse amplified by an amplifier to a finger of a user touching with the speaker and a microphone. The pulse signal is transmitted to the microphone with an attenuation characteristic that depends upon, for example, a transmission distance that depends upon the length and the thickness of the finger and characteristics of the living organism such as the thickness of the flesh and the bone density. A data synthesis processing section synthesizes the pulse and an impulse received by the microphone and amplified by an amplifier. A template recording section records a template to be used as a reference of discrimination in an authentication process together with a user ID and so forth. A comparison discrimination processing section refers to the template recording section to discriminate whether or not an impulse response inputted from a memory coincides with one of templates of registered users.

The invention relates to integrated high-speed remote sensing data receiving and processing equipment. When the equipment is used to receive data, satellite downlink serial data are converted into parallel data through serial-parallel conversion and are sent into a frame synchronous processing module for frame synchronous treatment, frame synchronization characters are searched, fault-tolerant parameters are calculated, frame synchronization signals are generated according to the frame length and the frame synchronization characters, and the frame synchronization signals and data frame are sent to a combined decode module. The combined decode module performs descramble, RS decoding and CRC check processing on the data frame according to the frame synchronization signals, and sends the data to a cache. The data in the cache are subjected to data synthesis and wave filtering processing to be sent into a data distribution module. The data distribution module sends the transmitted data into a load imaging and processing module for format analysis processing, performs classified processing on the data according to load category, and displays the data in real time. When the data are sent, uplink data are sent into the cache through a data distribution module, a combined coding module reads the data from the cache, performs scrambling, RS coding and CRC check processing on the data, sends the data into a parallel-serial conversion module, converts the data into the serial data and transmits the data to a satellite.

The invention provides a long-lead source transient electromagnetic ground-air detecting method which comprises the following steps of: detecting a deep geologic target body by a working device adopting a long-lead source transient electromagnetic terra-air detecting method to obtain transient electromagnetic observing data; converting the transient electromagnetic observing data into transient electromagnetic virtual wave data and obtaining a transient electromagnetic synthetic aperture data volume by adopting a multipoint data synthesis method to finish data reserve for carrying out fine explanation on the deep geologic target body; and obtaining a transient electromagnetic synthetic aperture imaging profile by adopting a Kirchhoff offset imaging method and a speed modeling method and processing and explaining the transient electromagnetic synthetic aperture data volume to finish fine detection on the deep geologic target body and obtain the information of the deep geologic target body. The long-lead source transient electromagnetic ground-air detecting method can be used for carrying out accurate exploration on the deep geologic target body, thereby obtaining the accurate and detailed geologic information of the deep geologic target body so as to achieve the aim of searching various deep geologic target bodies.

The present invention provides a meta-data display system capable of displaying meta data related to a shot video material along with the video material synchronously to the material. The meta-data display system includes an imaging apparatus for generating a video signal representing the video material. The meta-data display system also has a meta-data synthesis apparatus for extracting at least a part of meta data from the video signal including the meta data added for every frame and synthesizing the extracted part with the video signal. The meta-data display system also includes display apparatus each used for displaying a video signal including meta data synthesized therewith. In such a configuration, a part of meta data added to a video signal is extracted and converted into video data, which is then multiplexed in the video signal of every frame. Thus, the display apparatus are each capable of displaying not only an image represented by a video signal, but also meta data related to the image synchronously with the image.

A methodology and an apparatus for the detection of biological substances employing the integration of multiple functions and units designed into and implemented in the form of an individual silicon chip, described as a sensor unit. The deployment of a set of sensor units as a group results in a distributed detecting, discriminating, and alerting network. Distribution of the sensor units facilitates the on-the-spot detection of different biological substances such as viruses, bacteria, spores, allergens, and other toxins that can be suspended in multiple media (air, liquid, blood, etc.). Besides detection / sensing, the individual sensor units perform: data acquisition, data development, data storage, statistical analysis, and data transmission. A set of sensor units deployed in proximity to each other can be designated as a group and act as a distributed sensing network with consistent and reliable data flow to a router and further to a central computer for extended data synthesis, analysis, and decision support. The group deployment facilitates achieving enhanced security and wider sensing capability.

The invention discloses a self-adaptation flight path data correlation method comprising the following steps: establishing and maintaining a public data flight path table and a message queue, wherein the flight path table is mainly used for maintaining flight path data and comprises attribute information of flight paths per se, information of flight path points, information of sequential lengths, state information of an affiliated flight path, information of an affiliated fused flight path and the like, and the message queue comprises flight path batch numbers used for searching corresponding flight paths; and distinguishing correlation states of an input flight path into three states, namely undistributed state, undetermined distribution state and determined distribution state by automatically updating a flight path correlation sequential length M, thus enabling a flight path correlation algorithm to automatically adapt to different target distribution, different target motion states, and different tracking dimensions, tracking accuracy and tracking stability of different sensors and overcome adverse influences of the differences on correlation judgments. By applying the method, data synthesis is carried out only on input flight paths in the determined distribution state, so that the precision of the fused flight path is favorably maintained.

The application provides a method and device for generating interpretation data according to video data, a method and device for synthesizing a video and the interpretation data, and electronic equipment for synthesizing the video and the interpretation data. The method for generating the interpretation data according to the video data comprises the following steps: acquiring a video clip to be processed; identifying identifiable events and attribute parameters thereof from video frames to be processed of the video clip to be processed; and generating the interpretation data corresponding to the identifiable events according to the attribute parameters of the identifiable events. Thereby, the effect of saving the cost of program production can be achieved, and the role of enabling programsto be broadcast more quickly and timely can also be played.

The invention provides a turnout and track full-profile wear vision measurement instrument and a measurement method thereof. The turnout and track full-profile wear amount can be measured, and the whole-section turnout and steel track wear is analyzed; the measurement instrument comprises a hardware measurement system and measurement data analysis and processing software, the hardware measurementsystem comprises a laser vision measurement and acquisition device, a rotary encoder measurement and positioning device and a data transmission and power supply device, the measurement data analysis and processing software comprises a laser vision measurement and acquisition device calibration module, a laser vision measurement and acquisition device control module, a data acquisition image processing module, a measurement data synthesis module and a measurement data analysis module, and the software can be installed on a PC. The measurement instrument can achieve non-contact and automatic measurement, and the problems of large measurement time consumption and low efficiency of mechanical measurement tools are solved; meanwhile, the measurement instrument has high measurement precision, the measurement efficiency is greatly improved, and the digital management of railway maintenance is facilitated.

The invention relates to a malicious domain name detection method based on SMOTE and BI-LSTM. The method comprises the following steps of carrying out data preprocessing on domain names in a trainingset in order to obtain a domain name character sequence training set D; carrying out equalized data synthesis on the domain name character sequence training set D by utilizing an improved SMOTE algorithm in order to obtain an enhanced and equalized training set D'; constructing and initializing a neural network model comprising a BI-LSTM layer and a full-connection layer; converting the domain names in the training set D' into representation vectors with the fixed length; inputting the representation vectors of the domain names obtained in the step D into the BI-LSTM layer of the neural network model in order to obtain context vectors of the domain names; inputting he context vectors of the domain names into the full-connection layer of the neural network model in order to obtain a trainedneural network model; and converting the domain names to be judged into a character sequence, inputting the trained neural network model, and outputting a judgment result.

The invention provides a system for optimizing the screen handwriting following performance. The optimizing system comprises an input optimizing module and a drawing optimizing module. The input optimizing module is used for optimizing the tedious reading and distributing process procedure of an input event of an android system; in other words, an independent processing thread is built for a screen, and a screen coordinate event reported by a driver of the android system is directly processed through the independent thread. The drawing optimizing module is used for optimizing the bottom-layer memory allocation, data synthesis and display process procedure when the android system applies drawing; in other words, drawing processing of an application layer of the android system is transferred to a local layer of the android system to be achieved, an independent graph buffering area is applied to serve as canvas, the canvas is put to a hardware frame buffering area, and then coordinate data received through the canvas is drawn and displayed to the screen through a drawing engine SKIA. By means of the system, the input delay condition of the screen of the android system is improved, and the drawing speed can not be decreased along with improvement of the complexity of the content.

The invention relates to an aerosol optical thickness inversion method based on remote sensing data. The aerosol optical thickness inversion is conducted by the dark-object method and a 6S radiation transfer model. The aerosol optical thickness inversion method comprises data acquisition, data synthesis, angle data synthesis, cloud detection, aerosol inversion, aerosol type determination, surfacereflectivity determination, distribution map production based on AOD of the inversion, and relationship between AOD and PM2.5 establishment. MODIS AOD data obtained based on the method can be used asan effective supplementary means for ground monitoring of PM2.5 concentration.

The invention belongs to the technical field of information processing in fire-to-air management, and particularly relates to an ADS and radar information system error calibration algorithm used in fire-to-air management. The algorithm aims to both a single target in the air and multiple targets in the air. When the algorithm aims to a single target in the air, the algorithm comprises the steps that (a) a distributed multisensory system is used for receiving data and processing the data; (b) the data are sent to a fusion center, data conversion, time alignment, space alignment, track correlation and data synthesis are achieved according to track data of all sensor nodes, and then the optimal estimation of system errors is obtained. When the algorithm aims to multiple targets in the air, multi-target system error estimation and measurement data integration are achieved through recognition of the multiple targets in the air. The algorithm has the advantages that system errors can be effectively calibrated, monitoring precision is high, tracks which are detected by different radar devices and ADSs and belong to the same target are correlated, multiple radar data are fused and ADS data are further fused, accurate recognition and high-precision positioning and tracking of targets in the air are achieved, and unified and accurate management is realized.

The invention discloses a high-precision map road information rapid generation system and method based on space-time trajectory reconstruction. The system comprises a track reconstruction module, a lane determination module, a lane separation module, a lane boundary line drawing module and a multi-data synthesis fusion module, wherein the track reconstruction module performs reverse space-time reconstruction of a vehicle driving track by utilizing the vehicle track information acquired by a roadside sensing unit and a vehicle-mounted unit, the lane determination module determines a lane according to a distance between vehicle driving tracks, the lane separation module draws a lane separation line at a midpoint of connecting lines of middle points of the adjacent single lane tracks, the lane boundary line drawing module draws a lane boundary line according to the vehicle driving track direction and the driving speed by taking the lane separation line as the reference and combining the road standard width and various line widths, and the multi-data synthesis fusion module fuses the collected roadside sign information with a map. The system is advantaged in that problems that an existing diversion high-precision map drawing process is complex, and special conditions of a road surface cannot be reflected in time are solved.

The invention provides a visual compiler-based person and pose detection method. The method mainly comprises the steps of performing data synthesis of scene description; learning a net through synthesized data; defining the net by adopting basic blocks; and performing joint locating by using a pose net. According to the process, the scene description is used as an input of a visual compiler, a pedestrian detection system and a pose estimation system are trained by calibrated real data annotations, and the net is learnt through the synthesized data; then the net is defined by using two basic units including a residual module and a spatial belief module; and finally pedestrians are located by using the pose net. According to the method, the detected annotations, body part positions and segmented mask codes can be automatically obtained, the pedestrians are located by using a video camera, and the pedestrians are subjected to pose estimation and activity analysis; and the influence of illumination, shielding and the like on the detection is reduced, so that the identification efficiency is effectively improved.

Provided are a device and a method capable of efficiently performing a synthesis process of broadcast reception data and network reception data. Broadcast reception data received by a receiving device via a communication unit is set as a media source object corresponding to a processing object of an application executed by the receiving device under an application programming interface (API). The application executes a synthesis process of the broadcast reception data and network reception data received via a network as processing for the media source object. The application obtains a time offset corresponding to a time difference between an application time axis and a broadcast time axis on the basis of an API application process to execute a high-accuracy and low-delay data synthesis process.

The invention provides a data sharing method and system based on multi-cloud storage, and relates to the technical field of the cloud. The data sharing method based on the multi-cloud storage provided by the invention comprises the following steps: firstly splitting shared data needing to be shared into multiple fragmented data by a sharing end, and respectively storing the fragmented data into at least two target storage clouds; accessing the target storage clouds to acquire all fragmented data when a shared end wants to acquire the shared data, and then forming the complete shared data by use of data synthesis rule obtained in advance, namely, accomplishing the sharing process of the data. In the sharing process, the fragmented data is respectively stored in at least two target storage clouds so as to prevent the third-party from directly acquiring all data; and the third-party is difficult to know the data synthesis rule, so that the fragmented data cannot form the complete shared data even the fragmented data are acquired, and the data security is guaranteed.

The invention discloses a cognitive-technology-based video data synthesis method which belongs to the field of fault diagnosis systems of automatic systems. The method comprises the following steps of: (1) checking whether the line of a video monitoring system is normal or not; (2) carrying out data encryption and storing encrypted data into a data server of the video monitoring system; (3) implementing the positioning and the integration of video data and article electronic tag data; and (4) generating a text signal for specific characters requiring overlying in the video monitoring system by utilizing a corresponding relation of the video data and the article electronic tag data which are generated in the step (3) as well as temperature and humidity data, and meanwhile, finding the recorded text signal of the image in the data server of the video monitoring system; and comparing the two text signals. The method has the advantages of implementing the association of the video data and the data requiring overlaying, implementing data fake prevention, judging whether the line has unexpected cases or not and monitoring whether the work of monitoring flow operation staff is formal or not.

Three-dimensional data (120) is compressed at a high compression ratio without deteriorating resolution and accuracy, by computing a coupling coefficient from input three-dimensional data and a three-dimensional base data group obtained from a plurality of objects and outputting the coupling coefficient as compressed data (130). Specifically, the three-dimensional data (120) is input to corresponding point determination means (101). The corresponding point determination means (101) generates three-dimensional data to be synthesized (103) in which vertexes of the three-dimensional data (120) are made to correspond to vertexes of three-dimensional reference data (111) serving as a reference to determine association relationship between vertexes. Coefficient computation means (102) computes a coupling coefficient for coupling a three-dimensional base data group (112) used for synthesis of three-dimensional data to synthesize three-dimensional data to be synthesized (103), and outputs the computed coupling coefficient as the compressed data (130) of the three-dimensional data (120).

A data synthesis apparatus detects the start of a period of voice waveform data, stores the voice waveform data in a first storage device, starting with its part indicative of the start of the detected period. The apparatus stores in a second storage device musical-sound waveform data including information on pulses having a specified period, and then performs a convolution operation on the voice waveform data stored in the first storage device and the musical-sound waveform data stored in the second storage device, thereby outputting synthesized waveform data synchronized with the specified period of the musical-sound waveform data stored in the second storage device.

The invention relates to a malicious domain name detection method based on RNN-SPP network which comprises the following steps: carrying out data preprocessing on domain names in a training set to obtain a domain name character sequence training set D; performing equalization data synthesis on the domain name character sequence training set D by utilizing an improved SMOTE algorithm to obtain an equalization-enhanced training set D '; constructing and initializing a neural network model comprising a cyclic convolution RCNN based on spatial pyramid SPP pooling; converting the domain name in thetraining set D' into a representation vector with a fixed length; inputting the representation vector of the domain name obtained in the step D into the RCNN-SPP neural network model to obtain a feature vector of the domain name; inputting the context vector of the domain name into a full connection layer of the neural network model to obtain a trained neural network model; and converting the domain name to be judged into a character sequence, inputting the character sequence into the trained neural network model, and outputting a judgment result.