103 results about "Multislice" patented technology

A novel monolithic step-down dc-dc buck converter that uses two or more ("n") parallel slices to achieve a high output current with a small filter capacitor is provided. Each of the n slices may be operated with a phase difference of 360° / n. Each of the converter slices may be based on a synchronous rectifier topology to avoid the excessive power losses introduced by the diode component of conventional step-down buck converters. Hysteretic control may be used (with or without pulse-width modulation and pulse-frequency modulation) to provide an internal gate-drive waveform without the need to provide a dedicated clock signal or oscillator circuit. The hysteretic control is further refined using digital control techniques to enforce a brief dead time between the activation of each slice such that undesirable circulating currents are prevented. A significant advantage of the proposed multi-slice step-down dc-dc buck converter and its associated control is that the semiconductor switches, filter inductors and capacitor, and the control circuit may be fabricated as part of a single monolithic integrated circuit.

The present invention discloses a deep learning-based insulator identification method. The insulator identification method comprises the steps of pre-processing an aerial image, and secondly, extending the data via the methods, such as the geometric transformation, the contrast enhancement, an analog noise adding method, etc.; acquiring the insulator samples, aiming at the insulators of different types, classifying to acquire; determining a to-be-trained model structure; inputting the samples in the to-be-trained model, and continuously adjusting the weights and the bias parameters by the forward propagation and backward propagation methods, and finally determining an optimal model parameter, based on the trained model, taking a to-be-detected image as an input signal, and by the network multi-layer convolution, pooling and full-connection operations, obtaining a final detection identification result. According to the present invention, by a deep learning method, the insulator characteristics are learned continuously, a learning network model is determined, the different insulators are identified under different background environments, and support is provided for the electric power maintenance decisions.

Accommodative and vergence systems training method and multifocal ophthalmic lenses of Horizontal periodic or quasi-periodic Optical Power Stepless Alternating (HOPSA-lenses) therefor are proposed. Reading the text with HOPSA-lenses worn, while the head is stationary, provides a continuous alternating of accommodation and vergence strain / relaxation and, therefore, provides dynamic training of both systems. The method is applicable for human visual system therapy, eye diseases treatment / prevention and ophthalmology researches. The method enables combining the effective visual trainings with documents reading / processing, visual target examining, watching TV / video, playing computer games, etc., as well as combining the training with conventional vision correction. Several embodiments of HOPSA-lenses are disclosed, such as multi- and mono-cyclic, multi-layer, splitting the basic correction and training functions between lens' sides or layers, having left and right lenses' surfaces individually configured to provide the congruence of optical power for fixation, and / or the convergence invariability for fixation during the training.

Metrology methods and targets are provided, that expand metrological procedures beyond current technologies into multi-layered targets, quasi-periodic targets and device-like targets, without having to introduce offsets along the critical direction of the device design. Several models are disclosed for deriving metrology data such as overlays from multi-layered target and corresponding configurations of targets are provided to enable such measurements. Quasi-periodic targets which are based on device patterns are shown to improve the similarity between target and device designs, and the filling of the surroundings of targets and target elements with patterns which are based on device patterns improve process compatibility. Offsets are introduced only in non-critical direction and / or sensitivity is calibrated to enable, together with the solutions for multi-layer measurements and quasi-periodic target measurements, direct device optical metrology measurements.

A three-dimensional urban traffic road network global state prediction method under a large data environment is disclosed. Aiming at a coupling multi-layer large-scale urban traffic road network, a global traffic mode containing coupling information is defined, a method of extracting a main road network characteristic is provided, a high dimension state time sequence is converted into a discrete state type sequence, and difficulties of high dimension data in processing and prediction aspects are simplified. From aspects of traffic periodicity, delay performance and other characteristics, a historical state transfer information database containing a multielement traffic characteristic is constructed, mass data information and a data driving idea are fully used and a Markov transition probability theory is applied so as to realize multi-step global traffic state prediction. Through the global traffic prediction, a traffic state of each area can be acquired one time. Compared to an existing traffic prediction method, by using the method of the invention, a long-term prediction result is acquired rapidly and a good prediction output effect is possessed.

The invention discloses a hyperspectral remote sensing data classification method based on deep learning, and belongs to the technical field of hyperspectral data classification. The invention aims to solve a problem of low classification precision of a method for classifying hyperspectral remote sensing data with nonlinear characteristics. The hyperspectral remote sensing data classification method comprises the following steps: firstly, processing hyperspectral original data to obtain the spectral feature vector and the spatial feature information of the hyperspectral original data; then, integrating the spectral feature vector with the spatial feature information; confirming labeled samples by hyperspectral integrated data, selecting a training sample and a test sample from the labeled samples; Pre-training a multi-layer restricted Boltzmann machine which forms a deep network by the training sample; carrying out supervised learning to the network formed by the multi-layer restricted Boltzmann machine through the training sample; and inputting the test sample into the trimmed network formed by the multi-layer restricted Boltzmann machine to realize hyperspectral remote sensing data classification. The invention is used for the hyperspectral remote sensing data classification.

A system for mapping and localization includes first and second robotic devices and fiducial tags placed at predetermined positions of an environment. The quality of the first robotic device is better than the second robotic device. The first robotic device prepares a high resolution multilayer map, and the second robotic device uses the multilayer map for fast and accurate localization. The first robotic device has a visual sensor and a light detection and ranging (LIDAR) device, and is configured to prepare the multilayer map by: constructing a feature point cloud layer of the environment using images captured by the visual sensor; building a tag layer comprising identifications and locations of the tags; constructing an occupancy map using LIDAR scans; and forming the multilayer map having the feature point cloud layer, the tag layer, and the occupancy map layer, which share the same coordinate system.

The exposure device is able to supply only EUV radiation to a mask, while eliminating radiation other than the EUV radiation. A multi layer made from a plurality of Mo / Si pair layers is provided upon the front surface of a mirror, and blazed grooves are formed in this multi layer. Radiation which is incident from a light source device is incident upon this mirror, and is reflected or diffracted. Since the reflected EUV radiation (including diffracted EUV radiation) and the radiation of other wavelengths are reflected or diffracted at different angles, accordingly their directions of progression are different. By eliminating the radiation of other wavelengths with an aperture and / or a dumper, it is possible to irradiate a mask only with EUV radiation of high purity.

A cryosurgery system for application of medical-grade liquid nitrogen to a treatment area via a small, low pressure, open tipped catheter. The system includes a console, including a touch panel computer, a cryogen module, a suction module and an electronics module, and a disposable spray kit. Features include optional low cryogen flow setting to reduce the cryogen flow rate by 50%, improved cryogen flow consistency reducing pressure pulses and peaks, an integrated suction pump for improved consistency and self-checks, specified vent tube areas and corresponding maximum expected pressures during cryospray procedure; optional pressure sensing capability to monitor pressure during a treatment, and novel catheter designs of multilayer and flexible construction providing a variety of spray patterns.

The invention provides a parallel iteration-based grading and distributed video coding / decoding method and a system. The method comprises the steps as following: an original video frame is divided into a basic layer video sequence and a multi-layer strengthening layer video sequence; an H.264 coder is adopted to code the basic layer video sequence; a residual error prediction-based parallel PDWZ coder is adopted to code the strengthening layer video sequence; for the coded basic layer video sequence, an H.264 decoder is adopted to generate a low-quality low-resolution rate video sequence; for the decoded strengthening layer video sequence, a WZ decoder with iteration structure is adopted to generate a high-quality video sequence with a high resolution rate; and the multi-layer strengthening layer video sequence executes individual quantification and signal channel coding in the parallel PDWZ coder, and individual bit streams are respectively generated. The method provided by the embodiment of the invention improves the coding efficiency of the whole system and lowers dependency of the system compressing rate on GOP. The invention further provides a parallel iteration-based grading and distributed video coding / decoding system.

A cryosurgery system for application of medical-grade liquid nitrogen to a treatment area via a small, low pressure, open tipped catheter. The system includes a console, including a touch panel computer, a cryogen module , a suction module and an electronics module, and a disposable spray kit. Features include optional low cryogen flow setting to reduce the cryogen flow rate by 50%, improved cryogen flow consistency reducing pressure pulses and peaks, an integrated suction pump for improved consistency and self-checks, specified vent tube areas and corresponding maximum expected pressures during cryospray procedure; optional pressure sensing capability to monitor pressure during a treatment, and novel catheter designs of multilayer and flexible construction providing a variety of spray patterns.

A magnetic resonance method and system are provided for providing improved multi-band (MB) magnetic resonance imaging. The adaptive MB imaging can be achieved by providing one or more modified multi-band excitation pulse sequences that include at least either one nullified “dummy” slice within a slab that is not excited simultaneously with the other slices during a single multislice acquisition sequence, or one excitation slice group that utilizes a non-uniform slice spacing between simultaneously excited slices. Adaptive GRAPPA or slice-GRAPPA kernel sizes can also be used during image reconstruction to improve speed without excessive point spread blurring or MB reconstruction failure. A total leakage factor (TLF) can also be determined based on test images using modified MB excitation sequences, and used to improve the adaptive MB procedure.

Accommodative and vergence systems training method and multifocal ophthalmic lenses of Horizontal periodic or quasi-periodic Optical Power Stepless Alternating (HOPSA-lenses) therefor are proposed. Reading the text with HOPSA-lenses worn, while the head is stationary, provides a continuous alternating of accommodation and vergence strain / relaxation and, therefore, provides dynamic training of both systems. The method is applicable for human visual system therapy, eye diseases treatment / prevention and ophthalmology researches. The method enables combining the effective visual trainings with documents reading / processing, visual target examining, watching TV / video, playing computer games, etc., as well as combining the training with conventional vision correction. Several embodiments of HOPSA-lenses are disclosed, such as multi- and mono-cyclic, multi-layer, splitting the basic correction and training functions between lens' sides or layers, having left and right lenses' surfaces individually configured to provide the congruence of optical power for fixation, and / or the convergence invariability for fixation during the training.

DIR imaging of blood vessels by administering a series of DIR preparation pulse modules at a repetition interval short enough that at least two DIR preparation pulse modules generally occur within each RR interval, and by acquiring image data for a plurality of slices following each DIR module.

The invention provides a DWT- and Parametric t-SNE-based characteristic extracting method of motor imagery EEG(Electroencephalogram) signals. First, effective time and frequency ranges of EEG characteristics are determined by using a Wigner-Ville distribution and power spectrum; the EEG signals in a specific time and frequency segment is subjected to three-layer discrete wavelet decomposition and statistical characteristic quantity including the average value, the energy average value, the mean square error and the like are calculated and are taken as the time frequency characteristic of the EEG signals; at the same time, a parameterization t-SNE algorithm is utilized for performing non-linear characteristic mapping on said wavelet coefficients and embedded coordinates corresponding to a low-dimensional space are taken as the non-linear characteristic; the two characteristics are standardized and a characteristic vector including both the time frequency information and the non-linear information of the EEG signals in the specific time frequency segment is obtained. According to the invention, EEG characteristics of compactness and completeness are obtained and a method for solving a problem of poor generalization performance of a traditional manifold learning algorithm in pattern classification application through fitting a multilayer forward propagation neural network to nonlinear mapping is proposed, so that accuracy of pattern classification of MI-EEG signals is improved further.

A magnetic resonance imaging technique provides adequate MT effects by applying a shortened MT pulse and scan time. An MT pulse whose frequency is off-resonance to a region to be imaged of an object is first applied to he object, a spoiler pulse is applied to the object, and an echo signal is acquired from the region to be imaged. The duration of the MT pulse is set to a relatively shorter time than conventional. Preferably, the duration is less than 10 ms. The area of the MT pulse waveform is set to a specified value equivalent to the conventional. The MT pulse is preferably applicable to a two-dimensional scan using the multislice technique but can also be used in a three-dimensional scan.

The invention discloses an X-band high-gain broadband lens antenna based on a phase gradient multilayer super-surface structure. The antenna exhibits a broadband characteristic by using a microstrip antenna based on H-type slot coupling as a feed source. Furthermore, a super-surface structural unit having a phase gradient variation feature is designed and arranged in accordance with a parabolic focusing equation. Therefore, good low profile characteristics are obtained. A combined super-surface transmission array can convert a normally incident quasi-spherical wave into a plane wave, thereby obtaining a high-gain antenna. The lens antenna operates in the X-band with a center frequency of 10 GHz and the relative bandwidth reaches 20.6%. The antenna gain can be increased from 8.25 dB to 18.98 dB, and the side-lobe level of the super-surface focusing is less than -14.3dB. The simple three-dimensional structure and compact size provide a new method for implementing high-gain antennas, andthe antenna is good in application prospect in field of long-distance wireless communication.

A method for three-dimensional segmentation of a target in multislice images of volumetric data includes determining a center and a spread of the target by a parametric fitting of the volumetric data, and determining a three-dimensional volume by non-parametric segmentation of the volumetric data iteratively refining the center and spread of the target in the volumetric data.

An X-ray computer tomography apparatus includes an X-ray tube, an X-ray detector, a rotating mechanism, a reconstruction unit which reconstructs multislice or volume image data based on the projection data detected by the X-ray detector, a profile generating unit which generates a CT value profile in the slice direction for each pixel by using image data, a profile portion extraction unit which extracts a profile portion exceeding a predetermined threshold from each of the CT value profiles, a scattered radiation distribution estimation unit which estimates a scattered radiation distribution centered on the profile portion based on the CT value integral and width of the profile portion, and a scattered radiation correction unit which corrects the image data based on the estimated scattered radiation distribution.

The invention relates to the technical field of image processing, in particular to a method and device for smoothing seams of panoramic stitching images. By decomposing each panoramic partition image into multi-layer frequency bands, the fusion of images is carried out separately in different frequency bands, which can effectively avoid ghost images during image fusion, and achieve smoothness at the seams of panoramic stitching images, and due to the multi-layer frequency bands of images Decomposition and reconstruction are a reciprocal process, and using this method can guarantee the correctness of the resulting stitched image. The generation of panoramas meets the requirements of video surveillance with a large field of view. With the continuous expansion of the monitoring range, it is necessary to monitor the entire bridge, airport, a long section of road or high-rise buildings. The field of view of most cameras cannot reach this A large monitoring range has gradually formed a high-end demand. Smoothing the seams of panorama stitched images can make the panorama look smoother.

This invention discloses a real-time phased array 3D acoustic imaging sonar processing system and method. The system comprises a 3D sonar dot data collection module, a 3D sonar dot data uploading module and a real-time 3D sonar dot data processing module of PC terminal; the method includes the following content: make use of two-stage FPGA signal processing to realize real-time formation of electronic focusing wave beam, and acquire 3D sonar dot data; further transmit 3D sonar dot data to PC terminal for setting and acquisition of directives; PC terminal aims to receive 3D sonar dot data for real-time processing: This includes real-time single-frame multilayer reconstruction, image registration and splice, 3D visualization and real-time saving of data. This invention features in compact structure, high real-time precision, clear image, easy alternation and high extendibility, which has effectively realized the real-time 3D processing of sonar targets.

In a method and magnetic resonance (MR) apparatus to image a partial region of an examination subject by means of a multislice measurement, which partial region includes at least two measurement slices, and is located at least in part at the edge of a field of view of the magnetic resonance apparatus, for each voxel to be optimized that is located at the edge of the field of view, a gradient field is configured for each measurement slice of the partial region that is to be measured and is used to acquire magnetic resonance data in the multislice measurement. The gradient field is configured so as to cause a nonlinearity of the gradient field and a B0 field inhomogeneity to cancel at each of the aforementioned voxel to be optimized at the partial region at the edge of the field of view. An image of the partial region of the examination subject is determined from the magnetic resonance data acquired in this manner.

DIR imaging of blood vessels by administering a series of DIR preparation pulse modules at a repetition interval short enough that at least two DIR preparation pulse modules generally occur within each RR interval, and by acquiring image data for a plurality of slices following each DIR module.

An MRI method includes: performing a first data acquisition block of a pulse sequence to acquire a first MR data from a plurality of slices of a subject during a period of fully recovered longitudinal magnetization within the plurality of slices disposed at different locations in the subject; performing a second data acquisition block of the pulse sequence including a magnetization preparation module followed by a recovery period and an imaging sequence executed during the recovery period, to acquire a second MR data from the plurality of slices during the recovery period; and generating a T1 map of the subject based on the first MR data and the second MR data, of the plurality of slices.

In a method and system to adjust at least one shim current in a shim channel of a magnetic resonance apparatus and an associated RF center frequency for the radio-frequency system of the magnetic resonance apparatus during an interleaved multislice MR measurement of a moving examination subject, in which MR measurement at least two excitations are implemented for complete acquisition of the desired data of a slice of the examination subject that is to be measured, the multislice MR measurement is implemented such that phase discontinuities between measurement data acquired after individual excitations are avoided.

The invention discloses a fast directional multilevel algorithm for analyzing a planar microstrip circuit. The planar microstrip circuit structure is analyzed based on the fast directional multilevel algorithm. In the algorithm, a complex circuit is divided by a planar triangular surface element, and the fast directional multilevel algorithm is combined with a Rao-Wilton-Glisson (RWG) function to be applied to an electric field integral equation to ensure the calculating accuracy of a model; and by utilizing a principle that an impedance matrix formed by a moment method when the field-to-source point distance is long enough has good low-rank characteristic, fast directional multilevel calculation is adopted for a far field area, a Green function is unfolded by a low-rank expression, the expansion is only related to calculation of a kernel function so as to greatly reduce the calculating complexity of the multilayer microstrip circuit, and the calculating complicity and internal memory demand are reduced to O(NlogN) magnitude. A quadtree form is also adopted for grouping analysis of the planar microstrip multilayer circuit, the consumption of the internal memory is effectively reduced, the calculation result is accurate, the testing cost is low, and the fast directional multilevel algorithm can be widely applied to simulation analysis of complex circuits.

The invention discloses a magnetic resonance diffusion imaging method for integration and reconstruction based on a Gaussian diffusion model acting as an instance. The method comprises the steps that signal acquisition is performed on a tested target based on multilayer simultaneously excited preset sequences; phase estimation is performed on the acquired under-sampled signals through a parallel imaging technology; the Gaussian diffusion model is established through the estimated phase, the acquired under-sampled signals and a reference image without diffusion weight; the under-sampled signals of all the directions are integrated according to the Gaussian diffusion model, and a target equation is established; the target equation is iteratively solved by using a nonlinear conjugate gradient algorithm so as to obtain a diffusion tensor parameter; and a diffusion coefficient and a diffusion weight image are calculated according to the diffusion tensor parameter. Therefore, high acceleration acquisition of magnetic resonance diffusion tensor imaging can be realized so that the acquisition time can be effectively reduced, the diffusion tensor parameter can be accurately estimated to obtain the diffusion image of high signal-to-noise ratio and high resolution, and the requirement of clinical application can be met.

The invention discloses a path planning method for a non-continuous grid division three-dimensional point cloud. According to the method, the three-dimensional point cloud serves as input, a three-dimensional printing path serves as output, the non-continuous grid division three-dimensional point cloud is adopted for path planning, the precision and the calculation amount of path planning are controlled through a plurality of parameters, the parameters comprise two types, namely basic parameters and related technology parameters, the basic parameters comprise grid dimension parameters and gridspacing parameters, and the related technology parameters comprise grid center point sparsification parameters, line spacing parameters, boundary shrinkage parameters, multi-layer printing parametersand nozzle lifting parameters. The method has the advantages that path planning is directly carried out through the three-dimensional point cloud, 3D modeling is not needed, a non-continuous grid canreduce the calculation amount on the premise that the precision is guaranteed, the problems of different printing technologies, point cloud cavities, orthogonal paths and the like are taken into consideration in the parameter setting process, and the method can be applied to the various printing technologies.

The invention provides a magnetic resonance T2 * weighted rapid imaging method and device which relate to the magnetic resonance imaging technology field wherein the method comprises: according to single-band radio frequency pulses, generating multi-band radio frequency pulses; according to the fast parallel imaging technique based on controllable mixing and superposing, modulating the initial phases of the multi-band radio frequency pulses in different repetition times to generate the modulated multi-band radio frequency pulses; through the layer separation scanning and echo migration technique, generating the T2 * weighted imaging sequence based on multi-layer co-excitation technique and the layer separation scanning and echo migration technique; obtaining the multi-layer mixed and superposed images of a to-be-processed object; and using the reconstruction algorithm to reconstruct and process the multi-layer mixed and superposed images to determine the single layer T2 * weighted images. The method and apparatus of the invention can further accelerate the layer separation scanning and echo migration technique through the multi-layer co-excitation technique to realize faster T2 * weighted imaging to ensure the signal-to-noise ratio and reduce the influence of the artifacts and distortions on the image quality. The use of small flip angled radio frequency pulses effectively reduces the specific absorption rate.

A 3D image reconstructing method based on dual-source dual-spiral multi-layer spiral CT features that two x-ray sources, two structures of multi-row detectors and two spiral scan paths are used to increase the projection data acquisition speed. It includes such steps as calculating the projection data needed by reconstructing a planar image perpendicular to Z axis by using interpolation algorithm of the multi-layer projection data on two paths, reconstructing the planar image by 2D filter-reverse projection algorithm, and reconstructing a 3D image with a series of said planar images.