67results about How to "Increase sample" patented technology

The invention discloses a point cloud registration method for three-dimensional reconstruction. The point cloud registration method comprises the steps of S101, acquiring three-dimensional point clouds at multiple view angles, and adopting three-dimensional point clouds at two view angles as a source point set and a target point set respectively; S102, constructing a KD-tree; S103, figuring out the normal vectors of all points in the source point set; S104, calculating an average value of the included angles of the normal vectors; S105, classifying points in the source point set and setting amaximum resolution, wherein the initial resolution is 1; S106, calculating the sampling proportion of each stage in the source point set at the current resolution and extracting a sampling point; S107, figuring out the matching point of the sampling point in the target point set on the basis of the matching degree in the method; S108, calculating a rotation matrix and a translation matrix by usinga quaternion method; S109, converting the source point set to obtain a new source point set; S110, repeating the steps from S107 to S109 until a objective function is the minimum; S111, if a preset condition is met, ending the process; otherwise, adding 1 to the current resolution, and returning to the step S106. According to the invention, low-resolution matching points are used for rapidly completing the registration. The high-resolution matching points are used for improving the precision. Meanwhile, matching points are searched for through the matching degree. The registration speed and the registration precision of large-scale point clouds are greatly improved.

Sampling in the process of trans-raster distortion correction is described. The distortion parameterization is analyzed to determine the maximum inflation and deflation (magnification and minification) over the image. The maximum inflation is then used to determine the optimal resolution (dimensions in pixels) of the linear-projected image such that it is not undersampled by the output image. The maximum deflation, coupled with the optimal resolution determined in the inflation step, is then used to configure the filter used in the resampling process such that aliasing due to undersampling is minimized, while simultaneously controlling the computational burden of the filter.

The invention relates to a super-sensitivity method, primer group and kit for capturing whole genome of novel coronavirus. The method can conveniently and quickly amplify the whole genome of the novelcoronavirus by using a small amount of RNA, and can directly dock with a second-generation sequencing library-building reagent and a third-generation sequencing platform so as to obtain a whole genome sequence of the novel coronavirus.

A data acquisition system (DAS) (30) for a computed tomography (CT) scanner (12) includes a two-dimensional array (32) of detectors (34) which is arranged to detect x-rays produced by the CT scanner (12). Each detector (34) is divided into two sub-detectors (34a, 34b) along an axial direction (z). A high-speed switching circuit (40) combines selected adjacent sub-detector outputs (34a, 34b), e.g. combining a sub-detector n alternately with sub-detectors (n−1) and (n+1). The high-speed switching circuit (40) switches its configuration between DAS measurements to produce interlaced DAS output signals along the axial direction (z).

The present invention is an incremental umbrella sampling method to improve the performance of established sampling methods. It is sampling the state space by iteratively generating states _xi,t and their weighting factors represented by Formula (a) by fitting the sampling distribution function ρ_j(x) of the next iteration to at least one weighted property of the already sampled states. This means that ρ_j(x) is fitted to the product represented by Formula (b), in which Formula (a) is the weighting factor and O(_x,i) is a function respectively a property of the states x_i,t. The number of states x_i,t and the number of weighting factors (see Formula (a)) is incremented with each iteration. In order to have a consistent set of weighting factors (see Formula (a)), the weighting factors are recalculated in each iteration for all, respectively for a set of selected, states. By fitting ρ_j(x) in the state space it is possible to use all the information of Formula (a) and O(x_i,t) for the states x_i,t generated so far. The fitting step allows to use different fitting strategies. For example the fitting can bias the sampling away from areas where intensive sampling has been done in the preceding iterations, or the sampling can be directed along local gradients respectively towards local minima or maxima of one or several weighted properties. In each of the iterations, the sampling distribution function is fitted in a way to improve the overall sampling of the state space. The method supports multi-objective optimisations. State space integrals can be solved. It reduces the probability that the system is trapped. The invention is general. It can be used with different sampling methods, in particular with Monte Carlo sampling, Metropolis Monte Carlo sampling, or dynamic simulations. It can be combined with the concepts of simulated annealing and multicanonical sampling. It provides a general framework that can be adapted to the system and the observables of interest.

The invention discloses a pedestrian re-identification method based on an improved YOLOv3 network and feature fusion, and mainly solves the problems of low retrieval precision and low speed of specific pedestrians in a video monitoring scene in the prior art. According to the scheme, the method comprises the steps of 1) constructing a pedestrian picture data set; 2) establishing an improved YOLOv3network; 3) establishing a pedestrian re-identification network fusing global features and multi-scale local features; 4) training an improved YOLOv3 network and a pedestrian re-identification network by using the data set; 5) fusing the two networks trained in the step 2) and the step 3) to obtain a pedestrian re-identification system; and 6) inputting the monitoring video and the to-be-retrieved target pedestrian picture into a pedestrian re-identification system, retrieving the to-be-retrieved target pedestrian, and outputting a re-identification result of the target pedestrian. Accordingto the method, the sensitivity to pedestrians with different postures is enhanced, the retrieval speed and precision of pedestrian re-identification are improved, and the method can be used for regional security and protection, criminal investigation, video monitoring and behavior understanding.

Attenuation correction in SPECT studies such as cardiac function imaging is carried out using an iterative statistically-based transmission projection reconstruction algorithm that is capable of modeling overlapping transmission beams from a line source array of radiation emitters. Downscatter between emission and transmission photons is additively corrected for in the algorithm. Optimal line source spacing techniques and source collimation angle selection are derived to improve performance and reduce cost.

The invention relates to a verifiable multi-party kmeans federated learning method witha privacy protection function, and belongs to the technical field of data mining. The data is horizontally distributed on multiple users, and each user encrypts and uploads the respective data to the cloud server; the cloud server randomly selects an initial centroid, and calculates the square of the Euclidean distance between the data and the initial centroid by using a secure multiplication protocol and a secure distance calculation protocol; the cloud server performs distance comparison by using a security bit decomposition protocol and a security comparison protocol, and divides the data; each user updates the clustering center by using a secret sharing protocol, encrypts the clustering center and uploads the clustering center to the cloud server; and the cloud server calculates the distance between the new clustering center and the original clustering center, if the distance is smaller than a threshold value, clustering operation is ended, and otherwise, the clustering center is updated for next iteration.

The invention discloses a new energy lithium battery surface defect detection method based on adaptive deep learning. The method comprises the following steps: carrying out nonlinear mapping on a lithium battery surface grayscale image; transforming the decoupled irradiation component and reflection component to a frequency domain; performing filtering, inverse Fourier transform and exponential transform on the frequency domain data to obtain a reconstructed lithium battery image; based on morphological processing and background differencing, enhancing gray scale response at the defect; carrying out image segmentation and connected domain analysis and screening processing, and taking a result as a labeled image; designing an operator to simulate illumination details, and carrying out sample enhancement operation on the surface grayscale image of the lithium battery; training a deep convolutional neural network based on the enhanced sample image set and the labeled image; and achievinglithium battery surface defect detection based on the trained network. By utilizing the method, the detection efficiency can be improved and the false detection rate can be reduced in a lithium battery surface defect detection scene.

The invention provides an engine external characteristic data adjusting method and device, equipment and a storage medium. The engine external characteristic data adjusting method comprises the stepsthat the current actual output torque corresponding to a current rotation speed point is obtained; according to a preset mapping relation between each rotation speed point and the maximum output torque, the maximum output torque corresponding to the current actual output torque is determined; the current actual output torque is compared with the corresponding maximum output torque, so that whetherthe current actual output torque meets the engine torque adjustment requirement or not is determined; if the engine torque adjustment requirement is met, the external characteristic torque correctionvalue corresponding to the current actual output torque is determined; according to the external characteristic torque correction value, the external characteristic torque of the engine is adjusted,so that the current actual output torque meets the engine delivery requirement. By the adoption of the engine external characteristic data adjusting method and device, the method for conducting spot check and adjustment of the external characteristic data of the engine is easy to operate, risk of misoperation is avoided, the efficiency of spot check and adjustment is greatly improved.

A detection device includes a carrying unit, at least a reaction unit, a wire transporting unit, and a guiding unit. The carrying unit includes at least a hole, and the reaction unit is disposed on one side of the hole. The reaction unit includes at least a reaction area, which includes a hydrophilic material containing a detection reagent. The wire transporting unit is disposed on the other side of the hole and includes a sampling end and a transporting end. The guiding unit is in contact with the reaction unit and the transporting end. The guiding unit collects a test sample through the transporting end and provides at least a part of the test sample to react with the detection reagent. The detection device adopts the wire transporting unit to replace the pipette for collecting the test sample without directly contacting the detection area with the sample source.

The invention relates to the technical field of agricultural production, and particularly discloses a large greenhouse PD control system and method based on a neural network, and the system comprises an acquisition terminal which comprises a data acquisition control module and a plurality of sensor groups arranged in a greenhouse; the data acquisition control module is used for controlling the sensor group to acquire the internal environment information of the greenhouse in different time sequences; the user terminal is used for sending crop information of crops planted in the greenhouse to the server; an error feedback module is used for calculating an error feedback value according to the expected environment information and the internal environment information; a control parameter analysis module is used for performing control parameter analysis on the environment information and the expected environment information through a preset control parameter analysis model to obtain control parameters, and the control parameters comprise adjustment parameters of illumination intensity, temperature and humidity; and a PD control module is used for performing PD control on the environment adjusting equipment according to the control parameters and the error feedback values. By adopting the technical scheme of the invention, the time for controlling and adjusting the PD is shortened.

The present disclosure discloses a voice remote controller and a power supplying method thereof. The voice remote controller comprises: a noise gathering module for gathering ambient noise at the periphery of the voice remote controller by using a sound aperture array; a sound-electricity converting module for converting the ambient noise gathered by the noise gathering module into an electrical signal; and an energy storing module for storing the electrical signal converted by the sound-electricity converting module as electric energy and supplying power to the voice remote controller. According to the present disclosure, ambient noise is gathered by a sound aperture array and converted into electricity used as the working energy, thereby avoiding battery replacing or frequent charging. When a voice function of a voice remote controller is turned off, all of the gathered ambient noise is converted into electricity, thereby improving the energy utilization rate. By adjusting the number of opened sound apertures in the sound aperture array and changing the noise gathering angle, ambient noise can be separated from a voice signal of the user, and by inverting a noise waveform and superimposing the same onto the voice signal, ambient noise in the voice signal can be filtered out, thereby reducing the negative effects of ambient noise on the voice remote controller.

Apparatuses and methods for biopsying body vessels having small lumen, such as Fallopian tubes. In particular, described herein are apparatuses (e.g., systems and devices) and methods that evert a textile biopsying member to sample tissue from a vessel or conduit such as the fallopian tubes without applying substantial shear force.

The invention provides for a method of operating a magnetic resonance imaging system for imaging a subject. The method comprises acquiring (700) tagged magnetic resonance data (642) and a first portion (644) of fingerprinting magnetic resonance data by controlling the magnetic resonance imaging system with tagging pulse sequence commands (100). The tagging pulse sequence commands comprise a tagging inversion pulse portion (102) for spin labeling a tagging location within the subject. The tagging pulse sequence commands comprise a background suppression portion (104). The background suppression portion comprises MRF pulse sequence commands for acquiring fingerprinting magnetic resonance data according to a magnetic resonance fingerprinting protocol. The tagging pulse sequence commands comprise an image acquisition portion (106). The method comprises acquiring (702) control magnetic resonance data (646) and a second portion (648) of the fingerprinting magnetic resonance data by controlling the magnetic resonance imaging system with control pulse sequence commands. The control pulse sequence commands comprise a control inversion pulse portion (202). The control pulse sequence commands comprise the background suppression portion (104′). The control pulse sequence commands comprise the image acquisition portion (106). The method comprises reconstructing (704) tagged magnitude images (650) using the tagged magnetic resonance data. The method comprises reconstructing (706) a control magnitude images (652) using the control magnetic resonance data. The method comprises constructing (708) an ASL image by subtracting the control magnitude images and the tagged magnitude images from each other. The method comprises reconstructing (710) a series of magnetic resonance fingerprinting images (656) using the first portion of the fingerprinting magnetic resonance data and/or the second portion of the fingerprinting magnetic resonance data. The method comprises generating (712) at least one magnetic resonance parametric map (658) by comparing the series of magnetic resonance fingerprinting images with a magnetic resonance fingerprinting dictionary.