194results about How to "Avoid noise effects" patented technology

The invention discloses a method and a device for allocating and determining a quasi-co-location. The method comprises the steps of: acquiring all currently allocated CSI-RSs; allocating a corresponding QCL-CSI-RS for each set of CSI-RSs of all the CSI-RSs separately, wherein the QCL-CSI-RS is used for estimating a channel large scale feature parameter in conjunction with the CSI-RS; and allocating all the CSI-RSs and the QCL-CSI-RSs corresponding to various sets of CSI-RSs to a terminal. According to the method and the device provided by the invention, the CSI-RS can be prevented from suffering the influence of noise when being used for estimation of frequency deviation and frequency extension, and the CSI-RS can be directly used as a cell discovery signal.

The invention discloses an image type fire flame identification method. The method comprises the following steps of 1, image capturing; 2, image processing. The image processing comprises the steps of 201, image preprocessing; 202, fire identifying. The fire identifying comprises the steps that indentifying is conducted by the adoption of a prebuilt binary classification model, the binary classification model is a support vector machine model for classifying the flame situation and the non-flame situation, wherein the building process of the binary classification model comprises the steps of I, image information capturing;II, feature extracting; III, training sample acquiring; IV, binary classification model building; IV-1, kernel function selecting; IV-2, classification function determining, optimizing parameter C and parameter D by the adoption of the conjugate gradient method, converting the optimized parameter C and parameter D into gamma and sigma 2; V, binary classification model training. By means of the image type fire flame identification method, steps are simple, operation is simple and convenient, reliability is high, using effect is good, and the problems that reliability is lower, false or missing alarm rate is higher, using effect is poor and the like in an existing video fire detecting system under a complex environment are solved effectively.

The embodiments of the invention provide a fault diagnosis method and apparatus for a rolling bearing. The fault diagnosis method for the rolling bearing comprises the following steps: acquiring vibration signals of the rolling bearing, and solving a time-frequency spectrum of the vibration signals; according to the time-frequency spectrum, determining an optimal order component of the vibration signals; according to the optimal order component, performing angle re-sampling on the vibration signals to obtain angle domain signals; and performing envelope analyzing on the angle domain signals, and determining fault positions of the rolling bearing, wherein determination of the optimal order component comprises: setting an equivalent frequency conversion scope and searching for all order components from the time-frequency spectrum; and calculating order energy of the order components and determining the optimal order component from the equivalent frequency conversion scope. The fault diagnosis method and apparatus for the rolling bearing realize automatic and rapid determination of the optimal order component so as to provide a basis for fault diagnosis of the rolling bearing, improve the fault diagnosis accuracy and avoid interference of manual factors.

The invention discloses an image enhancement and partition method, which comprises the following steps of I, image enhancement: a processor and an image enhancement method based on fuzzy logic are adopted for performing enhancement processing on an image needing to be processed, and the process is as follows: i, converting from an image domain to a fuzzy domain: mapping the grey value of each pixel point of the image needing to be processed to a fuzzy membership degree of a fuzzy set according to a membership function (shown in the specifications); ii, performing fuzzy enhancement processing by utilizing a fuzzy enhancement operator in the fuzzy domain; iii, converting from the fuzzy domain to the image domain; II, image partition: partitioning a digital image subjected to enhancement processing, i.e. a to-be-partition image, according to an image partition method based on a three-dimensional fuzzy division maximum entropy. The method is simple in steps, reasonable in design, convenient to realize, good in processing effect and high in practical value, and an image enhancement and partition process can be simply, conveniently and quickly completed with high quality.

The invention discloses a fast sparse Bayesian learning based direction-of-arrival estimation method and mainly solves the problems of heavy computation and large location estimation error in the prior art. The method includes the implementation steps: (1) adopting antenna receivers to form a uniform linear array; (2) sampling space signals and computing an array covariance matrix R; (3) vectorizing R to obtain a sparse model vector y; (4) dividing space domain grids, and constructing an over-complete base phi(theta) according to the structure of the sparse model vector y; (5) establishing a sparse equation according to the sparse representation relation of the sparse model vector and the over-complete base; (6) defining a hyper-parameter vector alpha, and adopting a fast sparse Bayesian learning algorithm to solve the sparse equation; (7) drawing a magnitude spectrogram according to an optimal estimation value of alpha to obtain a direction-of-arrival value. By the method, estimation accuracy of target reconnaissance and passive location under the conditions of low signal to noise ratio and small snapshot number is improved, computational complexity is lowed, and the method can be applied to target reconnaissance and passive location.

The invention relates to a superpixel segmentation method based on the fuzzy theory. The superpixel segmentation method improves fitness of superpixel edges and original image edges, and segmented superpixels have an undiversified internal structure and uniform gray scale. The superpixel segmentation method comprises steps of 1) processing images and initializing cluster centers; 2) adding coordinate distances of pixel points into the fuzzy C-means clustering algorithm and establishing an objective function; 3) updating a membership matrix; 4) updating cluster centers with gray scale values and coordinates included; 5) repeating steps 3) and 4) until a termination condition that the change amount of the cluster center gray scale values is smaller than an artificially set threshold value or iteration times are larger than some artificially set value; 6) completing initial superpixel division based on the finally obtained membership matrix; and 7) performing post-processing to complete the final superpixel division by combining sets of points ,which are isolated and unavoidably exist in superpixels generated after the above six steps, with adjacent superpixels which have the maximum similarity.

The invention discloses a three-dimensional convolutional neural network-based workflow identification method. Only different process tasks are divided in advance and different action behaviors are tagged artificially in a video analysis process, so that the automation demand of intelligent manufacturing is not met. An inter-frame difference method with an adaptive threshold is proposed firstly and is mainly used for segmenting out a region of a moving object from a complex background, so that the time complexity of subsequent feature extraction and model training is lowered; secondly, a 3D convolutional neural network is improved to fully adapt to a factory environment with a plurality of monitoring devices, and for different views, a view pooling layer is adopted for fusing views of different angles according to weights; and finally, a new action division method is proposed for performing automatic division on continuous production actions in a video, so that an automated workflow identification process is realized.

The invention discloses a direction of arrival angle estimation method based on Sparse Bayesian learning. The direction of arrival angle estimation method mainly solves the problems in the prior art that the computation burden is heavy, the performance of the coherent signal source process is poor, and the errors in the passive location estimation are big. The direction of arrival angle estimation method comprises the following steps: (1) using an antenna receiver to form a uniform linear array, and sampling space signals to obtain observation data, (2) converting the observation data into real values and calculating a covariance matrix, (3) carrying out the mesh generation on airspace, and constructing a real value over-complete base, (4) establishing a sparse matrix equation according to the sparse presentation relationship of the covariance matrix and the over-complete base, (5) obtaining a most sparse solution of an unknown matrix variance through solving a matrix equation by employing the Sparse Bayesian learning, (6) drawing an amplitude spectrogram based on the one-to-one corresponding relation between the spares solution and the space angle, and obtaining the direction of arrival angle degree. According to the direction of arrival angle estimation method, the passive direction-determination calculating speed and the estimation performance on the signal direction angle when in fast and low number of beats are improved. The direction of arrival angle estimation method is applicable to the target reconnaissance and the passive direction-determination.

The invention discloses a multilayer bitmap color feature-based image retrieval method. In the method, fast clustering is performed on an image with rich color information to obtain rational statistical distribution centers of each color cluster, and based on the rational statistical distribution centers, features capable of reflecting color differences among different distribution layers of the image are extracted to perform image retrieval. The method comprises the following steps of: first performing meshing on a color space of the queried image, counting the numbers of pixel points in each mesh and selecting the mesh with a number local maximum; then quickly generating each color cluster and the rational statistical distribution centers thereof by adopting a novel distance optimization algorithm and an equal-average nearest neighbor algorithm search (ENNS) algorithm in a K-average clustering algorithm, and on the other hand, performing space sub-block division on the queried image and calculating a Gaussian-weighted color average of sub-blocks; next comparing the color average of the image sub-blocks with the rational statistical distribution centers of the color clusters to extract the features of a K-layer bitmap; and finally performing the matched searching of the image features by combining the similarity measurements of the rational statistical distribution centers of the color clusters and the bitmap.

The invention discloses a method for manufacturing thin film transistor array substrate and liquid crystal display panel, the method for manufacturing thin film transistor array substrate integrates the manufacturing of an optical sensor, a photosensitive dielectric layer is formed between a transparent conductive layer and a metal electrode to sense the external light. As the optical sensor uses the transparent conductive layer as an electrode, the external light can directly irradiate the photosensitive dielectric layer through the transparent conductive layer, dramatically increasing the sensitization area of the optical sensor and promoting the light sensing performance. In addition, another side of the photosensitive dielectric layer is metal electrode, therefore efficiently preventing the backlight from directly irradiating the photosensitive dielectric layer and avoiding the noise.

The invention discloses a detecting method for tobacco package and its setting. The method adopts the X ray linear scan technology to detecting the defects of packing, rolling and producing in the tobacco packaging process, which mainly is :appending a long narrow aperture in front of the X-ray generator so as to make the X ray forming a thin fan-shaped planar beam after transmitting it, adopting this fan-shaped planar beam to linearly scan the tobacco package, and using a detector which can receive X ray to receive the transmitted X-ray directly, the clairvoyance digital image can be receive by computer after processing, so as to recognize the short of bars or packs in a whole box, of packs or number in a whole bar, of number in a whole pack and other quality problems produced in the packing and rolling process so as to remove the unqualified products with eliminating equipment. The principle and structure of the invention is more rational and simple, high sensitive and capable of detecting dynamically compared to current techniques.

The invention provides a method for positioning a lip region in a color face image. The technical scheme comprises two steps of roughly positioning the lip region and accurately positioning the lip region. The step of roughly positioning the lip region particularly comprises the following steps of: processing the input color face image by a parallel line projection segmentation technique and processing the input color face image by a complexion detection technique at the same time; and performing OR operation on the obtained results to obtain a roughly positioning result of the lip region. The step of accurately positioning the lip region particularly comprises the following steps of: establishing a narrow band region at the periphery of lip edge characteristic points in the roughly positioning result; performing texture segmentation on the narrow band region by a closed-form solution segmentation technique; and matching a characteristic template of an active shape model with a texture segmentation result and outputting an accurately positioning result of the lip region through a series of iterative processes. By the method for positioning the lip region in the color face image, the lip region can still be positioned accurately under the condition that the image comprises noise.

The embodiment of the invention discloses a multi-national paper money serial number identification method based on a sorting machine, and the method employs a character matching template to recognize a to-be-recognized character. The method comprises the steps: firstly enabling an obtained paper money serial number image to be converted into a gray scale image; secondly determining a character recognition region through initial location, binaryzation and relocation; thirdly enabling the size of a character region to be consistent with the size of a matching template through normalization; and finally screening out matched characters through dot product, and completing the recognition. The method not only avoids the influence from noisy points, and can save a filtering step. The method is suitable for a plurality of current mainstream currency types: RMB, dollar, Euro, Hongkong dollar, and yen.

The invention discloses a two-step robust filtering method and system of a GNSS/INS integrated navigation system. The method comprises the following steps: constructing a fading factor matrix according to information, constructing a gain coefficient matrix according to residual errors, constructing a state equation and a measurement equation of the GNSS/INS integrated navigation system, and fusingGNSS/INS data according to a Kalman filtering tight coupling mode; detecting whether the dynamic model is abnormal or not, and if the dynamic model exceeds a set threshold range, entering time updating through fading factor matrix adjustment; detecting whether the observed quantity is normal or not, if the observed quantity exceeds a set threshold range, adjusting an observation vector through multiple channels of a gain coefficient matrix, and entering measurement updating; and updating a Kalman filtering process, and outputting a GNSS/INS integrated navigation result. According to the method, the redundant information of the integrated navigation system can be fully utilized, meanwhile, scalar factors are expanded to a diagonal matrix, the algorithm complexity is reduced, and the positioning precision, tracking performance and stability of the integrated navigation system are enhanced.

The invention discloses a method and system for detecting a microseism event. The method includes the steps that the optimal time-window length L obtained after a microseism detection sequence of a marked microseism wave is trained with the first-order liner regression function and the kriging interpolation method is obtained; a to-be-detected microseism data sequence in time order is obtained; afirst time window and a second time window following the first time window are established, wherein the lengths of the two time windows are both L; data filling is conducted on the left side and the right side of the microseism data sequence; a first piece of data, aligned with the center of the first time window, of the microseism data sequence is slid by two time windows till the center of the first time window is aligned with a final piece of data of the microseism data sequence; every time sliding of the step size of one piece of data is conducted, the data in the two time windows is subjected to chi-square testing to determine whether a microseism wave is generated at each piece of data in the microseism data sequence. By means of the method and system, the influences of noise on detection precision can be effectively avoided, and detection accuracy can be improved.

The invention discloses a high-velocity motion object pose vision measurement method based on structured light, belongs to the technical field of computer vision measurement, and relates to a large-view-field small-target high-velocity motion space pose measurement method. The measurement method comprises the following steps: shooting auxiliary crossed laser by left and right high-speed cameras by adopting vertical laser strips and auxiliary crossed laser, recognizing crossed points, and recording the imaging positions of the auxiliary crossed laser at left and right high-speed cameras respectively; and collecting images of a measured cylinder and the vertical laser stripes by utilizing left and right high-speed cameras, transmitting the collected images to a diaphragm workstation, binding and optimizing a target axis by utilizing the distance from the deformation parts of the laser stripes generated by the measured cylinder to the axis, thereby obtaining the position and pose information of the target object finally. According to the invention, the pose information of high-speed motion target objects with high precision without any treatment by utilizing the laser stripes can be obtained, the image collecting quality is improved, and the influence by noisy points is prevented effectively.

The invention discloses a traffic matrix estimation method which is based on combination of a linear programming method and Zernike-Moment. According to the traffic matrix estimation method, due to adoption of the linear programming method, an objective function is selected instead of prior information, and therefore the sensibility of a model on the prior information is eliminated. According to interpolation processing of the Zernike-Moment method, the problem of massive data loss which cannot be solved in a K-NN (K-Nearest Neighbor) algorithm is solved, the sensibility of the model on a missing value is eliminated, and meanwhile, the influence of noise in input information on an estimation result is eliminated. In addition, due to adoption of the Zernike-Moment method, a result of the linear programming method can be corrected by utilizing a traffic matrix estimation value obtained by using the LP (Linear Programming) method as well as structural similarity, redundant information and space-time restriction information among traffic matrix data, and furthermore the estimation accuracy of the traffic matrix is improved.

The invention discloses an image matching method based on improved Harris-Laplace and a scale invariant feature transform (SIFT) descriptor. The method comprises the following steps: 1, extracting key points on an image by adopting the improved Harris-Laplace and determining the main direction of the key points to generate feature points; 2, describing the feature points by using the SIFT descriptor; and 3, roughly matching and finely matching the feature points respectively by a best bin first (BBF) nearest neighbor search algorithm and a random sample consensus algorithm (RANSAC). The key points extracted by the improved Harris-Laplace not only have invariance on illumination change, rotation change and scale change, but also can effectively overcome the influence of noise, so that the method has a high-precision matching result.

The invention belongs to the unmanned aerial vehicle technical field and relates to a visual positioning obstacle avoidance system and method for an unmanned aerial vehicle. The system comprises a binocular vision system, a sensor module and a flight control system which are mounted on the unmanned aerial vehicle. The method includes the following steps that: the binocular vision system acquires the visual information of the flight environment of the unmanned aerial vehicle, and processes the visual information to obtain obstacle information; the sensor module acquires the state information ofthe unmanned aerial vehicle; the flight control system receives the obstacle information and the state information of the unmanned aerial vehicle, establishes a flight path, generates a flight control command and sends the flight control command to the unmanned aerial vehicle; and the unmanned aerial vehicle avoids an obstacle according to the flight control command so as to fly properly. According to the visual positioning obstacle avoidance system and method for the unmanned aerial vehicle of the invention adopted, visual information and the information of the sensor module are combined; the information of the flight environment is sensed; flight path control and path planning are performed to avoid the obstacle; and therefore, the problem of the visual obstacle avoidance of the unmanned aerial vehicle can be effectively solved, and the unmanned aerial vehicle can complete visual obstacle avoidance through using an airborne camera.

The invention discloses a SAR image change detection method based on a complex neural network, which inputs two original SAR gray-scale images and uses a preliminary difference map obtained by a traditional method as a preliminary label. According to the preliminary tags, some pixel blocks in the two original images are selected as candidate training samples by confidence detection. A complex network is constructed, which consists of three layers of all connection layers and two layers of multiple batches of normalized layers. The complex training samples are selected from the candidate training samples according to the stochastic proportional method to train the complex network. The training complex network is used to test the test complex samples constructed directly from two original SAR images, and the final change detection results are obtained. The invention not only makes full use of the advantages of the traditional change detection result, but also displays the characteristicsof the original data, so that the neural network can better learn the relationship between the two images, thereby obtaining a better change detection result.

The invention discloses a cross-domain image classification model construction method and device based on transfer learning. The cross-domain image classification model construction method performs image feature extraction on data of a source domain data set and a target domain data set by using a convolution layer of a pre-trained Inception-V3 model, and deletes the part, which is greatly different from the target domain, in the source domain data through the clustering algorithm, so as to reduce the noise influence possibly generated by the source domain data, solve the noise influence caused by the difference between the source domain and the target domain, improve the similarity between the source domain data and the target domain data, and reduce the noise influence of the source domain data. An attention mechanism is added to extract image feature information of more target domain data when model fine adjustment is carried out on the target domain. Therefore, the utilization rateof the target domain data is improved, and more image feature information of the target domain data is extracted, and the cross-domain image classification accuracy is improved, and the adaptabilityof the migration model to the target domain is enhanced, so that the accuracy of the finally constructed image classification model is ensured.

The invention relates to an unsupervised medical image segmentation method based on an adversarial network, belongs to medical care informatics, and particularly relates to the technical field of medical image segmentation. According to the technical scheme, the method comprises the following steps: firstly, randomly generating or utilizing a third-party data set to obtain a group of auxiliary masks according to shape prior information, and sending the auxiliary masks and an unlabeled training image into a cyclic consistency adversarial network to generate binary masks; and a discriminator based on variational self-encoding and a generator correction module based on discriminator feedback are utilized to improve the quality of the binary mask. And after the binary mask of the training image is obtained, iterative training is performed by using a noise weighted Dice loss function, so that a final high-precision segmentation model can be obtained. According to the method, the problem that the convolutional neural network needs a large number of manual annotations in the training process of medical image segmentation can be solved, the problems of low performance, poor robustness and the like of an unsupervised segmentation method are solved, and the performance of an unsupervised medical image segmentation algorithm is effectively improved.