54 results about "Fuzzy similarity" patented technology

The invention discloses a fine-grained vehicle type identification method based on weak supervision positioning and subclass similarity measurement, which comprises the following steps: 1) weak supervision positioning: using a pre-trained VGG-NET network to locate the image object, processing the mask map of convolution layer, and obtaining the boundary frame of the object. 2) constructing a fuzzysimilarity matrix: features of the pictures in the training set after the positioning of the picture are extracted by using B-CNN, and a fuzzy similarity matrix to measure the similarity of each subclass is obtained according to softmax classification results; 3) performing sampling to form a triple set: sampling to form a triple set on that basis of a fuzzy similarity matrix; 4) joint learning of the improved loss function: the improved triplet loss and the weighted softmax loss are used to restrict the distance between samples of the same sub-category and increase the distance between samples of different sub-categories. Compared with the original model, the invention is more accurate in positioning, and the classification accuracy is obviously improved, so that the vehicle target can be positioned well.

The invention discloses an image registering method based on a manifold subspace, comprising the following steps: (1) respectively reading in a target image and a floating image to be registered; (2) respectively calculating the phase consistency of the target image and the floating image to be registered in different scales and directions; (3) fusing information with the phase consistency in different scales and directions by using a manifold subspace method; (4) looking main-phrase consistency information of the target image and the floating image as fuzzy congregation, adopting nearness in fuzzy mathematical, and calculating the fuzzy similarity of the main-phrase consistency information; (5) using the fuzzy similarity of the main-phrase consistency information as an object function, and adopting powell algorithm to search a maximum value of the object function; finishing the registering when the fuzzy similarity reaches maximum. Under the conditions of lower image space resolution, noise influence, and the like, the image registering method has high registering precision and strong robustness.

The invention relates to a method for extracting a classification rule based on a fuzzy-rough model. Since fuzzy boundaries of continuous attribute values are not considered in the conventional continuous attribute discretization method, a data mining rule is not refined or accurate enough and important data information is easy to lose during discretization. The method for extracting the classification rule comprises the following steps of: performing attribute fuzzification on continuous attributes in an information sheet by using a membership function in a fuzzy set; and extracting parameters such as approaching precision approximate measure, rough approaching precision approximate measure, approaching precision classification quality measure, approaching precision relative classification measure and the like by using a rough set in a fuzzy similarity relation so as to establish an approaching approximate-based fuzzy-rough set reduction algorithm to solve the classification rule. In the method, each continuous attribute is added into an attribute reduction set in a descending order according to importance until the reduction condition is met, and particularly, the attribute reduction can be quickly solved when multiple condition attributes are available.

The invention relates to an assessment method for rolling bearing performance variation. The method comprises the following steps: in a service period of a rolling bearing, collecting data through a measuring system, and obtaining a time sequence of performance of the rolling bearing and dividing the time sequence into subsequences of G time segments evenly; reconstructing phase space of the subsequences of G time segments with a coordinate delay reconstruction method, and obtaining phase loci; establishing fuzzy similarity relations between two random phase loci, obtaining a fuzzy equivalence relation based on the phase space between the two phase loci with a transitive closure method, calculating an optimal fuzzy equivalence relation of G time segments phase loci, and obtaining an optimal fuzzy equivalence relation measurement u of the G time segments phase loci; comparing u with a first threshold u0 and a second threshold u1, and accessing variation state of the rolling bearing performance. According to the invention, abnormity appearance and performance failure degree of the rolling bearing in a service period are effectively assessed in real time in order to adopt corresponding measures timely and avoid generation of major accidents.

The invention provides a multi-resolution community discovering method based on fuzzy clustering. According to local interaction information of adjacent nodes, the structural similarity is introduced for measuring the fuzzy relation between the nodes, fuzzy transitivity of the fuzzy similarity between the nodes in a network topology is partially considered, fuzzy parameters are used for performing set cutting on a fuzzy transitivity matrix to obtain community structures under different resolutions, and therefore network communities can be discovered. Matrix transformation operation is adopted, a network community detection model based on fuzzy clustering is built, iterative optimization processes in a traditional method are reduced, the time complexity is lowered, a large number of experiments prove that the community structures in a network can be effectively revealed, the universality is strong, and the high application value is achieved; network structural analysis and community structural visualization can be effectively achieved.

The invention discloses a dissolved oxygen control method based on a fuzzy neural network. The method comprises the following steps: S1, modeling is carried out; S2, fuzzy identification is carried out, and a fuzzy identification method is adopted to identify the structure and parameters of an object model; S3, a clustering method is adopted to determine the number of rules; S4, a neural network is used for learning reasoning data of an expert, a fuzzy reasoning rule is acquired, and each connection weight is adjusted at the same time; S5, after a least square method is used for order identification, a transitive closure method is used for clustering analysis according to a fuzzy similarity relation; S6, a Smith predictor is introduced; and S7, two independent control circuits of blower pressure and oxygen dissolution are adopted. The wastewater treatment effects are good.

The invention provides a prediction data-based transformer early warning evaluation method and device. The method includes the following steps that: a prediction model of gas concentration is established on the basis of acquired training sample data; a fuzzy similarity matrix is established on the basis of prediction data obtained by the prediction model, so that an early warning threshold value can be obtained; and an early warning is reported or not reported according to the early warning threshold value and a national standard limit value. With the prediction data-based transformer early warning evaluation method and device provided by the technical schemes of the invention adopted, the early warning threshold having a plurality of standards is set according to the features of various types of early warning indexes, and with national standards adopted in combination, early warning results can satisfy actual application requirements to the greatest extent.

The invention discloses a collaborative filtering method based on integration of fuzzy weight similarity measurement and clustering. According to a user-item scoring matrix R<m x n>, three different similarity matrixes FCOS, FCOR and FADJ of users are respectively calculated by using fcos, fcor and fadj, and then according to a k-means algorithm and a cluster number kcluster, all users are clustered. A nearest neighbor set s (Ui) of users is determined and then scores are calculated and predicted by using r<i,c>; according to the above-mentioned strategy, the steps are repeated till scores of all user are predicated. By adopting the fuzzy similarity clustering IBCF\UBCF of the invention, the searching accuracy of the neighbor set s (Ui) is obviously improved; by fuzzifying score values and score deviations, the evaluation is closer to the real evaluation of the users to items; by adding fuzzy weight wc during similarity calculation, the similarity between the users tends to be more accurate and thus the performance of a recommender system is improved.

The invention comprises invention discloses an automatic segmentation method for fuzzy spectral clustering brain tumor images based on super pixel, comprising the following steps of : firstly, performing super pixel segmentation on a FLAIR mode image of magnetic resonance imaging containing brain tumors, and extracting gray histogram features of super pixel blocks as input of an algorithm, calculating a fuzzy similarity matrix of images through the input features; then performing clustering through NJW spectral clustering algorithm to obtain a final segmentation result. According to the automatic segmentation method for fuzzy spectral clustering brain tumor images based on super pixel, fuzzy theory is used to optimize similarity measurement mode of spectral clustering, fuzzy weight parameters are introduced in Gaussian distance measurement method of spectral clustering, and a fuzzy similarity measurement mode based on super pixel features is defined. The invention is an automatic imagesegmentation method, human intervention is not needed, and time complexity of spectral clustering algorithm is greatly reduced and segmentation accuracy can be improved by utilizing fuzzy spectral clustering algorithm based on super pixel.

The invention discloses a relevance feedback measuring method based on the classification and recognition of the fuzzy region characteristics of medical images, comprising the following steps: (1) cutting all medical images which are selected from a medical image database; (2) extracting hard characteristics of each cut region; (3) converting the hard characteristics into fuzzy characteristics which are stored into a character database; (4) selecting one medical image to be compared and extracting fuzzy characteristics of the medical image to be compared , and obtaining the fuzzy similarity of the medical image to be compared and medical images in the characteristic database, arraying the medical images the characteristic database according to the value of the fuzzy similarity, and outputting M images from high value to low value; (5) bringing the fuzzy characteristics of the M once output images into feedback treatment based on the fuzzy similarity to calculate, calculating the similarity of the medical images to be compared and all medical images in the characteristics database again, and outputting N images from high value to low value sequentially. The feedback measuring method can effectively pick needed medical images.

The invention discloses a multi-scale fuzzy measure and semi-supervised learning based SAR (Synthetic Aperture Radar) image identification method and solves the problem that the SAR image identification accuracy in the prior art is low. The multi-scale fuzzy measure and semi-supervised learning based SAR image identification method comprises the following steps of establishing an image library by segmenting an original SAR image and selecting image blocks with single targets; extracting characteristic vectors of the image blocks in the image library; classifying the selected image blocks into a plurality of categories, enabling corresponding characteristic vectors to be served as training samples, training a semi-supervised classifier and classifying the image library through the classifier; obtaining categories of inquire image blocks input by a user through a trained classifier; obtaining a category set of the image blocks through a confusion matrix; calculating a multi-scale area fuzzy similarity between the inquire image blocks and the image blocks belong to the set and returning the number of user required image blocks according to a sequence from large to small. The multi-scale fuzzy measure and semi-supervised learning based SAR image identification method can correct the classification error, is high in information identification accuracy and can be applied to simultaneous explain of a plurality of SAR images.

The invention discloses a rock mass structural plane dominant occurrence clustering analysis method based on a netting algorithm. The method comprises the following steps: (1) selecting an engineeringrock slope needing to be analyzed in the field; (2) carrying out polar coordinate transformation on the occurrence of the structural plane and projecting the occurrence into a spherical space; (3) calculating the similarity degree rij between occurrence samples by adopting the square of the sine value of the acute angle between the unit normal vectors of the structural plane, and constructing a fuzzy similarity matrix R of the occurrence of the structural plane; (4) transforming the fuzzy similarity matrix R of the structural plane occurrence; (5) constructing a lambda section matrix Rlambdaof the occurrence of the structural plane for the transformed fuzzy similar matrix R; (6) transforming the section matrix Rlambda of the occurrence of the structural plane; (7) clustering and groupingof structural plane occurrence of the transformed section matrix Rlambda; and (8) calculating effectiveness evaluation indexes under different grouping numbers according to a clustering result of thestructural plane occurrence, and determining an optimal grouping number in combination with engineering practice to obtain the structural plane dominant occurrence. According to the invention, the grouping result is more reasonable, and the dominant occurrence accords with objective reality more.

The invention discloses a fuzzy SVM feedback measuring method used for the target recognition of medical images, comprising the following steps: (1) regulating a window width and a window position of medical image data in a characteristic database and filtering; (2) extracting hard characteristics of the medical images processed by the step (1); (3) converting the hard characteristics extracted by the step (2) into the fuzzy characteristics which are stored into a characteristic database; (4) selecting one medical image to be compared and extracting the fuzzy characteristics of the medical image to be compared , and obtaining the fuzzy similarity of the medical image to be compared and medical images in the characteristic database, arraying the medical images in the characteristic database according to the value of the fuzzy similarity, and M images are output from high value to low value; (5) bringing the fuzzy characteristics of the M once output images into feedback treatment based on the fuzzy similarity to calculate, calculating the similarity of the medical images to be compared and all medical images in the characteristic database, and outputting N images from high value to low value sequentially. The feedback measuring method can effectively pick needed medical images.

The embodiment of the invention discloses a method for subjective evaluation results of image quality based on fuzzy clustering statistics. The core idea of this method is that the different subjective evaluation scores obtained from the image can be regarded as the matching degree between the image and each score. The fuzzy similarity matrix of image sequence is established by the similarity relation of sample image score, and fuzzy clustering analysis is realized. After the best classification of sample image is determined, the image sequence is the final evaluation result of image quality.This method is suitable for subjective evaluation of image quality with large sample size. The method also provides the dynamic clustering diagram of the sample image, and displays the clustering process of the image for the user to refer. The user can change the threshold value according to the use demand, and reclassify the image and score.

The invention relates to a method for positioning a temporary voltage drop source on line by adopting fuzzy similarity match. The method is characterized by comprising the following steps: establishing a node positive sequence voltage match index calculation model by taking the positive sequence voltage of existing nodes of the whole network as a characteristic quantity, and positioning the temporary voltage drop source by utilizing the fuzzy similarity of voltage match indexes and the match index similarity of monitored voltage vectors. The method disclosed by the invention can be used for accurately positioning the temporary voltage drop source by utilizing existing limited monitoring point information. The method disclosed by the invention can be used for calculating the voltage match index of the positive sequence voltage of a node during temporary voltage drop and matching and identifying by utilizing fuzzy correlation degree and a line fault characteristic set established off line, thereby being insensitive to data accuracy. Besides, the method disclosed by the invention can be used for integrally judging and identifying the temporary voltage drop source by simultaneously adopting the information of all monitoring nodes and the information of a network topology, is rich in data source and has certain fault tolerance, and thus high accuracy and universality are achieved in positioning the temporary voltage drop source by utilizing the method.

The invention relates to a point of interest recommending method and device. The method comprises the following steps: obtaining a differential privacy noise factor; determining a friend fuzzy similarity recommendation probability between users based on a set social relationship privacy protection algorithm according to the differential privacy noise factor; determining a radius of a virtual circle based on a set geographic position privacy protection algorithm according to a historical visitor number of a point of interest of a target region and an actual geographic position of a target user,wherein an area corresponding to the virtual circle is a privacy area of the user; According to the virtual circle, the geographic location distance recommendation probability between the users is determined, and according to the friend fuzzy similarity recommendation probability and the geographic location distance recommendation probability, the interest points are recommended to the users. Theinvention solves the problem that users' privacy information is exposed too much in the process of recommending points of interest, and solves the problem of privacy disclosure of users in a more friendly manner on the premise of recommending reasonable points of interest for users.

The invention provides a batch processing method for identifying defects of an X-ray directional instrument, belonging to the field of single crystal material processing. A method for detect defects in off-line batch single crystal is proposed, the eigenvector of the pendulum curve of single crystal is designed, so that the characteristics of the curve are abstracted, combined particle swarm optimization algorithm, and introduces the effective radius based on the density function, the traditional FCM algorithm is improved, the robustness of the lifting algorithm to the initialization of clustering centers is improved, falling into local optima is avoided, and interfering data is well filtered out, so as to realize clustering of batch data, only the clustering center eigenvector needs to bedetected. The defect types of the curves to be tested can be obtained according to the membership degree relation, wherein, the improved fuzzy transfer closure clustering algorithm proposed by the invention is included, the fuzzy similarity matrix is defined, the accuracy of the similarity calculation is ensured, and the invention provides a new idea and a realization mode for the crystal detection technology.

The invention discloses a comprehensive geomagnetic matching method based on geomagnetic information entropy and similarity measurement. The method comprises the following steps: 1, acquiring geomagnetic field intensity of the current position (x, y) of an aircraft, 2, obtaining a geomagnetic map covering (x, y) from a geomagnetic database, calculating the geomagnetic information entropy H (x, y)at (x, y) through an area A covering (x, y), judging whether (x, y) is located in an adapted area or not according to H (x, y), 3, setting a to-be-matched flight path template T with a mesh quantity of M*N, and a search area S, performing similarity matching on the template T on the geomagnetic map of the geomagnetic database to obtain a submap R with the maximum similarity to the template T, and4, performing kriging interpolation on the submap R to obtain a high precision and small scale geomagnetic map Map, and performing iterative solution on the geomagnetic map Map to obtain a matching value of the current position of the aircraft. The method utilizes a fuzzy similarity algorithm, selects a matching area, obtains an optimal matching point through an accurate related algorithm, and reduces the influence of interference noise on matching and positioning accuracy.

The invention belongs to the field of measuring the fuzzy similarity of an ontology concept in an intelligent semantic web and particularly relates to a method for measuring the fuzzy similarity of the ontology concept in the intelligent semantic web. The method comprises the following steps of: (A) extending the ontology concept into an explanation set, wherein the explanation set comprises an is an A subset, a relatedTo subset and a nearTo subset; (B) extending the explanation set of the ontology concept into a fuzzy explanation set; and (C) calculating the fuzzy similarity of the ontology concept according to the fuzzy explanation set. The method can simultaneously meet the three characteristics of the similarity among concepts, and further, supports the calculation of the similarity when a sub concept of a complicated ontology has a plurality of parent concepts.

The invention provides a method for predicting a message passing node based on the meeting probability of a target node. A fuzzy similarity matrix is established based on various relationship characteristics of nodes, different attributes of mobile nodes are deeply researched by forming the fuzzy similarity matrix, the social attribute change rule of the mobile nodes is mined, and the weights of the different attributes are dynamically and adaptively distributed. The social relationship and cooperation relationship of the nodes are further quantified. Finally, experiments verify that a good effect is achieved when the encountering model provided by the invention is used for screening the trusted node as the next hop node of data transmission, so that the data is always transmitted along the trusted cooperative node in the network, and meanwhile, the influence of malicious node non-cooperation on the performance of the network is reduced.

The invention provides a colorful morphological image processing method based on fuzzy similarity and relates to the image processing technology field. The method comprises steps that to-be-processed colorful images are made to correspond to the RGB colorful space, a fuzzy similarity measurement (FSM) function used for representing the similarity degree of two colorful vectors is determined, a pixel is taken as an index, structure units and a colorful vector set of the structure units are acquired in the RGB colorful space one by one, a least upper bound vector and a greatest lower bound vector of the colorful vector set are determined based on the FSM criterion, colorful morphological basic operation is constructed, and application to colorful images is further carried out. The method is advantaged in that the morphological idea is applied to colorful image processing, good stability and strong practicality are realized in actual colorful image processing, not only can colorful target smoothing be realized, but also the detail characteristic of homogenous region pixel inconsistency can be further excellently processed, and colorful image analysis and processing targets can be finally realized.

The invention discloses a hob state intelligent monitoring method of a numerical control hobbing machine. The method comprises the following steps that S1, B-axis vibration signals are collected in real time; S2, data segmentation is conducted on the B-axis vibration signals; S3, a hob state standard sample set X is constructed, and a simple initial feature vector f0 is extracted; S4, a hob statemutual K neighbor graph G is constructed, and feature selection is carried out to form a sample feature vector f; and S5, a hob state feature matrix F is constructed, a fuzzy similarity relation matrix R is built, a transmission closure t(R) is constructed and clustering analysis is carried out, and hob state recognition is realized. According to the method, the hob state mutual k neighbor graph is constructed, main shaft vibration signal processing of the numerical control hobbing machine is combined with an atlas theory, on-line real-time monitoring of the hob state can be realized, so thatthe hob replacement and the blade grinding can be carried out in time, expansion cracks and hob teeth fractures are avoided, the dependence on professional skills of operators can be reduced, and theintelligent process of the numerical control hobbing machine can be promoted.