82results about How to "Achieve division" patented technology

The invention provides an implementation method of a VPN (virtual private network) on basis of an SDN (software defined network). The method comprises steps as follows: a user uses a northbound interface provided through an SDN controller to send a service request for establishing the VPN; the SDN controller receives the service request which comes from the user and is used for establishing the VPN, and the service request comprises a service type and service parameters; the SDN controller performs packaging processing on a data packet from a local area network host according to the service type and the service parameters, and the data packet is sent to a designated foreign network through an OpenFlow exchanger; after an SDN controller of the foreign network receives the data packet, corresponding de-encapsulation processing is performed, and the network data packet of a sender is recognized; finally, the VPN is established between networks of two places. According to the method, the VPN is partitioned in an SDN service deployment manner; the problems of single configuration manner, complicated configuration and management and difficulty in maintenance of conventional VPN configuration are solved.

The application discloses the container configuration method and network transmission device of a virtual network system. The method comprises: performing virtual network interface card configuration in a main network space, adding a configured virtual network interface card to a container only corresponding to a virtual network interface card, and starting the APP service of the container in order to ensure that the container and the external network system of a physical machine can perform network communication. The main network space is a preset network space corresponding to the physical machine, wherein the process of performing virtual network interface card configuration in a main network space comprises: setting the MAC address of a virtual network interface card to be an MAC address same with the physical network interface card on the physical machine, distributing a single IP address to the virtual network interface card, and obtaining a configured virtual network interface. Accordingly, based on different IP addresses, the method realizes division of a virtual local area network, thereby effectively improving the isolation effect of the container, and increasing the virtualization degree of a virtual network system.

The invention provides a community detection method and system based on label propagation algorithm; the method comprises the following steps: obtaining source data, and initializing an initial relation network graph; ranking each node in the relation network graph in a random sequence, and traversing and updating all node labels in the relation network graph according to the ranked sequence; the node m update method comprises the following steps: determining the label with the most occurrence frequency of all nodes adjacent to the node m, if the labels are more than one, the adjacent nodes with the labels of most occurrence frequency can form an adjacent node set M; selecting the adjacent node connected with the node m and having the biggest vector weight value from the adjacent node set M to serve as the selected node, and updating the label of the selected node as a novel label of the node m; when termination conditions are satisfied, the network forum community dividing process is finished, and user IDs corresponding to nodes of same label belong to the same community. The novel method and system can more accurately divide the communities, thus obtaining community distribution conditions of the network forum.

Provided is a method for effectively segmenting a hyperspectral oil-spill image. The method comprises steps of: defining an initial level set function and other related functions; acquiring a new fitting item in combination with a Fisher criterion; constructing an edge stop function to obtain a new length item; performing improvement in combination with an end member extraction algorithm; introducing a level set regular item to prevent reinitialization of the level set function; minimizing an energy function to obtain an Euler-Lagrange equation; setting parameters; selecting a display band and an initial contour; displaying a segmentation result graph; calculating various segmentation precision evaluation indexes; comparing and evaluating the accuracy of the segmentation results. The method can classify a target area in a simulated hyperspectral image and a real hyperspectral image, and effectively segments the hyperspectral oil-spill image with boundary blur and noise, improves the segmentation accuracy of the hyperspectral image, obtain a more accurate classification effect, makes the parameter change more stable, makes the contour curve more accurate, obtains the continuous and closed boundary contour, and has higher precision of segmentation.

The invention relates to a region recognition method, a robot and a storage medium. The method comprises the steps that a first label image and a second label image are recognized; first position information corresponding to the first label image and second position information corresponding to the second label image are obtained; a virtual wall is arranged between the first label image and the second label image according to the first position information and the second position information; the region type of a region to be recognized is determined according to the first label image and thesecond label image; and the region to be recognized is the region divided by the virtual wall as a boundary and not including a robot. According to the method, the detection accuracy of the virtual wall can be improved, and further the recognition accuracy of the region to be recognized is improved.

The invention discloses a model generation method and device, and a lithology identification method and device. The model generation method comprises the steps of obtaining a training data set, wherein the training data set comprises multiple groups of instances, each group of instances comprises at least one piece of logging data and a data label, the at least one piece of logging data comprisesany one or more of a natural gamma curve, a natural potential curve, resistivity and a borehole diameter of the target area, and the data label is used for labeling lithology of the at least one pieceof logging data corresponding to different stratum depths; training a pre-constructed model by using the training data set to obtain a trained model, wherein the trained model is used for identifyinglithology of different stratum depths of the to-be-detected area. High automation of data processing and intellectualization of lithology identification can be achieved, the lithology identificationspeed and precision are improved, the labor intensity of interpreters is relieved, and the requirements for practical experience and professional technical knowledge of the interpreters are lowered.

The invention provides a multistage wheel disc rotor structure interactive design method driven by geometrical features and oriented to engineering requirements. The method comprises the following steps of 1, inputting control point coordinate values of a multistage wheel disc meridian plane geometrical structure; 2, building a multistage wheel disc structure axisymmetric finite element model forthe multistage wheel disc meridian plane geometrical structure; 3, analyzing the structure of the built multistage wheel disc structure axisymmetric finite element model; 4, obtaining the multistage wheel disc meridian plane geometrical structure which meets the design requirement and outputting a CAD model. Through the steps, the change of the control point coordinate values of the multistage wheel disc meridian plane geometrical structure can be automatically displayed, the working efficiency of designing a multistage wheel disc structure is improved to a great extent, the problems existingin the prior art during design that the time and labor cost are wasted when an optimizing algorithm is not steady, and artificial remodeling is needed once the scheme is changed during design in the prior art are solved, and finite element mesh dividing and multistage wheel disc structure design can be achieved.

The invention relates to the field of sea surface oil spill detection, and provides a quaternion-based sea surface oil spill identification method and system. The method comprises the steps of 1, collecting reflected polarized light reflected from a target object at different polarization angles, and converting the reflected polarized light into an electric signal, so as to obtain a polarization image of the target object, 2, obtaining a polarization parameter according to the polarization image, 3, projecting the polarization parameters into an HSV color model, and conducting normalization processing on the polarization parameters, 4, combining the quaternion with a Fresnel formula, and performing oil film identification according to a reflected light quaternion expression, and 5, verifying the equivalence relationship between the quaternion-form Poincare ball and Stokes, marking quaternion expressions of different oil films on the Poincare ball, and realizing identification of various oil films at different positions of the Poincare ball surface by the quaternion expressions of different oil films. Division of an oil film area and a water surface and identification and detection of types of spilled oil can be realized.

The invention discloses a container-based energy controller terminal software deployment and management method. The method comprises the following steps: firstly, configuring resources such as a container CPU, a memory and a disk space; then deploying and starting software in the container, newly establishing a network bridge and binding the network bridge with the container after receiving the container network configuration information in the container network resource mapping stage, modifying the routing of the container according to the configuration, and modifying the IP address mapping,port mapping and other information of the firewall of the terminal; and finally, carrying out state monitoring management on the container software. According to the invention, on the basis of the container, unified installation and deployment are carried out on energy controller terminal software, so that dynamic configuration of hardware resources and unified monitoring of a software running state are realized; meanwhile, the method meets the requirement for resource isolation in software operation, and the reliability and safety of the terminal are improved.

The invention relates to a gas valve part numerical control machining automatic programming system. The system comprises a drawing number input module (1) used for inputting a drawing number of to-be-machined gas valve parts, an automatic data retrieval module (2) used for retrieving the input drawing number to obtain a part structure list of the to-be-machined gas valve parts, a program automaticprogramming module (3) used for performing machining process scheduling and cutter selection according to the part structure list to generate a numerical control machining control program, and a program storage module (4) used for storing the generated numerical control machining control program. Compared with the prior art, according to the system, automatic generation of the numerical control machining control program of the gas valve parts is realized, the system is very suitable for single-piece/small-batch production of the gas valve parts, and the production efficiency of single-piece/small-batch production is greatly improved.

The invention discloses a network slice reservation method and device. The method comprises the following steps: generating a reservation template in response to a service reservation request initiated by a user; wherein the service reservation request comprises a user identifier, a service type, reservation time and a reservation position; judging whether the current slice can meet the service requirement of the user or not based on the service type, the reservation time, the reservation position and the historical use data of the wireless side slice, and obtaining a judgment result; and returning a reservation message based on the judgment result. According to the network slice reservation method, reasonable configuration and division of the network slices in the specific service scene are realized, so that the jamming phenomenon is avoided, and the user experience is improved.

The invention belongs to the technical field of rock stratum detection, and discloses a Deep learning and transfer learning-based rock stratum structure intelligent detection and classification method, and the method comprises the steps of obtaining the existing drilling image data and acoustic logging information, processing the obtained corresponding information, and constructing a feature database with a label; classifying and sorting into a training set and a test set which can be used for a deep neural network model according to a feature database; based on the acquired drilling image data and acoustic logging information, respectively constructing a sensing network; verifying the reliability of the trained sensing network through a field drilling test, and carrying out model feedback optimization; and performing intelligent detection and classification of the rock stratum structure based on the optimized sensing network. According to the invention, intelligent detection and real-time classification statistics of rock stratum structure characteristics are realized, and a basic condition is laid for practical application of an intelligent identification and real-time classification method of a deep rock stratum structure in field drilling exploration.

The invention relates to the technical field of image detection, and discloses a product packaging quality detection method based on deep learning, and the method comprises the steps: obtaining a product appearance image, carrying out the preprocessing of image graying and gray stretching of the product appearance image, and obtaining a preprocessed product appearance image; performing noise reduction processing on the preprocessed product appearance image by using a noise reduction algorithm based on similar image blocks to obtain a denoised product appearance image; processing the denoised product appearance image by using an image reconstruction algorithm to obtain a super-resolution product image; performing feature extraction on the super-resolution product image by using a feature extraction algorithm based on the sparse dictionary to obtain packaging features in the product image; and detecting the packaging quality of the product by using a packaging defect detection algorithmbased on the spatial pyramid pool. The invention further provides a product packaging quality detection system based on deep learning. According to the method, image detection is realized.

The invention relates to the technical field of food detection boxes, in particular to a portable food detection box. The portable food detection box comprises a food detection box body, a storage box is arranged on the surface of the food detection box body, a center column is arranged on a bottom plate of the storage box, a plug plate is arranged on the top surface of the center column, and connecting screw holes are formed in the top surfaces of the plug plate and the center column. A locking screw rod is inserted into the connecting screw hole, the locking screw rod is in threaded connection with the surface of a sealing plate, a butt joint groove is formed in the bottom surface of the sealing plate, and a plug plate is inserted into the butt joint groove; the portable food detection box has the beneficial effects that the storage boxes are additionally arranged on the two sides of the portable food detection box to collect and store sampling pipes, after the sampling pipes are inserted into limiting seats, supporting strips are expanded outwards, clamping heads extrude damping fins to be attached to the surfaces of the sampling pipes, the sampling pipes are prevented from bumping, the two storage boxes expand the sample collection space of the food detection box, and the locking screw rod can be driven to be screwed on the central column by screwing the handle, so that the sealing plate is fixed.

The present invention discloses a method for dividing AMC Zone of broadband wireless access system, which comprises the following procedures: 1. differentiating the style of each MS in broadband wireless system to construct AMC MS array and PUSC MS array; and 2. determining the style of a first MS scheduled by an MS scheduling mode, and dividing AMC Zone according to the determined result. The technical plan according to the invention is introduced with a scheduling mode scheduled by MS. The dividing of AMC Zone can be realized without looking through the whole connection and only executing corresponding scheduling according to the physical property of MS at the condition that the scheduling resource can not be obtained beforehand.

The invention discloses a brand word mining method, device and equipment and a readable storage medium, which relate to the technical field of Internet, and aim to generate a target brand word by utilizing the real condition of a store, realize the division of the store, ensure that the brand word mining process is consistent with the actual condition of the store, avoid the misleading to a user and ensure relatively the higher user viscosity. The method comprises the steps of determining a first segmented word string in a store name, and extracting the candidate brand words from the first segmented word string according to frequency data of the first segmented word string; determining an associated store of each candidate brand word in the plurality of candidate brand words, and extracting a real store from the associated stores according to the dish names of the associated stores; comparing the real stores of the plurality of candidate brand words, and calculating the store similarity among the plurality of candidate brand words; and merging the candidate brand words of which the store similarity is greater than a similarity threshold in the plurality of candidate brand words toobtain at least one target brand word.

The invention discloses a light emitting fence component and a light emitting fence having the light emitting fence component. The light emitting fence component comprises a light source and a bar type light transmitting piece, wherein the light transmitting piece is provided with an accommodating cavity for accommodating the light source; the light source emits light through the light transmitting piece; a projection and a slot are arranged at opposite two sides of the light transmitting piece in the length direction respectively; and the projection is used for being inserted in a slot of an adjacent light emitting fence component. The light source emits the light through the light transmitting piece, and the projection and the slot are arranged at the two sides of the light transmitting piece in the length direction respectively, so that the area division through the light emitting fence component and the light emitting fence is realized, and a better effect, in particular a colorful night scene light effect can be achieved under the condition of weak ambient light.

The invention discloses a hair styling device, and belongs to the technical field of hot-air styling devices. The hair styling device is characterized by comprising a shell (1); a plurality of air direction flow guide devices are arranged on the inner wall of the shell (1), a plurality of air channels are formed inside the shell (1) by the multiple air direction flow guide devices, an air inlet (6) is formed in an end of the shell (1), side air outlets (2) are respectively formed in two sides of the lower portion of the shell (1), the corresponding air direction flow guide devices are fixedly arranged on a side of each side air outlet (2), and a hot-air leading-out device is connected to the air inlet (6). The hair styling device has the advantages that air can be divided in different directions, so that drying actions can be carried out in different directions, hair drying, dividing and styling can be synchronized, other tools or the special styling cost charged by professional hairstyle shops can be omitted, the problem of trouble due to the fact that hair only can be dried after being divided in advance at every turn can be solved, and the hair styling device is simple and is easy to operate.