50results about How to "Solve correlation problems" patented technology

The invention provides an interest point information processing system, which comprises an original record library, a pre-treatment system, a record library to be called, a telephone investigation platform, a coordinate calculation subsystem, a quality inspection subsystem, a finished product record library, an operation record library and an operation system error-indication mechanism. Useless information of the interest point information in the original record library is removed by the pre-treatment system, the interest point information after being removed with the useless information is stored in the record library to be called, the telephone investigation platform confirms the authenticity of each information and collects relevant information of the interest point entity including the door address, a coordinate of the interest point is obtained through the coordinate calculation subsystem, the interest point information is quality checked by the quality inspection subsystem to enter the finished product record library and then to enter the operation record library to be operated on line, and the operation system error-indication mechanism compensates the missed information and feeds back the compensated information to the original record library. The interest point information processing system solves the contradiction between the comprehensiveness and the accuracy for collecting the traditional public service information and provides a crucial link to the formation of the entire public information service industrial chain.

The invention discloses an adjustable bituminous pavement microwave heat regeneration test device which comprises a bearing device and an outer framework connected with the bearing device. The bearing device is provided with a microwave heating device and at least one transverse and longitudinal movement adjusting device used for adjusting the microwave heating device to transversely and longitudinal move in the plane where the bearing device is located. The microwave heating device can be adjusted in the transverse and longitudinal directions, the height of the microwave heating device can be adjusted, the device can adapt to damaged bituminous pavements of different heating areas, different heights and irregular heights, and repair requirements in actual life are met. The regeneration test device is simple in structure, complete in function and convenient to operate.

The invention discloses a large-span prestressed groove-shaped laminated plate used as the top plate of a civil air defense basement. In the short range of a few seconds or tens of seconds, there are both wave compression and wave absorption, and the deformation ability and deformation recovery ability of this kind of component are much better under external nuclear explosion, that is, under static loading; ③ remove the large The flange of the load-bearing double T plate becomes a groove shape, which can withstand greater loads, reduce the height of the bottom section, reduce the cost, and it is easy to form a complete section height and shape with its own ventilation channel, and the prestressed large span saves steel bars and saves Pillars, thereby saving the parking area, the same area increases 10%∽20% of the parking space or saves 10%∽20% of the excavation area compared with the traditional method.

The invention relates to the field of image processing, and provides a method for image classification. The method for image classification comprises the steps of (1) obtaining images without being labeled, carrying out preprocessing on the obtained images without being labeled to generate a standardized matrix of the images without being labeled, (2) generating a dictionary primitive of dictionary learning and sparsity coefficients of the images without being labeled according to a prebuilt optimization function model of the dictionary learning and the standardized matrix, and (3) carrying out classification on the images without being labeled according to the sparsity coefficients of the images without being labeled. According to the method for image classification, the situations that each specific feature only represents a part of information, and each specific feature is only effective on part of image targets due to complexity of the images are avoided, meanwhile, the situations that a training model only contains information of labeled images, labeled information is extremely limited, and a large amount of information without being labeled is wasted are avoided, thus, the problem that an existing method for image classification is poor in validity and comprehensiveness is solved, and efficiency of image classification is improved.

The invention provides a device and method for executing vector inner product operation, and aims at cooperating with a corresponding instruction set to execute vector inner product operation. The device comprises an instruction storage unit, a register unit and a vector inner product unit, wherein the storage unit is internally stored with vectors; the register unit is internally stored with addresses for vector storage; and the vector inner product unit is used for obtaining a vector address in the register unit according to an assorted instruction, obtaining a corresponding vector in the storage unit according to the vector address and carrying out vector inner product operation according to the obtained vector so as to obtain an operation result. According to the device and method, vector data participating in the calculation is temporarily stored on a scratch pad memory, so that the data with different widths can be supported more flexibly and effectively in the vector inner product operation process and then the execution performance of applications comprising a large number of vector inner products is enhanced.

The invention provides a device and method for executing vector outer product operation. The device comprises a storage unit, a register unit and a vector outer product operation unit, wherein the storage unit is internally stored with vectors; the register unit is internally stored with addresses for vector storage; and the vector outer product operation unit is used for obtaining a vector address in the register unit according to an assorted instruction, obtaining a corresponding vector in the storage unit according to the vector address and carrying out vector outer product operation according to the obtained vector so as to obtain an operation result. According to the device and method, vector data participating in the calculation is temporarily stored on a scratch pad memory, so that the data with different widths can be supported more flexibly and effectively in the vector outer product operation process and then the execution performance of applications comprising vector outer products is enhanced.

The invention discloses a large-span prestressing force groove-shaped composite plate used as a top plate of an anti-aircraft basement of 6 grades and lower than 6 grades. The composite plate is characterized in that a groove-shaped composite plate with a large span, a high load, a narrow board width and a low rib height forms a functional member together with an upper composite layer. Due to the adoption of an elastic-plastic deformation resilience characterization coefficient of a prestressing force plate, the resistance of the member under the action of a short-term effect of a nuclear explosion shock wave is improved. Therefore, materials are substantially saved, and the floor height is reduced. A groove-shaped large load meets the using requirements of an underground garage; parking spaces are saved in a large span; due to the adoption of a groove-shaped narrow board, the manufacturing cost and the floor height can be reduced; and the composite plate has an appropriate rib height and can meet load bearing requirements and the requirement of forming an air channel on a rib, the floor height is not increased, and a groove-shaped horizontal air channel is easy to connect. When the composite plate is used for the top plate of the underground garage, a construction period can be accelerated, the floor height can be reduced, the manufacturing cost can be reduced, a column grid can be saved, the parking spaces can be increased, and the grade of the underground garage can be improved.

The invention discloses a large-span and large-load pre-stress groove-shaped composite slab serving as a top plate of a basement. The composite slab is characterized in that: the composite slab is a groove-shaped plate in appearance; and the groove-shaped plate and an upper part overlapping layer form a functional component. Due to the groove shape and the large load, the requirements on the utilization of an underground garage are met; due to the large span, the parking space can be saved; due to a groove-shaped narrow panel, the construction cost and the layer height are saved; the rib height is moderate, so the requirement on loading capacity and the requirement of forming a ventilating passage on a rib can be met, and the floor height is not increased; and a groove-shaped horizontal ventilating passage is easy to connect. When the composite slab is used on the top plate of the underground garage, the construction period can be reduced, the floor height is reduced, the cost is reduced, column nets are saved, the parking space is increased, and the grade of the underground garage is improved.

The invention discloses a tea disease identification method based on deep transfer learning, and the method comprises the following steps: a training stage: obtaining pictures of a plurality of varieties of tea leaves, and taking the pictures as a training data set; preprocessing the pictures in the training data set, wherein the preprocessing comprises random flipping, random clipping, random angle rotation, color jitter and noise addition; constructing a picture classification model; introducing SE Block into the DenseNet, introducing a channel attention mechanism through the SE Block, and constructing a feature channel weighted SE-DenseNet network model; inputting the training data set into the constructed SE-DenseNet network model, training the SE-DenseNet network model through transfer learning, and storing the trained model; and a test stage: inputting a to-be-identified tea image, performing classification identification on diseases in the tea image through the SE-DenseNet network model obtained through training to obtain disease features, and outputting a disease classification result through the classifier. According to the method, the recognition accuracy under the conditions of small samples and uneven sample distribution is improved, and the recognition accuracy and speed are higher than those of an original model.

The invention discloses an Internet of Things storage pledge supervision intelligent tray device and system, and the device comprises an intelligent tray which is integrated with a plurality of sensors, such as weight and acceleration sensors, and through combination with a fusion analysis algorithm, physical quantities, such as weight and position, of supervised goods can be continuously obtained, and abnormal behaviors, such as movement and inclination, can be monitored. The intelligent tray is connected to the Internet of Things gateway, management, information convergence and protocol conversion of the tray are achieved through the Internet of Things gateway, the gateway is connected to the cloud storage pledge service platform, and service operations such as management, warehouse-in and warehouse-out operation, supervision alarm and fault operation and maintenance of the gateway are achieved through the cloud storage pledge service platform. When risks such as unauthorized movement, stealing and robbing of goods occur, the system can timely inform supervisors and privileges in a short message or APP notification mode, the problems of existence and uniqueness of pledged goods and compliance of storage enterprise operation are solved, and an original supervision mode and system with excessive human intervention or too low efficiency are changed.

The invention discloses a traffic route generation method and device, computer equipment and a storage medium, and the method comprises the steps: screening obtained bus clearing data according to a potential passenger model, and obtaining a potential passenger data set according to the screened bus clearing data; clustering the bus stops according to a clustering algorithm, generating destinationclusters, arranging the destination clusters in an inverted order according to the number of passengers, and traversing and querying the destination clusters from the destination cluster with the first order; and determining the departure place cluster corresponding to the destination cluster according to the source information, and generating the traffic route according to the departure place cluster and the destination cluster, thereby solving the problem that the generation of the traffic route of the regular bus does not have rapidity, accuracy and comprehensiveness.

The invention discloses a multi-keyword ciphertext storage and retrieval method and system based on word vectors. The method comprises the following steps: a data owner represents a keyword of a plaintext document as an (n+1)-dimensional word vector, calculates and uploads a ciphertext index and an encrypted document corresponding to the plaintext document to a cloud server, and sends an index key, a decryption private key and model parameters to a data user; and the data user inquires a keyword set to generate a trap door, obtains a plurality of encrypted documents with the highest relevancyfrom the cloud server side, and obtains corresponding plaintext documents through decryption. According to the method, the implicit semantics of the words can be accurately obtained through the word vectors, so that the query accuracy can be improved; and the thought of an MRSC method is used for reference, so that the ciphertext query safety is guaranteed, and background attacks of enemies are prevented.

The invention provides a solid beverage for reinforcing gastrointestinal peristalsis and boosting defecation. The solid beverage comprises the following components in parts by weight: konjac glucomannan, resistant dextrin, psyllium seed husk powder, orange powder, Chinese dwarf cherry seed powder, sea-buckthorn powder, beta-glucan rice flour, yam flour, poria cocos powder, coix seed powder, hawthorn powder, agaric powder, Puer powder, polydextrose, probiotics powder, yeast and sucralose. A preparation method provided by the invention comprises the following steps: milling all the components into powder and uniformly mixing the powder, wherein beta-glucan rice is stir-fried at 100 DEG C for 2-4 minutes and then is milled into flour. Various foods are taken as main components of the solid beverage; the solid beverage is free from toxic reaction, is nonirritant and is suitable for long-term use; high dietary fiber and various food materials are combined with each other and the probioticsare adopted for reinforcing, so that the curative effects of adjusting intestinal function and relieving constipation are obvious.