59results about How to "Improve Data Management Efficiency" patented technology

The invention relates to an automatic classification system of aerial photographing line patrol images based on depth learning. The method comprises the following steps of: establishing a line patrolimage classification image library and a label library thereof; establishing a deep learning model, including VGGNet, ResNet, DenseNet, NasNet and other high performance classification models; training a classification model, and randomly performing data enhancement operation on the input data of each iteration in the training process, including rotation, filling cut and grayscale; integrating thefour trained models and obtaining the integrated model; setting classification rules, inputting the images to be detected into the integrated model, taking the three image categories with the first three confidence levels as the classification results of the images to be detected, and copying and storing the results to the servers of the three categories for archiving; continuously optimizing theclassification model, and periodically upgrading the performance of the automatic classification system.

The invention provides a test method for an open-loop fiber optic gyroscope, wherein the test method includes the following steps: Step 1) The original sampling data output by the gyroscope is transmitted to the signal amplifying circuit, and according to the speed of the turntable corresponding to the sampling point, judge The port output status of the acquisition card; step 2) the acquisition card performs sampling, and a value is smoothly output at 3200 points per second during sampling; step 3) data restoration; step 4) data collection, calculation, and preservation. The present invention also provides a test system for an open-loop fiber optic gyroscope, which includes a rate turntable, on which a gyroscope is arranged, and the lead-out plug on the gyroscope includes a gyroscope power supply end and a gyroscope output end, and the gyroscope power supply end The power control circuit is connected with the power supply, the gyro output is connected with the output of the acquisition card through the signal amplification circuit, the power control circuit and the signal amplification circuit are connected with the acquisition card, and the acquisition card is connected with the central control host computer.

The invention provides a real-time and non-location-limiting atmosphere particle substance measurement data generation system and aims to achieve a monitoring method of atmosphere particle substance in real time and high region resolution. The system comprises: a movable atmosphere particle substance measuring sensor which is used for acquiring continuous multi-point atmosphere particle substance data; a satellite positioning system receiver which is used for dynamically acquiring satellite positioning system information; a timer; a local information communication apparatus; an algorithm which is used for calibrating a measuring time and a measuring position of the atmosphere particle substance data; a data communicating service carrier; a remote communicating apparatus; a remote data server group which is compatible with cloud computing; a data management system which is used for analyzing and classifying the atmosphere particle substance data in which the measuring time and the measuring position are calibrated and is used for providing data saving, data reading, data downloading and data uploading through a program which is compatible with the cloud computing. The system, in a high-efficiency and low-cost manner, can achieve real-time generation of the atmosphere particle substance data and the generated atmosphere particle substance data has the high region resolution.

In accordance with embodiments, there are provided mechanisms and methods for conditionally performing garbage collection. These mechanisms and methods for conditionally performing garbage collection can enable reduced application delay, improved data management efficiency, enhanced customer response, etc.

The invention is suitable for the technical field of data management, and provides a data quality verification method and a system based on double helix management, and the method comprises the steps:obtaining a business demand and a data demand, carrying out the tracing of the business demand and the data demand, and generating business tracing information and data tracing information; formulating a service specification according to the service traceability information, performing data evaluation according to the data traceability information, and formulating a data optimization scheme according to an evaluation result; implementing business processing according to business specifications, and implementing data optimization according to a data optimization scheme; and performing data verification on the local database, and performing data replacement according to a verification result. According to the method, service requirements, service traceability, service specification formulation and service processing implementation are designed; a service closed-loop structure is formed, a data closed-loop structure is formed through data demand acquisition, data tracing, data evaluation and data optimization design, a double-helix management mechanism is formed through the service closed-loop structure and the data closed-loop structure, and the data management efficiency is improved.

The invention discloses an immediate processing method and device for a document. The method comprises the steps that 1 edit information and an initial text of a document are obtained; 2 the edit information of the document is added to an edit table of the document; 3 the edit table of the document is saved. Accordingly, the efficient data processing experience can be provided for a user.

The embodiment of the invention discloses a server index data acquisition method, a server index data acquisition device and a computer readable storage medium. The method comprises the following steps: acquiring various index data of a server according to a current sampling frequency; and calculating the linear change rate of each type of index data in a preset time period. The linear change rateof each type of index data reflects the change trend of the index data. The CPU utilization rate and the acquisition time of each acquisition thread reflect the load cost of the system. The current sampling frequency is adjusted according to each linear change rate and the CPU utilization rate and the acquisition time consumption of each acquisition thread, thereby realizing adaptive frequency acquisition of the server index data. By comprehensively considering the load cost of the system and the change rule of the index data, the adjusted sampling frequency is more suitable for actual requirements, the change characteristics of the index data can be truly reflected while the load of the system is reduced, and the data management efficiency of the system is effectively improved.

The invention relates to the technical field of massive data query management, in particular to a complicated query method based on attributes which is specially applicable to fabric data management and can effectively improve the fabric management efficiency. The method is characterized in that a data storage database is initially built, wherein each attribute corresponds to a field of an integer type; each attribute value is saved as the nth power of 2, and bitwise AND operation is used during query, wherein the query method comprises the steps that attribute values to be searched are summed during query, a retrieval value is obtained by means of & operation of SQL, query is conducted in the storage database through the retrieval value, and a fabric with the attributes to be searched is obtained. By means of the complicated query method based on the attributes, the query efficiency can be greatly improved, script complexity is reduced, the efficiency of fabric data management is improved, and thus the design and production efficiencies are improved.

The invention provides a spatial data indexing method integrating GeoHash and Quadtree. The method comprises the steps of creating an improved quadtree index structure and dynamically updating from bottom to top ,wherein the improved quadtree index structure comprises an improved quadtree, a direct access table and a Hash index table established on a space object, the improved quadtree adopts a region division rule based on GeoHash and quadtree similarity, a space region is divided according to the GeoHash, the quadtree with nodes in one-to-one correspondence with GeoHash codes is establishedaccording to a grid, and a direct access table is used for pointing to all intermediate nodes in the improved quadtree; and a corresponding space object record in the improved quadtree leaf child nodeis accessed by using a Hash index table based on a space object identifier based on a bottom-up dynamic updating algorithm, and when the space object sends a position updating request, a plurality ofupdating strategies are adopted according to the position change degree of the space object. According to an improved quadtree index structure, an extended bottom-up dynamic updating algorithm is provided, and frequent updating operation of indexes is supported.

The invention discloses a cloud data processing method based on workload, which comprises the following steps: a cloud data copy placement step: acquiring a historical workload, and adopting a transaction-driven partitioning strategy to complete the mapping of data items to a storage area; a cloud data replica locating step: after receiving the transaction request initiated by the application system, using a greedy algorithm to select the locating replica. A cloud data processing system based on workload is also disclosed. The invention supports the elastic expansion of the data storage model,at the same time, solves the performance degradation problem caused by the workload change, and ensures the efficient operation of the application.

The invention discloses a product data processing method and device. The method comprises the steps that in the design stage of a product, product configuration and design data of the product are constructed, and the design data is used for designing the product; in the production stage of the product, production data is constructed based on the product configuration, and the production data is used for guiding production of the product; and in the operation and maintenance stage of the product, operation and maintenance data is constructed based on the product configuration, and the operationand maintenance data is used for operating and maintaining the product. The technical problem of relatively low product data management efficiency caused by incapability of establishing effective connection due to mutual independence of all stages in a product life cycle in related technologies is solved.

The invention discloses a data management system and a technology transaction platform, and the system comprises a data management module which is used for the management of the basic data and user information of the technology transaction platform; a business management module which is used for managing the business application of the technology transaction platform and the interaction between the technology transaction parties; a resource management module which is used for gathering and integrating similar service parties and dynamically connecting service intermediary resources of all links of the whole transaction process so as to realize matching of requirements and resources; and a resource storage module which is used for storing various resource data related to the technology transaction platform. According to the invention, effective data management can be carried out on the technology transaction platform, accuracy and reliability of data interaction of the technology transaction platform are guaranteed, improvement of data management efficiency, resource utilization efficiency and operation reliability of the technology transaction platform is facilitated, and functions of the technology transaction platform are enriched. The system has high advancement, openness, expandability, stability and reliability.

The invention provides an intelligent supervision system for fish video recognition. The system comprises a business layer, a data layer and a presentation layer; the business layer comprises a user system, a business system and an algorithm system, wherein the user system is mainly used for managing platform user operation behaviors and information management, the business system is used for managing businesses among whole platform modules, and the algorithm system realizes fish video identification through a fish model clustering algorithm. According to the method, a fish model is fixed at an original point of a coordinate system, and an independent fish clustering color library is constructed by utilizing gray scale distribution statistics of fish schools and background targets according to the characteristics of color clustering and fish videos; and therefore, clustering time is reduced, and the efficiency is improved; and a clustering value is adaptively determined through the valley value of a normalized histogram, so that low-efficiency and invalid segmentation caused by artificial experience is avoided.

The invention discloses a cloud platform data management system. The system comprises a data input unit, a cloud platform, a data conversion module, a data feedback module, a data backup module, a classification identification unit, a hierarchical compression unit, a cloud storage unit, an extraction unit, a first encryption unit and a second encryption unit. The output end of the data input unitis connected with the input end of the cloud platform, and the output end of the cloud platform is connected with the input end of the data conversion module and the input end of the data backup module. The invention discloses a cloud platform data management system. Data information is input through the data input unit; the information is analyzed and processed through the cloud platform, classified management is carried out on the information through the classification identification unit and the hierarchical compression unit, encryption processing is carried out in cooperation with the first encryption unit and the second encryption unit, automatic management of the data is achieved, the data management efficiency is improved, the labor cost is reduced, and the safety performance is greatly improved.

The invention provides a data management method based on a decentralized autonomous organization, a server and a storage medium. The method comprises the steps of obtaining a data change request; judging whether the data change request is a preset data change request or a non-preset data change request; if the change member information is not the preset data change request, judging whether an organization member corresponding to the change member information meets a data change condition or not; if the data change condition is met, generating a non-preset data change confirmation instruction and sending the non-preset data change confirmation instruction to an organization member terminal with management authority for the non-preset data change request; obtaining a first feedback result ofthe organization member terminal having the management authority for the non-preset data change request through a block chain; and determining whether to execute the non-preset data change request according to the first feedback result. The data change types of the decentralized autonomous organization can be distinguished and classified for intelligent management, and the data management efficiency of the decentralized autonomous organization is improved.