CN109887098B - Web AR data presentation mode based on distributed computing - Google Patents
Web AR data presentation mode based on distributed computing Download PDFInfo
- Publication number
- CN109887098B CN109887098B CN201910116350.7A CN201910116350A CN109887098B CN 109887098 B CN109887098 B CN 109887098B CN 201910116350 A CN201910116350 A CN 201910116350A CN 109887098 B CN109887098 B CN 109887098B
- Authority
- CN
- China
- Prior art keywords
- model
- data
- client
- request
- server
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention discloses a web AR data presentation mode based on distributed computing, and relates to the technical field of data presentation modes; respectively storing and managing the model and the action data at a server side; according to the type of the request, the client sends the request to a nearby mobile edge server; the mobile edge server carries out requirement analysis according to the request of the user; the cloud computing side analyzes the interactive request of the client side, acquires the matching action and sends the matching action to the edge server; performing data calculation at a moving edge calculation end; the mobile edge server receives the model transmitted by the cloud computing terminal; and returning the calculation result data to the client to realize the loading function of the dynamic model. The operation efficiency of the client is greatly improved; the service is more flexible; the method is more suitable for the calculation response mode of augmented reality, reduces the GPU calculation cache pressure of the mobile terminal, and realizes relatively complex interaction model calculation; the computing pressure of the cloud is reduced, and the performance is obviously improved.
Description
Technical Field
The invention relates to the technical field of data presentation modes, in particular to a web AR data presentation mode based on distributed computing.
Background
With the advance of information-based construction and the development of intelligent mobile terminal devices, the augmented reality technology based on the intelligent mobile terminal brings huge development opportunities for industries, businesses and tourism. Meanwhile, the web-based mobile augmented reality technology is more widely applied to the field of mobile augmented reality due to the characteristics of portability, universality and the like.
However, compared with the traditional mobile augmented reality technology, the web-based mobile augmented reality brings convenience to the majority of users, and is different from the traditional mobile application due to the characteristics of rendering and interactive operation based on the browser, and the rendering interactive operation based on the browser needs to call the bottom operation of the intelligent mobile device and the interface of the real device by means of the inner core of the browser; the browser of the mobile terminal has certain limitation on scheduling of the memory and the operation resources, which indirectly influences the invoking of the calculation resources in the augmented reality rendering and interactive operation processes of the mobile terminal, and influences the construction of the augmented reality interactive experience and the service environment of the mobile web.
With the above problems, compression and transmission of model data are the main research objects in mobile augmented reality applications, especially for the introduction of network streaming of three-dimensional models. At present, relevant researchers at home and abroad have obtained certain research results aiming at network streaming transmission. Martin proposes a model-based adaptive star concept, and the models are compressed through different respective rates; ren Huiling, shenchun propose research on loading of model data in large-scale scenarios, and improvement of streaming algorithms, etc.
However, most of the current research focuses on model data compression and transmission loading modes in a PC-based application environment, the computing pressure of a client is mainly considered on a loading algorithm and the transmission and loading modes, and for mobile web augmented reality applications, a user considers various factors such as traffic cost and initial loading time in a business application process, so that further optimization needs to be performed on a network streaming transmission method from the aspects of initial loading efficiency, network transmission compression and the like.
Disclosure of Invention
Aiming at the defects and shortcomings of the prior art, the invention provides a web AR data presentation mode based on distributed computing, which has a simple structure, reasonable design and convenient use, and greatly improves the operation efficiency of a client; the service is more flexible; the method is more suitable for the computational response mode of augmented reality, reduces the GPU computation cache pressure of the mobile terminal, and realizes relatively complex interactive model computation; the computing pressure of the cloud is reduced, and the performance is obviously improved.
In order to realize the purpose, the invention adopts the technical scheme that: the operation steps are as follows:
1. respectively storing and managing the model and the action data at a server side; firstly, model data are converted into object data, and in conversion application, the model data are converted into combination of attribute objects according to definition of each key value in a model and are stored in a server;
2. according to the type of the request, the client sends the request to a nearby mobile edge server; when a client side makes a page request to a server side, firstly, a related request is sent to a nearby mobile edge server through an http protocol;
3. the mobile edge server carries out requirement analysis according to the request of the user; judging whether the model data is routed in advance, if so, requesting action data from a cloud server, and performing model operation after waiting for response; if the content routing is not performed, requesting a model and action data from the cloud, and performing model operation, namely, loading the model by a cloud server, receiving the multi-dimensional data model of the client by a server end, and performing rendering calculation through a model adaptation module and a model generation module;
4. the cloud computing side analyzes the interactive request of the client side, acquires the matching action and sends the matching action to the edge server; analyzing the interactive request of the client while the client sends the interactive request, inquiring and comparing in a semantic description library to obtain a matched action, and sending the matched action to the mobile edge server;
5. performing data calculation at a moving edge calculation end; the mobile edge server receives the model and the action method transmitted by the cloud computing end; performing composite operation on the received model and the action data in a program mode through a data structure in the model, and returning model response data to the server in a data format;
6. returning the calculation result data to the client to realize the loading function of the dynamic model; and the edge computing server returns the generated new model to the client in an interface form by a rapid computing method to realize the responding service.
Further, the specific content of performing respective storage management on the model and the action data at the server side in step 1 is as follows: based on a mass data storage model of cloud computing, according to a core computing mode MapReduce of the cloud computing and by means of an open source distribution parallel programming framework Hapdoop for realizing the MapReduce computing mode, the storage model and the cloud computing are combined together to realize the distributed storage of mass data, static model data required by a user is stored on a cloud computing end in such a way, the data is mapped into different blocks through a program of a Map function after being divided, the blocks are distributed to a computer group for processing to achieve the effect of distributed operation, the storage effect is realized, action data of the model is stored in a database, and the action data of the model and the model are respectively stored.
Further, the specific content of the request type in step 2 is: when the intelligent client requests, the client firstly needs to access the bottom information of the equipment, obtain the visual computation related equipment parameters of the equipment, such as resolution, screen size, CPU operation efficiency and other information, and construct a requested data model.
Further, the specific step of sending the request to the nearby mobile edge server in step 2 is: after the client sends a request to the server by an http protocol, the browser monitors request information returned by the server; after the request information is acquired, the status code of the return information is judged firstly, and when the status code is 200, the json format data of the request is captured and stored in the memory of the browser.
Further, the specific step of the mobile edge server performing the requirement analysis in step 3 is as follows:
3.1, after acquiring the request, the server analyzes the request data of the request client, performs operation, determines the request state information of the client, and determines the upper limit of the model data presented under the optimal network condition; gradually sending a model data node set to the client from M1 under the upper limit condition, returning the time consumed by the receiving end after the request is received, and calculating the network environment parameters of the client; the parameter is transmitted to the server end at the same time, and after receiving the parameter, the server end determines an upper limit cut-off value in the determined M sequence;
3.2, a multi-dimensional parameter model is constructed by parameters related to the loading operation of the client and the model, the parameters representing the operation capability, the network bandwidth, the interactive display interface and the like of the client are carried, the physical parameter model is used for requesting the server, the server analyzes the physical parameters of the client, the model suitable for the client is calculated, and the model is returned to the client for loading and rendering operation, so that the operation efficiency of the client is greatly improved.
Further, the specific steps of the cloud server in step 3 for loading the model are as follows: the server receives the multidimensional data model of the client, analyzes the model data, acquires physical parameters related to the current running state of the client, transmits the physical parameters to the model adaptation module, analyzes the physical parameters of the client by the model adaptation module, and transmits the model output weight to the model generation module; and the model generation module performs model data generation processing according to the weight value, and returns a processing result to the client in an interface mode for rendering calculation.
Further, the specific steps of the cloud computing side analyzing the interaction request of the client in the step 4 are as follows: the cloud computing center analyzes the operation request of the client, obtains a physical environment model of the client, exerts the high concurrency data advantage of cloud computing, calculates the weight of the client request model, and routes the weight to the mobile edge server nearest to the client through a network to perform corresponding model request computing; meanwhile, the server also needs to perform logical response of response according to the service request of the user, and returns the response result to the client to realize corresponding service scene construction.
Further, the specific step of obtaining the matching action and sending the matching action to the edge server in step 4 is: at the cloud end, firstly, storing a static model required by a service according to the service requirement; meanwhile, storing the action data of the model in a database; and performing semantic description on the model action while storing the model action data, analyzing the interaction request of the client while sending the interaction request by the client, performing query comparison in a semantic description library, acquiring a matched action, and sending the matched action to the mobile edge server.
Further, the specific step of the moving edge calculation end in step 5 performing data calculation is as follows:
5.1, moving edge calculation under a 5g application scene, and compared with traditional cloud calculation which takes centralization and large-scale as a core, the method is more suitable for a calculation response mode of mobile augmented reality; the method comprises the following steps of performing mobile edge computing, wherein computing is converted from a traditional cloud computing center server cluster to an edge closer to a mobile terminal, network resources occupied from a cloud computing center to the edge of a network and consumed transmission time are avoided in network circulation, and meanwhile, most of computing amount is converted from a center server to the edge of a mobile network, so that the computing capability of web augmented reality is improved to a certain extent; meanwhile, through a mode of carrying out cooperative operation through cloud computing and mobile edge computing, the GPU operation cache pressure of a mobile terminal is reduced, and the operation capability of a complex interaction model is realized; in an interface operation architecture mode based on mobile edge computing, a traditional center server cluster is used as a function center for client service request response, and a mobile edge server is used as a service center for model operation; firstly, a client sends an operation request carrying physical parameters of the client to a server under a corresponding service environment, a cloud computing center analyzes the operation request of the client to obtain a physical environment model of the client, exerts the high concurrency data advantage of cloud computing, calculates the weight of the request model of the client, and routes the request model to a mobile edge server nearest to the client through a network to perform corresponding model request computing; meanwhile, the server also needs to perform logical response of response according to the service request of the user, and returns the response result to the client to realize corresponding service scene construction;
5.2, referring to an operation mechanism of the CDN, the model data is routed to an edge server end close to an application place in advance, and transmission time is shortened; when a client has a request, the request is analyzed through a cloud computing terminal and then is sent to a mobile edge computing terminal, and a mobile edge computing server immediately checks whether the content requested by a user exists in a local cache or not, and if so, the mobile edge computing server directly serves the user; and if not, requesting the model and the action data from the cloud, performing model operation, and caching to the local. Thus, the request time is reduced, and the problem of network congestion is solved;
5.3, the model calculation module further samples and compresses the mapping of the model data and the three-dimensional imaging data according to the received weight, and returns the generated new model to the client in a form of an interface by a rapid calculation method to realize a response service; in the sampling and compression calculation of the model, on the realization of an edge server, a calculation method based on machine learning is adopted, certain preprocessing is firstly carried out on model compressed data, the model data is firstly preprocessed, and meanwhile, the model with high request rate is subjected to cloud calculation, storage and other operations according to the data probability requested by the model, so that the model can carry out quick model data response in the request process and is matched with the physical and network conditions of a client.
Further, the specific steps of implementing the loading function of the dynamic model in step 6 are as follows: according to the design of the optimization scheme, node data of the model is transmitted to the client in a dynamic control mode, in the traditional streaming transmission process, extra time consumption is brought to the time cost of a mode that a model data file needs to be transmitted after being preprocessed after being requested, and data with different threshold values M are stored in an independent data object mode in a dynamic model transmission mode based on an interface;
the specific loading process of the model of the client is as follows: the whole loading process of a model is divided into two parts: one is the process that the client acquires model data in the JD format from the server, namely the process of network transmission; js usable data is converted into data capable of being used by three through JDLoader plug-in after Json data is obtained and then displayed on a client, and the process is irrelevant to a network and is only calculated by the client;
the process of loading the model actions of the client is as follows: aiming at the condition that a model is loaded as required, the method is realized by separating model data and animation data in a JD file at the rear end, loading the model data and displaying the model data, enabling a user to see static model data information on a client, then initiating a requirement to a server according to the requirement of the user to obtain a matched model action, then loading the animation data as required, and extracting an animation data format which is converted into an animation data format which can be used by three. In addition to single model loading, there is loading content for multi-model scenarios, and the traditional approach is to take the network to load all included models simultaneously by default when the page is opened. The selection of the packaging function enables model files to be loaded in a queue, one model file is loaded and displayed at the same time, and after one model is finished, the display work of the next model is carried out.
After the method is adopted, the invention has the beneficial effects that: the web AR data presentation mode based on distributed computing greatly improves the operation efficiency of the client; the service is more flexible; the method is more suitable for the calculation response mode of augmented reality, reduces the GPU calculation cache pressure of the mobile terminal, and realizes relatively complex interaction model calculation; the cloud computing system has the advantages of being simple in structure, reasonable in arrangement, low in manufacturing cost and the like.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of the client model specific loading in the present invention.
FIG. 2 is a flow chart of loading model actions at a client in the present invention.
FIG. 3 is a flow diagram of data storage in a browser memory in accordance with an embodiment.
Detailed Description
The invention will be further described with reference to the accompanying drawings.
The technical scheme adopted by the specific implementation mode is as follows: the operation steps are as follows:
1. respectively storing and managing the model and the action data at a server side; firstly, model data are converted into object data, and in conversion application, the model data are converted into combination of attribute objects according to definition of each key value in a model and are stored in a server; the specific steps of the model distributed storage management are as follows:
taking three-dimensional city model data storage as an example, a large scale map sheet dividing method based on a topological relation model, and uniformly naming and coding divided three-dimensional model data; the MongoDB fragment cluster server is constructed by means of strong mass data organization and high-efficiency multi-concurrent access function of a non-relational database MongoDB; the method comprises the steps of performing unit division on three-dimensional City model data, modeling by adopting a rule modeling software City Engine to obtain a three-dimensional City model, and performing a data storage experiment by means of non-relational database software MongoDB;
2. according to the type of the request, the client sends the request to a nearby mobile edge server; when a client side makes a page request to a server side, firstly, a related request is sent to a nearby mobile edge server through an http protocol; taking an Android platform as an example, an Android client wants to establish a communication mode with a server, the Android client adopts an HTTP communication mode, and the HTTP communication mode is divided into a get mode and a post mode, when the client sends a request to the server, a data block, namely request information, is transmitted to the server, a status code of return information is determined according to the request information, and data is stored in a browser memory (see fig. 3);
3. the mobile edge server carries out requirement analysis according to the request of the user; judging whether the model data is routed in advance, if so, requesting action data from a cloud server, and performing model operation after waiting for response; if the content routing is not performed, requesting a model and action data from the cloud end, and performing model operation, wherein the specific method comprises the following steps: because the mobile edge computing server is arranged close to the information source and the mobile computing service is close to the terminal equipment, when the mobile edge server receives a request of a client, the data processing is simply carried out locally, all data or information is not required to be uploaded to the cloud, the corresponding model can be directly called if the requested model data is preloaded, the preloading of the model is not completed, and the mobile edge server sends a request to the cloud to acquire the model data, action and other information stored by the cloud; if the content routing is not carried out, the cloud server carries out loading of the model; the server side receives the multi-dimensional data model of the client side, and the rendering calculation is carried out through the model adaptation module and the model generation module;
4. the cloud computing side analyzes the interactive request of the client side, acquires the matching action and sends the matching action to the edge server; the method comprises the following steps that when a client sends an interactive request, the interactive request of the client is analyzed, query comparison is carried out in a semantic description library, a matched action is obtained and sent to a mobile edge server, in a specific application example, the cloud computing side analyzes the interactive request of the client, and the obtained matched action is sent to the edge server, and the specific steps of:
the client sends an HTTP request header, and the server processes according to a response DoGet ()/Dopost () method according to GET/POST at the time of the request; after processing the request, the response object obtains a PriterWriter output stream object out, and outputs the data in a format of Accpt (Accpt) submitted at the client, such as according to response.setcontentType ("text/html; charset = gb 2312'), in a form of print (); the server inquires the action matched with the model in the database according to the request content of the client, sends data information to the edge server, then the server closes the connection, and the client analyzes the return response head to recover the page;
5. performing data calculation at a moving edge calculation end; the method comprises the steps that a mobile edge server receives a model and an action method transmitted by a cloud computing end; through a data structure in the model, the received model and action data are subjected to compound operation in a program mode, model response data are returned to a server in a data format, and by taking the application scene of ' 5G ' with low power consumption and large connection ' as an example, mobile edges are deployed at the mobile edges, so that the wireless network and the internet are effectively fused together, functions of computing, storing, processing and the like are added at the wireless network side, a mobile edge cloud is constructed, and an information technology service environment and cloud computing capability are provided. The forwarding and processing time in data transmission is reduced, the end-to-end time delay is reduced, the requirement of low time delay is further met, and the power consumption is reduced;
6. returning the calculation result data to the client to realize the loading function of the dynamic model; the edge computing server returns the generated new model to the client in an interface form by a rapid computing method to realize a response service; the specific implementation method is as follows:
according to the design of the optimization scheme, node data of the model is transmitted to the client in a dynamic control mode, in the traditional streaming transmission process, model data files need to be preprocessed after being requested and then transmitted in a mode of bringing extra time consumption on time cost, in the specific implementation mode, a dynamic model transmission mode based on an interface is adopted, and data with different threshold values M are stored in an independent data object mode;
the model specific loading process of the client (see fig. 1), the whole loading process of one model is divided into two parts: one is that the client end obtains model data in JD format from the server, namely the process of network transmission; the second is that Json data is obtained and then converted into data which can be used by three.js through JDLoader plug-in, and then the data is displayed on a client, and the process is irrelevant to a network and is only calculated by the client;
procedure for model action loading of client (see fig. 2): aiming at the condition that a model is loaded as required, the method is realized by separating model data and animation data in a JD file at the back end, loading the model data and displaying the model data, enabling a user to see static model data information on a client, then initiating a requirement to a server according to the requirement of the user to obtain a matched model action, then loading the animation data as required, converting the animation data into an animation data format which can be used by three.
In addition to single model loading, there is loading content for multi-model scenarios, and the traditional approach is to take the network to load all included models simultaneously by default when the page is opened. The selection of the packaging function enables model files to be loaded in a queue, one model file is loaded and displayed at the same time, and after one model is finished, the display work of the next model is carried out.
The beneficial effects of the embodiment are as follows:
1. in an augmented reality application scene, strict limitations are performed on the aspects of the calculation amount, the capacity and the like of the model, and the physical parameter model is used for sending a request to the server, so that the operation efficiency of the client is greatly improved;
2. compared with the traditional mode of providing services in a file form, the web augmented reality model service mode based on the interface mode is more flexible in service;
3. the improvement of the debit phone operation based on the mobile edge calculation can be more suitable for the calculation response mode of the augmented reality, the GPU calculation cache pressure of the mobile terminal is reduced, and the relatively complex interactive model operation is realized;
4. in a mobile edge computing mode, a model response computing mode is lowered from a cloud end to a mobile edge end which is closer to a client end, so that the computing pressure of the cloud end is reduced;
5. the cache operation pressure of the client can be relieved to a certain extent in a single-action response mode; especially, the multi-action interaction mode oriented to complex business scenes can have obvious improvement on the performance.
The above description is only for the purpose of illustrating the technical solutions of the present invention and not for the purpose of limiting the same, and other modifications or equivalent substitutions made by those skilled in the art to the technical solutions of the present invention should be covered within the scope of the claims of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (10)
1. A web AR data presentation mode based on distributed computing is characterized in that: the operation steps are as follows:
(1) Respectively storing and managing the model and the action data in a cloud server; firstly, model data are converted into object data, and in conversion application, the model data are converted into combination of attribute objects according to definition of each key value in a model and are stored in a cloud server;
(2) According to the type of the request, the client sends the request to a nearby mobile edge server; when a client requests a page from a cloud server, firstly, a related request is sent to a nearby mobile edge server through an http protocol;
(3) The mobile edge server carries out demand analysis according to the request of the user; judging whether the model data is routed in advance, if so, requesting action data from the cloud server, and performing model operation after waiting for response; if the pre-routing is not carried out, the mobile edge server requests a model and action data from the cloud server to carry out model operation, namely, the cloud server carries out model loading, receives the multi-dimensional data model of the client and carries out rendering calculation through the model adaptation module and the model generation module;
(4) The cloud server analyzes the interaction request of the client, acquires a matching action and sends the matching action to the mobile edge server; analyzing the interactive request of the client while the client sends the interactive request, inquiring and comparing in a semantic description library to obtain a matched action, and sending the matched action to the mobile edge server;
(5) Performing data calculation on the mobile edge server; after receiving the model and the action data transmitted by the cloud server, the mobile edge server; performing composite operation on the received model and the action data in a program mode through a data structure in the model, and returning model response data to the client in a data format;
(6) Returning the calculation result data to the client to realize the loading function of the dynamic model; and the mobile edge server returns the generated new model to the client side in an interface form by a rapid calculation method, so that a response service is realized.
2. A distributed computing based web AR data presentation as claimed in claim 1, wherein: the specific contents of performing respective storage management on the model and the action data in the cloud server in the step (1) are as follows: the method comprises the steps that a mass data storage model based on cloud computing is combined with the cloud computing according to a core computing mode MapReduce of the cloud computing and by means of an open source distribution parallel programming framework Hapdoop of the MapReduce computing mode, distributed storage of mass data is achieved, static model data required by a user are stored on a cloud server in the mode, the data are mapped into different blocks through a program of a Map function after being divided, the blocks are distributed to a computer group to be processed, the effect of distributed operation is achieved, the storage effect is achieved, action data of the model are stored in a database, and the action data of the model and the model are stored respectively.
3. A distributed computing based web AR data presentation as claimed in claim 1, wherein: the specific content of the request type in the step (2) is as follows: when the intelligent client requests, the client firstly needs to access the bottom information of the equipment, obtain the visual computation related equipment parameters of the equipment, namely the resolution, the screen size and the CPU operation efficiency, and construct a requested data model.
4. A distributed computing based web AR data presentation as claimed in claim 1, wherein: the specific steps of sending the request to the nearby mobile edge server in the step (2) are as follows: after the client sends a request to the server by an http protocol, the browser monitors request information returned by the server; after the request information is acquired, the status code of the return information is judged firstly, and when the status code is 200, the json format data of the request is captured and stored in the memory of the browser.
5. A distributed computing based web AR data presentation as claimed in claim 1, wherein: the specific steps of the mobile edge server in the step (3) for analyzing the requirements are as follows:
(3.1) after the server acquires the request, analyzing the request data of the request client, performing operation, determining the request state information of the client, and determining the upper limit of the model data presented under the optimal network condition; gradually sending a model data node set to the client from M1 under the upper limit condition, returning the time consumed by the receiving end after the request is received, and calculating the network environment parameters of the client; the parameter is simultaneously transmitted to a cloud server, and after the cloud server receives the parameter, an upper limit cut-off value is determined in the determined M sequence;
and (3.2) constructing a multi-dimensional parameter model by using parameters related to the loading operation of the client and the model, carrying the operational capability, the network bandwidth and the interactive display interface which represent the client, requesting the cloud server by using the physical parameter model, analyzing the physical parameters of the client by using the cloud server, calculating the model suitable for the client, and returning the model suitable for the client to the client for loading and rendering operation.
6. A distributed computing based web AR data presentation as claimed in claim 1, wherein: the specific steps of the cloud server in the step (3) for loading the model are as follows: the cloud server receives a multi-dimensional data model of the client, analyzes the model data, acquires physical parameters related to the current operation state of the client, transmits the physical parameters to the model adaptation module, analyzes the physical parameters of the client by the model adaptation module, and transmits the model output weight to the model generation module; and the model generation module performs model data generation processing according to the weight value, and returns a processing result to the client in an interface mode for rendering calculation.
7. The web AR data presentation based on distributed computing of claim 1, wherein: the specific steps of the cloud server analyzing the interaction request of the client in the step (4) are as follows: the cloud server analyzes the operation request of the client to obtain a physical environment model of the client, exerts the high concurrency data advantage of cloud computing, calculates the weight of the client request model, and routes the weight to the mobile edge server nearest to the client through a network to perform corresponding model request computing; meanwhile, the cloud server also needs to perform logical response of response according to the service request of the user, and returns the response result to the client to realize corresponding service scene construction.
8. A distributed computing based web AR data presentation as claimed in claim 1, wherein: the specific step of obtaining the matching action and sending the matching action to the mobile edge server in the step (4) is as follows: the cloud server firstly stores a static model required by the service according to the service requirement; meanwhile, storing the action data of the model in a database; and carrying out semantic description on the model action while storing the model action data, analyzing the interaction request of the client while sending the interaction request by the client, inquiring and comparing in a semantic description library to obtain a matched action, and sending the matched action to the mobile edge server.
9. A distributed computing based web AR data presentation as claimed in claim 1, wherein: the specific steps of the mobile edge server in the step (5) for data calculation are as follows:
(5.1) firstly, the client sends an operation request carrying physical parameters of the client to a cloud server in a corresponding service environment, the cloud server analyzes the operation request of the client to obtain a physical environment model of the client, exerts the advantage of high concurrency data of cloud computing, calculates the weight of the request model of the client, and routes the request model to a mobile edge server nearest to the client through a network to perform corresponding model request computation; meanwhile, the cloud server also needs to perform logical response of response according to the service request of the user, and returns the response result to the client to realize corresponding service scene construction;
(5.2) routing the model data to a mobile edge server which is close to the application place in advance by referring to an operation mechanism of the CDN, so that the transmission time is reduced; when a client has a request, the request is analyzed by the cloud server and then sent to the mobile edge server, and the mobile edge server immediately checks whether the local cache has the content requested by the user or not, and if so, the mobile edge server directly serves the user; if not, requesting the model and the action data from the cloud server, performing model operation, and caching to the local;
(5.3) the model calculation module further samples and compresses the mapping of the model data and the three-dimensional imaging data according to the received weight, and returns the generated new model to the client in an interface form by a rapid calculation method to realize a response service; regarding the sampling and compression calculation of the model, on the realization of an edge server, a calculation method based on machine learning is adopted, the compressed data of the model is preprocessed firstly, the model data is preprocessed firstly, and meanwhile, the model with high request rate is subjected to cloud calculation and storage operation according to the data probability requested by the model, so that the model can perform quick model data response in the request process and the physical and network conditions of a client side are matched.
10. A distributed computing based web AR data presentation as claimed in claim 1, wherein: the specific steps for realizing the loading function of the dynamic model in the step (6) are as follows: according to the design of an optimization scheme, node data of a model is transmitted to a client in a response mode in a dynamic control mode, in the traditional streaming transmission process, extra time consumption is brought to the model data file in the time cost mode of transmission after preprocessing the file after request, and data with different threshold values M are stored in the form of independent data objects in a dynamic model transmission mode based on an interface;
the specific loading process of the model of the client is as follows: the whole loading process of a model is divided into two parts: one is that the client acquires model data in the JD format from the cloud server, i.e. the process of network transmission; the second is that Json data is obtained and then converted into data which can be used by three.js through JDLoader plug-in, and then the data is displayed on a client, and the process is irrelevant to a network and is only calculated by the client;
the process of loading the model action of the client is as follows: aiming at the condition that the model is loaded according to the requirement, the implementation method is that the model data and the animation data in the JD file are separated at the back end, the model data are loaded and displayed firstly, a user can see static model data information on a client firstly, then the requirement is initiated to a cloud server according to the requirement of the user to obtain a matched model action, then the animation data are loaded according to the requirement, and an animation data format which can be used by the thread is extracted and converted from the JDLoader and the model is moved by a special animation loading method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910116350.7A CN109887098B (en) | 2019-02-13 | 2019-02-13 | Web AR data presentation mode based on distributed computing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910116350.7A CN109887098B (en) | 2019-02-13 | 2019-02-13 | Web AR data presentation mode based on distributed computing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109887098A CN109887098A (en) | 2019-06-14 |
| CN109887098B true CN109887098B (en) | 2023-01-17 |
Family
ID=66928133
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910116350.7A Active CN109887098B (en) | 2019-02-13 | 2019-02-13 | Web AR data presentation mode based on distributed computing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109887098B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110677683B (en) * | 2019-09-30 | 2022-03-04 | 北京奇艺世纪科技有限公司 | Video storage and video access method and distributed storage and video access system |
| CN111212107B (en) * | 2019-12-10 | 2022-05-13 | 中移（杭州）信息技术有限公司 | Service processing method for CDN platform and CDN system |
| CN111212128B (en) * | 2019-12-30 | 2021-05-04 | 北京邮电大学 | Edge-assisted-based collaborative Web AR multi-user service system and method |
| CN111260540B (en) * | 2020-01-13 | 2023-06-13 | 成都卓影科技股份有限公司 | 2.5D conversion engine of 2D-3D under 5G network |
| CN111339192A (en) * | 2020-02-21 | 2020-06-26 | 深圳供电局有限公司 | Distributed edge computing data storage system |
| CN111400381A (en) * | 2020-02-28 | 2020-07-10 | 北京致医健康信息技术有限公司 | Data storage method and device, computer equipment and storage medium |
| CN114169365A (en) * | 2020-11-25 | 2022-03-11 | 摩赛恩科技(苏州)有限公司 | Mass spectrum data processing system and method |
| CN114791893B (en) * | 2021-12-15 | 2023-05-09 | 许磊 | Serialization system for random data access |
| CN114745260B (en) * | 2022-03-09 | 2024-04-02 | 优刻得科技股份有限公司 | Method, device, equipment and storage medium for enhancing computing power of content distribution network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103999434A (en) * | 2011-12-21 | 2014-08-20 | 瑞典爱立信有限公司 | Devices and methods using network load data in mobile cloud accelerator context to optimize network usage by selectively deferring content delivery |
| CN107222468A (en) * | 2017-05-22 | 2017-09-29 | 北京邮电大学 | Augmented reality processing method, terminal, cloud server and edge server |
| US9912718B1 (en) * | 2011-04-11 | 2018-03-06 | Viasat, Inc. | Progressive prefetching |
| CN108847981A (en) * | 2018-06-26 | 2018-11-20 | 咸宁职业技术学院 | Distributed computer cloud computing processing method |
-
2019
- 2019-02-13 CN CN201910116350.7A patent/CN109887098B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9912718B1 (en) * | 2011-04-11 | 2018-03-06 | Viasat, Inc. | Progressive prefetching |
| CN103999434A (en) * | 2011-12-21 | 2014-08-20 | 瑞典爱立信有限公司 | Devices and methods using network load data in mobile cloud accelerator context to optimize network usage by selectively deferring content delivery |
| CN107222468A (en) * | 2017-05-22 | 2017-09-29 | 北京邮电大学 | Augmented reality processing method, terminal, cloud server and edge server |
| CN108847981A (en) * | 2018-06-26 | 2018-11-20 | 咸宁职业技术学院 | Distributed computer cloud computing processing method |
Non-Patent Citations (1)
| Title |
|---|
| 移动云计算――移动增强现实技术和服务;吕强等;《中兴通讯技术》(第02期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109887098A (en) | 2019-06-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109887098B (en) | Web AR data presentation mode based on distributed computing | |
| US10447772B2 (en) | Managed function execution for processing data streams in real time | |
| CN109327509B (en) | Low-coupling distributed streaming computing system of master/slave architecture | |
| Zhang et al. | Toward transcoding as a service: energy-efficient offloading policy for green mobile cloud | |
| CN111368520A (en) | Editing method and device for electronic forms | |
| CN103092626B (en) | Based on the rich internet application process of Flex and web3D technology | |
| CN113902866B (en) | Double-engine driven digital twin system | |
| US9747382B1 (en) | Measuring page value | |
| CN113569117B (en) | Cloud platform system and method based on industrial internet big data service | |
| CN102857565A (en) | Intelligent clothes trying-on system based on cloud computing | |
| Huang et al. | DeepAdapter: A collaborative deep learning framework for the mobile web using context-aware network pruning | |
| CN112291728A (en) | Private industry application platform implementation architecture based on 5G network | |
| CN112464123A (en) | Water quality monitoring data visualization system and method based on micro-service | |
| WO2024021476A1 (en) | Data processing method and apparatus, electronic device and storage medium | |
| CN115550335A (en) | Distributed mobile WEB3D rendering system and method | |
| CN111680799A (en) | Method and apparatus for processing model parameters | |
| CN109617960B (en) | Attribution separation-based web AR data presentation method | |
| WO2022104611A1 (en) | Data distribution system and data distribution method | |
| CN114513512B (en) | Interface rendering method and device | |
| CN115186210A (en) | Web3D rendering and loading optimization method based on multiple granularities | |
| CN115866331A (en) | Video frame extraction analysis method, device, equipment and storage medium | |
| CN113448649A (en) | Redis-based home page data loading server and method | |
| CN111464640A (en) | Dynamic interaction protocol of virtual network equipment | |
| CN111046228B (en) | Video processing method based on stream calculation | |
| CN115361032B (en) | Antenna unit for 5G communication |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |