An edge computing video analysis algorithm access method

Abstract

The application provides an edge computing visual analysis algorithm access method, comprising the following steps: collecting video data of a monitoring target through a preset video collection node; identifying a delay parameter of the video data of the monitoring target through the video data; uploading the video data and the delay parameter to a preset edge computing server to construct an edge computing model; collecting configuration data of the video collection node connected to the edge computing model and dynamically configuring the delay parameter through the configuration data; wherein the configuration data at least comprises accuracy, resource consumption and time delay; and connecting the edge computing device corresponding to the video collection node after dynamic configuration to the edge computing model.

Description

Technical Field

[0001] This invention relates to the field of edge computing technology, and in particular to a method for accessing edge computing video analysis algorithms. Background Technology

[0002] Currently, the access efficiency of edge computing devices is constrained by the engineering efficiency of edge computing hardware providers, and the process standardization is low. Access needs to be handled individually by edge computing hardware providers based on the specific details of the algorithm SDK. The access process for both edge computing device providers and algorithm providers is cumbersome, involving streaming, stream retrieval and analysis, and the degree of freedom and efficiency of the algorithms are relatively low. Summary of the Invention

[0003] This invention provides an edge computing video analysis algorithm access method to address the issues mentioned in the background technology.

[0004] This technical solution provides a method for accessing edge computing video analytics algorithms, including:

[0005] Video data of the monitored targets is collected through preset video acquisition nodes;

[0006] The video data is used to identify the latency parameters of the video data of the monitored target;

[0007] The video data and latency parameters are uploaded to a preset edge computing server to construct an edge computing model;

[0008] The configuration data of the video acquisition nodes connected to the edge computing model is collected, and the latency parameter is dynamically configured using the configuration data; wherein,

[0009] The configuration data includes at least accuracy, resource consumption, and latency;

[0010] Connect the edge computing devices corresponding to the dynamically configured video acquisition nodes to the edge computing model.

[0011] As one embodiment of this technical solution, the step of uploading the video data and latency parameters to a preset edge computing server to construct an edge computing model includes:

[0012] By using preset variable indicators, multiple video data variables contained in the video data are filtered to determine the first input variable group;

[0013] The delay parameters are compared with the predetermined delay conditions, and a second group of input variables that meet the predetermined conditions is selected.

[0014] The correspondence between the first input variable group and the second input variable group is extracted, and an edge computing model is constructed based on the correspondence.

[0015] As one embodiment of this technical solution, the configuration data of the video acquisition nodes connected to the edge computing model is used to dynamically configure the latency parameters, including:

[0016] Based on the video acquisition nodes connected to the edge computing model, the user's request information is collected; wherein...

[0017] The activity information includes at least the request time, request content, and configured video data;

[0018] Based on the request information, determine whether to dynamically configure the delay parameter, and if the determination result is to dynamically configure, calculate the configuration data of the video acquisition node;

[0019] The delay parameters are dynamically configured using the configuration data.

[0020] As one embodiment of this technical solution, the edge computing device corresponding to the dynamically configured video acquisition node is connected to the edge computing model using a unified access method, which includes:

[0021] It uses a unified standard interface to receive YUV video streams from video acquisition nodes;

[0022] The YUV video stream is transmitted to a preset program for format processing, generating JSON analysis results; wherein,

[0023] YUV is the color encoding method used in video streams;

[0024] JSON is a lightweight data exchange format;

[0025] The JSON analysis results are formatted and output to the corresponding file.

[0026] This technical solution provides an edge computing video analytics algorithm access system, including:

[0027] The video data module is used to collect video data of the monitored target through preset video acquisition nodes;

[0028] The delay parameter module is used to identify the delay parameters of the video data of the monitored target based on the video data.

[0029] The edge computing model construction module is used to upload the video data and latency parameters to a preset edge computing server to construct an edge computing model;

[0030] A dynamic configuration module is used to collect configuration data from video acquisition nodes connected to the edge computing model, and to dynamically configure the latency parameters using this configuration data; wherein,

[0031] The configuration data includes at least accuracy, resource consumption, and latency;

[0032] The access module is used to connect the edge computing devices corresponding to the dynamically configured video acquisition nodes to the edge computing model.

[0033] As one embodiment of this technical solution, the step of uploading the video data and latency parameters to a preset edge computing server to construct an edge computing model includes:

[0034] The first input variable group unit is used to filter multiple video data variables contained in the video data through preset variable indicators to determine the first input variable group;

[0035] The second input variable group unit is used to compare the delay parameter with the predetermined parameter delay condition and filter the second input variable group that meets the predetermined condition.

[0036] An edge computing model unit is used to extract the correspondence between the first input variable group and the second input variable group, and to construct an edge computing model through the correspondence.

[0037] As one embodiment of this technical solution, the configuration data of the video acquisition nodes connected to the edge computing model is used to dynamically configure the latency parameters, including:

[0038] The user request information collection unit is used to collect user request information based on the video acquisition nodes connected to the edge computing model; wherein...

[0039] The activity information includes at least the request time, request content, and configured video data;

[0040] The configuration data calculation unit is used to determine whether to dynamically configure the delay parameter based on the request information, and to calculate the configuration data of the video acquisition node when the determination result is to perform dynamic configuration.

[0041] The dynamic configuration unit is used to dynamically configure the delay parameters using the configuration data.

[0042] As one embodiment of this technical solution, the edge computing device corresponding to the dynamically configured video acquisition node is connected to the edge computing model using a unified access method, which includes:

[0043] The receiving unit is used to receive YUV video streams from video acquisition nodes using a unified standard interface;

[0044] The format processing unit is used to transmit the YUV video stream to a preset program for format processing and generate JSON analysis results; wherein,

[0045] YUV is the color encoding method used in video streams;

[0046] JSON is a lightweight data exchange format;

[0047] The output unit is used to format the JSON analysis results and output the corresponding file.

[0048] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0049] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0050] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0051] Figure 1 This is a flowchart of an edge computing video analysis algorithm access method in an embodiment of the present invention. Detailed Implementation

[0052] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0053] Example 1:

[0054] This invention provides a method for accessing edge computing video analytics algorithms, including:

[0055] Video data of the monitored targets is collected through preset video acquisition nodes;

[0056] The video data is used to identify the latency parameters of the video data of the monitored target;

[0057] The video data and latency parameters are uploaded to a preset edge computing server to construct an edge computing model;

[0058] The configuration data of the video acquisition nodes connected to the edge computing model is collected, and the latency parameter is dynamically configured using the configuration data; wherein,

[0059] The configuration data includes at least accuracy, resource consumption, and latency;

[0060] Connect the edge computing devices corresponding to the dynamically configured video acquisition nodes to the edge computing model.

[0061] The working principle and beneficial effects of the above technical solution are as follows:

[0062] This invention provides a method for accessing edge computing video analytics algorithms, including:

[0063] Video data of the monitored target is collected through preset video acquisition nodes; the latency parameters of the video data of the monitored target are identified through the video data; the video data and latency parameters are uploaded to a preset edge computing server to construct an edge computing model; the configuration data of the video acquisition nodes connected to the edge computing model is collected, and the latency parameters are dynamically configured through the configuration data; wherein, the configuration data includes at least accuracy, resource consumption and latency; the edge computing device corresponding to the dynamically configured video acquisition node is connected to the edge computing model.

[0064] Example 2:

[0065] In one embodiment, uploading the video data and latency parameters to a preset edge computing server to construct an edge computing model includes:

[0066] By using preset variable indicators, multiple video data variables contained in the video data are filtered to determine the first input variable group;

[0067] The delay parameters are compared with the predetermined delay conditions, and a second group of input variables that meet the predetermined conditions is selected.

[0068] The correspondence between the first input variable group and the second input variable group is extracted, and an edge computing model is constructed based on the correspondence.

[0069] The working principle and beneficial effects of the above technical solution are as follows:

[0070] This technical solution uploads the video data and latency parameters to a preset edge computing server to construct an edge computing model, including: filtering multiple video data variables contained in the video data through preset variable indicators to determine a first input variable group; comparing the latency parameters with predetermined latency conditions and filtering a second input variable group that meets the predetermined conditions; extracting the correspondence between the first input variable group and the second input variable group, and constructing an edge computing model through the correspondence.

[0071] Example 3:

[0072] In one embodiment, the acquisition of configuration data for the video acquisition nodes connected to the edge computing model, and the dynamic configuration of the latency parameter using the configuration data, include:

[0073] Based on the video acquisition nodes connected to the edge computing model, the user's request information is collected; wherein...

[0074] The activity information includes at least the request time, request content, and configured video data;

[0075] Based on the request information, determine whether to dynamically configure the delay parameter, and if the determination result is to dynamically configure, calculate the configuration data of the video acquisition node;

[0076] The delay parameters are dynamically configured using the configuration data.

[0077] The working principle and beneficial effects of the above technical solution are as follows:

[0078] This technical solution collects configuration data from video capture nodes connected to the edge computing model, and dynamically configures the latency parameter using the configuration data. This includes: collecting user request information based on the video capture nodes connected to the edge computing model; wherein the activity information includes at least request time, request content, and configured video data; determining whether to dynamically configure the latency parameter based on the request information, and calculating the configuration data of the video capture nodes when the determination result is yes; and dynamically configuring the latency parameter using the configuration data.

[0079] Example 4:

[0080] In one embodiment, the connection of the edge computing device corresponding to the dynamically configured video acquisition node to the edge computing model adopts a unified access method, which includes:

[0081] It uses a unified standard interface to receive YUV video streams from video acquisition nodes;

[0082] The YUV video stream is transmitted to a preset program for format processing, generating JSON analysis results; wherein,

[0083] YUV is the color encoding method used in video streams;

[0084] JSON is a lightweight data exchange format;

[0085] The JSON analysis results are formatted and output to the corresponding file.

[0086] The working principle and beneficial effects of the above technical solution are as follows:

[0087] This technical solution connects the edge computing devices corresponding to dynamically configured video acquisition nodes to the edge computing model using a unified access method. This access method includes: receiving YUV video streams from the video acquisition nodes using a unified standard interface; transmitting the YUV video streams to a preset program for format processing to generate JSON analysis results; wherein YUV is the color encoding method used in the video stream; JSON is a lightweight data exchange format; and formatting the JSON analysis results and outputting them to a corresponding file.

[0088] Example 5:

[0089] This technical solution provides an edge computing video analytics algorithm access system, including:

[0090] The video data module is used to collect video data of the monitored target through preset video acquisition nodes;

[0091] The delay parameter module is used to identify the delay parameters of the video data of the monitored target based on the video data.

[0092] The edge computing model construction module is used to upload the video data and latency parameters to a preset edge computing server to construct an edge computing model;

[0093] A dynamic configuration module is used to collect configuration data from video acquisition nodes connected to the edge computing model, and to dynamically configure the latency parameters using this configuration data; wherein,

[0094] The configuration data includes at least accuracy, resource consumption, and latency;

[0095] The access module is used to connect the edge computing devices corresponding to the dynamically configured video acquisition nodes to the edge computing model.

[0096] The working principle and beneficial effects of the above technical solution are as follows:

[0097] This technical solution provides an edge computing video analysis algorithm access system, comprising: a video data module for collecting video data of a monitored target through preset video acquisition nodes; a latency parameter module for identifying latency parameters of the monitored target's video data using the video data; an edge computing model construction module for uploading the video data and latency parameters to a preset edge computing server to construct an edge computing model; a dynamic configuration module for collecting configuration data of the video acquisition nodes connected to the edge computing model and dynamically configuring the latency parameters using the configuration data; wherein the configuration data includes at least accuracy, resource consumption, and latency; and an access module for connecting the edge computing devices corresponding to the dynamically configured video acquisition nodes to the edge computing model.

[0098] Example 6:

[0099] In one embodiment, uploading the video data and latency parameters to a preset edge computing server to construct an edge computing model includes:

[0100] The first input variable group unit is used to filter multiple video data variables contained in the video data through preset variable indicators to determine the first input variable group;

[0101] The second input variable group unit is used to compare the delay parameter with the predetermined parameter delay condition and filter the second input variable group that meets the predetermined condition.

[0102] An edge computing model unit is used to extract the correspondence between the first input variable group and the second input variable group, and to construct an edge computing model through the correspondence.

[0103] The working principle and beneficial effects of the above technical solution are as follows:

[0104] This technical solution uploads the video data and latency parameters to a preset edge computing server to construct an edge computing model, including: a first input variable group unit, used to filter multiple video data variables contained in the video data through preset variable indicators to determine a first input variable group; a second input variable group unit, used to compare the latency parameters with predetermined latency conditions and filter a second input variable group that meets the predetermined conditions; and an edge computing model unit, used to extract the correspondence between the first input variable group and the second input variable group, and construct an edge computing model through the correspondence.

[0105] Example 7:

[0106] In one embodiment, the acquisition of configuration data for the video acquisition nodes connected to the edge computing model, and the dynamic configuration of the latency parameter using the configuration data, include:

[0107] The user request information collection unit is used to collect user request information based on the video acquisition nodes connected to the edge computing model; wherein...

[0108] The activity information includes at least the request time, request content, and configured video data;

[0109] The configuration data calculation unit is used to determine whether to dynamically configure the delay parameter based on the request information, and to calculate the configuration data of the video acquisition node when the determination result is to perform dynamic configuration.

[0110] The dynamic configuration unit is used to dynamically configure the delay parameters using the configuration data.

[0111] The working principle and beneficial effects of the above technical solution are as follows:

[0112] This technical solution collects configuration data of video capture nodes connected to the edge computing model, and dynamically configures the latency parameters using the configuration data. It includes a user request information collection unit, used to collect user request information based on the video capture nodes connected to the edge computing model; wherein the activity information includes at least request time, request content, and configuration video data.

[0113] A configuration data calculation unit is used to determine whether to dynamically configure the delay parameter based on the request information, and to calculate the configuration data of the video acquisition node when the determination result is to perform dynamic configuration; a dynamic configuration unit is used to dynamically configure the delay parameter based on the configuration data.

[0114] Example 8:

[0115] In one embodiment, the connection of the edge computing device corresponding to the dynamically configured video acquisition node to the edge computing model adopts a unified access method, which includes:

[0116] The receiving unit is used to receive YUV video streams from video acquisition nodes using a unified standard interface;

[0117] The format processing unit is used to transmit the YUV video stream to a preset program for format processing and generate JSON analysis results; wherein,

[0118] YUV is the color encoding method used in video streams;

[0119] JSON is a lightweight data exchange format;

[0120] The output unit is used to format the JSON analysis results and output the corresponding file.

[0121] The working principle and beneficial effects of the above technical solution are as follows:

[0122] This technical solution connects the dynamically configured video acquisition nodes to the edge computing devices in the edge computing model using a unified access method. This access method includes: a receiving unit for receiving YUV video streams from the video acquisition nodes using a unified standard interface; a format processing unit for transmitting the YUV video streams to a preset program for format processing, generating JSON analysis results; wherein YUV is the color encoding method used in the video stream; and JSON is a lightweight data exchange format; and an output unit for formatting the JSON analysis results and outputting the corresponding file.

[0123] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for accessing edge computing video analysis algorithms, characterized in that, include: Video data of the monitored targets is collected through preset video acquisition nodes; The video data is used to identify the latency parameters of the video data of the monitored target; The video data and latency parameters are uploaded to a preset edge computing server to construct an edge computing model; The configuration data of the video acquisition nodes connected to the edge computing model is collected, and the latency parameter is dynamically configured using the configuration data; wherein, The configuration data includes at least accuracy, resource consumption, and latency; Connect the edge computing devices corresponding to the dynamically configured video acquisition nodes to the edge computing model; The step of uploading the video data and latency parameters to a preset edge computing server to construct an edge computing model includes: By using preset variable indicators, multiple video data variables contained in the video data are filtered to determine the first input variable group; The delay parameters are compared with the predetermined delay conditions, and a second group of input variables that meet the predetermined conditions is selected. Extract the correspondence between the first input variable group and the second input variable group, and construct an edge computing model based on the correspondence. The method of connecting the edge computing devices corresponding to the dynamically configured video acquisition nodes to the edge computing model adopts a unified access method, which includes: It uses a unified standard interface to receive YUV video streams from video acquisition nodes; The YUV video stream is transmitted to a preset program for format processing, generating JSON analysis results; wherein, YUV is the color encoding method used in video streams; JSON is a lightweight data exchange format; The JSON analysis results are formatted and output to the corresponding file.

2. The edge computing video analysis algorithm access method according to claim 1, characterized in that, The configuration data of the video acquisition nodes connected to the edge computing model is used to dynamically configure the latency parameters, including: Based on the video acquisition nodes connected to the edge computing model, the user's request information is collected; wherein... The request information includes at least the request time, request content, and configuration video data; Based on the request information, determine whether to dynamically configure the delay parameter, and if the determination result is to dynamically configure, calculate the configuration data of the video acquisition node; The delay parameters are dynamically configured using the configuration data.

3. An edge computing video analytics algorithm access system, characterized in that, include: The video data module is used to collect video data of the monitored target through preset video acquisition nodes; The delay parameter module is used to identify the delay parameters of the video data of the monitored target based on the video data. The edge computing model construction module is used to upload the video data and latency parameters to a preset edge computing server to construct an edge computing model; A dynamic configuration module is used to collect configuration data from video acquisition nodes connected to the edge computing model, and to dynamically configure the latency parameters using this configuration data; wherein, The configuration data includes at least accuracy, resource consumption, and latency; The access module is used to connect the edge computing devices corresponding to the dynamically configured video acquisition nodes to the edge computing model. The step of uploading the video data and latency parameters to a preset edge computing server to construct an edge computing model includes: The first input variable group unit is used to filter multiple video data variables contained in the video data through preset variable indicators to determine the first input variable group; The second input variable group unit is used to compare the delay parameter with the predetermined parameter delay condition and filter the second input variable group that meets the predetermined condition. An edge computing model unit is used to extract the correspondence between the first input variable group and the second input variable group, and to construct an edge computing model through the correspondence. The method of connecting the edge computing devices corresponding to the dynamically configured video acquisition nodes to the edge computing model adopts a unified access method, which includes: The receiving unit is used to receive YUV video streams from video acquisition nodes using a unified standard interface; The format processing unit is used to transmit the YUV video stream to a preset program for format processing and generate JSON analysis results; wherein, YUV is the color encoding method used in video streams; JSON is a lightweight data exchange format; The output unit is used to format the JSON analysis results and output the corresponding file.

4. The edge computing video analysis algorithm access system according to claim 3, characterized in that, The configuration data of the video acquisition nodes connected to the edge computing model is used to dynamically configure the latency parameters, including: The user request information collection unit is used to collect user request information based on the video acquisition nodes connected to the edge computing model; wherein... The request information includes at least the request time, request content, and configuration video data; The configuration data calculation unit is used to determine whether to dynamically configure the delay parameter based on the request information, and to calculate the configuration data of the video acquisition node when the determination result is to perform dynamic configuration. The dynamic configuration unit is used to dynamically configure the delay parameters using the configuration data.

Images