CN110087041A - Video data processing and transmission method and system based on the base station 5G - Google Patents
Video data processing and transmission method and system based on the base station 5G Download PDFInfo
- Publication number
- CN110087041A CN110087041A CN201910364933.1A CN201910364933A CN110087041A CN 110087041 A CN110087041 A CN 110087041A CN 201910364933 A CN201910364933 A CN 201910364933A CN 110087041 A CN110087041 A CN 110087041A
- Authority
- CN
- China
- Prior art keywords
- video
- data
- video data
- transmission
- sensor
- 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.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/42—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Abstract
The video data processing and transmission method that the present invention relates to a kind of based on the base station 5G, comprising: choose acquisition intersection, the video sensor in the acquisition set is accessed into the base station 5G as mist node；The video data of sensor acquisition is obtained by the mist node, and boil down to transmits data；The transmission data are directly transmitted or other mist node-node transmissions by communicating to connect with the mist node are to cloud data center.
Description
Technical field
The invention belongs to internet of things field, and in particular to a kind of using the video data processing of mist computing technique and biography
Transmission method and system.
Background technique
In recent years, with the rapid development of the technologies such as Internet of Things, deep learning, artificial intelligence, intelligent traffic monitoring technology
Huge practical application value is shown in different social sectors.In order to preferably safeguard the traffic order in city
Sequence, traffic department are mounted with many monitoring devices on urban road, and have accessed urban traffic control center.Pass through cloud meter
Calculate, the integrated application of big data and artificial intelligence, at present at home some cities may be implemented to a certain extent from
Dynamicization traffic administration.Such as " city ET brain " platform of Alibaba is achieved that and carries out intelligent traffic management in Hangzhou.
The core of intelligent traffic monitoring technology is to be taken using the relevant technologies of computer vision to monitoring camera
Video is handled and is analyzed, and extracts the objects such as vehicle, pedestrian, road present in video, and on this basis to object
The behavior of body is judged, realizes automatic discrimination break in traffic rules and regulations and automation traffic administration.However a large amount of camera passes through
Network insertion urban traffic control center, so that existing cloud computing and web-transporting device face huge pressure.To the north of
For the city of capital, the whole city possesses traffic monitoring camera 300,000 altogether, and the data volume generated daily is up to 30PB, and 80% or more
Data be real time data.The existing technology based on cloud computing needs these data all uploading to cloud data center, it
Concentrate the processing and analysis for carrying out video again afterwards, this makes the network of cloud data center bear sizable pressure.In order to solve
This problem, Alibaba Co is using the method for distributed cloud computing, and distribution is established in different sections respectively in city
Cloud node accelerate the response speed of calculating to reduce data transfer delay.However, such method needs to change now
The topological structure of some Traffic Surveillance Video transmission networks, and distributed cloud node there is construction and maintenance cost is higher asks
Topic.
With the development of communication technology, the mode of wireless communication is also developed from 4G communication to 5G communication.4G to 5G's turns
Become, other than being substantially improved of message transmission rate, the base station the 5G base station ratio 4G possesses stronger storage and computing capability, can be with
Handle the task of a part of video analysis and processing.In the environment of 5G communication network, the framework of current cloud computing will be to mist
The direction of calculating changes, so that the storage of the base station 5G is preferably utilized with computing capability.Not with distributed cloud computing
It is same, transmission of video is carried out in 5G wireless network using mist calculating and is not necessarily to change the topological structure of network, without additional
Construction and maintenance cost.However, how the transmission for carrying out urban transportation monitor video, at present industry are calculated using mist in 5G network
Boundary there is no concrete scheme.
Summary of the invention
For in the prior art, Traffic Surveillance Video transmits the low problem of real-time, and the present invention passes through in the middle part of the base station 5G
It affixes one's name to mist and calculates transmission of video framework, to meet the needs of traffic monitoring real time data processing and analysis.
Specifically, video data of the invention processing and transmission method include: the selecting video sensor access base station 5G,
Using the base station 5G as mist node；It is screened by the video data that the mist node acquires the video sensor, and will screening
Video data compression out is transmission data；The transmission data are directly transmitted or route transmission is to cloud data center.
Video data processing of the present invention and transmission method, wherein the process of selecting video sensor specifically includes:
According to video sensor CjTo mist node BiThe delay D of transmitting video datai,jWith code rate Ri,j, obtaining indicates video sensor Cj
Connect mist node BiServiceability Index Ui,j；Wherein,Umax=ni∑j∈MUi,j, ni=| M |min,|M|min+
1,...,|M|max, niExpression is currently accessed mist node BiVideo sensor quantity, | M |maxWith | M |minRespectively mist node
BiIt can access the maximum capacity and minimum capacity of video sensor；With Serviceability Index Ui,jObtain video sensor CjPlace sensor
The gross effect U of set C, to have maximum total utility UmaxSet of sensors be acquisition set C'；It will be in acquisition set C'
Video sensor access the mist node.
Video data processing of the present invention and transmission method, wherein the process for compressing the video data specifically includes:
Target Segmentation is carried out to video data V by Video Segmentation, to obtain the video data V of quasi- compressiontrans={ Vi|Fi=
1 }, wherein ViVideo-data fragment when feature object T, F are detected for i-thiIt is special for the video after video data V Target Segmentation
Sign；By video data VtransIt is divided into M × N number of grid, with the grid to video data VtransDown-sampling is carried out, compression is formed
Data；Wherein the height of grid and width are respectivelyWithMeet minEARAPWhenAndEARAPFor the loss function of video data compression,
Ω is about video data VtransW × H dimensional feature matrix, and if only if VijThe element Ω of Ω when ∈ Tij=1, whenWhen T
Ωij=0, W × H is the resolution ratio of video data V, and W' × H' is the compressed resolution ratio of video data V, M, N, W, H, W', H'
For positive integer；With the grid to video data VtransDown-sampling is carried out, compressed data is formed, it willWithBe added compressed data with
Form the transmission data.
Video data processing of the present invention and transmission method, wherein the process for transmitting the transmission data includes: by this
Transmission data are divided into multiple data blocks；Each data block is directly transferred to cloud data center, or passes through at least one mist node
Transfer is with route transmission to cloud data center；Wherein, according between mist node transmission bandwidth and delay, choose the mist section of transfer
Point.
The present invention also proposes a kind of video data processing and Transmission system based on the base station 5G, comprising: sensor accesses mould
Block makees the base station 5G for the access of selecting video sensor, using the base station 5G as mist node；Data processing module, for by being somebody's turn to do
The video data that mist node acquires the video sensor screens, and is transmission number by the video data compression filtered out
According to；Data transmission module, for directly transmitting the transmission data or route transmission is to cloud data center.
Video data processing of the present invention and Transmission system, wherein the sensor AM access module includes: sensor effect
With module is obtained, for according to video sensor CjTo mist node BiThe delay D of transmitting video datai,jWith code rate Ri,j, obtain table
Show video sensor CjConnect mist node BiServiceability Index Ui,j；Wherein,Umax=ni∑j∈MUi,j, ni=|
M|min,|M|min+1,...,|M|max, niExpression is currently accessed the base station 5G BiVideo sensor quantity, | M |maxWith | M |minPoint
It Wei not mist node BiIt can access the maximum capacity and minimum capacity of video sensor；Sensor intersection chooses module, for effect
With index Ui,jObtain video sensor CjThe gross effect U of place set of sensors C, to have maximum total utility UmaxSensor
Collection is combined into acquisition set C'；Video sensor in acquisition set C' is accessed into the mist node.
Video data processing of the present invention and Transmission system, wherein the data processing module includes: video data choosing
Modulus block, for carrying out Target Segmentation to video data V by Video Segmentation, to obtain the video data V of quasi- compressiontrans
={ Vi|Fi=1 }, wherein ViVideo-data fragment when feature object T, F are detected for i-thiFor video data V Target Segmentation
Video features afterwards；Compressed data generation module is used for video data VtransIt is divided into M × N number of grid, with the grid to view
Frequency is according to VtransDown-sampling is carried out, compressed data is formed；Wherein the height of grid and width are respectivelyWithMeet
minEARAPWhenAndEARAPFor the loss function of video data compression,Ω is about video data VtransW × H dimensional feature matrix, when
And if only if VijThe element Ω of Ω when ∈ Tij=1, whenΩ when Tij=0, W × H is the resolution ratio of video data V, and W' × H' is
The compressed resolution ratio of video data V, M, N, W, H, W', H' are positive integer；Data generation module is transmitted, for the grid pair
Video data VtransDown-sampling is carried out, compressed data is formed, it willWithCompressed data is added to form the transmission data.
Video data processing of the present invention and Transmission system, wherein the data transmission module includes: transmission data point
Block module, for the transmission data to be divided into multiple data blocks；Block data transmission module, for directly passing each data block
The cloud data center is transported to, or by least one mist node transfer with route transmission to the cloud data center；Wherein, according to mist
Transmission bandwidth and delay between node choose the mist node of transfer.
The present invention also proposes a kind of readable storage medium storing program for executing, is stored with executable instruction, and the executable instruction is for executing such as
Video data processing and transmission method above-mentioned based on the base station 5G.
The present invention also proposes that a kind of mist calculates video data acquiring system, comprising: video sensor, for acquiring video counts
According to；Mist node, the mist node are the base station 5G, including processor and readable storage medium storing program for executing as claimed in claim 9, wherein
The processor calls and executes the executable instruction in the readable storage medium storing program for executing, to obtain the video data, to the video data
Rear boil down to transmission data are analyzed and processed, and the transmission data are sent to cloud data center；Cloud data center, for connecing
The transmission data are received, and are decompressed to obtain the video data.
Detailed description of the invention
Fig. 1 is the base station 5G Traffic Surveillance Video transmission network architecture diagram of the invention.
Fig. 2 is the video compression algorithm process schematic of the invention based on video features.
Fig. 3 is 5G base station collaboration schematic diagram of transmission process of the invention.
Fig. 4 is the processing of the base station 5G video data and Transmission system structural schematic diagram of the invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with attached drawing, the present invention is mentioned
The video data processing calculated based on the base station 5G mist and transmission method out is further described.It should be appreciated that this place is retouched
The specific implementation method stated is only used to explain the present invention, is not intended to limit the present invention.
The purpose of the present invention is solve Traffic Surveillance Video by disposing mist calculating transmission of video framework in the base station 5G
Transmit the low problem of real-time.
To achieve the goals above, the present invention provides a kind of low latency video transmission method calculated based on mist, comprising:
Mist is disposed on the base station 5G and calculates framework, becomes mist node, and maintenance accesses the video sensor of the mist node
The information of (such as traffic camera of the embodiment of the present invention), receives and processes monitor video data, is last transmitted to cloud data
Center；
It includes sensor AM access module, data processing module and data transmission module that the mist of the base station 5G, which calculates framework,；
The characteristic information of sensor AM access module collection video sensor, the video code rate shot including video sensor,
The information such as resolution ratio, position access the video sensor equipment of the base station based on these information dynamic managements；
Data processing module is analyzed and is handled to the video data that video sensor transmits, including video motion is special
Sign, Content Feature Extraction, Video compression, it is therefore an objective to reduce the data volume for uploading to cloud data center, while be cloud data
Share a part of calculating task in center；
Data transmission module is responsible between the processing of the video data between base station and cloud data center and transmission and base station
Cooperation transmission, for guarantee video sensor obtain Traffic Surveillance Video transmission real-time highest.
At least one example of the invention provides a kind of mist calculating Traffic Surveillance Video Transmission system for the base station 5G.
By realizing traffic camera access-in management, video analysis processing and cooperation between base stations transmission on the base station 5G, so that largely
The video data of traffic camera can access cloud data center with lower delay, and then reduce the response of traffic monitoring task
Time.Example deployment of the invention does not need the additional network equipment, does not change traffic video and connect on the existing base station 5G
Enter the topological structure of network, practical application value with higher.
For in the prior art, Traffic Surveillance Video transmits the low problem of real-time, and the present invention passes through in the middle part of the base station 5G
It affixes one's name to mist and calculates transmission of video framework, to meet the needs of traffic monitoring real time data processing and analysis.
Specifically, mist of the invention calculating video transmission method includes:
1) acquisition intersection is chosen, the video sensor in acquisition set is accessed to the base station 5G as mist node；
2) video data of sensor acquisition is obtained by mist node, and boil down to transmits data；
3) transmission data are directly transmitted or other mist node-node transmissions by communicating to connect with mist node is into cloud data
The heart.
Wherein, the process for choosing acquisition set specifically includes: according to video sensor CjTo mist node BiTransmitting video data
Delay Di,jWith code rate Ri,j, obtaining indicates video sensor CjConnect mist node BiServiceability Index Ui,j；Wherein,ni=| M |min,|M|min+1,...,|M|max, niExpression is currently accessed mist node
BiVideo sensor quantity, | M |maxWith | M |minRespectively mist node BiIt can access the maximum capacity and most of video sensor
Low capacity；With Serviceability Index Ui,jObtain video sensor CjThe gross effect U of place set of sensors C, to have maximum total utility
UmaxSet of sensors be acquisition set C'；The video sensor in set C' will be acquired and access mist node.
The process for compressing the video data specifically includes: Target Segmentation is carried out to video data V by Video Segmentation,
To obtain the video data V of quasi- compressiontrans={ Vi|Fi=1 }, wherein ViVideo counts when feature object T are detected for i-th
According to section, FiFor the video features after video data V Target Segmentation；By video data VtransIt is divided into M × N number of grid, with the net
Lattice are to video data VtransDown-sampling is carried out, compressed data is formed；Wherein the height of grid and width are respectivelyWithIt is full
Sufficient minEARAPWhenAndEARAPFor the loss function of video data compression,Ω is about video data VtransW × H dimensional feature matrix, when
And if only if VijThe element Ω of Ω when ∈ Tij=1, whenΩ when Tij=0, W × H is the resolution ratio of video data V, and W' × H' is
The compressed resolution ratio of video data V, M, N, W, H, W', H' are positive integer；With the grid to video data VtransAdopt
Sample forms compressed data, willWithCompressed data is added to form transmission data.
The process for transmitting the transmission data includes: that transmission data are divided into multiple data blocks；Each data block is directly passed
It is defeated, or by one/multiple mist node transfers with route transmission to cloud data center；Wherein, according to the transmission between mist node
The mist node of transfer is chosen in bandwidth and delay.
The present invention also proposes a kind of video data processing and Transmission system based on the base station 5G, comprising:
The base station 5G is made in sensor AM access module, the access of selecting video sensor, and using the base station 5G as video data
The mist node of processing and Transmission system；
Data processing module is screened by the video data that mist node acquires video sensor, and will be filtered out
Video data compression be transmission data；
Data transmission module directly transmits transmission data or other mist node roads by communicating to connect with the mist node
By being transmitted to cloud data center.
Wherein sensor AM access module includes:
Sensor effectiveness obtains module, for according to video sensor CjTo mist node BiThe delay D of transmitting video datai,j
With code rate Ri,j, obtaining indicates video sensor CjConnect mist node BiServiceability Index Ui,j；Wherein,ni=| M |min,|M|min+1,...,|M|max, niExpression is currently accessed the base station 5G
BiVideo sensor quantity, | M |maxWith | M |minRespectively mist node BiIt can access the maximum capacity and most of video sensor
Low capacity；
Sensor intersection chooses module, for Serviceability Index Ui,jObtain video sensor CjPlace set of sensors C's
Gross effect U, to have maximum total utility UmaxSet of sensors be acquisition set C'；By the video in acquisition set C'
Sensor accesses the mist node.
Data processing module specifically includes:
Video data chooses module, quasi- to obtain for carrying out Target Segmentation to video data V by Video Segmentation
The video data V of compressiontrans={ Vi|Fi=1 }, wherein ViVideo-data fragment when feature object T, F are detected for i-thiFor
Video features after video data V Target Segmentation；
Compressed data generation module is used for video data VtransIt is divided into M × N number of grid, with the grid to video counts
According to VtransDown-sampling is carried out, compressed data is formed；Wherein the height of grid and width are respectivelyWithMeet minEARAPWhenAndEARAPFor the loss function of video data compression,Ω is about video data VtransW × H dimensional feature matrix, when
And if only if VijThe element Ω of Ω when ∈ Tij=1, whenΩ when Tij=0, W × H is the resolution ratio of video data V, and W' × H' is
The compressed resolution ratio of video data V, M, N, W, H, W', H' are positive integer；
Data generation module is transmitted, is used for the grid to video data VtransDown-sampling is carried out, compressed data is formed, it willWithCompressed data is added to form transmission data.
Data transmission module includes:
Deblocking module is transmitted, is divided into multiple data blocks for the transmission data；
Block data transmission module for each data block to be directly transferred to cloud data center, or passes through at least one
The transfer of mist node is with route transmission to the cloud data center；Wherein, according between mist node transmission bandwidth and delay, in selection
The mist node turned.
The present invention also proposes a kind of readable storage medium storing program for executing, is stored with executable instruction, and the executable instruction is for holding this hair
The bright video data based on the base station 5G handles and transmission method.
The present invention also proposes that a kind of mist calculates video data acquiring system, comprising: video sensor, for acquiring video counts
According to；Mist node, mist node of the invention are the base station 5G, and the base station 5G includes processor and readable storage medium storing program for executing, wherein processor tune
With and execute the executable instruction in readable storage medium storing program for executing, to obtain video data, pressed after being analyzed and processed to video data
Transmission data are condensed to, and are sent to cloud data center；Cloud data center for receiving transmission data, and is decompressed to be regarded
Frequency evidence.
One is provided first before method of the invention is discussed in detail in order to facilitate the working method for understanding of the invention
The possible application scenarios of the present invention.By taking Beijing as an example, it is more than 300,000 that Beijing, which has disposed traffic monitoring camera, and institute
There is camera to access network by way of wired network, these cameras are predominantly located at urban road two sides and each crossing
Place.Pass through network after camera shooting video and upload to traffic control center and is stored.The video of each camera shooting
It has differences, these differences are mainly manifested in the following aspects:
(1) the shooting coverage area of camera is different.Due to the location of different camera difference, and for shooting
Act of violating regulations it is also different, there is also differences for the road conditions range of video that single camera can photograph.Such as the prison positioned at crossing
The range that control camera is often shot is larger, and the range that the unidirectional camera for being located at road edge is then shot is smaller.
(2) video parameter of camera shooting has differences.Since the camera within the scope of the whole city is disposed in batches,
It is had differences in the specification of video parameter, there are in code rate, resolution ratio and frame per second for the video of different camera shootings
Difference.
Other than above-mentioned difference, Traffic Surveillance Video also has the characteristics that amount of redundant information is big.Traffic Surveillance Video
It is largely street or road background in picture, and to monitoring violating the regulations, the effective traffic information of vehicle flow monitor task
Often only occupy a small amount of part.
Based on These characteristics, taken the photograph firstly the need of base station according to the selection of remaining bandwidth, calculating and storage resource in the present invention
As the total delay minimum that head is managed, and the camera video that realization accesses in the system transmits；Secondly in a base station to traffic
Monitor video is analyzed and is handled, and carries out feature extraction and compression to video；Finally in terms of the cooperation transmission between base station, often
The video that processing is completed after the completion of video processing in one base station uploads cloud, needs first to transmit the video to other when necessary
Base station, then it is transmitted to cloud, it is minimum come the delay for guaranteeing that video uploads with this.
To achieve the goals above, the present invention is firstly the need of the access-in management for realizing camera.Base station is firstly the need of measurement
The delay of thecamera head video and code rate.Consider the base station 5G Traffic Surveillance Video transmission network as shown in Figure 1, enables BiIt indicates
The base station 5G that number is i and as transmission network mist node, CjThe camera for being j for number, Ri,jFor camera CjTransmit video
To base station BiCode rate, Di,jFor camera CjVideo is transmitted to base station BiDelay.Di,jWith Ri,jIt can be surveyed in real time by base station
It measures.The available total bandwidth of base station is limited, if this upper limit is Rmax, therefore single base station can not manage it is all
Camera.In order to enable the selection to access camera of base station is more reasonable, following utility function is employed herein
Measure camera CjAccess base station BiEffectiveness Ui,j:
In above formula, niExpression is currently accessed base station BiCamera quantity.The access effect of camera defined in formula (1)
Base station can be construed to function and manages camera as much as possible, and shoots that video code rate is higher and transmission delay is lower
Camera can be preferred and manage.
After the utility function for defining camera access base station, base station can be according to the access effectiveness to camera
It carries out being selectively accessing management, which can be indicated with following equation:
In above-mentioned formula, set M indicates to number set by the camera that base station is chosen and is managed.Due to actual
In network, hardly occur for two different camera Ri,j/Di,jEqual situation, therefore above-mentioned optimization problem is one
A 01 knapsack optimization problem, the optimization problem are proved to be np complete problem, i.e., it is multiple for multinomial can not to find a kind of complexity
Miscellaneous algorithm solves, therefore provides a kind of didactic greedy algorithm here and carry out come the model defined to formula (2) (3)
It solves:
Step 1: calculation base station BiMaximum capacity | M |maxWith minimum capacity | M |min, calculation method is as follows:
Step 11: initialization | M |max=0, | M |min=0
Step 12: enabling C1=C, C2=C, from camera set C1With C2In select code rate maximum and the smallest camera Cmax
With Cmin
Step 13:| M |max=| M |max+ 1, | M |min=| M |min+ 1, and by CmaxWith CminFrom C1、C2In leave out
Step 14: step 11- step 13 is repeated, untilAnd
Step 15: at this time | M |maxWith | M |minValue be base station BiMaximum and minimum capacity
Step 2: taking ni=| M |min,|M|min+1,...,|M|max, execute the following steps:
Step 21: according to the U of each equipmenti,jValue priority that equipment is selected be ranked up
Step 22: the priority that selection is obtained according to previous step, according to the sequential selection n of priority from high to lowiIt is a
Equipment is as the cluster tool C' that base station is added
Step 23: judging whether the sum of code rate of all devices in C' is greater than the code rate limitation R of base stationmaxIf being no more than
RmaxStep 25 is then skipped to, step 24 is otherwise carried out
Step 24: enabling Δ R=∑j∈C'Ri,j-Rmax, the camera C that be selected into eachi,j∈ C',
Step 25: the base station total utility n after calculating selection camerai∑j∈MUi,j, and save the value and current choosing of total utility
The camera set C' selected
Step 26: step 21- step 25 is repeated, until ni=| M |max
Step 3: the cluster tool with maximum total utility is selected from the camera sequence of sets obtained in step 2,
Then the camera in the set is included in management by base station, and group network process is completed.
Algorithm above illustrates working process and principle of the base station as base station to camera progress access-in management, is only one
The special case of a group of network process, networking described in this patent is not limited to the description of above-mentioned algorithm, but is believed using related camera device
Breath, the technology that camera access-in management is carried out on the base station 5G belong to the coverage area of this patent.
The present invention needs to utilize base to achieve the purpose that the transmission of video data volume for reducing base station and traffic surveillance and control center
The computing resource stood to carry out video feature extraction and analysis, and then the view for transmitting needs according to the feature in monitor video
Frequency is compressed.Specific step is as follows for it, as shown in Fig. 2, the video for enabling V indicate that camera acquires, Shi Changwei tV, point of video
Resolution is W × H, carries out step for the video V that base station receives:
Step 1: using Video Segmentation to video carry out target segmentation, partitioning algorithm can be used with
MobilenetV2 is the DeeplabV3 of front end, for reducing the calculation amount of Video segmentation and accelerating to analyze speed；
Step 2: F being enabled to indicate by the video features after Video Segmentation, Fi=1, when i-th detection
There is vehicle in video, primary video segmentation is carried out every the t time, then [t available at this timeV/ t] a feature, while V also by
It is divided into [tV/ t] section.In this technique, the video-frequency band set V that object end is transmittedtransIt is determined by following formula:
Vtrans={ Vi|Fi=1 } (4)
Obviously, | Vtrans|≤| V |, therefore this technology can reduce the data volume that object end is visually transmitted.
Step 3: after carrying out step 2, the characteristic pattern for indicating kind of object included in video can be generated, according to friendship
The task of logical monitoring, can by video vehicle and the objects such as pedestrian be set as feature object, enable T expression feature object, Ω table
Show the eigenmatrix of video, Ω is the matrix of W × H dimension, and if only if Vij∈ T, Ωij=1, otherwise Ωij=0；
Step 4: according to eigenmatrix Ω, follow the steps below:
Step 41: video being divided into the grid of a M × N, is set according to the transmittability of object end equipment compressed
Video resolution is W' × H'；
Step 42: the loss function for establishing video compress is as follows
In the loss functionRespectively indicate the high col width with j-th of grid of row of i-th of grid；
Step 43: being solved to drag using serializing Novel Algorithm, obtain the size of each grid
minEARAP (6)
Step 44: down-sampling being carried out to original video according to the size of each grid, to reduce the size of video, and is protected
Precision is not lost in characteristics of syndrome region, so that the data volume of video be greatly reduced；
Step 45: by the size of above-mentioned video gridIt is transmitted together with compressed video, root when decoding
Original video can be restored by up-sampling according to sizing grid to video.
Algorithm above illustrate the monitor video that base station acquires camera as base station carry out the process of data compression with
And principle, this is belonged to using the algorithm that video feature extraction method carries out self-adapting compressing to video resolution in the base station 5G
The coverage area of patent.
The transmission of video of low latency is carried out to guarantee base station to cloud, also needs to carry out cooperation transmission between base station.It should
The principle of cooperation transmission is that the video for sending over camera carries out piecemeal, then again directly passes the video of each piecemeal
It is defeated by cloud or is transferred to cloud again by the transfer of one or more base stations, biography to reach increase handling capacity and is reduced with this
The effect of defeated delay.
The process of base station collaboration transmission as shown in figure 3, the video received can be segmented by base station, obtain again later with
Connected base station bandwidth and delay, select the node that bandwidth is high and delay is low to be transmitted.
The embodiment of the present invention also provides mist and calculates video data acquiring system.As shown in figure 4, mist of the invention calculates video
Data collection system includes at least one video sensor, at least one mist node, and the cloud data center in cloud, in
In the embodiment of the present invention, video sensor is traffic camera or other video capture devices, the present invention not as
Limit, video sensor acquires video data, and is sent to the mist node for communicating connection, and each video sensor only accesses one
A mist node, a mist node can connect one or more video sensors；Mist node includes processor and readable storage medium
Matter, readable storage medium storing program for executing are stored with meter executable instruction and realize above-mentioned base when executable instruction is executed by the processor of mist node
In the video data processing of the base station 5G and transmission method；Mist node direct communication is connected to cloud data center, or passes through other mists
Node switching is to cloud data center；Cloud data center first solves it after the transmission data for receiving the transmission of mist node
Compression, is then further processed again.Those of ordinary skill in the art will appreciate that all or part in the above method
Step can instruct related hardware (such as processor) to complete by program, and described program can store in readable storage medium storing program for executing
In, such as read-only memory, disk or CD.One or more collection also can be used in all or part of the steps of above-described embodiment
It is realized at circuit.Correspondingly, each module in above-described embodiment can take the form of hardware realization, such as pass through integrated electricity
Its corresponding function is realized on road, can also be realized in the form of software function module, such as is stored in by processor execution
Program/instruction in memory realizes its corresponding function.The embodiment of the present invention be not limited to any particular form hardware and
The combination of software.
Although the present invention has been disclosed by way of example above, it is not intended to limit the present invention., any technical field
In those of ordinary skill can make several modifications and improvements without departing from the spirit and scope of the present invention, therefore it is of the invention
Protection scope should be defined by the scope of the appended claims.
Claims (10)
1. a kind of video data processing and transmission method based on the base station 5G characterized by comprising
Selecting video sensor accesses the base station 5G, using the base station 5G as mist node；
The video data compression screened by the video data that the mist node acquires the video sensor, and will filtered out
To transmit data；
The transmission data are directly transmitted or route transmission is to cloud data center.
2. video data processing as described in claim 1 and transmission method, which is characterized in that the process of selecting video sensor
It specifically includes:
According to video sensor CjTo mist node BiThe delay D of transmitting video datai,jWith code rate Ri,j, obtaining indicates video sensing
Device CjConnect mist node BiServiceability Index Ui,j；Wherein,Umax=ni∑j∈MUi,j, ni=| M |min,|M|min+
1,...,|M|max, niExpression is currently accessed mist node BiVideo sensor quantity, | M |maxWith | M |minRespectively mist node
BiIt can access the maximum capacity and minimum capacity of video sensor；
With Serviceability Index Ui,jObtain video sensor CjThe gross effect U of place set of sensors C, to have maximum total utility Umax
Set of sensors be acquisition set C'；Video sensor in acquisition set C' is accessed into the mist node.
3. video data processing as described in claim 1 and transmission method, which is characterized in that compress the process of the video data
It specifically includes:
Target Segmentation is carried out to video data V by Video Segmentation, to obtain the video data V of quasi- compressiontrans={ Vi|Fi
=1 }, wherein ViVideo-data fragment when feature object T, F are detected for i-thiFor the video after video data V Target Segmentation
Feature；
By video data VtransIt is divided into M × N number of grid, with the grid to video data VtransDown-sampling is carried out, compression is formed
Data；Wherein the height of grid and width are respectivelyWithMeet minEARAPWhenAndEARAPFor the loss function of video data compression,Ω is about video data VtransW × H dimensional feature matrix, when
And if only if VijThe element Ω of Ω when ∈ Tij=1, whenWhen Ωij=0, W × H is the resolution ratio of video data V, W' × H'
For the compressed resolution ratio of video data V, M, N, W, H, W', H' are positive integer；
With the grid to video data VtransDown-sampling is carried out, compressed data is formed, it willWithCompressed data is added to be formed
The transmission data.
4. video data processing as described in claim 1 and transmission method, which is characterized in that transmit the process of the transmission data
Include:
The transmission data are divided into multiple data blocks；
Each data block is directly transferred to the cloud data center, or extremely should by least one mist node transfer with route transmission
Cloud data center；Wherein, according between mist node transmission bandwidth and delay, choose the mist node of transfer.
5. a kind of video data processing and Transmission system based on the base station 5G characterized by comprising
Sensor AM access module makees the base station 5G for the access of selecting video sensor, using the base station 5G as mist node；
Data processing module, the video data for being acquired by the mist node to the video sensor screen, and will sieve
The video data compression selected is transmission data；
Data transmission module, for directly transmitting the transmission data or route transmission is to cloud data center.
6. video data processing as claimed in claim 5 and Transmission system, which is characterized in that the sensor AM access module packet
It includes:
Sensor effectiveness obtains module, for according to video sensor CjTo mist node BiThe delay D of transmitting video datai,jAnd code
Rate Ri,j, obtaining indicates video sensor CjConnect mist node BiServiceability Index Ui,j；Wherein,Umax=ni
∑j∈MUi,j, ni=| M |min,|M|min+1,...,|M|max, niExpression is currently accessed the base station 5G BiVideo sensor quantity, |
M|maxWith | M |minRespectively mist node BiIt can access the maximum capacity and minimum capacity of video sensor；
Sensor intersection chooses module, for Serviceability Index Ui,jObtain video sensor CjTotal effect of place set of sensors C
U is answered, to have maximum total utility UmaxSet of sensors be acquisition set C'；By the video sensor in acquisition set C'
Access the mist node.
7. video data as claimed in claim 5 processing and Transmission system, which is characterized in that the data processing module includes:
Video data chooses module, for carrying out Target Segmentation to video data V by Video Segmentation, to obtain quasi- compression
Video data Vtrans={ Vi|Fi=1 }, wherein ViVideo-data fragment when feature object T, F are detected for i-thiFor video
Video features after data V Target Segmentation；
Compressed data generation module is used for video data VtransIt is divided into M × N number of grid, with the grid to video data
VtransDown-sampling is carried out, compressed data is formed；Wherein the height of grid and width are respectivelyWithMeet minEARAPWhenAndEARAPFor the loss function of video data compression,Ω is about video data VtransW × H dimensional feature matrix, when
And if only if VijThe element Ω of Ω when ∈ Tij=1, whenWhen Ωij=0, W × H is the resolution ratio of video data V, W' × H'
For the compressed resolution ratio of video data V, M, N, W, H, W', H' are positive integer；
Data generation module is transmitted, is used for the grid to video data VtransDown-sampling is carried out, compressed data is formed, it willWithCompressed data is added to form the transmission data.
8. video data as claimed in claim 5 processing and Transmission system, which is characterized in that the data transmission module includes:
Deblocking module is transmitted, for the transmission data to be divided into multiple data blocks；
Block data transmission module for each data block to be directly transferred to the cloud data center, or passes through at least one mist
Node transfer is with route transmission to the cloud data center；Wherein, according between mist node transmission bandwidth and delay, choose transfer
Mist node.
9. a kind of readable storage medium storing program for executing, is stored with executable instruction, which appoints for executing Claims 1 to 4 such as
Video data processing and transmission method described in one based on the base station 5G.
10. a kind of mist calculates video data acquiring system characterized by comprising
Video sensor is communicatively coupled to mist node, for acquiring video data；
Mist node, the mist node are the base station 5G, including processor and readable storage medium storing program for executing as claimed in claim 9, wherein
The processor calls and executes the executable instruction in the readable storage medium storing program for executing, to obtain the video data, to the video data
Rear boil down to transmission data are analyzed and processed, and the transmission data are sent to cloud data center；
Cloud data center for receiving the transmission data, and is decompressed to obtain the video data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910364933.1A CN110087041B (en) | 2019-04-30 | 2019-04-30 | Video data processing and transmitting method and system based on 5G base station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910364933.1A CN110087041B (en) | 2019-04-30 | 2019-04-30 | Video data processing and transmitting method and system based on 5G base station |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110087041A true CN110087041A (en) | 2019-08-02 |
| CN110087041B CN110087041B (en) | 2021-01-08 |
Family
ID=67418373
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910364933.1A Active CN110087041B (en) | 2019-04-30 | 2019-04-30 | Video data processing and transmitting method and system based on 5G base station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110087041B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111565303A (en) * | 2020-05-29 | 2020-08-21 | 深圳市易链信息技术有限公司 | Video monitoring method, system and readable storage medium based on fog calculation and deep learning |
| CN112398158A (en) * | 2020-10-27 | 2021-02-23 | 国网经济技术研究院有限公司 | Distributed collection device and method for operation indexes of hybrid high-voltage direct-current power transmission system |
| CN113203439A (en) * | 2021-05-07 | 2021-08-03 | 南京邮电大学 | Master-slave dynamic edge sensor ad hoc network system for water information detection |
| CN115150404A (en) * | 2022-06-09 | 2022-10-04 | 安徽天元通信发展有限公司 | 5G base station information analysis processing method and system based on integrated structure |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101316279A (en) * | 2008-07-09 | 2008-12-03 | 南京邮电大学 | Subjective interest driven wireless multimedia sensor network design method |
| CN103024400A (en) * | 2011-12-19 | 2013-04-03 | 北京捷成世纪科技股份有限公司 | Video compression fault-tolerant transmission method and system based on network |
| CN107731011A (en) * | 2017-10-27 | 2018-02-23 | 中国科学院深圳先进技术研究院 | A kind of harbour is moored a boat monitoring method, system and electronic equipment |
| CN107808122A (en) * | 2017-09-30 | 2018-03-16 | 中国科学院长春光学精密机械与物理研究所 | Method for tracking target and device |
| CN108377264A (en) * | 2018-02-05 | 2018-08-07 | 江苏大学 | Vehicular ad hoc network quorum-sensing system data report De-weight method |
| US20190044818A1 (en) * | 2018-01-12 | 2019-02-07 | Intel Corporation | Self-adjusting data processing system |
| CN109547541A (en) * | 2018-11-12 | 2019-03-29 | 安徽师范大学 | Mist calculates the node low overhead collaboration method under environment based on filtering and distribution mechanism |
-
2019
- 2019-04-30 CN CN201910364933.1A patent/CN110087041B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101316279A (en) * | 2008-07-09 | 2008-12-03 | 南京邮电大学 | Subjective interest driven wireless multimedia sensor network design method |
| CN103024400A (en) * | 2011-12-19 | 2013-04-03 | 北京捷成世纪科技股份有限公司 | Video compression fault-tolerant transmission method and system based on network |
| CN107808122A (en) * | 2017-09-30 | 2018-03-16 | 中国科学院长春光学精密机械与物理研究所 | Method for tracking target and device |
| CN107731011A (en) * | 2017-10-27 | 2018-02-23 | 中国科学院深圳先进技术研究院 | A kind of harbour is moored a boat monitoring method, system and electronic equipment |
| US20190044818A1 (en) * | 2018-01-12 | 2019-02-07 | Intel Corporation | Self-adjusting data processing system |
| CN108377264A (en) * | 2018-02-05 | 2018-08-07 | 江苏大学 | Vehicular ad hoc network quorum-sensing system data report De-weight method |
| CN109547541A (en) * | 2018-11-12 | 2019-03-29 | 安徽师范大学 | Mist calculates the node low overhead collaboration method under environment based on filtering and distribution mechanism |
Non-Patent Citations (3)
| Title |
|---|
| BETHANABHOTLA, D；CAIRE, G；NEELY, MJ: "《Adaptive Video Streaming for Wireless Networks With Multiple Users and Helpers》", 《IEEE TRANSACTIONS ON COMMUNICATIONS》 * |
| ZHANG, Y: "《Multi-Dimensional Payment Plan in Fog Computing with Moral Hazard》", 《2016 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATION SYSTEMS (ICCS)》 * |
| 罗家木; 陈雍君; 陈渝江; 邱实: "《基于5G无线传感网络的智慧管廊综合监控系统设计》", 《电子测量技术》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111565303A (en) * | 2020-05-29 | 2020-08-21 | 深圳市易链信息技术有限公司 | Video monitoring method, system and readable storage medium based on fog calculation and deep learning |
| CN112398158A (en) * | 2020-10-27 | 2021-02-23 | 国网经济技术研究院有限公司 | Distributed collection device and method for operation indexes of hybrid high-voltage direct-current power transmission system |
| CN112398158B (en) * | 2020-10-27 | 2022-11-01 | 国网经济技术研究院有限公司 | Distributed collection device and method for operation indexes of hybrid high-voltage direct-current power transmission system |
| CN113203439A (en) * | 2021-05-07 | 2021-08-03 | 南京邮电大学 | Master-slave dynamic edge sensor ad hoc network system for water information detection |
| CN115150404A (en) * | 2022-06-09 | 2022-10-04 | 安徽天元通信发展有限公司 | 5G base station information analysis processing method and system based on integrated structure |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110087041B (en) | 2021-01-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110087041A (en) | Video data processing and transmission method and system based on the base station 5G | |
| US9251425B2 (en) | Object retrieval in video data using complementary detectors | |
| CN109635748B (en) | Method for extracting road characteristics in high-resolution image | |
| CN110390246A (en) | A kind of video analysis method in side cloud environment | |
| CN106682592A (en) | Automatic image recognition system and method based on neural network method | |
| CN110659391A (en) | Video detection method and device | |
| CN108280902A (en) | The document handling method and device of vehicle-mounted monitoring equipment, vehicle-mounted monitoring equipment | |
| CN112184625A (en) | Pavement defect identification method and system based on video deep learning | |
| CN113642403B (en) | Crowd abnormal intelligent safety detection system based on edge calculation | |
| CN110009675A (en) | Generate method, apparatus, medium and the equipment of disparity map | |
| CN112085031A (en) | Target detection method and system | |
| CN116343103B (en) | Natural resource supervision method based on three-dimensional GIS scene and video fusion | |
| CN112309068A (en) | Forest fire early warning method based on deep learning | |
| CN108509495A (en) | The processing method and processing device of seismic data, storage medium, processor | |
| CN112668675B (en) | Image processing method and device, computer equipment and storage medium | |
| CN106993163A (en) | A kind of video monitoring system based on motion image detection | |
| CN106529497A (en) | Image acquisition device positioning method and device | |
| CN113971666A (en) | Power transmission line machine inspection image self-adaptive identification method based on depth target detection | |
| CN112597995A (en) | License plate detection model training method, device, equipment and medium | |
| GB2598640A (en) | Processing of images captured by vehicle mounted cameras | |
| CN114095725B (en) | Method and system for judging whether camera is abnormal | |
| CN114639084A (en) | Road side end vehicle sensing method based on SSD (solid State disk) improved algorithm | |
| CN114022425A (en) | Vehicle detection method and device with scene self-adaption function | |
| CN114639076A (en) | Target object detection method, target object detection device, storage medium, and electronic device | |
| KR20200122900A (en) | Vehicle tracking system based on surveillance videos |
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 |