CN117651123A - Multi-path video stream processing method and system based on camera - Google Patents
Multi-path video stream processing method and system based on camera Download PDFInfo
- Publication number
- CN117651123A CN117651123A CN202410127877.0A CN202410127877A CN117651123A CN 117651123 A CN117651123 A CN 117651123A CN 202410127877 A CN202410127877 A CN 202410127877A CN 117651123 A CN117651123 A CN 117651123A
- Authority
- CN
- China
- Prior art keywords
- video
- self
- video stream
- path
- target
- 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
- 238000003672 processing method Methods 0.000 title claims abstract description 5
- 230000005540 biological transmission Effects 0.000 claims abstract description 259
- 238000004458 analytical method Methods 0.000 claims abstract description 52
- 238000000034 method Methods 0.000 claims abstract description 51
- 238000012545 processing Methods 0.000 claims abstract description 39
- 238000005457 optimization Methods 0.000 claims abstract description 27
- 238000009499 grossing Methods 0.000 claims abstract description 19
- 238000009826 distribution Methods 0.000 claims abstract description 18
- 238000012544 monitoring process Methods 0.000 claims abstract description 15
- 238000000605 extraction Methods 0.000 claims abstract description 13
- 230000010354 integration Effects 0.000 claims abstract description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 33
- 230000006835 compression Effects 0.000 claims description 25
- 238000007906 compression Methods 0.000 claims description 25
- 230000008901 benefit Effects 0.000 claims description 17
- 238000005259 measurement Methods 0.000 claims description 17
- 239000013598 vector Substances 0.000 claims description 17
- 238000011156 evaluation Methods 0.000 claims description 15
- 238000003012 network analysis Methods 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000013139 quantization Methods 0.000 claims description 12
- 230000002457 bidirectional effect Effects 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 230000004927 fusion Effects 0.000 claims description 6
- 230000008447 perception Effects 0.000 claims description 5
- 230000003044 adaptive effect Effects 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000005096 rolling process Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 description 27
- 238000005516 engineering process Methods 0.000 description 8
- 230000008859 change Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000003860 storage Methods 0.000 description 4
- 230000015556 catabolic process Effects 0.000 description 3
- 238000006731 degradation reaction Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 230000000903 blocking effect Effects 0.000 description 2
- 238000010924 continuous production Methods 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000002068 genetic effect Effects 0.000 description 2
- 238000001303 quality assessment method Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000000547 structure data Methods 0.000 description 2
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005206 flow analysis Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000000700 time series analysis Methods 0.000 description 1
Landscapes
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The application relates to the technical field of video stream processing and discloses a multichannel video stream processing method and system based on a camera. The method comprises the following steps: node configuration and link are carried out on a plurality of cameras, an initial self-organizing network is generated, multi-state video coding integration is carried out, and a first self-organizing network is generated; path planning and distribution are carried out, an optimal non-overlapping path is determined, multipath video stream propagation is carried out, and an initial multipath video stream is obtained; performing real-time video processing and SCTP concurrent multipath transmission to obtain a target multipath video stream; performing content analysis, feature extraction and video transmission parameter optimization to obtain a first video transmission parameter combination; performing video smoothing and dynamic bandwidth allocation to obtain a second video transmission parameter combination; video transmission quality monitoring and self-adaptive adjustment are carried out based on the second video transmission parameter combination, and a second self-organizing network is obtained.
Description
Technical Field
The present disclosure relates to the field of video stream processing technologies, and in particular, to a method and a system for processing multiple video streams based on a camera.
Background
With the continuous development of technology and the rising of applications such as social media and remote collaboration, a multi-channel video stream processing technology is becoming a field of great concern. The background of this technology stems from the need for multiple cameras to capture, process and transmit video streams simultaneously to enable a wider range of video applications.
However, current multi-path video stream processing still faces some important issues, one of which is how to effectively manage configuration and collaboration among multiple cameras to obtain high quality multi-path video streams. In addition, challenges in real-time video processing, transmission, and quality monitoring, etc. need to be resolved to ensure the stability and quality of video transmission. How to optimize video parameters in different application scenarios is also a problem that remains to be studied in depth.
Disclosure of Invention
The application provides a multichannel video stream processing method and system based on cameras, and the multichannel video stream transmission efficiency and transmission quality of a plurality of cameras are improved.
In a first aspect, the present application provides a method for processing multiple video streams based on a camera, where the method for processing multiple video streams based on a camera includes:
Node configuration and link are carried out on a plurality of cameras to generate an initial self-organizing network, multi-state video coding integration is carried out on the initial self-organizing network, and a first self-organizing network is generated;
respectively planning and distributing paths of the cameras according to the first self-organizing network, determining an optimal non-overlapping path of each camera, and carrying out multipath video stream propagation on the cameras according to the optimal non-overlapping path to obtain an initial multipath video stream;
performing real-time video processing on the initial multi-path video stream to obtain a standard multi-path video stream, and performing SCTP concurrent multi-path transmission on the standard multi-path video stream to obtain a target multi-path video stream;
performing content analysis and feature extraction on the target multipath video stream to obtain a target video feature set, and performing video transmission parameter optimization through the target video feature set to obtain a first video transmission parameter combination;
performing video smoothing and dynamic bandwidth allocation on the first video transmission parameter combination to obtain a second video transmission parameter combination;
and performing video transmission quality monitoring and self-adaptive adjustment on the first self-organizing network based on the second video transmission parameter combination to obtain a second self-organizing network.
In a second aspect, the present application provides a camera-based multi-path video stream processing system, including:
the configuration module is used for carrying out node configuration and link on a plurality of cameras to generate an initial self-organizing network, and carrying out multi-state video coding integration on the initial self-organizing network to generate a first self-organizing network;
the planning module is used for respectively planning and distributing paths of the cameras according to the first self-organizing network, determining an optimal non-overlapping path of each camera, and carrying out multipath video stream propagation on the cameras according to the optimal non-overlapping path to obtain an initial multipath video stream;
the transmission module is used for carrying out real-time video processing on the initial multi-path video stream to obtain a standard multi-path video stream, and carrying out SCTP concurrent multi-path transmission on the standard multi-path video stream to obtain a target multi-path video stream;
the optimizing module is used for carrying out content analysis and feature extraction on the target multipath video stream to obtain a target video feature set, and carrying out video transmission parameter optimization through the target video feature set to obtain a first video transmission parameter combination;
The distribution module is used for carrying out video smoothing and dynamic bandwidth distribution on the first video transmission parameter combination to obtain a second video transmission parameter combination;
and the adjusting module is used for carrying out video transmission quality monitoring and self-adaptive adjustment on the first self-organizing network based on the second video transmission parameter combination to obtain a second self-organizing network.
In the technical scheme provided by the application, the camera nodes are configured by using a dynamic network algorithm so as to generate an initial self-organizing network. By means of the node efficiency index and the path diversity measurement, self-organizing optimization of the network is achieved, and high efficiency and diversity of the network are guaranteed. And selecting video coding parameters according to the efficiency and path diversity of the network nodes by adopting self-adaptive quantization parameter calculation so as to optimize video transmission effect and reduce transmission bandwidth consumption. Through path planning and distribution, multipath video stream propagation is realized, and concurrent multipath transmission is performed by using an SCTP protocol, so that the reliability and the bandwidth utilization rate of video transmission are effectively improved. Through content analysis and feature extraction, dynamic video coding, self-adaptive throttling control and content perception compression ratio adjustment are carried out according to content feature weight data, so that the efficiency and quality of video transmission are improved. Through video smoothing analysis and bandwidth demand analysis, smoothness prediction and dynamic bandwidth allocation of video transmission are realized so as to adapt to the change of network conditions, and continuity and stability of video streams are ensured. By using multidimensional video transmission state data and an self-organizing network analysis model, monitoring and self-adapting adjustment of video transmission quality are realized, and the multi-path video stream transmission efficiency and transmission quality of a plurality of cameras are improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained based on these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic diagram of one embodiment of a method for processing multiple video streams based on a camera in an embodiment of the present application;
fig. 2 is a schematic diagram of one embodiment of a camera-based multi-path video stream processing system according to an embodiment of the present application.
Detailed Description
The embodiment of the application provides a method and a system for processing multiple paths of video streams based on cameras. The terms "first," "second," "third," "fourth" and the like in the description and in the claims of this application and in the above-described figures, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
For ease of understanding, the following describes a specific flow of an embodiment of the present application, referring to fig. 1, and one embodiment of a method for processing multiple video streams based on a camera in the embodiment of the present application includes:
step 101, performing node configuration and link on a plurality of cameras to generate an initial self-organizing network, and performing multi-state video coding integration on the initial self-organizing network to generate a first self-organizing network;
it may be understood that the execution body of the present application may be a camera-based multi-path video stream processing system, and may also be a terminal or a server, which is not limited herein. The embodiment of the present application will be described by taking a server as an execution body.
Specifically, node configuration is performed on a plurality of cameras based on a dynamic network algorithm. By dynamically determining the position and role of cameras in the network using dynamic network algorithms, such as algorithms based on behavioral patterns or environmental feedback, it is ensured that each camera can participate in the operation of the entire network in the most efficient manner. Each camera becomes a network node that can be intelligently adjusted according to the network state and its own conditions. And linking the network nodes corresponding to each camera to form an initial self-organizing network. The connection between the nodes is dynamically established according to the real-time condition of the network and the capabilities of the camera nodes, and aims to optimize the data transmission path of the whole network and reduce potential bottlenecks. This process requires accurate calculations to ensure stability and efficiency of the network, as well as flexible adjustments to accommodate environmental changes or node state changes. To further optimize network performance, ad hoc network efficiency index calculations and path diversity metrics are performed. The performance and contribution of each node in the network are known by calculating the node efficiency index of each network node, thereby adjusting the inefficient or problematic nodes. Meanwhile, the path diversity measurement ensures the robustness of the network, and the path can be quickly adjusted when facing node faults or network congestion by evaluating and optimizing the path diversity in the network, so that the stable transmission of the video stream is ensured. And performing video coding self-adaptive parameter calculation according to the node efficiency index and the path diversity index to obtain self-adaptive quantization parameter data. These data are based on specific coding parameters of the current condition of the network, such as quantization level, frame rate and resolution, etc., which will be used to guide the video coding process to adapt to the changes and demands of the network condition. By the method, the coding efficiency and the network transmission efficiency can be improved to the maximum extent without sacrificing the video quality. And according to the self-adaptive quantization parameter data, performing multi-state video coding integration on the initial self-organizing network to generate a first self-organizing network which is more efficient and stable. The process not only improves the quality and efficiency of video transmission, but also lays a foundation for the expandability and adaptability of the whole network.
102, respectively planning and distributing paths of a plurality of cameras according to a first self-organizing network, determining an optimal non-overlapping path of each camera, and carrying out multipath video stream propagation on the plurality of cameras according to the optimal non-overlapping path to obtain an initial multipath video stream;
specifically, a multipath benefit evaluation is performed on the first ad hoc network. By comprehensively considering factors such as bandwidth, delay, packet loss rate and the like of each path in the network, the transmission benefit of the whole network is macroscopically grasped, and multipath benefit evaluation data which comprehensively reflects the network state is obtained. Meanwhile, path disjoint measurement is carried out, namely, the fault tolerance and the resource utilization rate of the network are evaluated by analyzing the overlapping degree of all paths in the network, and further path disjoint measurement data are obtained. These two items of data together constitute a comprehensive assessment of network status and performance. A route stability factor for the first ad hoc network is calculated based on the multipath benefit evaluation data and the path-disjoint metric data. The route stability factor is an index for comprehensively reflecting the stability and reliability of the network, and not only considers the transmission benefit of the network, but also synthesizes the mutual influence and conflict degree between paths. And calculating the multipath load balancing index of the first self-organizing network according to the route stability factor. The index reflects the balance degree of each path load when the network carries out multipath transmission in the current state, and is a key for ensuring the efficient and stable operation of the network. A good load balance index can not only reduce network congestion and conflict, but also improve the throughput and response speed of the network. And planning and distributing paths of the cameras according to the multipath load balancing indexes. And determining an optimal non-overlapping path for each camera through an optimization algorithm, such as a genetic algorithm or a particle swarm optimization algorithm, by considering the position, the network state, the degree of non-intersection among paths and other factors of each camera. This not only makes maximum use of network resources, but also avoids interference and collisions between different video streams. And according to the determined optimal non-overlapping paths, carrying out multipath video stream propagation on the plurality of cameras. Video data captured by each camera is transmitted to a destination along a pre-planned path. Because the paths are optimized, the transmission of the video data can be ensured to be efficient and stable, and meanwhile, due to the adoption of multi-path transmission, even if a certain path has a problem, the video data can be continuously transmitted through other paths, so that the transmission reliability is greatly improved. The resulting initial multi-path video stream will have high quality and high stability.
Step 103, performing real-time video processing on the initial multi-path video stream to obtain a standard multi-path video stream, and performing SCTP concurrent multi-path transmission on the standard multi-path video stream to obtain a target multi-path video stream;
specifically, the initial multi-path video stream is subjected to real-time video quality analysis, and the definition, color accuracy, frame rate and erroneous or lost frames of each video stream are evaluated. Real-time video quality data is obtained by employing algorithms such as signal-to-noise ratio calculations or image quality assessment models. And then, carrying out integrated compression on the initial multi-path video stream according to the real-time video quality data. Integrated compression is a process aimed at reducing the amount of video data while maintaining video quality, which typically involves the removal of redundant information, the application of efficient coding techniques, and the like. For example, h.264 or h.265 coding techniques may be used to dynamically adjust coding parameters, such as quantization parameters, intra-prediction and motion estimation strategies, etc., based on real-time video quality data to obtain a standard multi-channel video stream. This not only reduces the burden of data transmission, but also ensures the integrity and definition of the video content. Video stream selection analysis is performed by SCTP protocol. SCTP is a reliable transport protocol supporting multipath transmission and congestion control, suitable for the transmission of video streams. And analyzing and determining the path most suitable for transmitting each video stream according to the current network condition, the performance and the load of each path and other information. In this process, the system generates video stream adaptive adjustment factors, which are performance estimates for each path at a particular point in time, that will guide subsequent transmission path selection and adjustment. And optimizing the bandwidth utilization rate of the standard multipath video stream and carrying out concurrent multipath transmission according to the video stream self-adaptive adjustment factor. Bandwidth utilization optimization is a key for ensuring the maximum video transmission quality under limited network resources, and relates to the problems of dynamically adjusting the code rate of a video stream, solving network congestion, avoiding packet loss and the like. Concurrent multipath transmission improves the reliability and efficiency of transmission by transmitting video data using multiple paths simultaneously. And adjusting transmission strategies, such as switching paths, adjusting transmission rates or changing video coding parameters, in real time according to the video stream self-adaptive adjustment factors so as to ensure that each video stream can be transmitted in an optimal state, and finally obtaining the target multipath video stream. 104, performing content analysis and feature extraction on the target multipath video stream to obtain a target video feature set, and performing video transmission parameter optimization through the target video feature set to obtain a first video transmission parameter combination;
Specifically, content analysis is performed on the target multipath video stream, key content information such as scene description, object identification and activity detection is extracted from the video stream through image processing and analysis technology, and detailed video stream content data is obtained through the content analysis. And carrying out content characteristic weight distribution on the content data of the video stream, and distributing different importance levels for different content characteristics. For example, for security surveillance video, the identification of moving objects is more important than the static background, so the characteristics of moving objects will be given a higher weight. Video feature extraction is performed on video stream content data through a preset graph rolling network (GCN). GCN is an efficient deep learning model suitable for processing graph structure data, such as a network of pixel relationships in video frames. Through the GCN, high-dimensional features representing the video content are extracted, which capture essential attributes and structural information of the video content, forming an initial set of video features. And then, carrying out feature weight distribution on the content feature weight data according to the content feature weight data, thereby obtaining a target video feature set. The influence of important features is strengthened through weighting processing, and meanwhile less important features are restrained, so that the feature set is ensured to reflect key information of video content more accurately. And (3) carrying out dynamic video coding optimization on the target multipath video stream according to the target video feature set, dynamically adjusting coding parameters such as code rate, resolution, compression level and the like according to the complexity and characteristics of video content, ensuring the video quality and simultaneously reducing the data quantity to the greatest extent. And performing self-adaptive throttling control, and dynamically adjusting the transmission rate according to the network condition and the characteristics of video content so as to avoid network congestion and video jamming. And (3) adjusting the content-aware compression ratio, and adjusting the compression ratio according to the importance and complexity of the video content so as to ensure the definition of the important content and improve the overall compression efficiency. The optimized dynamic video coding parameters, the network congestion adjustment transmission rate and the content characteristic adjustment compression ratio are combined to generate a first video transmission parameter combination of the target multipath video stream. The parameter combination is a comprehensive configuration which comprehensively reflects the characteristics of video content, network conditions and transmission targets, and guides the system to carry out high-efficiency, stable and high-quality video transmission, thereby ensuring the optimization and performance improvement of the whole video stream processing process.
Step 105, performing video smoothing and dynamic bandwidth allocation on the first video transmission parameter combination to obtain a second video transmission parameter combination;
specifically, video smoothing analysis is performed on the target multipath video stream according to the first video transmission parameter combination, smoothness in the video transmission process is predicted, and consistency in video playing are evaluated. And obtaining a video transmission smoothness predicted value through the technologies such as an autoregressive model or a machine learning prediction algorithm. After the predicted value of the video transmission smoothness is obtained, the predicted value is compared with a preset target value of the video transmission smoothness to obtain a target comparison result. This preset target value is set based on user experience and quality of service requirements and represents an ideal video transmission smoothness. By comparing the predicted value and the target value, the difference between the current video transmission state and the ideal state is known, and the difference is the key basis for subsequent parameter adjustment. And carrying out video smoothing parameter adjustment on the first video transmission parameter combination according to the target comparison result so as to obtain a smoothed video transmission parameter combination. The smoothness of the video stream is optimized by means of dynamically adjusting video coding parameters, frame rate, buffer zone size and the like, the blocking and delay are reduced, and the user experience is improved. This process requires a comprehensive consideration of network conditions, video content characteristics, and user requirements, with precise control algorithms to adjust the parameters. And meanwhile, carrying out bandwidth demand analysis on the target multipath video stream through a preset bandwidth demand prediction model. The model predicts future bandwidth demands based on historical data, real-time network status, video content characteristics, and other factors to obtain target bandwidth demand data. Therefore, bandwidth allocation is performed more accurately, and each video stream can obtain enough bandwidth for high-quality transmission. And dynamically adjusting bandwidth allocation of the available bandwidth to the target multipath video stream according to the target bandwidth demand data. By means of dynamically adjusting network route, changing video coding bit rate or applying network congestion control technology, high-efficiency utilization of network resources is ensured, and video quality degradation caused by insufficient bandwidth is avoided. And carrying out parameter fusion on the smooth video transmission parameter combination and the dynamic bandwidth allocation parameter to obtain a second video transmission parameter combination. The fusion is a process of comprehensively considering video quality, network condition and user requirements, and various parameters are adjusted to the optimal state through an optimization algorithm, such as a multi-objective optimization or decision tree algorithm, so that high efficiency, smoothness and high quality of video transmission are ensured.
And step 106, performing video transmission quality monitoring and self-adaptive adjustment on the first self-organizing network based on the second video transmission parameter combination to obtain a second self-organizing network.
Specifically, video transmission is performed on the target multipath video stream based on the second video transmission parameter combination, and multidimensional video transmission state data are obtained. These data include multiple dimensions of transmission delay, packet loss rate, bandwidth utilization, and video quality, which reflect the real-time state and performance of video transmission. Then, video transmission quality analysis is carried out on the multidimensional video transmission state data, and a plurality of video transmission quality characteristic indexes are extracted from the complex state data through statistical analysis and a machine learning method. These characteristic indices describe the quality and performance of video transmissions from different perspectives. And performing feature index coding and vector conversion on the plurality of video transmission quality feature indexes, and converting the video transmission quality feature indexes into video transmission quality feature vectors. The characteristic indexes of various types and scales are unified into a standardized mathematical representation form, so that the characteristic indexes are convenient to process and analyze in a subsequent analysis model. The video transmission quality feature vector is input into a preset self-organizing network analysis model, wherein the model comprises a bidirectional GRU network, a unidirectional GRU network and a ReLU function. The bidirectional GRU network can capture the front-back dependency of time series data, the unidirectional GRU network focuses on the front-to-back time dependency, and the ReLU function is used for increasing nonlinearity and improving the expression capability of the model. And analyzing the video transmission quality feature vector through the self-organizing network analysis model to obtain a target probability prediction value. This predictor is a prediction of future network status and video transmission quality, reflecting problems and performance conditions that occur in video transmission under current parameters and network conditions. And inquiring a preset self-adaptive adjustment strategy list according to the predicted value, so as to obtain the target self-adaptive adjustment strategy which is most suitable for the current situation. This strategy includes various measures such as adjusting video coding parameters, changing transmission paths, or reallocating bandwidth. And carrying out self-adaptive adjustment on the first self-organizing network according to the target self-adaptive adjustment strategy so as to cope with network changes and ensure video transmission quality. This adjustment is a dynamic and continuous process that requires real-time monitoring of network conditions and video transmission performance, and rapid and accurate adjustments based on predictions of real-time data and analytical models. By self-adaptive adjustment, the system can effectively cope with network fluctuation and uncertainty, ensure the stability and quality of video transmission, and finally obtain the optimized and improved second self-organizing network.
In the embodiment of the application, the camera node is configured by using a dynamic network algorithm to generate an initial self-organizing network. By means of the node efficiency index and the path diversity measurement, self-organizing optimization of the network is achieved, and high efficiency and diversity of the network are guaranteed. And selecting video coding parameters according to the efficiency and path diversity of the network nodes by adopting self-adaptive quantization parameter calculation so as to optimize video transmission effect and reduce transmission bandwidth consumption. Through path planning and distribution, multipath video stream propagation is realized, and concurrent multipath transmission is performed by using an SCTP protocol, so that the reliability and the bandwidth utilization rate of video transmission are effectively improved. Through content analysis and feature extraction, dynamic video coding, self-adaptive throttling control and content perception compression ratio adjustment are carried out according to content feature weight data, so that the efficiency and quality of video transmission are improved. Through video smoothing analysis and bandwidth demand analysis, smoothness prediction and dynamic bandwidth allocation of video transmission are realized so as to adapt to the change of network conditions, and continuity and stability of video streams are ensured. By using multidimensional video transmission state data and an self-organizing network analysis model, monitoring and self-adapting adjustment of video transmission quality are realized, and the multi-path video stream transmission efficiency and transmission quality of a plurality of cameras are improved.
In a specific embodiment, the process of executing step 101 may specifically include the following steps:
(1) Based on a dynamic network algorithm, carrying out node configuration on a plurality of cameras to obtain network nodes corresponding to each camera;
(2) Node linking is carried out on the network nodes corresponding to each camera, and an initial self-organizing network corresponding to a plurality of cameras is generated;
(3) Performing self-organizing network efficiency index calculation on the initial self-organizing network to obtain node efficiency indexes of each network node in the initial self-organizing network;
(4) Carrying out path diversity measurement on the initial self-organizing network to obtain path diversity indexes of the initial self-organizing network;
(5) Performing video coding self-adaptive parameter calculation on the initial self-organizing network according to the node efficiency index and the path diversity index to obtain self-adaptive quantization parameter data;
(6) And carrying out multi-state video coding integration on the initial self-organizing network according to the self-adaptive quantization parameter data to generate a first self-organizing network.
Specifically, the node configuration is performed on the plurality of cameras, so that each camera can play a role in the network and effectively transmit video streams. By dynamic network algorithms, such as dynamic clustering algorithms based on signal strength or geographical location information, these algorithms can dynamically assign cameras to optimal network node locations based on their real-time data. And linking the nodes to generate a corresponding initial self-organizing network. This process requires consideration of the distance between nodes, signal strength, and obstructions, and uses a shortest path algorithm such as a minimum spanning tree or graph theory to determine the best connection between nodes. And calculating the self-organizing network efficiency index of the initial self-organizing network by analyzing the topological structure of the network, the connection quality of the nodes, the data transmission efficiency and other factors. The node efficiency index of each network node is a measure of the performance of the node in the network, including its transmission rate, connection stability, processing power, etc. Meanwhile, path diversity measurement is carried out on the initial self-organizing network. Path redundancy and backup options in the network are analyzed to evaluate the robustness and fault tolerance capabilities of the network. The path diversity index is an important metric that reflects the recovery capability of the network in the face of node failure or data transmission interference. For example, paths from one camera node to another node, and redundancy and backup of those paths are analyzed, and then a path diversity index is calculated, ensuring that the network continues to operate stably even if a problem occurs at a certain node or path. And performing video coding self-adaptive parameter calculation according to the node efficiency index and the path diversity index. By analyzing the performance and capacity of the network, parameters of video coding, such as code rate and resolution, are then dynamically adjusted according to the complexity and importance of the video content, so as to maximize video quality while ensuring efficient operation of the network. For example, the system may find that some nodes may transmit higher quality video due to their better location and connection quality, while other nodes need to reduce video quality to maintain stable transmission. And carrying out multi-state video coding integration on the initial self-organizing network according to the self-adaptive quantization parameter data to generate a first self-organizing network. Each video stream in the network is optimally encoded by applying various video coding techniques and algorithms, such as h.264 or h.265, in combination with the previously calculated adaptive parameters.
In a specific embodiment, the process of executing step 102 may specifically include the following steps:
(1) Performing multipath benefit evaluation on the first self-organizing network to obtain multipath benefit evaluation data, and performing path disjoint measurement on the first self-organizing network to obtain path disjoint measurement data;
(2) Calculating a route stability factor of the first self-organizing network according to the multipath benefit evaluation data and the path disjoint measurement data;
(3) Calculating a multipath load balancing index of the first self-organizing network according to the route stability factor;
(4) Path planning and distribution are carried out on a plurality of cameras according to the multipath load balancing index, and the optimal non-overlapping path of each camera is determined;
(5) And carrying out multipath video stream propagation on the cameras according to the optimal non-overlapping paths to obtain an initial multipath video stream.
Specifically, the first self-organizing network is subjected to multipath benefit evaluation, and the transmission efficiency, delay, bandwidth utilization and other relevant performance indexes of each path in the network are analyzed. Such evaluation is typically accomplished by collecting real-time network performance data and applying an analytical method such as a weight graph algorithm or network flow analysis. The result of the multipath benefit evaluation is multipath benefit evaluation data detailing the performance and benefit of each path in the network. For example, the evaluation involves analyzing the paths from the cameras to the central server, evaluating their transmission rates and reliability to determine which paths can provide the most efficient video streaming. And carrying out path disjoint measurement on the first self-organizing network, evaluating the diversity and redundancy of paths in the network, and ensuring that the network can still keep stable operation when facing node or link faults. Path disjoint metrics generally involve calculating the degree of overlap between paths in a network using algorithms such as shortest path algorithms or network topology analysis to identify individual paths in the network. For example, a path disjoint metric will help determine whether video data can bypass through other paths to avoid transmission disruption when one path fails. And calculating the route stability factor of the first self-organizing network according to the multipath benefit evaluation data and the path disjoint measurement data. The route stability factor is an index for comprehensively reflecting the stability and reliability of the network, comprehensively considers the benefits and diversity of paths, and ensures that the network has enough elasticity to cope with faults while efficiently transmitting data. For example, calculating a route stability factor involves comprehensively considering the transmission efficiency of each path and the path redundancy in the network to ensure that the system can operate stably even when some nodes or paths are problematic. And calculating the multipath load balancing index of the first self-organizing network according to the route stability factor. This index reflects the degree of balancing of the path loads when the network is performing multipath transmission in the current state. A good load balance index can not only reduce network congestion and conflict, but also improve the throughput and response speed of the network. Calculating the multipath load balancing index involves evaluating the load conditions of each path at different times and under different conditions to dynamically adjust the transmission strategy of video data, ensuring that the video streams of all cameras can be efficiently and stably transmitted. And carrying out path planning and distribution on the cameras according to the multipath load balancing index, and determining the optimal non-overlapping path of each camera. By applying a path optimization algorithm, such as a genetic algorithm or an ant colony optimization algorithm, an efficient and stable independent transmission path is found for each camera. According to the current network condition and the positions of the cameras, an optimal path is dynamically allocated to each camera, so that video data of each camera can be rapidly and reliably transmitted to a central server. And carrying out multipath video stream propagation on the cameras according to the optimal non-overlapping paths to obtain an initial multipath video stream. This process requires the network to dynamically adjust the transmission path of the video stream while monitoring the performance of each path in real time to cope with network changes and congestion.
In a specific embodiment, the process of executing step 103 may specifically include the following steps:
(1) Real-time video quality analysis is carried out on the initial multipath video stream, and real-time video quality data are obtained;
(2) Integrating and compressing the initial multi-path video stream according to the real-time video quality data to obtain a standard multi-path video stream;
(3) Carrying out video stream path selection analysis on the standard multipath video stream through an SCTP protocol to obtain a video stream self-adaptive adjustment factor;
(4) And optimizing the bandwidth utilization rate of the standard multipath video stream and carrying out concurrent multipath transmission according to the video stream self-adaptive adjustment factor to obtain the target multipath video stream.
Specifically, real-time video quality analysis is performed on the initial multipath video stream, and real-time video quality data is obtained. By introducing video quality assessment algorithms such as PSNR (peak signal to noise ratio) or SSIM (structural similarity index), sharpness, frame rate, color accuracy, transmission errors, etc. of each path of video stream are monitored and assessed. And carrying out integrated compression on the initial multi-path video stream according to the real-time video quality data to obtain a standard multi-path video stream. Integrated compression aims to reduce the size of video data for ease of transmission and storage while maintaining video quality as much as possible. And carrying out video stream flow path selection analysis on the standard multipath video streams through SCTP (stream control transmission protocol) to obtain video stream self-adaptive adjustment factors. SCTP is a reliable transport protocol supporting multipath transport and network failover, and is suitable for video streaming, which requires high reliability and real-time data transmission. Video stream path selection analysis aims to select the optimal video transmission path according to the current conditions of the network, such as the bandwidth, delay and packet loss rate of the path. The adaptive adjustment factor is generated according to the analysis results and guides the dynamic path selection and transmission adjustment of the video stream. For example, when a path is reduced in bandwidth due to congestion, the system can automatically switch the video stream to other paths with better quality by utilizing the multipath characteristic of SCTP, so as to ensure the continuity and quality of video transmission. And optimizing the bandwidth utilization rate of the standard multipath video stream and carrying out concurrent multipath transmission according to the video stream self-adaptive adjustment factor so as to obtain the target multipath video stream. Bandwidth utilization optimization aims to ensure that the quality of video transmission is maximized within the available bandwidth, and the code rate or resolution of the video stream is dynamically adjusted based on current network conditions. The concurrent multipath transmission utilizes the multipath characteristic of SCTP to transmit video data in parallel through multiple paths so as to enhance the reliability and efficiency of transmission.
In a specific embodiment, the process of executing step 104 may specifically include the following steps:
(1) Content analysis is carried out on the target multipath video stream, and corresponding video stream content data is obtained;
(2) Content characteristic weight distribution is carried out on the content data of the video stream, so that content characteristic weight data are obtained;
(3) Extracting video characteristics of the video stream content data through a preset graph rolling network to obtain an initial video characteristic set;
(4) Performing feature weight distribution on the initial video feature set according to the content feature weight data to obtain a target video feature set;
(5) According to the target video feature set, carrying out dynamic video coding optimization on the target multipath video stream to obtain dynamic video coding parameters;
(6) According to the target video feature set, performing self-adaptive throttling control on the target multipath video stream to obtain a network congestion adjustment transmission rate;
(7) According to the target video feature set, content perception compression ratio adjustment is carried out on the target multipath video stream, and content feature adjustment compression ratio is obtained;
(8) And generating a first video transmission parameter combination of the target multi-path video stream according to the dynamic video coding parameters, the network congestion adjustment transmission rate and the content characteristic adjustment compression ratio.
Specifically, content analysis is performed on the target multipath video stream, and corresponding video stream content data is obtained. Various content in the video, including scenes, objects, characters, actions, etc., are identified and understood by video analysis techniques, such as deep learning or computer vision algorithms. And distributing the content characteristic weight of the video stream content data to obtain the content characteristic weight data. Content feature weight assignment is typically based on the nature of the video content and the goal of transmission optimization, with the weight of each feature determined by analyzing the importance of the video content and the need for transmission. Video feature extraction is performed on the video stream content data through a preset graph rolling network (GCN) to obtain an initial video feature set. The GCN is an effective deep learning model suitable for processing graph structure data. In video analytics, the GCN may capture complex relationships between video frames, such as interactions between objects or time-series changes in the scene, thereby extracting high-dimensional features that describe the video content. And then, carrying out feature weight distribution on the initial video feature set according to the content feature weight data so as to obtain a target video feature set. The influence of the most important and most informative parts of the video content is intensified by a weighting process while suppressing those features that are less important or irrelevant. And (3) carrying out dynamic video coding optimization on the target multipath video stream according to the target video feature set, and dynamically adjusting parameters of video coding, such as code rate, resolution and the like, according to the complexity and characteristics of video content so as to ensure efficient operation of a network while maximizing video quality. And performing self-adaptive throttling control according to the target video feature set to obtain the network congestion adjustment transmission rate. When network conditions change, such as bandwidth fluctuation or route change, the transmission rate of the video stream is dynamically adjusted to avoid network congestion and video jamming. For example, if network congestion or bandwidth degradation is detected, the system may temporarily reduce the transmission rate of some video streams to ensure that all video streams are transmitted stably. And adjusting the content-aware compression ratio of the target multipath video stream according to the target video feature set to obtain the content feature adjustment compression ratio. The compression ratio is adjusted according to the importance and complexity of the video content to ensure the definition of the important content while improving the overall compression efficiency. For example, if the video shows an important event, such as an accident or emergency, the system may decrease the compression ratio to improve video quality; whereas for those videos that are displaying normal, the system increases the compression ratio to reduce the amount of data. And integrating the dynamic video coding parameters, the network congestion adjustment transmission rate and the content characteristic adjustment compression rate to generate a first video transmission parameter combination of the target multipath video stream. This combination of parameters is a comprehensive configuration that fully reflects the video content characteristics, network conditions, and transmission objectives, and will guide the system in efficient, stable, and high quality video transmission.
In a specific embodiment, the process of executing step 105 may specifically include the following steps:
(1) According to the first video transmission parameter combination, video smoothing analysis is carried out on the target multipath video stream, and a video transmission smoothness predicted value is obtained;
(2) Comparing the video transmission smoothness predicted value with a preset video transmission smoothness target value to obtain a target comparison result;
(3) Performing video smoothing parameter adjustment on the first video transmission parameter combination according to the target comparison result to obtain a smoothed video transmission parameter combination;
(4) Carrying out bandwidth demand analysis on the target multipath video stream through a preset bandwidth demand prediction model to obtain target bandwidth demand data;
(5) Performing available bandwidth dynamic adjustment bandwidth allocation on the target multipath video stream according to the target bandwidth demand data to obtain dynamic bandwidth allocation parameters;
(6) And carrying out parameter fusion on the smooth video transmission parameter combination and the dynamic bandwidth allocation parameter to obtain a second video transmission parameter combination.
Specifically, video smoothing analysis is performed on the target multipath video stream according to the first video transmission parameter combination, and a video transmission smoothness predicted value is obtained. For example, monitoring and evaluating transmission continuity, frame rate stability, and delay and jitter of the video stream, etc. Through these analyses, the smoothness of the video transmission is predicted based on real-time network conditions and video content characteristics, forming a video transmission smoothness prediction value. And comparing the video transmission smoothness predicted value with a preset video transmission smoothness target value to obtain a target comparison result. The preset video transmission smoothness target value is set based on the desired quality of service and user experience, and represents the ideal video transmission smoothness. And comparing the predicted value with the target value to obtain a gap between the current video transmission state and the ideal state, wherein the gap is a key basis for subsequent parameter adjustment. And carrying out video smoothing parameter adjustment on the first video transmission parameter combination according to the target comparison result so as to obtain a smoothed video transmission parameter combination. The adjustment is realized by means of dynamically adjusting video coding parameters, frame rate, buffer size and the like, and aims to optimize the smoothness of a video stream, reduce the blocking and delay and improve the user experience. This process requires a comprehensive consideration of network conditions, video content characteristics, and user requirements, and various parameters are adjusted by a control algorithm. For example, if the system detects that the video transmission is not smooth enough, the buffer size may be increased or the frame rate may be decreased to reduce the jitter and improve the continuity of the video stream. And meanwhile, carrying out bandwidth demand analysis on the target multipath video stream through a preset bandwidth demand prediction model. The model predicts future bandwidth demands based on historical data, real-time network status, video content characteristics, and other factors to obtain target bandwidth demand data. And dynamically adjusting bandwidth allocation of the available bandwidth to the target multipath video stream according to the target bandwidth demand data. By means of dynamically adjusting network route, changing video coding bit rate or applying network congestion control technology, high-efficiency utilization of network resources is ensured, and video quality degradation caused by insufficient bandwidth is avoided. And carrying out parameter fusion on the smooth video transmission parameter combination and the dynamic bandwidth allocation parameter to obtain a second video transmission parameter combination. The fusion is a process of comprehensively considering video quality, network condition and user requirements, and various parameters are adjusted to the optimal state through an optimization algorithm, such as a multi-objective optimization or decision tree algorithm, so that high efficiency, smoothness and high quality of video transmission are ensured.
In a specific embodiment, the process of executing step 106 may specifically include the following steps:
(1) Video transmission is carried out on the target multipath video stream based on the second video transmission parameter combination, and multidimensional video transmission state data are obtained;
(2) Performing video transmission quality analysis on the multidimensional video transmission state data to obtain a plurality of video transmission quality characteristic indexes;
(3) Performing feature index coding and vector conversion on a plurality of video transmission quality feature indexes to obtain video transmission quality feature vectors;
(4) Inputting the video transmission quality feature vector into a preset self-organizing network analysis model, wherein the self-organizing network analysis model comprises: bidirectional GRU networks, unidirectional GRU networks, and ReLU functions;
(5) Performing self-organizing network analysis on the video transmission quality feature vector through a self-organizing network analysis model to obtain a target probability prediction value, and inquiring a preset self-adaptive adjustment strategy list according to the target probability prediction value to obtain a target self-adaptive adjustment strategy;
(6) And carrying out self-adaptive adjustment on the first self-organizing network according to the target self-adaptive adjustment strategy to obtain a second self-organizing network.
Specifically, video transmission is performed on the target multipath video stream based on the second video transmission parameter combination, and multidimensional video transmission state data are obtained. These data typically include multiple dimensions of frame rate, bandwidth utilization, delay, packet loss rate, etc., which collectively reflect the real-time state and performance of the video transmission. And carrying out video transmission quality analysis to obtain a plurality of video transmission quality characteristic indexes. Key features describing video transmission quality are extracted from complex state data by a variety of analysis techniques, such as time series analysis, statistical modeling, or machine learning algorithms. And performing feature index coding and vector conversion on the multiple video transmission quality feature indexes to obtain video transmission quality feature vectors. The characteristic indexes of various types and scales are unified into a standardized mathematical representation form, so that the characteristic indexes are convenient to process and analyze in a subsequent analysis model. The video transmission quality feature vector is input into a preset self-organizing network analysis model, wherein the model comprises a bidirectional GRU network, a unidirectional GRU network and a ReLU function. The bidirectional GRU network can capture the front-back dependency of time series data, the unidirectional GRU network focuses on the front-to-back time dependency, and the ReLU function is used for increasing nonlinearity and improving the expression capability of the model. These components together form a powerful analytical model that can effectively process and analyze video transmission quality feature vectors, capturing patterns and trends therein. And analyzing the video transmission quality feature vector through the self-organizing network analysis model to obtain a target probability prediction value. This predictor is a prediction of future network status and video transmission quality, reflecting problems and performance conditions that occur in video transmission under current parameters and network conditions. And inquiring a preset self-adaptive adjustment strategy list according to the predicted value, so as to obtain the target self-adaptive adjustment strategy which is most suitable for the current situation. This strategy includes various measures such as adjusting video coding parameters, changing transmission paths, or reallocating bandwidth. And carrying out self-adaptive adjustment on the first self-organizing network according to the target self-adaptive adjustment strategy so as to obtain a second self-organizing network. This adjustment is a dynamic and continuous process that requires real-time monitoring of network conditions and video transmission performance, and rapid and accurate adjustments based on predictions of real-time data and analytical models. Through the self-adaptive adjustment, the system can effectively cope with network fluctuation and uncertainty, ensure the stability and quality of video transmission, and finally obtain the optimized and improved second self-organizing network.
The method for processing multiple video streams based on the camera in the embodiment of the present application is described above, and the following describes a multiple video stream processing system based on the camera in the embodiment of the present application, please refer to fig. 2, and one embodiment of the multiple video stream processing system based on the camera in the embodiment of the present application includes:
the configuration module 201 is configured to perform node configuration and link on a plurality of cameras, generate an initial self-organizing network, and perform multi-state video coding integration on the initial self-organizing network to generate a first self-organizing network;
the planning module 202 is configured to plan and allocate paths for the multiple cameras according to the first ad hoc network, determine an optimal non-overlapping path of each camera, and perform multipath video stream propagation for the multiple cameras according to the optimal non-overlapping path, so as to obtain an initial multipath video stream;
the transmission module 203 is configured to perform real-time video processing on the initial multi-path video stream to obtain a standard multi-path video stream, and perform SCTP concurrent multi-path transmission on the standard multi-path video stream to obtain a target multi-path video stream;
the optimizing module 204 is configured to perform content analysis and feature extraction on the target multi-path video stream to obtain a target video feature set, and perform video transmission parameter optimization through the target video feature set to obtain a first video transmission parameter combination;
An allocation module 205, configured to perform video smoothing and dynamic bandwidth allocation on the first video transmission parameter combination to obtain a second video transmission parameter combination;
and the adjusting module 206 is configured to monitor and adaptively adjust video transmission quality of the first ad hoc network based on the second video transmission parameter combination, so as to obtain a second ad hoc network.
By the cooperative cooperation of the above-mentioned individual components, the camera nodes are configured by using a dynamic network algorithm to generate an initial ad hoc network. By means of the node efficiency index and the path diversity measurement, self-organizing optimization of the network is achieved, and high efficiency and diversity of the network are guaranteed. And selecting video coding parameters according to the efficiency and path diversity of the network nodes by adopting self-adaptive quantization parameter calculation so as to optimize video transmission effect and reduce transmission bandwidth consumption. Through path planning and distribution, multipath video stream propagation is realized, and concurrent multipath transmission is performed by using an SCTP protocol, so that the reliability and the bandwidth utilization rate of video transmission are effectively improved. Through content analysis and feature extraction, dynamic video coding, self-adaptive throttling control and content perception compression ratio adjustment are carried out according to content feature weight data, so that the efficiency and quality of video transmission are improved. Through video smoothing analysis and bandwidth demand analysis, smoothness prediction and dynamic bandwidth allocation of video transmission are realized so as to adapt to the change of network conditions, and continuity and stability of video streams are ensured. By using multidimensional video transmission state data and an self-organizing network analysis model, monitoring and self-adapting adjustment of video transmission quality are realized, and the multi-path video stream transmission efficiency and transmission quality of a plurality of cameras are improved.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, systems and units may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are merely for illustrating the technical solution of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.
Claims (8)
1. The multichannel video stream processing method based on the camera is characterized by comprising the following steps of:
node configuration and link are carried out on a plurality of cameras to generate an initial self-organizing network, multi-state video coding integration is carried out on the initial self-organizing network, and a first self-organizing network is generated;
respectively planning and distributing paths of the cameras according to the first self-organizing network, determining an optimal non-overlapping path of each camera, and carrying out multipath video stream propagation on the cameras according to the optimal non-overlapping path to obtain an initial multipath video stream;
Performing real-time video processing on the initial multi-path video stream to obtain a standard multi-path video stream, and performing SCTP concurrent multi-path transmission on the standard multi-path video stream to obtain a target multi-path video stream;
performing content analysis and feature extraction on the target multipath video stream to obtain a target video feature set, and performing video transmission parameter optimization through the target video feature set to obtain a first video transmission parameter combination;
performing video smoothing and dynamic bandwidth allocation on the first video transmission parameter combination to obtain a second video transmission parameter combination;
and performing video transmission quality monitoring and self-adaptive adjustment on the first self-organizing network based on the second video transmission parameter combination to obtain a second self-organizing network.
2. The method for processing multiple video streams based on cameras according to claim 1, wherein the steps of performing node configuration and linking on the cameras to generate an initial ad hoc network, performing multi-state video coding integration on the initial ad hoc network to generate a first ad hoc network, and include:
based on a dynamic network algorithm, carrying out node configuration on a plurality of cameras to obtain network nodes corresponding to each camera;
Node linking is carried out on the network nodes corresponding to each camera, and an initial self-organizing network corresponding to the cameras is generated;
performing self-organizing network efficiency index calculation on the initial self-organizing network to obtain node efficiency indexes of each network node in the initial self-organizing network;
performing path diversity measurement on the initial self-organizing network to obtain a path diversity index of the initial self-organizing network;
performing video coding self-adaptive parameter calculation on the initial self-organizing network according to the node efficiency index and the path diversity index to obtain self-adaptive quantization parameter data;
and carrying out multi-state video coding integration on the initial self-organizing network according to the self-adaptive quantization parameter data to generate a first self-organizing network.
3. The method for processing multiple video streams based on cameras according to claim 1, wherein the steps of respectively planning and distributing paths of the multiple cameras according to the first ad hoc network, determining an optimal non-overlapping path of each camera, and performing multiple-path video stream propagation on the multiple cameras according to the optimal non-overlapping path to obtain an initial multiple video stream, include:
Performing multipath benefit evaluation on the first self-organizing network to obtain multipath benefit evaluation data, and performing path disjoint measurement on the first self-organizing network to obtain path disjoint measurement data;
calculating a route stability factor of the first self-organizing network according to the multipath benefit evaluation data and the path disjoint measurement data;
calculating a multipath load balancing index of the first self-organizing network according to the route stability factor;
carrying out path planning and distribution on the cameras according to the multipath load balancing index, and determining an optimal non-overlapping path of each camera;
and carrying out multipath video stream propagation on the cameras according to the optimal non-overlapping paths to obtain initial multipath video streams.
4. The method for processing multiple video streams based on a camera according to claim 1, wherein the performing real-time video processing on the initial multiple video streams to obtain a standard multiple video stream, and performing SCTP concurrent multipath transmission on the standard multiple video stream to obtain a target multiple video stream includes:
performing real-time video quality analysis on the initial multi-path video stream to obtain real-time video quality data;
Integrating and compressing the initial multi-path video stream according to the real-time video quality data to obtain a standard multi-path video stream;
performing video stream path selection analysis on the standard multipath video stream through an SCTP protocol to obtain a video stream self-adaptive adjustment factor;
and optimizing the bandwidth utilization rate of the standard multipath video stream and carrying out concurrent multipath transmission according to the video stream self-adaptive adjustment factor to obtain a target multipath video stream.
5. The method for processing multiple video streams based on a camera according to claim 1, wherein the performing content analysis and feature extraction on the target multiple video streams to obtain a target video feature set, and performing video transmission parameter optimization through the target video feature set to obtain a first video transmission parameter combination includes:
content analysis is carried out on the target multipath video stream, and corresponding video stream content data is obtained;
performing content characteristic weight distribution on the video stream content data to obtain content characteristic weight data;
extracting video characteristics of the video stream content data through a preset graph rolling network to obtain an initial video characteristic set;
Performing feature weight distribution on the initial video feature set according to the content feature weight data to obtain a target video feature set;
according to the target video feature set, carrying out dynamic video coding optimization on the target multipath video stream to obtain dynamic video coding parameters;
according to the target video feature set, performing self-adaptive throttling control on the target multipath video stream to obtain a network congestion adjustment transmission rate;
according to the target video feature set, content perception compression ratio adjustment is carried out on the target multipath video stream, and content feature adjustment compression ratio is obtained;
and generating a first video transmission parameter combination of the target multi-path video stream according to the dynamic video coding parameter, the network congestion adjustment transmission rate and the content characteristic adjustment compression ratio.
6. The method for processing multiple video streams based on a camera according to claim 1, wherein said performing video smoothing and dynamic bandwidth allocation on the first video transmission parameter combination to obtain a second video transmission parameter combination includes:
according to the first video transmission parameter combination, video smoothing analysis is carried out on the target multipath video stream to obtain a video transmission smoothness predicted value;
Comparing the video transmission smoothness predicted value with a preset video transmission smoothness target value to obtain a target comparison result;
performing video smoothing parameter adjustment on the first video transmission parameter combination according to the target comparison result to obtain a smoothed video transmission parameter combination;
carrying out bandwidth demand analysis on the target multipath video stream through a preset bandwidth demand prediction model to obtain target bandwidth demand data;
performing available bandwidth dynamic adjustment bandwidth allocation on the target multipath video stream according to the target bandwidth demand data to obtain dynamic bandwidth allocation parameters;
and carrying out parameter fusion on the smooth video transmission parameter combination and the dynamic bandwidth allocation parameter to obtain a second video transmission parameter combination.
7. The method for processing multiple video streams based on a camera according to claim 1, wherein the performing video transmission quality monitoring and adaptive adjustment on the first ad hoc network based on the second video transmission parameter combination to obtain a second ad hoc network includes:
video transmission is carried out on the target multipath video stream based on the second video transmission parameter combination, and multidimensional video transmission state data are obtained;
Performing video transmission quality analysis on the multidimensional video transmission state data to obtain a plurality of video transmission quality characteristic indexes;
performing feature index coding and vector conversion on the plurality of video transmission quality feature indexes to obtain video transmission quality feature vectors;
inputting the video transmission quality feature vector into a preset self-organizing network analysis model, wherein the self-organizing network analysis model comprises the following components: bidirectional GRU networks, unidirectional GRU networks, and ReLU functions;
performing self-organizing network analysis on the video transmission quality feature vector through the self-organizing network analysis model to obtain a target probability prediction value, and inquiring a preset self-adaptive adjustment strategy list according to the target probability prediction value to obtain a target self-adaptive adjustment strategy;
and carrying out self-adaptive adjustment on the first self-organizing network according to the target self-adaptive adjustment strategy to obtain a second self-organizing network.
8. A camera-based multi-path video stream processing system, the camera-based multi-path video stream processing system comprising:
the configuration module is used for carrying out node configuration and link on a plurality of cameras to generate an initial self-organizing network, and carrying out multi-state video coding integration on the initial self-organizing network to generate a first self-organizing network;
The planning module is used for respectively planning and distributing paths of the cameras according to the first self-organizing network, determining an optimal non-overlapping path of each camera, and carrying out multipath video stream propagation on the cameras according to the optimal non-overlapping path to obtain an initial multipath video stream;
the transmission module is used for carrying out real-time video processing on the initial multi-path video stream to obtain a standard multi-path video stream, and carrying out SCTP concurrent multi-path transmission on the standard multi-path video stream to obtain a target multi-path video stream;
the optimizing module is used for carrying out content analysis and feature extraction on the target multipath video stream to obtain a target video feature set, and carrying out video transmission parameter optimization through the target video feature set to obtain a first video transmission parameter combination;
the distribution module is used for carrying out video smoothing and dynamic bandwidth distribution on the first video transmission parameter combination to obtain a second video transmission parameter combination;
and the adjusting module is used for carrying out video transmission quality monitoring and self-adaptive adjustment on the first self-organizing network based on the second video transmission parameter combination to obtain a second self-organizing network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410127877.0A CN117651123B (en) | 2024-01-30 | 2024-01-30 | Multi-path video stream processing method and system based on camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410127877.0A CN117651123B (en) | 2024-01-30 | 2024-01-30 | Multi-path video stream processing method and system based on camera |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117651123A true CN117651123A (en) | 2024-03-05 |
| CN117651123B CN117651123B (en) | 2024-04-05 |
Family
ID=90043765
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410127877.0A Active CN117651123B (en) | 2024-01-30 | 2024-01-30 | Multi-path video stream processing method and system based on camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117651123B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9867112B1 (en) * | 2016-11-23 | 2018-01-09 | Centurylink Intellectual Property Llc | System and method for implementing combined broadband and wireless self-organizing network (SON) |
| CN111654869A (en) * | 2020-05-13 | 2020-09-11 | 中铁二院工程集团有限责任公司 | Wireless network ad hoc network method |
| CN114025330A (en) * | 2022-01-07 | 2022-02-08 | 北京航空航天大学 | Air-ground cooperative self-organizing network data transmission method |
| CN116016000A (en) * | 2022-11-29 | 2023-04-25 | 深圳市瀚晖威视科技有限公司 | Video base station multi-network intelligent access method |
-
2024
- 2024-01-30 CN CN202410127877.0A patent/CN117651123B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9867112B1 (en) * | 2016-11-23 | 2018-01-09 | Centurylink Intellectual Property Llc | System and method for implementing combined broadband and wireless self-organizing network (SON) |
| CN111654869A (en) * | 2020-05-13 | 2020-09-11 | 中铁二院工程集团有限责任公司 | Wireless network ad hoc network method |
| CN114025330A (en) * | 2022-01-07 | 2022-02-08 | 北京航空航天大学 | Air-ground cooperative self-organizing network data transmission method |
| CN116016000A (en) * | 2022-11-29 | 2023-04-25 | 深圳市瀚晖威视科技有限公司 | Video base station multi-network intelligent access method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117651123B (en) | 2024-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111212038B (en) | Open data API gateway system based on big data artificial intelligence | |
| CN109525299B (en) | Satellite communication system and communication method for adaptive coding modulation optimization | |
| EP3528433B1 (en) | Data analyses using compressive sensing for internet of things (iot) networks | |
| GB2593180A (en) | A distributed network traffic data decomposition method | |
| Lekharu et al. | Deep learning based prediction model for adaptive video streaming | |
| Sani et al. | SMASH: A supervised machine learning approach to adaptive video streaming over HTTP | |
| CN116627631A (en) | Resource scheduling method, system, electronic equipment and storage medium | |
| Huang et al. | Machine learning for broad-sensed internet congestion control and avoidance: A comprehensive survey | |
| CN117651123B (en) | Multi-path video stream processing method and system based on camera | |
| Tong et al. | SDN-based application-aware segment routing for large-scale network | |
| Houidi et al. | Multi-agent graph convolutional reinforcement learning for intelligent load balancing | |
| CN113518039A (en) | Deep reinforcement learning-based resource optimization method and system under SDN architecture | |
| GB2500245A (en) | Rate Optimisation for Scalable Video Transmission | |
| CN117221295A (en) | Low-delay video transmission system based on edge calculation and network slicing | |
| CN106406082B (en) | System control method, device, controller and control system | |
| CN114500561B (en) | Power Internet of things network resource allocation decision-making method, system, equipment and medium | |
| Naresh et al. | Sac-abr: Soft actor-critic based deep reinforcement learning for adaptive bitrate streaming | |
| CN116132361A (en) | QoS and QoE-based network resource optimization adjustment method and system | |
| US11558263B2 (en) | Network device association with network management system | |
| Kumar et al. | Adaptive video streaming over HTTP using stochastic bitrate prediction in 4G wireless networks | |
| CN117651288B (en) | Signal transmission method, device and equipment of intelligent outdoor television and storage medium | |
| CN117440135B (en) | Image transmission method and system based on Beidou satellite communication | |
| Knapińska et al. | Prediction of multiple types of traffic with a novel evaluation metric related to bandwidth blocking | |
| Zuo | A deep learning approach to downlink user throughput prediction in cellular networks | |
| CN117729164B (en) | Dynamic bandwidth allocation system for four-port gigabit network card |
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 |