A high-reliability network video broadcast control system

Abstract

This application relates to the field of video compression and transmission technology, specifically to a highly reliable network video broadcast control system. The system includes: a data acquisition module for acquiring network bandwidth and the number of videos being transmitted during each transmission period; a parameter correction module for obtaining corrected quantization parameters, including: extracting each connected component of each frame in the network video; obtaining the matching connected component, change factor, and contour evaluation value corresponding to each connected component; calculating the inter-frame difference and feature richness of each frame; and calculating the network instability, compression coefficient, and corrected quantization parameters for each transmission period; and a compression and transmission module for real-time compression and transmission of the network video using the corrected quantization parameters and the HEVC algorithm. This application improves the transmission reliability and broadcast control smoothness of network video in complex scenarios such as weak networks by dynamically adjusting the quantization parameters.

Description

Technical Field

[0001] This application relates to the field of video compression and transmission technology, specifically to a highly reliable network video broadcasting and control system. Background Technology

[0002] With the rapid development of network technology, ultra-high-definition video streams are widely used in real-time scenarios such as telemedicine, online education, and cloud video conferencing, which also places high demands on the reliability of network video transmission.

[0003] Because the network environment is dynamic and prone to sudden bandwidth fluctuations and packet loss, traditional high-efficiency video coding technology HEVC, while reducing bitrate through inter-frame prediction during real-time compression of network video, relies primarily on block residuals and rate-distortion optimization for adjusting quantization parameters. It fails to adequately consider minor local changes in video content, such as fade-in / fade-out situations and network fluctuations during transmission. This leads to unreasonable quantization parameter settings, making it difficult to balance smooth playback and compression quality under weak network conditions, thus affecting the reliability of network video transmission and the stability of playback control. Summary of the Invention

[0004] To address the aforementioned technical issues, a highly reliable network video broadcasting and control system is provided to resolve existing problems.

[0005] The solution to the technical problem of this application is to provide a highly reliable network video broadcast control system, the system comprising:

[0006] The data acquisition module is used to collect the network bandwidth at each moment during the transmission of network video, as well as the number of videos being transmitted during each transmission period.

[0007] The parameter correction module is used to obtain the corrected quantization parameters, including;

[0008] The network video is divided into multiple video segments, and the video segments corresponding to each transmission time period are obtained; all video segments are processed by frame segmentation, and each connected component of each frame image is extracted;

[0009] Analyze the matching of any connected component in each frame with the connected components in the previous frame, obtain the matching connected component corresponding to the given connected component, determine the existence of the matching connected component corresponding to the given connected component, and determine the change factor of the given connected component; determine the contour evaluation value of the given connected component by the contour change between the given connected component and its corresponding matching connected component, and the grayscale change of the pixels at the upper boundary of the given connected component in the neighborhood.

[0010] Based on the difference in similarity between each frame and its adjacent frames, the inter-frame difference of each frame is calculated. Combined with the contour evaluation value and the change factor, the feature richness of each frame is obtained.

[0011] The changes in network bandwidth and the number of videos being transmitted under all previous transmission periods were analyzed, as well as the fluctuation of network bandwidth. The network instability of each transmission period was calculated. The element richness of all frames in the video segment was combined to determine the compression coefficient of each transmission period. The quantization parameters of the HEVC algorithm were corrected to obtain the corrected quantization parameters for each transmission period.

[0012] The compression and transmission module is used to compress and transmit network video in real time using the corrected quantization parameters and the HEVC algorithm.

[0013] Preferably, the process of obtaining the matching connected component corresponding to any connected component is as follows: matching any connected component in each frame of the image with all connected components in the previous frame of the image, and recording the connected component in the previous frame of the image that matches the any connected component as the matching connected component corresponding to the any connected component.

[0014] Preferably, determining the change factor of any connected component includes: if the any connected component exists in the previous frame image where the matched connected component exists, then the change factor of the any connected component is assigned a preset first value; otherwise, the change factor of the any connected component is assigned a preset second value, wherein the preset second value is greater than the preset first value.

[0015] Preferably, determining the contour evaluation value of any connected component includes:

[0016] Calculate the Hu moment value of each connected component and its corresponding matched connected component; denot the difference between the Hu moment values ​​of each connected component and its corresponding matched connected component as the relative difference.

[0017] Extract the edge contours of each connected component, analyze the dispersion of gray values ​​of all pixels in the local range of the edge pixels on the edge contours, and calculate the edge dispersion.

[0018] The contour evaluation value is the ratio of the relative difference to the edge dispersion.

[0019] Preferably, the calculation of edge dispersion includes:

[0020] A local window of a preset size is constructed with any edge pixel on the edge contour of each connected region as the center; the dispersion of the gray values ​​of all pixels within the local window is calculated and denoted as the local dispersion.

[0021] The edge dispersion is the mean of the local dispersion of all edge pixels on the edge contour of each connected region.

[0022] Preferably, the calculation of the inter-frame difference of each frame image includes:

[0023] Calculate the image similarity between each frame and its previous and next frames, and denot them as the first similarity and the second similarity, respectively.

[0024] The inter-frame difference is the difference between the first similarity and the second similarity.

[0025] Preferably, obtaining the feature richness of each frame of image includes:

[0026] Count the number of all connected components in each frame of the image whose change factor is a preset second value;

[0027] Calculate the average of the contour evaluation values ​​of all connected components in each frame where the change factor is a preset first value; calculate the product of the average value and the inter-frame difference.

[0028] The element richness is the sum of the product value and the number of elements.

[0029] Preferably, the calculation process for the network instability is as follows:

[0030] The network bandwidth at all times in all transmission periods preceding each transmission period is compiled into a bandwidth sequence; the number of videos being transmitted in all transmission periods preceding each transmission period is compiled into a transmission sequence; the distance between the bandwidth sequence and the transmission sequence is calculated and denoted as the trend difference; the dispersion of all elements in the bandwidth sequence is calculated and denoted as the bandwidth fluctuation.

[0031] The network instability is the product of the trend difference and the bandwidth fluctuation.

[0032] Preferably, determining the compression coefficient for each transmission period includes:

[0033] The mean of the feature richness of all frames in the video segment is used as the information importance of each transmission period.

[0034] The compression coefficient is the normalized result of the ratio of network instability to information importance.

[0035] Preferred, the first Each transmission period corresponds to the corrected quantization parameters. The calculation formula is: ,in, For the first Each transmission period corresponds to the quantization parameters before correction. For the first Compression factor for each transmission period.

[0036] This application has at least the following beneficial effects:

[0037] This application analyzes the contour changes of connected components between two adjacent frames and the blurring of connected component boundaries to calculate the contour evaluation value of each connected component in each frame. Its advantages include considering the possibility of fade-in / fade-out scenes in connected components corresponding to objects in each frame, reflecting the possibility of loss or distortion of detailed information contained in the connected component, and assessing that when compressing the connected component, the quantization parameter should be reduced to improve the compression quality of the video. Secondly, it calculates the inter-frame difference of each frame, which considers the difference in image similarity between each frame and its preceding and following frames, reflecting the possibility of unique content contained in that frame. By determining the change factor of each connected component, its advantages include considering the situation where each connected component appears suddenly in each frame, reflecting the situation where the connected component is newly added content to the image. Finally, it obtains the feature richness of each frame, which has the following advantages: The approach considers scenarios where connected components in the image exhibit fade-in / fade-out behavior, as well as the degree to which these components represent newly added content. This reflects the importance of the detailed information contained in the frame. During subsequent compression and transmission, the quality of the compressed video image should be improved to avoid over-compression and information loss. Calculating network instability at each transmission stage is beneficial because it allows for a comprehensive assessment of network instability by analyzing network bandwidth fluctuations and how these fluctuations are affected by the number of video transmissions. Determining the compression coefficient for each transmission stage and correcting the quantization parameters of the HEVC algorithm yields the corrected quantization parameters for each stage. This allows for dynamic adjustment of quantization parameters based on the importance of the video content and network instability, improving the reliability and smoothness of network video transmission in complex scenarios such as weak networks. Attached Figure Description

[0038] The following description, in conjunction with the accompanying drawings, provides a more detailed explanation of a highly reliable network video broadcasting and control system based on this application.

[0039] Figure 1 A block diagram of a highly reliable network video broadcast control system provided in one embodiment of this application;

[0040] Figure 2 This is a block diagram illustrating the implementation of a parameter correction module according to one embodiment of this application;

[0041] Figure 3 This is a flowchart illustrating the process of obtaining modified quantization parameters according to one embodiment of this application. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of a highly reliable network video broadcasting and control system, in conjunction with the accompanying drawings and implementation examples, is provided. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0044] Please see Figure 1 The diagram illustrates a block diagram of a highly reliable network video broadcast control system according to an embodiment of this application. The system includes: a data acquisition module, a parameter correction module, and a compression and transmission module.

[0045] The data acquisition module is used to collect the network bandwidth at each moment during the transmission of network video, as well as the number of videos being transmitted during each transmission period.

[0046] This embodiment takes online video for course teaching as an example to compress and transmit the online video. Due to the dynamic changes in the network environment, such as fluctuations in network bandwidth and the number of videos being transmitted, the transmission efficiency of the course teaching video will be affected. Therefore, during the transmission of online video, the network bandwidth at each moment in each transmission period and the number of videos being transmitted in each transmission period are collected in real time.

[0047] In this embodiment, the network bandwidth sampling frequency is 60Hz and the duration of one transmission period is 1 second. Therefore, the number of videos being transmitted per second is counted. As for other implementation methods, the implementer can set it according to the actual situation.

[0048] Thus, we obtain the number of videos being transmitted in each transmission period, the network bandwidth at each time, and the connected components of each frame in the course teaching video.

[0049] The parameter correction module is used to obtain the corrected quantization parameters.

[0050] Furthermore, the implementation block diagram of the parameter correction module provided in this application is as follows: Figure 2 As shown.

[0051] (1) Divide the network video into multiple video segments and obtain the video segments corresponding to each transmission time period; perform frame segmentation on all video segments and extract each connected component of each frame image; analyze the matching situation of any connected component in each frame image with the connected component in the previous frame image, obtain the matching connected component corresponding to the any connected component, determine the existence of the matching connected component corresponding to the any connected component, and determine the change factor of the any connected component; determine the contour evaluation value of the any connected component by the change of the contour between the any connected component and its corresponding matching connected component, and the gray value change of the pixel point on the upper boundary of the any connected component in the neighborhood.

[0052] High Efficiency Video Coding (HEVC), also known as H.265, is an advanced video compression standard that provides higher quality video at the same or lower bitrate through more efficient compression algorithms. In HEVC coding, the quantization parameter is a crucial parameter used to control video compression quality and bitrate. The magnitude of the quantization parameter directly affects the degree of compression and image quality. A larger quantization parameter results in a higher compression ratio, lower compressed video quality, smaller video file size, faster transmission speed, and more stable transmission. Conversely, a smaller quantization parameter results in a lower compression ratio, retains more image details, produces higher compressed video quality, but a larger video file size and slower transmission speed.

[0053] Traditional methods for video encoding typically rely on block residual analysis and rate-distortion optimization to adjust quantization parameters. However, due to local changes between video frame contents, such as the fade-in / fade-out phenomenon of objects, the area of ​​objects changes and the edges become blurred in fade-in / fade-out scenes. Block residual analysis may produce large errors, leading to unreasonable quantization parameter settings, which in turn affects the video compression and transmission control effects.

[0054] Based on the above analysis, the online video is preprocessed as follows:

[0055] The network video is divided into multiple video segments according to the duration of the transmission period, and the video segments corresponding to each transmission period are obtained.

[0056] It should be noted that the duration of the transmission period in this embodiment is 1 second, meaning that one video segment corresponds to one second.

[0057] Then, the video segment is divided into frames and grayscale to obtain each frame image, and connected component extraction is performed on each frame image to obtain each connected component of each frame image.

[0058] In this embodiment, a region growing algorithm is used to extract connected components. The region growing algorithm is a well-known technology and will not be described in detail here. As other implementation methods, implementers can use other methods of the prior art, such as the seed filling method. This embodiment does not impose any special restrictions on this. Secondly, the frame segmentation process and grayscale processing are well-known technologies and will not be described in detail here. The frame rate is set to 60fps during frame segmentation, which means 60 frames per second. Thus, each second of video segment corresponds to 60 frames of images. As other implementation methods, implementers can set this according to the actual situation.

[0059] Secondly, the smaller the grayscale difference of pixels at the upper edge of each connected component within a local region, the greater the likelihood of blurred boundaries of the connected component. Therefore, by analyzing the contour changes of corresponding connected components between two adjacent frames, as well as the grayscale changes of pixels at the upper edge of the connected components, the contour evaluation value is calculated, specifically as follows:

[0060] For any connected component in each frame of the image, match it with all connected components in the previous frame of the image. The connected component in the previous frame of the image that matches any connected component is denoted as the matching connected component corresponding to any connected component.

[0061] In this embodiment, the SIFT (Scale Invariant Feature Transform) algorithm is used for matching. The SIFT algorithm is a well-known technology and will not be described in detail here. As other implementation methods, implementers can use other methods of the prior art, such as the SURF algorithm, etc. This embodiment does not impose any special restrictions on this.

[0062] If any connected component in each frame of an image has a matching connected component in the previous frame of an image, then the change factor of any connected component is assigned a preset first value; otherwise, the change factor of any connected component is assigned a preset second value, wherein the preset second value is greater than the preset first value.

[0063] In this embodiment, the first preset value is 0 and the second preset value is 1. In other implementation methods, the implementer can set the value according to the actual situation.

[0064] It should be noted that the change factor is 1, indicating that the connected component appears suddenly and did not exist in the previous frame, thus having high importance.

[0065] Calculate the Hu moment value of each connected component and its corresponding matched connected component in each frame of the image;

[0066] It should be noted that the Hu moment value is obtained through the Hu moment function, which is a well-known technique and will not be elaborated upon here.

[0067] The difference between the Hu moment values ​​of each connected component and its corresponding matched connected component in each frame of the image is denoted as the relative difference.

[0068] In this embodiment, the absolute value of the difference between the Hu moment values ​​between each connected component and its corresponding matching connected component in each frame image is denoted as the relative difference.

[0069] It should be noted that Hu moments are a feature description method based on geometric moments, which can maintain the invariance of images under transformations such as rotation, translation, and scaling. The larger the relative difference, the more gradual the contours of the two connected components change between adjacent frames. If no gradual change occurs, the relative difference should be 0. Therefore, for each connected component in each frame that does not have a matching connected component, since the connected component is newly added and no gradual change occurs between adjacent frames, the relative difference of the connected component is set to 0, that is, the relative difference of the connected component with a change factor of 1 is assigned to 0.

[0070] Extract the edge contours of each connected component; construct a local window of a preset size centered on any edge pixel on the edge contour of each connected component;

[0071] In this embodiment, the Canny edge detection algorithm is used to detect edges and obtain edge contours. The Canny edge detection algorithm is a well-known technology and will not be described in detail here. As other implementation methods, implementers can use other methods of existing technology, such as the Sobel operator, etc. This embodiment does not impose any special restrictions on this. Secondly, the size of the local window is 5×5. As other implementation methods, implementers can set it according to the actual situation.

[0072] Calculate the degree of dispersion of the gray values ​​of all pixels within the local window of any edge pixel, and denot it as the local dispersion.

[0073] In this embodiment, the degree of dispersion is measured by calculating the standard deviation of the gray values ​​of all pixels within the local window of any edge pixel. As other implementation methods, implementers may use other methods of the prior art, such as variance, coefficient of variation, etc. This embodiment does not impose any special restrictions on this.

[0074] The mean of the local discreteness of all edge pixels on the edge contour of each connected region is denoted as the edge discreteness.

[0075] The ratio of the relative difference to the edge dispersion is used as the contour evaluation value of each connected component in each frame of the image.

[0076] It should be noted that since the relative difference of a connected component with a change factor of 1 is 0, the contour evaluation value of that connected component is also 0. The smaller the edge dispersion, the smoother the grayscale change of the edge pixels on the edge contour of the connected component, and the more blurred its contour edge. The larger the contour evaluation value, the greater the possibility that the contour of the connected component changes and the edge becomes blurred, and the more likely the fade-in and fade-out phenomenon may occur. In this case, the encoding quality of the connected component is not high, which will lead to information loss or distortion. Compared with the fixed objects in the video image, the attention of the connected component should be greater, and the quantization parameter should be smaller to ensure video quality.

[0077] At this point, the contour evaluation values ​​of each connected component in each frame of the image are obtained.

[0078] (2) Based on the difference in similarity between each frame and adjacent frames, calculate the inter-frame difference of each frame and combine the contour evaluation value and the change factor to obtain the feature richness of each frame.

[0079] Furthermore, the video transitions in the course are relatively slow to ensure that the content is easily understood by the viewer. Therefore, there are often more redundant frames or information in each frame. Thus, by analyzing the redundancy of information in each frame, appropriate quantization parameters can be set for compression encoding.

[0080] Secondly, if a frame is not a redundant frame, meaning it possesses unique content features, then the differences between adjacent frames are significant. Therefore, the inter-frame difference is calculated by assessing the difference in similarity between each frame and the frames immediately preceding and following it. Specifically:

[0081] Calculate the image similarity between each frame and its previous frame, and denote it as the first similarity.

[0082] Calculate the image similarity between each frame and the next frame, and denote it as the second similarity.

[0083] In this embodiment, the image similarity calculation uses the normalization cross-correlation (NCC) algorithm for measurement. The calculation of the normalization cross-correlation algorithm is a well-known technique and will not be described in detail here. As other implementation methods, implementers can use other methods of the prior art, such as hash similarity. As other implementation methods, implementers can set them according to the actual situation.

[0084] The difference between the first similarity and the second similarity is used as the inter-frame difference of each frame image;

[0085] In this embodiment, the absolute value of the difference between the first similarity and the second similarity is used as the inter-frame difference of each frame image.

[0086] It should be noted that the greater the inter-frame difference, the greater the difference between the video content before and after each frame.

[0087] Furthermore, if the first The frame image is a redundant frame, its first... If the frame image is not redundant, it will lead to the calculation of the first frame. The differences between frames of an image can be significant, leading to misjudgment of the first frame. The importance of the content in each frame image. Therefore, by analyzing the number of newly added connected components in each frame image, and combining the contour evaluation value and inter-frame difference, the feature richness is calculated, specifically as follows:

[0088] Count the number of all connected components in each frame of the image whose change factor is a preset second value;

[0089] Calculate the average of the contour evaluation values ​​of all connected components in each frame of the image for which the change factor is a preset first value;

[0090] Calculate the product of the average value and the inter-frame difference, and use the sum of the product and the number as the feature richness of each frame image;

[0091] It should be noted that the larger the number, the more newly added objects there are in the frame, reflecting the greater importance of the frame. The larger the average value, the greater the possibility that objects corresponding to connected components in the frame will exhibit a fade-in / fade-out phenomenon. The greater the element richness, the greater the possibility that a fade-in / fade-out phenomenon or new additions will occur in the frame, reflecting the higher the importance of the frame, the more attention it needs, and the smaller the quantization parameter should be set to ensure the clear visibility of the video image content and guarantee the quality of the video image.

[0092] Thus, the feature richness of each frame of the image is obtained.

[0093] (3) Analyze the changing trends of network bandwidth and the number of videos being transmitted in all previous transmission periods, as well as the fluctuation of network bandwidth, calculate the network instability of each transmission period, combine the element richness of all frames in the video segment, determine the compression coefficient of each transmission period, correct the quantization parameters of the HEVC algorithm, and obtain the corrected quantization parameters for each transmission period.

[0094] Furthermore, to ensure high reliability of network video transmission, the quantization parameters need to be adjusted according to network conditions. When the network conditions are good and the number of videos being transmitted is small, the quantization parameters can be appropriately reduced to ensure video quality. When the network conditions are poor and the number of videos being transmitted is large, the quantization parameters need to be appropriately increased to ensure the stability of video transmission and broadcast control.

[0095] Therefore, by analyzing the fluctuations in network bandwidth at different times during each transmission period, and the correlation between the number of videos being transmitted and network bandwidth during all previous transmission periods, and combining the feature richness of all frames of the video segment corresponding to each transmission period, the compression coefficient is calculated as follows:

[0096] The mean of the feature richness of all frames in the video segment corresponding to each transmission period is used as the information importance of each transmission period.

[0097] It should be noted that the greater the importance of the information, the more important the video segment corresponding to that transmission period is. When compressing and encoding it, a smaller quantization parameter should be set to ensure the quality of the video image.

[0098] The network bandwidth at all times in all transmission periods preceding each transmission period is used to form a bandwidth sequence;

[0099] The number of videos being transmitted in all transmission periods preceding each transmission period is used to form a transmission sequence;

[0100] Calculate the distance between the bandwidth sequence and the transmission sequence, and denot it as the trend difference;

[0101] In this embodiment, since the sequence lengths of the bandwidth sequence and the transmission sequence value are inconsistent, the distance is measured by calculating the DTW distance between the bandwidth sequence and the transmission sequence. The calculation of the DTW distance is a well-known technique and will not be described in detail here.

[0102] Calculate the degree of dispersion of all elements in the bandwidth sequence, denoted as bandwidth volatility;

[0103] In this embodiment, the degree of dispersion is calculated by calculating the variance of all elements in the bandwidth sequence. As another implementation method, the implementer may use other methods of the prior art, such as the coefficient of variation, etc. This embodiment does not impose any special restrictions on this.

[0104] The product of the trend difference and the bandwidth fluctuation is taken as the network instability for each transmission period.

[0105] The normalized result of the ratio of network instability to information importance is used as the compression coefficient for each transmission period.

[0106] In this embodiment, the tanh function is used for normalization. The tanh function is a well-known technique and will not be described in detail here. The tanh function ensures that the compression coefficient ranges within a certain range. As another implementation method, the implementer may use other methods of the prior art, such as the sigmoid function, etc., and this embodiment does not impose any special restrictions on this; it should be noted that when using the sigmoid function for normalization processing, since the value range of the sigmoid function is... The normalization result needs to be transformed so that the compression coefficient can range from [value missing]. Within this embodiment, the transformation process is as follows: assuming the ratio of the network instability to the information importance is denoted as... ,but The range of values ​​is Therefore, The result is used as the normalized compression coefficient.

[0107] It should be noted that the greater the bandwidth fluctuation, the larger the network bandwidth fluctuation during the transmission period, and the more unstable the network. The greater the trend difference, the less likely the network bandwidth fluctuation is caused by changes in the number of transmitted videos, reflecting the inherent high instability of the network bandwidth itself. The greater the network instability, the worse the network condition during the transmission period, and the quantization parameter should be appropriately increased to ensure the stability of video transmission and broadcast control. The greater the compression coefficient, the less effective information the corresponding video segment contains and the more unstable the network. In this case, the quantization parameter should be increased to improve the compression effect and ensure the stability of video transmission and broadcast control.

[0108] Furthermore, based on the compression coefficient, the quantization parameters of the HEVC algorithm are corrected, specifically as follows:

[0109] The formula for calculating the corrected quantization parameters for each transmission period is as follows:

[0110] ;

[0111] in, For the first Each transmission period corresponds to the corrected quantization parameters. For the first Each transmission period corresponds to the quantization parameters before correction. For the first Compression factor for each transmission period.

[0112] It should be noted that the block residual analysis method is used to determine the quantization parameters of the HEVC algorithm during the transmission of network video, which are then used as the quantization parameters before correction. The block residual analysis method is a well-known technique and will not be described in detail here.

[0113] The flowchart for obtaining the modified quantization parameters provided in this application is as follows: Figure 3 As shown.

[0114] Thus, the compression coefficient for each transmission period is obtained;

[0115] The compression and transmission module is used to compress and transmit network video in real time using the corrected quantization parameters and the HEVC algorithm.

[0116] Based on the corrected quantization parameters, the HEVC algorithm is used to compress and transmit network video in real time, ensuring high reliability of network video transmission and playback, and realizing control over the transmission and playback of network video.

[0117] It should be noted that the HEVC algorithm is a well-known technology and will not be elaborated upon here.

[0118] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0119] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0120] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this application, without departing from the content of the technical solution of this application, shall fall within the protection scope of the technical solution of this application.

Claims

1. A highly reliable network video broadcasting and control system, characterized in that, The system includes: The data acquisition module is used to collect the network bandwidth at each moment during the transmission of network video, as well as the number of videos being transmitted during each transmission period. The parameter correction module is used to obtain the corrected quantization parameters, including; The network video is divided into multiple video segments, and the video segments corresponding to each transmission time period are obtained; all video segments are processed by frame segmentation, and each connected component of each frame image is extracted; Analyze the matching of any connected component in each frame with the connected components in the previous frame, obtain the matching connected component corresponding to the given connected component, determine the existence of the matching connected component corresponding to the given connected component, and determine the change factor of the given connected component; determine the contour evaluation value of the given connected component by the contour change between the given connected component and its corresponding matching connected component, and the grayscale change of the pixels at the upper boundary of the given connected component in the neighborhood. Based on the difference in similarity between each frame and its adjacent frames, the inter-frame difference of each frame is calculated. Combined with the contour evaluation value and the change factor, the feature richness of each frame is obtained. The changes in network bandwidth and the number of videos being transmitted under all previous transmission periods were analyzed, as well as the fluctuation of network bandwidth. The network instability of each transmission period was calculated. The element richness of all frames in the video segment was combined to determine the compression coefficient of each transmission period. The quantization parameters of the HEVC algorithm were corrected to obtain the corrected quantization parameters for each transmission period. The compression and transmission module is used to compress and transmit network video in real time using the corrected quantization parameters and the HEVC algorithm.

2. The highly reliable network video broadcast control system as described in claim 1, characterized in that, The process of obtaining the matching connected component corresponding to any connected component is as follows: match any connected component in each frame of the image with all connected components in the previous frame of the image, and record the connected component in the previous frame of the image that matches any connected component as the matching connected component corresponding to any connected component.

3. The highly reliable network video broadcast control system as described in claim 1, characterized in that, The step of determining the change factor of any connected component includes: if any connected component exists in the previous frame of the image where the matched connected component exists, then the change factor of any connected component is assigned a preset first value; otherwise, the change factor of any connected component is assigned a preset second value, wherein the preset second value is greater than the preset first value.

4. The highly reliable network video broadcast control system as described in claim 1, characterized in that, Determining the contour evaluation value of any connected component includes: Calculate the Hu moment value of each connected component and its corresponding matched connected component; denot the difference between the Hu moment values ​​of each connected component and its corresponding matched connected component as the relative difference. Extract the edge contours of each connected component, analyze the dispersion of gray values ​​of all pixels in the local range of the edge pixels on the edge contours, and calculate the edge dispersion. The contour evaluation value is the ratio of the relative difference to the edge dispersion.

5. A highly reliable network video broadcast control system as described in claim 4, characterized in that, The calculation of edge dispersion includes: A local window of a preset size is constructed with any edge pixel on the edge contour of each connected region as the center; the dispersion of the gray values ​​of all pixels within the local window is calculated and denoted as the local dispersion. The edge dispersion is the mean of the local dispersion of all edge pixels on the edge contour of each connected region.

6. The highly reliable network video broadcast control system as described in claim 1, characterized in that, The calculation of the inter-frame difference of each image frame includes: Calculate the image similarity between each frame and its previous and next frames, and denot them as the first similarity and the second similarity, respectively. The inter-frame difference is the difference between the first similarity and the second similarity.

7. A highly reliable network video broadcast control system as described in claim 3, characterized in that, The process of obtaining the feature richness of each frame of image includes: Count the number of all connected components in each frame of the image whose change factor is a preset second value; Calculate the average of the contour evaluation values ​​of all connected components in each frame where the change factor is a preset first value; calculate the product of the average value and the inter-frame difference. The element richness is the sum of the product value and the number of elements.

8. A highly reliable network video broadcast control system as described in claim 1, characterized in that, The calculation process for the network instability is as follows: The network bandwidth at all times in all transmission periods preceding each transmission period is compiled into a bandwidth sequence; the number of videos being transmitted in all transmission periods preceding each transmission period is compiled into a transmission sequence; the distance between the bandwidth sequence and the transmission sequence is calculated and denoted as the trend difference; the dispersion of all elements in the bandwidth sequence is calculated and denoted as the bandwidth fluctuation. The network instability is the product of the trend difference and the bandwidth fluctuation.

9. A highly reliable network video broadcast control system as described in claim 1, characterized in that, Determining the compression coefficient for each transmission period includes: The mean of the feature richness of all frames in the video segment is used as the information importance of each transmission period. The compression coefficient is the normalized result of the ratio of network instability to information importance.

10. A highly reliable network video broadcast control system as described in claim 1, characterized in that, No. Each transmission period corresponds to the corrected quantization parameters. The calculation formula is: ,in, For the first Each transmission period corresponds to the quantization parameters before correction. For the first Compression factor for each transmission period.

Images