A real-time video processing method based on high concurrency scenarios

Abstract

The present invention provides a real-time video processing method based on a high-concurrency scenario, including: step S1, inputting any real-time video into a video processing module in a concurrent coordination module, performing initial identification marking, and comparing the marked item with a historical concurrent coordination database to determine the mapping between the real-time video and the historical concurrent coordination database; step S2, rendering or compressing the bullet screen of the real-time video according to the mapped historical concurrent coordination database; step S3, the real-time video after the bullet screen is rendered calculates the position where the video bullet screen appears in the video according to the content of the video and the current time, and according to the calculated bullet screen position and the status information of the existing bullet screen, the bullet screen is placed in the original video stream. The real-time video processing method of the present invention improves the viewing experience of users when watching videos, reduces the probability of video occlusion, and at the same time, reduces the occurrence of video freeze.

Description

Technical Field

[0001] The invention belongs to the field of video monitoring, and in particular relates to a real-time video processing method based on high-concurrency scenarios. Background Art

[0002] The development of various technologies has provided great convenience to people's lives. In order to meet people's life needs, it is very necessary to realize low-latency and real-time interactive video scenes in the field of video surveillance. Especially in the live broadcast process, there is often an increase in real-time barrage, which requires reasonable distribution of real-time video barrage in order to increase the system load.

[0003] At present, a large number of barrages often appear in real-time videos, resulting in a large number of occlusions in real-time videos, which greatly affects the user experience. Moreover, the barrage rendering process of real-time videos is difficult to adapt to the traffic conditions of real-time videos, which increases the video freeze phenomenon. Staff need to take the initiative to perform maintenance and processing, which consumes a lot of manpower and maintenance costs. Summary of the invention

[0004] In view of this, the present invention aims to propose a real-time video processing method based on a high-concurrency scenario to solve the problem that a large number of barrages often appear in real-time videos, resulting in a large number of occlusions in real-time videos, which greatly affects the user experience. Moreover, the barrage rendering process of real-time videos is difficult to adapt to the traffic conditions of real-time videos, which increases the video freeze phenomenon, and requires staff to actively perform maintenance and processing, which consumes a lot of manpower costs and maintenance costs.

[0005] To achieve the above object, the technical solution of the present invention is achieved as follows:

[0006] The present invention provides a real-time video processing method based on a high-concurrency scenario, comprising: step S1, inputting any real-time video into a video processing module in a concurrent coordination module, performing initial identification marking, and comparing the marked item with a historical concurrent coordination database to determine the mapping between the real-time video and the historical concurrent coordination database;

[0007] Step S2, rendering or compressing the bullet screen of the real-time video according to the mapped historical concurrent coordination database;

[0008] Step S3, the real-time video after the bullet screen rendering calculates the position where the video bullet screen appears in the video according to the content of the video and the current time, and according to the calculated bullet screen position and the status information of the existing bullet screen, the bullet screen is placed into the original video stream;

[0009] In step S2, it includes: step S201, for any real-time video mapped to the first shunt database, the real-time video is shunted to the shunt processor, and the bullet screen rendering module in the shunt processor performs hierarchical rendering according to the bullet screen situation, and then compresses it;

[0010] Step S202: for any real-time video mapped to the second traffic diversion database, video diversion warning processing is performed on the real-time video, that is, whether to perform diversion processing is determined according to the real-time traffic detection situation, and step S201 is performed on the real-time video that has been diverted, and re-judgment processing is performed on the real-time video that continues to perform video diversion warning processing;

[0011] Step S203: for any real-time video mapped to the third offload database, real-time monitoring of bullet screens and adjustment of the bullet screen coverage ratio are performed on the real-time video;

[0012] Step S204: Summarize and classify any real-time video data whose mapping information has been changed, and update the data in the historical concurrent coordination database according to the summary result.

[0013] Further, in step S1, a marking processing unit is provided in the video processing module, and the marking processing unit performs initial identification and marking on any real-time video, and the initial identification and marking includes identifying and marking the real-time video; there are three types of marking items in the initial identification and marking process, including video subject, user name, and sending place, and the marking processing unit is provided with a marking identification pointer corresponding to the marking item;

[0014] The tag identification pointer includes a subject tag identification pointer corresponding to the tag item being the video subject, a user name tag identification pointer corresponding to the tag item being the user name, and an origin tag identification pointer corresponding to the origin tag item;

[0015] After any of the real-time videos is initially identified and marked, a number of mark identification pointers generate an initial identification package, which enters the historical concurrent coordination database to determine the mapping between the real-time video and the historical concurrent coordination database.

[0016] Furthermore, the historical concurrent coordination database is provided with a diversion database after the marked items are marked, and the diversion database includes a first diversion database, a second diversion database, and a third diversion database; any diversion data diversion database is provided with a key identification pointer after the marked items are integrated again;

[0017] The key identification pointer in the first traffic diversion database has at least two marking items, the key identification pointer in the second traffic diversion database has only one marking item, and the third traffic diversion database has no key identification pointer;

[0018] The key identification pointer in the diversion database is updated according to the data storage duration in the historical concurrent coordination database.

[0019] Further, in step S201, when it is determined that the real-time video is mapped to the first shunt database, the video processing module is triggered to perform video shunt processing on it. During the video shunt processing, the bullet screen request is distributed to several shunt processors of the video shunt through the load balancer. Any shunt processor is equipped with its corresponding shunt bullet screen rendering module. Any shunt bullet screen rendering module determines to select the corresponding rendering level for rendering according to the bullet screen situation in the shunt video, and then compresses it;

[0020] The diversion barrage rendering module is provided with three rendering levels, including basic rendering, standard rendering, and advanced rendering; the basic rendering performs basic rendering for ordinary barrages in the diverted video, the standard rendering performs standard rendering for special barrages in the diverted video, and the advanced rendering performs advanced rendering for advanced barrages in the diverted video.

[0021] Further, in step S202, when it is determined that the real-time video is mapped to the second diversion database, the video processing module is triggered to perform video diversion warning processing on this. The video diversion unit is provided with a diversion warning duration. The video diversion unit in the video processing module monitors the real-time traffic within the diversion warning duration on this real-time video, and determines whether to perform diversion processing on this real-time video according to the real-time traffic;

[0022] The video splitting unit is provided with a first data flow level, a second data flow level, and a third data flow level, wherein the first data flow level is the largest, and the second data flow level is smaller than the first data flow level and larger than the third data flow level;

[0023] When the real-time traffic is greater than or equal to the first data traffic level, the mapping result is changed to the first traffic distribution database, and step S201 is performed on the real-time video;

[0024] When the real-time traffic is greater than or equal to the second data traffic level, and less than or equal to the first data traffic level, the video diversion warning processing continues, and after the main bullet screen rendering module in the main server performs bullet screen rendering, the rendering result is compressed; the main bullet screen rendering module is provided with three rendering levels including basic rendering, standard rendering, and advanced rendering; the basic rendering performs basic rendering for ordinary bullet screens in the diverted video, the standard rendering performs standard rendering for special bullet screens in the diverted video, and the advanced rendering performs advanced rendering for advanced bullet screens in the diverted video;

[0025] When the real-time traffic is less than the first data traffic level, the video diversion warning processing continues. If the result is still the same next time, the mapping result of the real-time video is changed to the third diversion database, and the real-time video is processed in step S203.

[0026] Further, in step S203, a bullet screen monitoring unit is provided in the main server for the real-time video, and any bullet screen comprehensive evaluation item of the bullet screen monitoring unit is provided with its corresponding evaluation coefficient, and the evaluation coefficient includes a first evaluation coefficient set for the number of bullet screens, a second evaluation coefficient set for the bullet screen speed, and a third evaluation coefficient set for the bullet screen density. The bullet screen monitoring unit performs a comprehensive calculation of the real-time bullet screen score based on the number of bullet screens, the bullet screen speed, the bullet screen density, and the evaluation coefficients for the bullet screen of the real-time video. The bullet screen monitoring unit adjusts the bullet screen coverage ratio of the main bullet screen rendering module provided in the main server according to the real-time bullet screen score; the bullet screen monitoring unit is provided with a first adjustment threshold and a second adjustment threshold for the bullet screen coverage ratio, and the first adjustment threshold is less than the second adjustment threshold. The main bullet screen rendering module is provided with an initial bullet screen coverage ratio and a bullet screen adjustment number;

[0027] When the real-time bullet comment score is greater than the first adjustment threshold and less than or equal to the second adjustment threshold, continue to perform bullet comment detection on the real-time video;

[0028] When the real-time barrage score is greater than the second adjustment threshold, the initial coverage ratio of the barrage will be reduced, and the real-time video will continue to be detected, and the number of barrage adjustments will be reduced. If the real-time barrage score has not been reduced to between the first adjustment threshold and the second adjustment threshold when the number of barrage adjustments is zero, the mapping result of the real-time video will be changed to the second diversion database, and step S202 will be processed.

[0029] Furthermore, the first evaluation coefficient is proportional to the number of barrages, the second evaluation coefficient is proportional to the speed of the barrage, and the third evaluation coefficient is proportional to the density of the barrage.

[0030] Furthermore, in step S204, when step S202 or step S203 is performed, data is recorded for any real-time video whose mapping relationship has been changed when its mapping relationship is changed, and the mark item corresponding to the real-time video is integrated into the historical concurrent coordination database and recorded in the diversion database after the change.

[0031] Compared with the prior art, the real-time video processing method based on high-concurrency scenarios described in the present invention has the following advantages:

[0032] (1) The present invention preliminarily marks the real-time video in any high-concurrency scenario through a marking pointer. Thus, the video is initially identified and marked. In the process of marking and identifying, facing the huge amount of data information of the real-time video, basic information is used to lock the video theme, user name, and place of origin involved in the real-time video, so as to further determine and lock more important information, thereby achieving the effect of optimizing the user experience.

[0033] (2) The present invention further determines and locks the more important information in the video subject, user name, and origin, and establishes a more refined diversion database with the relevant information to optimize the subsequent high-concurrency scenario diversion process. The key identification pointer in the diversion database is updated according to the storage duration, ensuring that the diversion database can be iteratively updated over time and can better adapt to the key information types in different time periods.

[0034] (3) In the present invention, when the real-time video is mapped to the first shunt database, the video data needs to be shunted for the real-time video to actively reduce the load of the main server. When the shunting server processes the bullet screen, different bullet screen rendering conditions are graded and the bullet screen rendering process of the real-time video is optimized to improve the video freeze phenomenon caused by a large number of shunts during the video processing process. After the bullet screen is graded and rendered, it is compressed and delivered to reduce the amount of data transmission and improve the performance and response speed of the bullet screen delivery in the real-time video.

[0035] (4) The present invention performs real-time bullet screen detection on ordinary real-time videos with few key contents, and adjusts the bullet screen coverage ratio according to the monitoring process when a large number of bullet screens may appear and cover the original video scene, thereby improving the adjustment of the bullet screen coverage by the bullet screen rendering module, enhancing the user's video viewing experience, reducing the probability of video occlusion, and at the same time, reducing the occurrence of video freezes. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0037] In the attached picture:

[0038] Figure 1 The figure is a schematic diagram of the overall flow of the real-time video processing method based on a high-concurrency scenario according to an embodiment of the present invention;

[0039] Figure 2 The figure is a schematic diagram of the judgment result branch processing in step S2 of the real-time video processing method based on a high-concurrency scenario according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0041] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and the like are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", and the like may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0042] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood by specific circumstances.

[0043] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0044] See also Figure 1-2 As shown, this embodiment provides a real-time video processing method based on a high-concurrency scenario, including: step S1, inputting any real-time video into a video processing module in a concurrent coordination module, performing initial identification marking, and comparing the marked item with a historical concurrent coordination database to determine the mapping between the real-time video and the historical concurrent coordination database;

[0045] Step S2, rendering or compressing the bullet screen of the real-time video according to the mapped historical concurrent coordination database;

[0046] Step S3, the real-time video after the bullet screen rendering calculates the position where the video bullet screen appears in the video according to the content of the video and the current time, and according to the calculated bullet screen position and the status information of the existing bullet screen, the bullet screen is placed into the original video stream;

[0047] In step S2, it includes: step S201, for any real-time video mapped to the first shunt database, the real-time video is shunted to the shunt processor, and the bullet screen rendering module in the shunt processor performs hierarchical rendering according to the bullet screen situation, and then compresses it;

[0048] Step S202: for any real-time video mapped to the second traffic diversion database, video diversion warning processing is performed on the real-time video, that is, whether to perform diversion processing is determined according to the real-time traffic detection situation, and step S201 is performed on the real-time video that has been diverted, and re-judgment processing is performed on the real-time video that continues to perform video diversion warning processing;

[0049] Step S203: for any real-time video mapped to the third offload database, real-time monitoring of bullet screens and adjustment of the bullet screen coverage ratio are performed on the real-time video;

[0050] Step S204: Summarize and classify any real-time video data whose mapping information has been changed, and update the data in the historical concurrent coordination database according to the summary result.

[0051] Specifically, in this embodiment, in step S1, a marking processing unit is provided in the video processing module, and the marking processing unit performs initial identification and marking on any real-time video, and the initial identification and marking includes identifying and marking the real-time video; there are three types of marking items in the initial identification and marking process, including video subject, user name, and place of issuance, and the marking processing unit is provided with a marking identification pointer corresponding to the marking item;

[0052] The tag identification pointer includes a subject tag identification pointer corresponding to the tag item being the video subject, a user name tag identification pointer corresponding to the tag item being the user name, and an origin tag identification pointer corresponding to the origin tag item;

[0053] After any of the real-time videos is initially identified and marked, a number of mark identification pointers generate an initial identification package, which enters the historical concurrent coordination database to determine the mapping between the real-time video and the historical concurrent coordination database.

[0054] In this embodiment, the real-time video in any high-concurrency scenario is preliminarily marked by the marking pointer. Thus, the video is initially identified and marked. In the process of marking and identifying, facing the huge data information of the real-time video, the basic information is used to lock the video theme, user name, and place of origin involved in the real-time video, so as to further determine and lock more important information, thereby achieving the effect of optimizing the user experience.

[0055] Specifically, in this embodiment, a diversion database after key marking for the marked items is provided in the historical concurrent coordination database, and the diversion database includes a first diversion database, a second diversion database, and a third diversion database; any diversion data diversion database is provided with a key identification pointer after the marked items are integrated again;

[0056] The key identification pointer in the first traffic diversion database has at least two marking items, the key identification pointer in the second traffic diversion database has only one marking item, and the third traffic diversion database has no key identification pointer;

[0057] The key identification pointer in the diversion database is updated according to the data storage duration in the historical concurrent coordination database.

[0058] In the embodiment, the more important information is further determined and locked for the video subject, user name, and place of origin, and a more refined diversion database is established for the relevant information to optimize the subsequent high-concurrency diversion process. The key identification pointer in the diversion database is updated according to the storage duration to ensure that the diversion database can be iteratively updated over time and can better adapt to the key information types in different time periods.

[0059] Specifically, in this embodiment, in step S201, when it is determined that the real-time video is mapped to the first shunt database, the video processing module is triggered to perform video shunt processing on it. During the video shunt processing, the barrage request is distributed to several shunt processors of the video shunt through the load balancer. Any shunt processor is equipped with its corresponding shunt barrage rendering module. Any shunt barrage rendering module determines to select the corresponding rendering level for rendering according to the barrage situation in its shunt video, and then compresses it;

[0060] The diversion barrage rendering module is provided with three rendering levels, including basic rendering, standard rendering, and advanced rendering; the basic rendering performs basic rendering for ordinary barrages in the diverted video, the standard rendering performs standard rendering for special barrages in the diverted video, and the advanced rendering performs advanced rendering for advanced barrages in the diverted video.

[0061] When the network request traffic in a high-concurrency scenario is large, the server's resource consumption will also increase, which will lead to a decrease in server processing capacity, thereby limiting the availability of high-level bullet screen rendering. Therefore, when the real-time video is mapped to the first diversion database, the video data needs to be diverted to the real-time video to actively reduce the load of the main server. When the diversion server processes the bullet screen, different bullet screen rendering situations are graded to optimize the bullet screen rendering process of the real-time video to improve the video stuttering caused by a large number of diversions during the video processing process. After the bullet screen is graded and rendered, it is compressed and delivered to reduce the amount of data transmission and improve the performance and response speed of the bullet screen delivery in the real-time video.

[0062] Specifically, in this embodiment, in step S202, when it is determined that the real-time video is mapped to the second shunt database, the video processing module is triggered to perform video shunt warning processing on this, and the video shunt unit is provided with a shunt warning duration. The video shunt unit in the video processing module monitors the real-time traffic within the shunt warning duration for this real-time video, and determines whether this real-time video is shunt processed according to the real-time traffic;

[0063] The video splitting unit is provided with a first data flow level, a second data flow level, and a third data flow level, wherein the first data flow level is the largest, and the second data flow level is smaller than the first data flow level and larger than the third data flow level;

[0064] When the real-time traffic is greater than or equal to the first data traffic level, the mapping result is changed to the first traffic distribution database, and step S201 is performed on the real-time video;

[0065] When the real-time traffic is greater than or equal to the second data traffic level, and less than or equal to the first data traffic level, the video diversion warning processing continues, and after the main bullet screen rendering module in the main server performs bullet screen rendering, the rendering result is compressed; the main bullet screen rendering module is provided with three rendering levels including basic rendering, standard rendering, and advanced rendering; the basic rendering performs basic rendering for ordinary bullet screens in the diverted video, the standard rendering performs standard rendering for special bullet screens in the diverted video, and the advanced rendering performs advanced rendering for advanced bullet screens in the diverted video;

[0066] When the real-time traffic is less than the first data traffic level, the video diversion warning processing continues. If the result is still the same next time, the mapping result of the real-time video is changed to the third diversion database, and the real-time video is processed in step S203.

[0067] The ordinary barrage is an ordinary text barrage, the special barrage is a functional barrage, and the advanced barrage is an advanced barrage with animation effects.

[0068] Specifically, in the present embodiment, in step S203, a bullet screen monitoring unit is provided in the main server for the real-time video, and any bullet screen comprehensive evaluation item of the bullet screen monitoring unit is provided with its corresponding evaluation coefficient, and the evaluation coefficient includes a first evaluation coefficient set for the number of bullet screens, a second evaluation coefficient set for the bullet screen speed, and a third evaluation coefficient set for the bullet screen density. The bullet screen monitoring unit performs a comprehensive calculation of the real-time bullet screen score by the number of bullet screens, the bullet screen speed, the bullet screen density, and the evaluation coefficient for the bullet screen of the real-time video. The bullet screen monitoring unit adjusts the bullet screen coverage ratio of the main bullet screen rendering module provided in the main server according to the real-time bullet screen score; the bullet screen monitoring unit is provided with a first adjustment threshold and a second adjustment threshold for the bullet screen coverage ratio, and the first adjustment threshold is less than the second adjustment threshold; the main bullet screen rendering module is provided with an initial bullet screen coverage ratio and a bullet screen adjustment number;

[0069] When the real-time bullet comment score is greater than the first adjustment threshold and less than or equal to the second adjustment threshold, continue to perform bullet comment detection on the real-time video;

[0070] When the real-time bullet screen score is greater than the second adjustment threshold, the initial bullet screen coverage ratio is reduced, and the bullet screen detection is continued for the real-time video, and the number of bullet screen adjustments is reduced;

[0071] If the real-time barrage score has not dropped to between the first adjustment threshold and the second adjustment threshold when the number of barrage adjustments is zero, the mapping result of the real-time video is changed to the second diversion database, and step S202 is performed.

[0072] Specifically, in this embodiment, the first evaluation coefficient is proportional to the number of barrages, the second evaluation coefficient is proportional to the speed of the barrage, and the third evaluation coefficient is proportional to the density of the barrage.

[0073] This embodiment performs real-time barrage detection on ordinary real-time videos with little key content, and adjusts the barrage coverage ratio according to the monitoring process for situations where a large number of barrages may appear and cover the original video scene, thereby improving the barrage rendering module's adjustment of the barrage coverage situation, enhancing the user's video viewing experience, reducing the probability of video occlusion, and at the same time, reducing the occurrence of video freezes.

[0074] Specifically, in this embodiment, in step S204, when step S202 or step S203 is performed, data is recorded for any real-time video whose mapping relationship has been changed when its mapping relationship is changed, and the mark item corresponding to the real-time video is integrated into the historical concurrent coordination database and recorded in the diversion database after the change.

[0075] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A real-time video processing method based on a high-concurrency scenario, characterized in that: The method comprises: step S1, inputting any real-time video into the video processing module in the concurrent coordination module, performing initial identification marking, and comparing the marked item with the historical concurrent coordination database to determine the mapping between the real-time video and the historical concurrent coordination database; Step S2, rendering or compressing the bullet screen of the real-time video according to the mapped historical concurrent coordination database; Step S3, the real-time video after the bullet screen rendering calculates the position where the video bullet screen appears in the video according to the content of the video and the current time, and according to the calculated bullet screen position and the status information of the existing bullet screen, the bullet screen is placed into the original video stream; In step S2, it includes: step S201, for any real-time video mapped to the first shunt database, the real-time video is shunted to the shunt processor, and the bullet screen rendering module in the shunt processor performs hierarchical rendering according to the bullet screen situation, and then compresses it; Step S202: for any real-time video mapped to the second traffic diversion database, video diversion warning processing is performed on the real-time video, that is, whether to perform diversion processing is determined according to the real-time traffic detection situation, and step S201 is performed on the real-time video that has been diverted, and re-judgment processing is performed on the real-time video that continues to perform video diversion warning processing; Step S203: for any real-time video mapped to the third offload database, real-time monitoring of bullet screens and adjustment of the bullet screen coverage ratio are performed on the real-time video; Step S204: Summarize and classify any real-time video data whose mapping information has been changed, and update the data in the historical concurrent coordination database according to the summary result.

2. The real-time video processing method based on high concurrency scenarios according to claim 1 is characterized in that: In step S1, a marking processing unit is provided in the video processing module, and the marking processing unit performs initial identification and marking on any real-time video, and the initial identification and marking includes identifying and marking the real-time video; there are three types of marking items in the initial identification and marking process, including video subject, user name, and sending place, and the marking processing unit is provided with a marking identification pointer corresponding to the marking item; The tag identification pointer includes a subject tag identification pointer corresponding to the tag item being the video subject, a user name tag identification pointer corresponding to the tag item being the user name, and an origin tag identification pointer corresponding to the origin tag item; After any of the real-time videos is initially identified and marked, a number of mark identification pointers generate an initial identification package, which enters the historical concurrent coordination database to determine the mapping between the real-time video and the historical concurrent coordination database.

3. The real-time video processing method based on high concurrency scenarios according to claim 2 is characterized in that: The historical concurrent coordination database is provided with a diversion database after the marked items are marked, and the diversion database includes a first diversion database, a second diversion database, and a third diversion database; any diversion data diversion database is provided with a key identification pointer after the marked items are integrated again; The key identification pointer in the first traffic diversion database has at least two marking items, the key identification pointer in the second traffic diversion database has only one marking item, and the third traffic diversion database has no key identification pointer; The key identification pointer in the diversion database is updated according to the data storage duration in the historical concurrent coordination database.

4. The real-time video processing method based on high concurrency scenarios according to claim 3 is characterized in that: In step S201, when it is determined that the real-time video is mapped to the first shunt database, the video processing module is triggered to perform video shunt processing on it. During the video shunt processing, the bullet screen request is distributed to several shunt processors of the video shunt through the load balancer. Any shunt processor is equipped with its corresponding shunt bullet screen rendering module. Any shunt bullet screen rendering module determines to select the corresponding rendering level for rendering according to the bullet screen situation in its shunt video, and then compresses it; The diversion barrage rendering module is provided with three rendering levels, including basic rendering, standard rendering, and advanced rendering; the basic rendering performs basic rendering for ordinary barrages in the diverted video, the standard rendering performs standard rendering for special barrages in the diverted video, and the advanced rendering performs advanced rendering for advanced barrages in the diverted video.

5. The real-time video processing method based on high-concurrency scenarios according to claim 4 is characterized in that: In step S202, when it is determined that the real-time video is mapped to the second diversion database, the video processing module is triggered to perform video diversion warning processing on this. The video diversion unit is provided with a diversion warning duration. The video diversion unit in the video processing module monitors the real-time traffic within the diversion warning duration for this real-time video, and determines whether to perform diversion processing on this real-time video according to the real-time traffic; The video splitting unit is provided with a first data flow level, a second data flow level, and a third data flow level, wherein the first data flow level is the largest, and the second data flow level is smaller than the first data flow level and larger than the third data flow level; When the real-time traffic is greater than or equal to the first data traffic level, the mapping result is changed to the first traffic distribution database, and step S201 is performed on the real-time video; When the real-time traffic is greater than or equal to the second data traffic level, and less than or equal to the first data traffic level, the video diversion warning process continues, and after the main bullet screen rendering module in the main server renders the bullet screen, the rendering result is compressed; The main bullet screen rendering module is provided with three rendering levels, including basic rendering, standard rendering, and advanced rendering; the basic rendering performs basic rendering for ordinary bullet screens in the diverted video, the standard rendering performs standard rendering for special bullet screens in the diverted video, and the advanced rendering performs advanced rendering for advanced bullet screens in the diverted video; When the real-time traffic is less than the first data traffic level, the video diversion warning processing continues. If the result is still the same next time, the mapping result of the real-time video is changed to the third diversion database, and the real-time video is processed in step S203.

6. The real-time video processing method based on high-concurrency scenarios according to claim 1 is characterized in that: In step S203, a bullet screen monitoring unit is provided in the main server for the real-time video, and any bullet screen comprehensive evaluation item of the bullet screen monitoring unit is provided with its corresponding evaluation coefficient, and the evaluation coefficient includes a first evaluation coefficient set for the number of bullet screens, a second evaluation coefficient set for the bullet screen speed, and a third evaluation coefficient set for the bullet screen density. The bullet screen monitoring unit calculates a real-time bullet screen score by comprehensively calculating the number of bullet screens, the bullet screen speed, the bullet screen density, and the evaluation coefficients for the bullet screen of the real-time video. The bullet screen monitoring unit adjusts the bullet screen coverage ratio of the main bullet screen rendering module provided in the main server according to the real-time bullet screen score; the bullet screen monitoring unit is provided with a first adjustment threshold and a second adjustment threshold for the bullet screen coverage ratio, and the first adjustment threshold is less than the second adjustment threshold; the main bullet screen rendering module is provided with an initial bullet screen coverage ratio and a bullet screen adjustment number; When the real-time bullet comment score is greater than the first adjustment threshold and less than or equal to the second adjustment threshold, continue to perform bullet comment detection on the real-time video; When the real-time bullet screen score is greater than the second adjustment threshold, the initial bullet screen coverage ratio is reduced, and the real-time video bullet screen detection is continued, and the number of bullet screen adjustments is reduced by one; If the real-time barrage score has not dropped to between the first adjustment threshold and the second adjustment threshold when the number of barrage adjustments is zero, the mapping result of the real-time video is changed to the second diversion database, and step S202 is performed.

7. The real-time video processing method based on high-concurrency scenarios according to claim 6 is characterized in that: The first evaluation coefficient is proportional to the number of barrages, the second evaluation coefficient is proportional to the speed of the barrages, and the third evaluation coefficient is proportional to the density of the barrages.

8. The real-time video processing method based on high-concurrency scenarios according to any one of claims 5-6, characterized in that: In step S204, when step S202 or step S203 is performed, data is recorded for any real-time video whose mapping relationship has been changed when its mapping relationship is changed, and the mark item corresponding to the real-time video is integrated into the historical concurrent coordination database and recorded in the diversion database after the change.

Images