Intelligent conference management system and intelligent conference management method

Abstract

The invention discloses an intelligent conference management system and an intelligent conference management method, and relates to the technical field of video conferences, and the intelligent conference management system comprises the following steps: detecting whether the flow consumption generated by a current video conference exceeds a preset flow consumption threshold in real time; if yes, sending a video compression request to a video compression module; when receiving a video compression request, the video compression module obtains character segmentation images in continuous video frame images in a current video conference based on a semantic segmentation technology, and generates a target compressed video based on the character segmentation images; the video decompression module decompresses the target compressed video to obtain a decompressed video frame image; and generating and displaying a target video frame image based on the decompressed video frame image. By applying the technical scheme, the technical problems that the video quality is reduced and the stability and the response speed of the video conference are influenced when the network quality of the online video conference is poor in the prior art are solved.

Description

Technical Field

[0001] This application relates to the technical field of video conferencing, and particularly to an intelligent conference management system and an intelligent conference management method. Background Art

[0002] With the rapid development of Internet technology, users can hold online meetings by using Internet audio and video conferencing systems. For example, they can access the video conferencing system through mobile terminal software to hold remote meetings with others. Among them, remote video conferencing can usually be applied to a conference management system based on a computer and a network. In this conference management system, when a user turns on the camera, the user's video data can be collected and uploaded to the server. The server then forwards the user's video data to other participants, so that other participants can see the user's video image. At the same time, the user can receive the video data of other participants returned by the server and display the video images of other participants.

[0003] Currently, when conducting an online video conference based on a conference management system, the quality of the conference usually depends on the current network quality of the terminal where the video conference is held. Therefore, if the network quality of the terminal deteriorates during the user's online video conference, it will lead to low video quality and affect the stability and response speed of the video conference at the same time. Summary of the Invention

[0004] The main purpose of this application is to provide an intelligent conference management system and an intelligent conference management method, aiming to solve the technical problem that the video quality will be reduced when the network quality of the existing technology in an online video conference is poor, and at the same time, it will affect the stability and response speed of the video conference.

[0005] To achieve the above object, this application proposes an intelligent conference management system, and the system includes: a conference management module, a video compression module, and a video decompression module;

[0006] The conference management module is used to detect in real time whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold;

[0007] The conference management module is further used to, if so, send a video compression request to the video compression module;

[0008] The video compression module is used to, when receiving the video compression request, obtain a person segmentation image in the continuous video frame images in the current video conference based on semantic segmentation technology, and generate a target compressed video based on the person segmentation image;

[0009] The video decompression module is used to decompress the target compressed video to obtain decompressed video frame images;

[0010] The video decompression module is further configured to generate and display a target video frame image based on the decompressed video frame image.

[0011] In one embodiment, the video compression module is configured to, when receiving the video compression request, obtain in real time consecutive video frame images in the current video conference;

[0012] The video compression module is further configured to obtain a person segmentation image and a background segmentation image in the video frame image based on semantic segmentation technology;

[0013] The video compression module is further configured to determine a person key area image in the person segmentation image, and determine a person non-key area image based on the person key area image;

[0014] The video compression module is further configured to generate a video to be compressed based on the person key area image;

[0015] The video compression module is further configured to compress the video to be compressed to generate a target compressed video;

[0016] The video decompression module is further configured to generate and display a target video frame image based on the decompressed video frame image, the background segmentation image, and the person non-key area image.

[0017] In one embodiment, the video decompression module is further configured to search in a preset virtual background image database for a background image to be replaced corresponding to the background segmentation image;

[0018] The video decompression module is further configured to match the person non-key area image with all images in a preset non-key area image database to obtain a non-key area image to be replaced;

[0019] The video decompression module is further configured to generate and display a target video frame image based on the decompressed video frame image, the background image to be replaced, and the non-key area image to be replaced.

[0020] In one embodiment, the video decompression module 30 is further configured to obtain a decompressed person key area image based on the decompressed video frame image;

[0021] The video decompression module is further configured to merge the background image to be replaced, the non-key area image to be replaced, and the decompressed person key area image to generate a target video frame image;

[0022] The video decompression module is further configured to display the target video frame image in the current video conference.

[0023] In one embodiment, the system further includes: a meeting creation module;

[0024] The meeting creation module is configured to, when receiving a meeting creation request, determine the meeting start time and the meeting end time of the meeting to be created based on the meeting creation request;

[0025] The meeting creation module is further configured to predict the traffic to be consumed by the current participating terminal based on the meeting start time and the meeting end time;

[0026] The meeting creation module is further configured to determine the total traffic consumption of the current participating terminal in the current video conference;

[0027] The meeting creation module is further configured to determine the remaining traffic to be reserved for the current participating terminal based on the total traffic consumption;

[0028] The meeting creation module is further configured to generate a meeting creation result corresponding to the meeting to be created based on the traffic to be consumed and the remaining traffic to be reserved.

[0029] In one embodiment, when the remaining traffic to be reserved is greater than the traffic to be consumed, the meeting creation module is further configured to generate and display an entry link corresponding to the meeting to be created based on the meeting start time and the meeting end time;

[0030] When the remaining traffic to be reserved is less than the traffic to be consumed, the meeting creation module is further configured to determine a target participating terminal from the remaining participating terminals in the current video conference;

[0031] The meeting creation module is further configured to send the meeting creation request to the target participating terminal, so that the target participating terminal generates and displays an entry link corresponding to the meeting to be created based on the meeting start time and the meeting end time.

[0032] In addition, to achieve the above object, the present application further proposes an intelligent conference management method based on the intelligent conference management system described above, and the method includes:

[0033] Real-time detection of whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold;

[0034] If so, obtaining a person segmentation image in consecutive video frame images in the current video conference based on semantic segmentation technology, and generating a target compressed video based on the person segmentation image;

[0035] Decompressing the target compressed video to obtain decompressed video frame images;

[0036] Generating and displaying target video frame images based on the decompressed video frame images.

[0037] In one embodiment, the steps of obtaining a person segmentation image from consecutive video frame images in the current video conference based on semantic segmentation technology and generating a target compressed video based on the person segmentation image include:

[0038] Obtain consecutive video frame images in the current video conference in real time;

[0039] Obtain a person segmentation image and a background segmentation image from the video frame images based on semantic segmentation technology;

[0040] Determine the key region image of the person in the person segmentation image, and determine the non-key region image of the person based on the key region image of the person;

[0041] Generate a video to be compressed based on the key region image of the person;

[0042] Compress the video to be compressed to generate a target compressed video;

[0043] The steps of generating and displaying a target video frame image based on the decompressed video frame image include:

[0044] Generate and display a target video frame image based on the decompressed video frame image, the background segmentation image, and the non-key region image of the person.

[0045] In one embodiment, the steps of generating and displaying a target video frame image based on the decompressed video frame image, the background segmentation image, and the non-key region image of the person include:

[0046] Search for a background image to be replaced corresponding to the background segmentation image in a preset virtual background image database;

[0047] Match the non-key region image of the person with all the images in a preset non-key region image database to obtain a non-key region image to be replaced;

[0048] Generate and display a target video frame image based on the decompressed video frame image, the background image to be replaced, and the non-key region image to be replaced.

[0049] In one embodiment, the steps of generating and displaying a target video frame image based on the decompressed video frame image, the background image to be replaced, and the non-key region image to be replaced include:

[0050] Obtain a decompressed key region image of the person based on the decompressed video frame image;

[0051] Merge the to-be-replaced background image, the to-be-replaced non-critical region image, and the decompressed key region image of the person to generate a target video frame image;

[0052] Display the target video frame image in the current video conference.

[0053] This application provides an intelligent conference management system. This application discloses that the conference management module real-time detects whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold; if so, it sends a video compression request to the video compression module; when the video compression module receives the video compression request, it obtains the person segmentation image in the continuous video frame images in the current video conference based on semantic segmentation technology, and generates a target compressed video based on the person segmentation image; the video decompression module decompresses the target compressed video to obtain a decompressed video frame image; generates and displays a target video frame image based on the decompressed video frame image; since when the traffic consumption generated by the current video conference in the present invention exceeds the preset traffic consumption threshold, it obtains the person segmentation image in the continuous video frame images in the current video conference based on semantic segmentation technology to generate a target compressed video, and displays it after decompressing the target compressed video, thus solving the technical problem that when the network quality of the online video conference in the prior art is poor, it will lead to a decrease in video quality, and at the same time affect the stability and response speed of the video conference. Description of the Drawings

[0054] The drawings here are incorporated into the specification and form a part of this specification, showing embodiments consistent with this application, and are used together with the specification to explain the principles of this application.

[0055] To more clearly illustrate the technical solutions in the embodiments of this application or in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0056] Figure 1 It is a structural block diagram provided for the first embodiment of the intelligent conference management system of this application;

[0057] Figure 2 It is a structural block diagram provided for the second embodiment of the intelligent conference management system of this application;

[0058] Figure 3 It is a flowchart provided for the first embodiment of the intelligent conference management method based on the above intelligent conference management system of this application;

[0059] Figure 4 It is the overall flowchart of the intelligent conference management method based on the above intelligent conference management system of this application.

[0060] The realization of the purpose, functional features and advantages of this application will be further described in combination with embodiments with reference to the accompanying drawings. Specific embodiments

[0061] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of this application and are not used to limit this application.

[0062] To better understand the technical solutions of this application, the following will be described in detail in combination with the accompanying drawings of the specification and specific embodiments.

[0063] The main solution of the embodiment of this application is: a conference management module, a video compression module and a video decompression module are provided in the system; the conference management module is used to detect in real time whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold; the conference management module is also used to, if so, send a video compression request to the video compression module; the video compression module is used to, when receiving the video compression request, obtain the person segmentation image in the continuous video frame images in the current video conference based on semantic segmentation technology, and generate a target compressed video based on the person segmentation image; the video decompression module is used to decompress the target compressed video to obtain decompressed video frame images; the video decompression module is also used to generate and display target video frame images based on the decompressed video frame images.

[0064] Since in the prior art when conducting an online video conference based on a conference management system, the quality of the conference usually depends on the current network quality of the terminal where the video conference is conducted. If the network quality is poor during the user's online video conference, it will result in low video quality and at the same time affect the stability and response speed of the video conference.

[0065] This application provides a solution. When the traffic consumption generated by the current video conference exceeds the preset traffic consumption threshold, it can obtain the person segmentation image in the continuous video frame images in the current video conference based on semantic segmentation technology to generate a target compressed video, and display it after decompressing the target compressed video, thus solving the technical problem that in the prior art, when the network quality of an online video conference is poor, it will lead to a decrease in video quality and at the same time affect the stability and response speed of the video conference.

[0066] Refer to Figure 1 , Figure 1 It is the structural block diagram provided by Embodiment 1 of the intelligent conference management system of this application.

[0067] As Figure 1 shown, the intelligent conference management system of this embodiment includes: a conference management module 10, a video compression module 20 and a video decompression module 30.

[0068] The conference management module 10 is used to detect in real time whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold.

[0069] It can be understood that the above-mentioned current video conference can be the currently ongoing video conference; the above-mentioned traffic consumption can be the traffic consumed by transmitting video, audio and other data through the network during the development of the video conference.

[0070] It should be understood that the above-mentioned preset traffic consumption threshold can be the maximum traffic value that the current video conference is allowed to consume of the participating terminals. In this embodiment, the preset traffic threshold can be set according to the actual situation. For example, the preset traffic threshold is set to 1 GB (Gigabyte).

[0071] The conference management module 10 is also used to, if so, send a video compression request to the video compression module 20.

[0072] It should be noted that the above-mentioned video compression request can be a request for instructing the video compression module 20 to compress the video file. Among them, the video compression module 20 can reduce the number of bytes of the video file through a specific compression technology, so as to save storage space and speed up the transmission speed, thereby improving the quality of the video conference.

[0073] In this embodiment, when the user's mobile terminal conducts a video conference, corresponding traffic consumption will be generated. If the traffic consumption of the terminal is too much, it will cause excessive network pressure, thereby affecting the quality of the video conference. Therefore, in this embodiment, the conference management module 10 can obtain the traffic consumption generated by the current video conference in real time, and detect whether the traffic consumption generated by the current video conference exceeds the maximum traffic value allowed to be consumed. If the current traffic consumption exceeds the maximum traffic value allowed to be consumed, a video compression request can be sent to the video compression module 20 to instruct the video compression module 20 to compress the video stream of the current video conference, and while ensuring the video quality, reduce the network pressure, thereby improving the stability of the video conference.

[0074] The video compression module 20 is used to, when receiving the video compression request, obtain the person segmentation image in the continuous video frame images of the current video conference based on the semantic segmentation technology, and generate a target compressed video based on the person segmentation image.

[0075] It should be understood that the above-mentioned video frame images can be a series of static images continuously transmitted by the current video conference captured by the camera. In practical applications, these images can be quickly displayed at a certain frame rate (such as 24 frames per second, 30 frames per second, etc.), so as to form a dynamic and continuous video picture at the receiving end. Among them, the higher the frame rate, the smoother the action during video playback.

[0076] It should be noted that the above semantic segmentation technology can be a technology for recognizing and segmenting objects in an image by classifying each pixel in the image. Correspondingly, the above person segmentation image can be an image obtained by recognizing and segmenting the person area in the video frame image through the semantic segmentation technology, including the head area, limb area, clothing occlusion area, etc. of the person.

[0077] It should be noted that the above target compressed video can be a video formed by compressing partial area images of the person segmentation images in consecutive video frame images. In practical applications, the video compression module 20 can compress the person area in the person segmentation image of the video frame image, and finally form the target compressed video based on the compressed image. In this embodiment, the video compression module 20 can recognize the person area in the video frame image based on the semantic segmentation technology, and perform targeted compression processing on the person area to obtain the target compressed image, so as to relieve the network pressure while ensuring the video quality, which is beneficial to maintaining the stability of the video conference.

[0078] In this embodiment, the video compression module 20 is configured to, when receiving the video compression request, obtain consecutive video frame images in the current video conference in real time.

[0079] The video compression module 20 is further configured to obtain the person segmentation image and the background segmentation image in the video frame image based on the semantic segmentation technology.

[0080] It can be understood that the above background segmentation image can be an image obtained by recognizing and segmenting the background area in the video frame image through the semantic segmentation technology. In this embodiment, the background segmentation image is also an image composed of the areas other than the person area in the video frame image.

[0081] The video compression module 20 is further configured to determine the person key area image in the person segmentation image, and determine the person non-key area image based on the person key area image.

[0082] It should be noted that the above-mentioned key region image of the person can be an image composed of the head region and the limb region of the person in the person segmentation image; correspondingly, the above-mentioned non-key region image of the person can be an image composed of the clothing occlusion region of the person in the person segmentation image. In this embodiment, the video compression module 20 can use a human key point tracking algorithm (such as AlphaPose) to identify each joint of the human body in the person segmentation image, such as facial regions and limb regions like the nose, eyes, shoulders, wrists, etc., and then determine these regions as the key regions of the person in the person segmentation image, and form a key region image of the person from these regions. Then, the video compression module 20 can remove the key region image of the person in the person segmentation image, so as to obtain a non-key region image of the person.

[0083] The video compression module 20 is further configured to generate a video to be compressed based on the key region image of the person.

[0084] It should be understood that the above-mentioned video to be compressed is the video that the video compression module 20 needs to compress.

[0085] The video compression module 20 is further configured to compress the video to be compressed to generate a target compressed video.

[0086] In this embodiment, the video compression module 20 can determine the video formed by the key region images of the human body in the consecutive video frame images in the current video conference as the video to be compressed, and compress the video to be compressed to generate a target compressed video.

[0087] The video decompression module 30 is configured to decompress the target compressed video to obtain decompressed video frame images.

[0088] It can be understood that the above-mentioned decompressed video frame images can be consecutive frame images obtained after decompressing the target compressed video. In this embodiment, the decompressed video frame images can be the key region images of the human body in the video frame images in the current video conference obtained by the video decompression module 30 after decompressing the target compressed video.

[0089] The video decompression module 30 is further configured to generate and display target video frame images based on the decompressed video frame images.

[0090] Specifically, the video decompression module 30 is further configured to generate and display target video frame images based on the decompressed video frame images, the background segmentation image, and the non-key region image of the person.

[0091] In practical applications, after the video compression module 20 generates the target compressed video, it can send the target compressed video to the video decompression module 30. After receiving the target compressed video, the video decompression module 30 can decompress the target compressed video to obtain the decompressed video frame images. Finally, it can generate the target video frame images based on the background segmentation images and the non-critical region images of the characters in the decompressed video frame images, and display the target video frame images.

[0092] Further, the video decompression module 30 is further configured to search for the background image to be replaced corresponding to the background segmentation image from a preset virtual background image database; match the non-critical region image of the character with all the images in the preset non-critical region image database to obtain the non-critical region image to be replaced; generate and display the target video frame images based on the decompressed video frame images, the background image to be replaced, and the non-critical region image to be replaced.

[0093] It should be noted that the above-mentioned preset virtual background image database can be a database storing several low-pixel virtual background images. Correspondingly, the above-mentioned background image to be replaced can be the virtual background image in the preset virtual background image database used to replace the background segmentation image.

[0094] It should be noted that the above-mentioned preset non-critical region image database can be a database storing several low-pixel images of the clothing occlusion regions of the human body. Correspondingly, the above-mentioned non-critical region image to be replaced can be the image in the preset non-critical region image database used to replace the non-critical region image of the character.

[0095] In practical applications, since the background image and the human body clothing occlusion area in the video frame image of the current video conference require a large amount of video memory, the traffic consumed during the video conference is relatively large, which will further cause greater network pressure. Therefore, in this embodiment, a preset virtual background image database can be established in advance based on several blank backgrounds of different sizes with low pixels, and at the same time, a preset non-critical area image database can be established based on several images of human body clothing occlusion areas with different postures and low pixels. Subsequently, during the video conference, if there is a video compression requirement, the video compression module 20 can send the generated target compressed image, the background segmentation image in the video frame image, and the non-critical area image of the person in the person segmentation image to the video decompression module 30, so that after receiving these images from the video compression module 20, the video decompression module 30 can search the preset virtual background image database for a background image with the same image size as it according to the image size of the background segmentation image, and determine the background image to be replaced. At the same time, the video decompression module 30 can identify the posture information of the human body in the image according to the non-critical area image of the person, and match the posture information of the human body with all the images in the preset non-critical area image database to finally determine the non-critical area image to be replaced. Then, a target video frame image can be generated and displayed based on the decompressed video frame image, the background image to be replaced, and the non-critical area image to be replaced.

[0096] Specifically, the video decompression module 30 is further configured to obtain a decompressed person key area image based on the decompressed video frame image; merge the background image to be replaced, the non-critical area image to be replaced, and the decompressed person key area image to generate a target video frame image; and display the target video frame image in the current video conference.

[0097] It can be understood that the above decompressed person key area image can be the person key area image in the decompressed video frame image of the current video conference. In this embodiment, the video compression module 20 sends the target compressed video obtained by compressing the person key area image to the video decompression module 30, and then the video decompression module 30 can decompress the target compressed video to obtain a plurality of consecutive decompressed video frame images, and determine the decompressed person key area image from the decompressed video frame images. Then, the video decompression module 30 can merge the background image to be replaced, the non-critical area image to be replaced, and the decompressed person key area image to generate a target video frame image, and finally display the target video frame image in the current video conference.

[0098] In this embodiment, it is disclosed that the conference management module real-time detects whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold; if so, it sends a video compression request to the video compression module; when the video compression module receives the video compression request, it obtains the person segmentation images in the continuous video frame images of the current video conference based on semantic segmentation technology, and generates a target compressed video based on the person segmentation images; the video decompression module decompresses the target compressed video to obtain decompressed video frame images; and generates and displays target video frame images based on the decompressed video frame images; since in this embodiment, when the traffic consumption generated by the current video conference exceeds the preset traffic consumption threshold, it obtains the person segmentation images in the continuous video frame images of the current video conference based on semantic segmentation technology to generate a target compressed video, and displays it after decompressing the target compressed video, thus solving the technical problem in the prior art that when the network quality of an online video conference is poor, it will cause the video quality to decrease, and at the same time affect the stability and response speed of the video conference.

[0099] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar content as that in the above-mentioned first embodiment can be referred to the above introduction and will not be repeated hereinafter. On this basis, please refer to Figure 2 , Figure 2 which is the structural block diagram provided by the second embodiment of the intelligent conference management system of the present application.

[0100] Based on the first embodiment of the above-mentioned intelligent conference management system, the second embodiment of the present invention based on the intelligent conference management system is proposed.

[0101] As Figure 2 shown, the system further includes: a conference creation module 40.

[0102] The conference creation module 40 is configured to, when receiving a conference creation request, determine the conference start time and conference end time of the conference to be created based on the conference creation request.

[0103] It can be understood that the above-mentioned conference creation request may be a request for instructing the conference creation module 40 to create a new video conference. The above-mentioned conference start time and conference end time may be the start time and end time of the video conference created this time, respectively.

[0104] In this embodiment, if a user needs to create a new video conference, a conference creation request can be generated in the system based on the start time and end time of the conference. After the conference creation module 40 in the system receives the conference creation request, it can parse the conference creation request to determine the conference start time and conference end time of the conference to be created.

[0105] The conference creation module 40 is further configured to predict the traffic to be consumed by the current participating terminal based on the conference start time and the conference end time.

[0106] It should be understood that the above-mentioned current parameter terminal can be the terminal that initiates the video conference this time; the traffic to be consumed can be the traffic that the video conference created this time needs to consume from the start to the end of the current parameter terminal. In practical applications, conducting a video conference on a terminal usually consumes the traffic of the terminal. Among them, the traffic consumed by each video conference is usually related to the duration of the video conference. In this embodiment, the conference creation module 40 can determine the target duration of the video conference based on the start time and end time of the video conference created this time, and predict the traffic that the current parameter terminal needs to consume to conduct this video conference based on the target duration, so as to obtain the above-mentioned traffic to be consumed. In addition, the traffic consumption of the video conference is also related to information such as video content and the number of participants. Therefore, the conference creation module 40 can also predict the traffic that the current parameter terminal needs to consume to conduct this video conference based on the target duration, video content and the number of participants of the video conference to be created, and then determine the traffic to be consumed by the current participating terminal.

[0107] The conference creation module 40 is further configured to determine the total traffic consumption of the current participating terminal in the current video conference.

[0108] It should be noted that the above-mentioned total traffic consumption can be the total traffic that the current video conference needs to consume from the start to the end of the current parameter terminal. In this embodiment, the conference creation module 40 can first obtain the current traffic consumption of the participating terminal at the current moment in the current video conference, and determine the current duration of the current video conference based on the start time of the current video conference, so as to determine the average traffic consumed by the participating terminal per minute in the current video conference based on the current traffic consumption and the current duration, and predict the total traffic consumption of the current participating terminal in the current video conference based on the average traffic and the end time of the current video conference.

[0109] The conference creation module 40 is further configured to determine the remaining traffic to be consumed corresponding to the current participating terminal based on the total traffic consumption.

[0110] It should be understood that the above-mentioned remaining traffic to be consumed can be the remaining traffic of the current parameter terminal after completing the current video conference. In this embodiment, the conference creation module 40 can obtain all the available total traffic of the current participating terminal before conducting the current video conference, and determine the remaining traffic of the current participating terminal after completing the current video conference based on the available total traffic and the total traffic consumption in the current video conference, so as to obtain the above-mentioned remaining traffic to be consumed.

[0111] The conference creation module 40 is further configured to generate a conference creation result corresponding to the to-be-created conference based on the to-be-consumed traffic and the to-be-remaining traffic.

[0112] It can be understood that the above conference creation result can be a result used to represent successful or failed conference creation.

[0113] Furthermore, the conference creation module 40 is further configured to, when the to-be-remaining traffic is greater than the to-be-consumed traffic, generate and display an access link corresponding to the to-be-created conference based on the conference start time and the conference end time; when the to-be-remaining traffic is less than the to-be-consumed traffic, determine a target participating terminal from the remaining participating terminals of the current video conference; and send the conference creation request to the target participating terminal, so that the target participating terminal generates and displays an access link corresponding to the to-be-created conference based on the conference start time and the conference end time.

[0114] It should be understood that the above access link is the only website address used to join the to-be-created conference.

[0115] It should be noted that the above remaining participating terminals can be all terminals other than the current participating terminal in the current video conference; correspondingly, the above target participating terminal can be a terminal among the remaining participating terminals whose remaining traffic can meet the traffic requirements of the to-be-created video conference and allows hosting the conference. In this embodiment, if the to-be-remaining traffic of the current participating terminal is greater than the to-be-consumed traffic, it means that the remaining traffic of the current participating terminal can support the development of the to-be-created video conference. At this time, an access link for the to-be-created conference can be generated and displayed, so that other participating terminals can join the new conference through this access link after the current video conference ends; if the to-be-remaining traffic of the current participating terminal is less than the to-be-consumed traffic, it means that the remaining traffic of the current participating terminal is not enough to support the development of the to-be-created video conference. Therefore, to avoid additional costs, at this time, the conference creation module 40 can determine, from all terminals other than the current participating terminal in the current video conference, terminals whose remaining traffic can meet the traffic requirements of the to-be-created video conference and allow hosting the conference, and determine these terminals as target participating terminals. Then, the conference creation request can be sent to these target participating terminals. At this time, the first target participating terminal that receives the conference creation request can generate and display an access link for the to-be-created conference based on the start time and end time of the to-be-created conference.

[0116] In this embodiment, it is disclosed that when the meeting creation module receives a meeting creation request, it determines the meeting start time and meeting end time of the meeting to be created based on the meeting creation request; predicts the traffic to be consumed by the current participating terminal based on the meeting start time and meeting end time; determines the total traffic consumption of the current participating terminal in the current video conference; determines the remaining traffic to be corresponding to the current participating terminal based on the total traffic consumption; and generates a meeting creation result corresponding to the meeting to be created based on the traffic to be consumed and the remaining traffic to be, so that a new meeting can be reasonably created based on the user's meeting creation request, improving the practicability of the system and the user experience.

[0117] Based on the above embodiments of the intelligent conference management system, the first embodiment of the intelligent conference management method based on the intelligent conference management system of the present invention is proposed.

[0118] Reference Figure 3 , Figure 3 is a schematic flowchart provided for the first embodiment of the intelligent conference management method based on the above intelligent conference management system of the present application.

[0119] As Figure 3 shown, in this embodiment, the above intelligent conference management method based on the intelligent conference management system includes the following steps:

[0120] Step S10: Real-time detect whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold.

[0121] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device or an intelligent conference management device equipped with video conference software that can implement the above functions. Hereinafter, the intelligent conference management device is taken as an example (hereinafter referred to as the device) to illustrate this embodiment and the following embodiments.

[0122] Step S20: If so, obtain the person segmentation image in the continuous video frame images in the current video conference based on the semantic segmentation technology, and generate a target compressed video based on the person segmentation image.

[0123] Step S30: Decompress the target compressed video to obtain decompressed video frame images.

[0124] Step S40: Generate and display target video frame images based on the decompressed video frame images.

[0125] Further, the step S20 includes: obtaining continuously video frame images in the current video conference in real time; obtaining a person segmentation image and a background segmentation image in the video frame image based on semantic segmentation technology; determining a person key area image in the person segmentation image, and determining a person non-key area image based on the person key area image; generating a video to be compressed based on the person key area image; and compressing the video to be compressed to generate a target compressed video.

[0126] Correspondingly, the step S40 includes:

[0127] Step S40': generating and displaying a target video frame image based on the decompressed video frame image, the background segmentation image, and the person non-key area image.

[0128] Further, the step S40' includes:

[0129] Step S40'a: searching for a background image to be replaced corresponding to the background segmentation image from a preset virtual background image database.

[0130] Step S40'b: matching the person non-key area image with all images in a preset non-key area image database to obtain a non-key area image to be replaced.

[0131] Step S40'c: generating and displaying a target video frame image based on the decompressed video frame image, the background image to be replaced, and the non-key area image to be replaced.

[0132] Further, the step S40'c includes: obtaining a decompressed person key area image based on the decompressed video frame image; merging the background image to be replaced, the non-key area image to be replaced, and the decompressed person key area image to generate a target video frame image; and displaying the target video frame image in the current video conference.

[0133] In specific implementation, refer to Figure 4 , Figure 4 is the overall flowchart of the intelligent conference management method based on the above intelligent conference management system. As Figure 4As shown, when a video conference is held between terminals, the device can obtain the traffic consumption generated by the current video conference in real time and determine whether the traffic consumption generated by the current video conference exceeds the maximum allowed traffic consumption. If so, it can perform semantic segmentation on consecutive video frame images in the current video conference based on semantic segmentation technology to segment the person region image and the background region image in the video frame image. Then, it can compress the key region image of the person in the person region image to generate a target compressed video. After that, the device can respectively determine the replacement images corresponding to the background region image and the non-key region image of the person in the person region image, and merge these replacement images and the decompressed person region image to obtain the video frame image to be displayed, and display the video frame image to be displayed.

[0134] In this embodiment, it is disclosed to detect in real time whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold; if so, obtain the person segmentation image in consecutive video frame images in the current video conference based on semantic segmentation technology, and generate a target compressed video based on the person segmentation image; decompress the target compressed video to obtain a decompressed video frame image; generate and display a target video frame image based on the decompressed video frame image; since in this embodiment, when the traffic consumption generated by the current video conference exceeds the preset traffic consumption threshold, the person segmentation image in consecutive video frame images in the current video conference is obtained based on semantic segmentation technology to generate a target compressed video, and the target compressed video is decompressed and then displayed, thus solving the technical problem in the prior art that when the network quality of an online video conference is poor, the video quality will be reduced, and at the same time, the stability and response speed of the video conference are affected.

[0135] It should be noted that the above examples are only for understanding this application and do not constitute a limitation to the intelligent conference management method based on the above intelligent conference management system. Based on this technical concept, more forms of simple transformation are within the protection scope of this application.

Claims

1. An intelligent conference management system, characterized in that, The system described above includes: a conference management module, a video compression module, and a video decompression module; The conference management module is used to detect in real time whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold; The conference management module is further used to, if so, send a video compression request to the video compression module; The video compression module is used to, when receiving the video compression request, obtain a person segmentation image from consecutive video frame images in the current video conference based on semantic segmentation technology, and generate a target compressed video based on the person segmentation image; The video decompression module is used to decompress the target compressed video to obtain decompressed video frame images; The video decompression module is further used to generate and display target video frame images based on the decompressed video frame images; 2. The system according to claim 1, wherein The video compression module is used to, when receiving the video compression request, obtain consecutive video frame images in the current video conference in real time; The video compression module is further used to obtain a person segmentation image and a background segmentation image from the video frame images based on semantic segmentation technology; The video compression module is further used to determine a person key area image in the person segmentation image, and determine a person non-key area image based on the person key area image; The video compression module is further used to generate a video to be compressed based on the person key area image; The video compression module is further used to compress the video to be compressed to generate a target compressed video; The video decompression module is further used to generate and display target video frame images based on the decompressed video frame images, the background segmentation image, and the person non-key area image; 3. The system according to claim 2, wherein The video decompression module is further used to find a background image to be replaced corresponding to the background segmentation image from a preset virtual background image database; The video decompression module is further used to match the person non-key area image with all images in a preset non-key area image database to obtain a non-key area image to be replaced; The video decompression module is further used to generate and display target video frame images based on the decompressed video frame images, the background image to be replaced, and the non-key area image to be replaced; 4. The system according to claim 2, wherein The video decompression module 30 is further used to obtain a decompressed person key area image based on the decompressed video frame images; The video decompression module is further used to merge the background image to be replaced, the non-key area image to be replaced, and the decompressed person key area image to generate a target video frame image; The video decompression module is further used to display the target video frame images in the current video conference; 5. The system according to any one of claims 1 to 4, characterized in that, The system further includes: a conference creation module; The conference creation module is used to, when receiving a conference creation request, determine the start time and end time of the conference to be created based on the conference creation request; The conference creation module is further used to predict the traffic to be consumed by the current participating terminals based on the start time and end time of the conference; The conference creation module is further configured to determine the total traffic consumption of the current participating terminal in the current video conference; The conference creation module is further configured to determine the remaining traffic to be consumed corresponding to the current participating terminal based on the total traffic consumption; The conference creation module is further configured to generate a conference creation result corresponding to the to-be-created conference based on the to-be-consumed traffic and the remaining traffic to be consumed; 6. The system according to claim 5, wherein The conference creation module is further configured to generate and display an entry link corresponding to the to-be-created conference based on the conference start time and the conference end time when the remaining traffic to be consumed is greater than the to-be-consumed traffic; The conference creation module is further configured to determine a target participating terminal from the remaining participating terminals in the current video conference when the remaining traffic to be consumed is less than the to-be-consumed traffic; The conference creation module is further configured to send the conference creation request to the target participating terminal, so that the target participating terminal generates and displays an entry link corresponding to the to-be-created conference based on the conference start time and the conference end time; 7. An intelligent conference management method based on the intelligent conference management system according to any one of claims 1 to 6, characterized in that, The method includes: Real-time detection of whether the traffic consumption generated by the current video conference exceeds a preset traffic consumption threshold; If so, obtain a person segmentation image in the continuous video frame images in the current video conference based on semantic segmentation technology, and generate a target compressed video based on the person segmentation image; Decompress the target compressed video to obtain decompressed video frame images; Generate and display target video frame images based on the decompressed video frame images; 8. The method according to claim 7, wherein The step of obtaining a person segmentation image in the continuous video frame images in the current video conference based on semantic segmentation technology and generating a target compressed video based on the person segmentation image includes: Real-time acquisition of continuous video frame images in the current video conference; Obtain a person segmentation image and a background segmentation image in the video frame image based on semantic segmentation technology; Determine the person key area image in the person segmentation image, and determine the person non-key area image based on the person key area image; Generate a to-be-compressed video based on the person key area image; Compress the to-be-compressed video to generate a target compressed video; The step of generating and displaying target video frame images based on the decompressed video frame images includes: Generate and display target video frame images based on the decompressed video frame images, the background segmentation image, and the person non-key area image; 9. The method according to claim 8, wherein The step of generating and displaying target video frame images based on the decompressed video frame images, the background segmentation image, and the person non-key area image includes: Search for a to-be-replaced background image corresponding to the background segmentation image in a preset virtual background image database; Match the person non-key area image with all the images in a preset non-key area image database to obtain a to-be-replaced non-key area image; Generate and display target video frame images based on the decompressed video frame images, the to-be-replaced background image, and the to-be-replaced non-key area image.

10. The method according to claim 8, characterized in that The step of generating and displaying a target video frame image based on the decompressed video frame image, the background image to be replaced, and the non-critical area image to be replaced includes: Obtaining a decompressed human key area image based on the decompressed video frame image; Merging the background image to be replaced, the non-critical area image to be replaced, and the decompressed human key area image to generate a target video frame image; Displaying the target video frame image in the current video conference.

Images