CN109089127A - A kind of video-splicing method, apparatus, equipment and medium - Google Patents

Abstract

The present invention discloses a video splicing method, device, equipment and medium. The method includes: using feature matching and/or using barrage information analysis to determine N exciting video segments from the target video, where N is greater than 1; Extracting the N exciting video segments from the target video; splicing the N exciting video segments into one video to form a spliced video; when receiving the request sent by the client for requesting to obtain the spliced video, sending The spliced video is played to the client. The method, device, equipment and medium provided by the present application can solve the technical problems in the prior art of watching live video of historical games, wasting viewers’ viewing time and causing the viewers to have a low probability of obtaining exciting video clips. The technical effect of saving viewing time is realized.

Description

A video splicing method, device, equipment and medium

technical field

The present invention relates to the field of computer technology, in particular to a video splicing method, device, equipment and media.

Background technique

At present, with the advancement of network communication technology and the speed-up of broadband network, webcasting has been more and more developed and applied. In order for users not to miss the wonderful live video of the host, video websites often record and provide the historical live video of the host for users to watch.

In the game live broadcast, there are often some wonderful game plots, for example, video clips of successful kills in killing games, video clips of successful collection in collection games, or wedding video clips in dating games, etc. These exciting video segments are often the most exciting and most worth watching in the game live broadcast. In order to watch these exciting scenes, the audience users often need to start from the beginning of the historical live broadcast video and watch the entire video in order to ensure that they will not miss it. These highlights. This will cause viewers to waste more time watching videos that they are not interested in, and it is easy to miss wonderful video moments.

It can be seen that watching the live video of historical games in the prior art has the technical problems of wasting the audience's viewing time and causing the audience to have a low probability of obtaining exciting video clips.

Contents of the invention

The present invention provides a video splicing method, device, equipment and medium, which are used to solve the technical problems in the prior art of watching live video of historical games, which waste viewers' viewing time and cause the viewers to have a low probability of obtaining exciting video clips.

In a first aspect, the present invention provides a video splicing method, comprising:

Using feature matching and/or adopting barrage information analysis to determine N wonderful video segments from the target video, N is greater than 1;

Extracting the N highlight video segments from the target video;

Splicing the N exciting video segments into one video to form a spliced video;

When receiving the request sent by the client for obtaining the spliced video, sending the spliced video to the client for playing.

Optionally, the determining N exciting video segments in the target video includes: setting feature information according to the video category of the target video; performing feature matching on the target video to determine the A target frame whose feature information matches; according to the target frame and preset highlight video interception rules, determine the N highlight video segments in the target video, wherein the highlight video segment includes the target frame, and The wonderful video interception rule corresponds to the feature information; or, obtain the target video and barrage information, and the barrage information includes the barrage quantity information of the target video in the historical playback process; according to the barrage information, The N exciting video segments in the target video whose barrage conditions meet the preset requirements are determined.

Optionally, the extracting the N exciting video segments from the target video includes: acquiring attribute information of the target video; judging whether the target video needs to use time stamp accurate extraction mode according to the attribute information ; If necessary, video decoding is carried out to the target video; and according to the interception rule of the wonderful video, according to the time stamp information of the decoded target video, extract the N wonderful video segments from the decoded target video; If not, then search the video unit whose corresponding timestamp information is closest to the timestamp information of the target frame in the undecoded target video, wherein the target video includes N video units, and N is greater than 1 Positive integer; determine and extract the highlight video segment according to the closest video unit.

Optionally, the splicing the N exciting video segments into one video to form a spliced video includes: splicing the N exciting video segments into one video, and inserting a prompt video before each exciting video segment, The prompt video is used to describe the exciting video segment to be played to form a spliced video; or, the N exciting video segments are spliced into one video, and an interval video is inserted between every two exciting video segments, the interval The video is used to represent that the previous exciting video segment has been played and the next exciting video segment is about to be played to form a spliced video; or, the N exciting video segments are spliced into one video, and at the beginning of each exciting video segment A video segment is superimposed with playback prompt information, and the prompt information is used to describe the wonderful video segment being played to form a spliced video.

Optionally, the determining the N exciting video segments in the target video is implemented at the GCR-Work layer; the extracting the N exciting video segments from the target video is implemented at the Media-Worker layer.

In a second aspect, a video splicing device is provided, including:

Determining unit, for adopting feature matching and/or adopting barrage information analysis to determine N exciting video segments from the target video, N is greater than 1;

An extracting unit, configured to extract the N highlight video segments from the target video;

A splicing unit, configured to splice the N exciting video segments into one video to form a spliced video;

A sending unit, configured to send the spliced video to the client for playing when receiving a request from the client for obtaining the spliced video.

Optionally, the splicing unit is also used to: splice the N exciting video segments into one video, and insert a prompt video before each exciting video segment, the prompt video is used to describe the exciting video segment to be played , forming a spliced video; or, splicing the N wonderful video segments into one video, and inserting an interval video between every two exciting video segments, the interval video is used to represent that the previous exciting video segment has played and ended and The next wonderful video segment is about to play to form a spliced video; or, the N wonderful video segments are spliced into one video, and a prompt message is superimposed in the initial segment video of each wonderful video segment, and the prompt information is used It is used to describe the exciting video segment being played to form a spliced video.

Optionally, the extracting unit is further configured to: obtain attribute information of the target video; judge whether the target video needs to adopt a time stamp accurate extraction mode according to the attribute information; if necessary, perform Video decoding; and according to the wonderful video interception rule, according to the timestamp information of the decoded target video, extract the N wonderful video segments from the decoded target video; if not needed, then in the undecoded target video Find the video unit whose corresponding timestamp information is closest to the timestamp information of the target frame, wherein the target video includes N video units, and N is a positive integer greater than 1; according to the closest video unit Determine and extract the wonderful video segment.

One or more technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages:

The method, device, device and medium provided in the embodiments of the present application determine and extract N exciting video segments from the target video by using feature matching and/or using barrage information analysis, and splicing the N exciting video segments For one video, form a mosaic video, and when receiving a request from the client for obtaining the mosaic video, send the mosaic video to the client for playback, so that the audience can watch the entire object without complete For the video, you only need to watch the mosaic directly to see all the most exciting video clips that you don't want to miss, which effectively saves the audience's viewing time and enables the audience to get all the wonderful video clips in a short period of time.

Further, according to the attribute information of the target video, it is judged whether the target video needs to adopt the accurate time stamp extraction mode, and when necessary, the video decoding is performed on the target video, and then extracted according to the time stamp information of the decoded target video The wonderful video segment, when not needed, directly searches the video unit whose corresponding time stamp information is closest to the time stamp information of the target frame in the undecoded target video, to extract the wonderful video segment, effectively reducing It shortens the video extraction time that does not require accurate extraction, and also ensures the extraction accuracy of some videos that require accurate extraction.

The above description is only an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention, it can be implemented according to the contents of the description, and in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable , the specific embodiments of the present invention are enumerated below.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

Fig. 1 is the flow chart of video mosaic method in the embodiment of the present invention;

Fig. 2 is a schematic diagram of extraction without using the time stamp accurate extraction mode in the embodiment of the present invention;

3 is a schematic structural diagram of a video splicing device in an embodiment of the present invention;

4 is a schematic structural diagram of an electronic device in an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a storage medium in an embodiment of the present invention.

Detailed ways

The embodiment of the present application provides a video splicing method, device, equipment, and medium to solve the problems in the prior art of watching live video of historical games, which waste the audience's viewing time and cause the audience to have a low probability of obtaining exciting video clips. question. It realizes the technical effect of saving the viewer's viewing time and enabling the viewer to obtain all the exciting video clips in a relatively short period of time.

The general idea of the technical solution in the embodiment of the application is as follows:

Determine and extract N exciting video segments from the target video using feature matching and/or bullet chat information analysis, and splice the N exciting video segments into one video to form a spliced video, when receiving the client sending When requesting to obtain the spliced video, send the spliced video to the client for playback, so that the viewer does not need to watch the entire target video in its entirety, but only needs to directly watch the spliced video to see all the latest spliced videos. Wonderful video clips that they don't want to miss, which effectively saves the audience's viewing time and enables the audience to obtain all the wonderful video clips in a relatively short period of time.

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

Embodiment one

This embodiment provides a video splicing method, as shown in Figure 1, including:

Step S101, using feature matching and/or using barrage information analysis to determine N exciting video segments from the target video, where N is greater than 1;

Step S102, extracting the N exciting video segments from the target video;

Step S103, splicing the N exciting video segments into one video to form a spliced video;

Step S104, when receiving the request from the client for obtaining the spliced video, sending the spliced video to the client for playing.

In this embodiment of the application, the method can be applied to a server, or to a viewer or an anchor, and there is no limitation here. The implementation device can be an electronic device such as a smart phone, a desktop computer, a notebook, or a tablet computer. This is also not limited.

Below in conjunction with Fig. 1, the specific implementation steps of the method provided by this embodiment are described in detail:

Firstly, step S101 is executed to determine N highlight video segments from the target video by feature matching and/or barrage information analysis, where N is greater than 1.

For adopting feature matching to determine N exciting video segments from the target video, its specific implementation method is:

Set feature information according to the video category of the target video; carry out feature matching to the target video, to determine the target frame matching the feature information in the target video; intercept according to the target frame and the preset wonderful video A rule for determining the N highlight video segments in the target video, wherein the highlight video segments include the target frame, and the highlight video interception rule corresponds to the feature information.

Specifically, first, feature information is set according to the video category of the target video.

It should be noted that the target video may be a video uploaded by the anchor; it may also be a video stored by the server during the previous live broadcast; it may also be a live video currently being broadcast. If the target video is a live video that is currently being broadcast live, the method provided by this implementation is to perform real-time target frame matching and highlight video segment extraction on the received live video stream during the live broadcast process.

In the specific implementation process, the video category of the target video is different, and the corresponding feature information is also different. The feature information can be voice feature information or image feature information, which is not limited here. Examples are given below:

The first type, feature information is image feature information.

That is, according to the video category of the target video, the feature information corresponding to the video category is determined from a preset feature information library, the feature information is information extracted from a wonderful image, and the wonderful image is the video The image in the video corresponding to the category. That is to say, some similar highlight video clips in the target video often have the same highlight image frames by default, and the feature information may be common image features extracted from these highlight image frames.

For example, when the target video is a game video including a killing plot, the characteristic information is set as information extracted from a game killing success screen. Specifically, after a successful kill, an image prompting the successful kill is often displayed on the video, such as the words "KO", or the words "the number of people plus 1", or the pattern of blood spots, etc., and these image features can be used as characteristic information.

When the target video is a game video including a captured plot, the characteristic information is set as information extracted from a successfully captured picture. Specifically, after the acquisition is successful, an image indicating the success of the acquisition is often displayed on the video, such as the word "plus 1", or the pattern of the collected item, etc., and these image features can be used as feature information.

The second type, feature information is speech feature information.

That is, according to the video category of the target video, the feature information corresponding to the video category is determined from a preset feature information database, and the feature information is information extracted from a video audio file. That is to say, there is often some voice information that is the same by default for some similar highlight video clips in the target video, and the feature information may be a common voice feature extracted from the voice information.

For example, when the target video is a game video including a killing plot, the feature information is set to be information extracted from the voice of a successful game killing. Specifically, after the kill is successful, the video will often play a voice prompting the success of the kill, such as the pronunciation of "KO", or the pronunciation of "killed successfully", or the pronunciation of screams such as "wow", you can use these voices features as feature information.

When the target video is a lottery video, the characteristic information is set as voice information extracted from the lottery video. Specifically, when the lottery is drawn, a voice prompting the lottery is often played along with the video, such as specific music, or voices such as "revealing soon", and these voice features can be used as feature information.

Of course, in a specific implementation process, the feature information is not limited to the above two types, and may also be time information, which is not limited here and will not be listed one by one.

In the specific implementation process, according to the video type and video content requirements, multiple or one type of feature information can be set for a target video, so that multiple or one type of exciting video segments can be extracted later.

Then, feature matching is performed on the target video to determine a target frame in the target video that matches the feature information.

In the specific implementation process, the feature information is different, and the corresponding matching methods are also different:

If the feature information is image feature information, the feature information is image-matched with each frame image of the target video, or the feature information is image-matched with the interval frame images of the target video. If the feature information corresponds to an image, it is determined that the frame is the target frame. For example, assuming that the feature information is a blood spot pattern, when a frame containing the blood spot pattern is matched, the frame image is used as the target frame.

If the feature information is voice feature information, the feature information is audio-matched with the audio file of the target video, and when a certain audio is matched to the feature information, it is determined that the frame corresponding to the audio is the target frame, specifically In other words, the frame corresponding to the audio here is the frame whose time stamp information is consistent with the time stamp information of the audio file at this place. For example, assuming that the feature information is the pronunciation of "successfully killed", when an audio file containing audio is matched, the frame with the same time stamp as the audio file is used as the target frame.

Certainly, the methods for performing feature matching are not limited to the above two methods, which are not limited here and will not be listed one by one.

Next, determine a highlight video segment in the target video according to the target frame and preset highlight video interception rules, the highlight video segment includes the target frame, the highlight video interception rule and the feature information correspond.

In the embodiment of the present application, according to the preset highlight video clipping rules, determine the playing time between the start frame of the highlight video segment and the target frame, and the distance between the end frame of the highlight video segment and the target frame. The playback duration of the frame, wherein, in the target video, the playback position of the start frame is located before the target frame or equal to the target frame, and the playback position of the end frame is located after the target frame or equal to The target frame.

Specifically, the wonderful video interception rule corresponds to the feature information, which means that for different feature information, there will be corresponding wonderful video interception rules, for example:

Assuming that this feature information is the information that indicates the success of the game kill in the game video including the killing plot, considering that the wonderful aiming and killing probably occurs about 1 minute before the successful kill, you can set this type of feature information The corresponding highlight video interception rule is: determine the video between 60s before the target frame and the target frame as the highlight video segment.

Assuming that the feature information is the information that represents the start of the lottery in the lottery video, considering that the lottery time is about 180s, you can set the exciting video interception rule corresponding to this type of feature information as: determine the video between the start of the target frame and the next 180s For great video segments.

Of course, in addition to determining the highlight video terminal duration and the time position relationship with the target frame through the above-mentioned feature information type, so as to determine the highlight video interception rule, there are other methods for determining the highlight video interception rule. For example, multiple feature information can also be set, and the video between target frames corresponding to certain two feature information is used as a highlight video segment. For example, assuming that for a lottery video, feature information A representing the start of the lottery and feature information B representing the end of the lottery are set, and the feature information A corresponds to the target frame A, and the feature information B corresponds to the target frame B, then the corresponding The highlights video interception rule is: determine the video between the target frame A and the target frame B as the highlights video segment.

For adopting barrage information analysis to determine N exciting video segments from the target video, its specific implementation method is:

Acquiring target video and barrage information, the barrage information including the barrage quantity information of the target video during historical playback; according to the barrage information, it is determined that the barrage situation in the target video meets the preset requirements The N exciting video segments of .

Specifically, firstly, the target video and barrage information are acquired, and the barrage information includes information on the number of barrages of the target video during historical playback.

It should be noted that the target video may be a video uploaded by the anchor; it may also be a video stored by the server during the previous live broadcast; it may also be a live video currently being broadcast. If the target video is a live video that is currently being broadcast live, the method provided by this implementation is to perform real-time barrage information acquisition and judgment on the received live video stream during the live broadcast, and extract exciting video segments.

In a specific implementation process, the bullet chat information may include bullet chat quantity information, bullet chat content information, bullet chat sender number information, bullet chat send word count information, etc. for each frame of the target video obtained during the live video broadcast.

Then, according to the barrage information, determine the exciting video segments in the target video whose barrage conditions meet preset requirements.

In the embodiment of the present application, the highlight video segment is determined by first determining the target frame in the target video that meets the preset requirements according to the barrage information, and then according to the target frame and the preset highlight video segment. The interception rule is to determine a highlight video segment in the target video, and the highlight video segment includes the target frame.

The preset requirement may be that the number of bullet chats displayed when the target frame is played is greater than a preset value, or the number of bullet chats is increased faster than a preset value, which is not limited here.

In the embodiment of the present application, according to the barrage information, there are many ways to determine the exciting video segment, and three examples are listed below:

The first one is that the number of barrage is greater than the preset value.

That is, according to the barrage information, determine the exciting video segments in the target video whose number of barrages is greater than a preset number.

Specifically, it is possible to first determine the corresponding frames when the number of bullet chats displayed is greater than the preset value, and then extract these frames and arrange them into wonderful video segments in chronological order.

Determining the exciting video segment according to the number of bullet chats being greater than the preset value can effectively determine the exciting video segment with high user participation.

The second type is the target frame with the largest number of barrage.

That is, according to the barrage information, determine the target frame with the largest number of barrages in the target video, and then determine the highlight video segment according to the target frame, and the highlight video segment includes the target frame.

Specifically, in order to avoid the problem of discontinuous highlight video segments caused by only extracting some frames, you can first determine the target frame with the largest number of barrages in the target video or the number of barrages is greater than a certain value, and then combine the target frame and its Videos within a period of time before and after are regarded as highlight video segments. For example, the target frame and the 30s before and after the video can be taken as the highlight video segment.

The third type is the increase in the number of barrage.

That is, according to the barrage information, it is determined in the target video that the number of barrages increases faster than a preset speed in the exciting video segment

Specifically, the growth rate of the number of bullet chats corresponding to each frame can be determined according to the number of bullet chats in each frame and the frames before and after it, and the frame whose bullet chatting number growth rate is greater than the preset speed is used as the target frame, and these frames are extracted by pressing The videos are arranged in chronological order as highlights. For example, you can set the bullet chatting growth rate corresponding to each frame equal to the ratio of the number of bullet chatting displayed in the frame behind the frame divided by the number of bullet chatting displayed in this frame, or set the bullet chatting growth rate corresponding to each frame as the frame The ratio of the total number of bullet chatter displayed for the next 5 consecutive seconds divided by the total number of bullet chatter displayed for 5 consecutive seconds before the frame is not limited here.

By determining the exciting video segment according to the growth rate of the number of barrage, it is possible to effectively determine the key video segment that stimulates users to send a large number of barrage.

Of course, the method for determining the highlight video segment is not limited to the above three methods, and the highlight video segment can also be determined according to the total number of words sent by the barrage, which is not limited here and will not be listed one by one.

Then, step S102 is executed to extract the N highlight video segments from the target video.

In the specific implementation process, the highlight video segment is determined, that is, the start time stamp and the end time stamp of the highlight video segment can be determined, so as to extract the start time stamp and the end time stamp from the target video Highlight video segments between timestamps.

Considering that the extraction of exciting video segments consumes more computing and processing resources, this embodiment also provides an extraction method with low resource consumption, which is described in detail as follows:

Please refer to Figure 2. Since the target video is a live video that is being broadcast live or historically live, its video transmission is interspersed with video units and audio units. Each video unit and audio unit has its corresponding time stamp information. Therefore, this embodiment does not decode the target video segment, and directly pulls the live stream through steps S201 to S204. After de-streaming and multiplexing, search for the corresponding timestamp information and the time stamp of the target frame in the undecoded target video. Stamp the video unit with the closest information, determine and extract the wonderful video segment according to the closest video unit, and then stream multiplex, synthesize and save the extracted wonderful video segment through steps S205-S206. For example, as shown in Figure 2, assuming that the time stamps of video unit 3 and video unit 4 are closest to the time stamp information of the determined highlight video segment, then de-streaming and multiplexing extracts the video unit 3 and video unit 4 , and after extracting the time stamp information and the audio unit corresponding to the video unit 3 and the video unit 4, stream multiplexing is performed to synthesize the video unit and the audio unit, thereby forming the extracted complete exciting video segment.

By adopting the method for extracting wonderful video segments, since the entire video does not need to be decoded, more computing and processing resources can be saved, and the processing speed can be improved.

Further, considering that some exciting video segments have strict requirements on time, it is also possible to set and obtain the attribute information of the target video before extracting the exciting video segment; judge whether the target video needs to be used according to the attribute information. Stamp the precise extraction mode; if necessary, video decoding is carried out to the target video; and according to the interception rule of the wonderful video, according to the time stamp information of the decoded target video, extract the wonderful video from the decoded target video segment; if not needed, search for the video unit whose corresponding timestamp information is closest to the timestamp information of the target frame in the undecoded target video, wherein the target video includes N video units, and N is greater than A positive integer of 1; determine and extract the highlight video segment according to the closest video unit.

That is, according to the situation of the wonderful video segment corresponding to each type of feature information, the staff will pre-set in the attribute information of the target video whether to use the time stamp accurate extraction mode to represent the extraction information. For example, if the time stamp accurate extraction is required, set the attribute The number after the Ti identifier of the information is 1, and if it is not necessary to use time stamp to accurately extract, set the number after the Ti identifier of the attribute information to 0. Before the subsequent extraction, judge whether the target video needs to use the time stamp accurate extraction mode according to the preset extraction information in the attribute information. If necessary, first decode the target video, and then perform accurate extraction according to the time stamp of each frame, such as If it is not needed, the target video is not decoded, and the low-resource consumption extraction is directly performed according to the time stamp of each video unit.

Then, step S103 is executed to splice the N exciting video segments into one video to form a spliced video.

In the specific implementation process, there can be multiple methods for splicing the multiple exciting video segments, and three examples are listed below:

The N exciting video segments can be spliced into one video, and a prompt video is inserted before each exciting video segment, and the prompt video is used to describe the exciting video segment to be played to form a spliced video. That is, insert a pre-prepared reminder video before each highlight video segment, which may include: the playing time information of the highlight video to be played next in the original target video, the highlight video to be played next A description of the content of the video or the type of video content of the next cool video, etc.

The N wonderful video segments can also be spliced into one video, and an interval video is inserted between every two exciting video segments, and the interval video is used to represent that the previous exciting video segment has been played and the next exciting video segment has been played It will be played to form a spliced video. That is, a pre-prepared interval video is inserted before each highlight video segment, and the interval video can be a blank video, a preset subtitle video, or a self-introduction video of the anchor.

It is also possible to splice the N wonderful video segments into one video, and superimpose playing prompt information in the initial segment video of each wonderful video segment, and the prompt information is used to describe the wonderful video segment being played to form a spliced video . That is, in order not to increase redundant video playback time, a preset prompt message is synthesized in the first frame or multiple frames of each exciting video segment. The prompt message can be a prompt picture or a prompt voice, which will not be described here. limit. If the prompt information is a prompt picture, it may be in a picture-in-picture manner, or may be in a semi-transparent covering manner, which is not limited here.

Of course, in the specific implementation process, the ways of splicing videos are not limited to the above three, and the multiple wonderful videos can also be seamlessly spliced according to the order of time stamps to reduce the playback and processing time. No longer list them one by one.

Next, step S104 is executed, when the request for acquiring the spliced video from the client is received, the spliced video is sent to the client for playing.

In the embodiment of this application, after forming the mosaic video, the opening link of the mosaic video can also be placed on the host room page corresponding to the target video in the live broadcast website, so that the audience can directly trigger the opening link to choose to play The spliced video.

Of course, in the embodiment of the present application, it is also possible to mark the exciting video segment on the playback progress bar.

That is, after determining the exciting video segment, obtain the playing time information of the exciting video segment in the target video; according to the playing time information, correspond to the playing time information on the playing progress bar of the target video Mark the wonderful video segment at the target position of the .

Specifically, the highlight video segment may be marked at a target position corresponding to the playback time information on the playback progress bar of the target video by adopting a gross point method. The color of the target position corresponding to the playing time information on the playing progress bar can be changed, or the width of the progress bar can be changed, or marking lines can be added for marking.

When an operation acting on the target position is received, a picture or video representing the exciting video segment is displayed. Specifically, it may be displaying the image of the target frame, or displaying other pictures in the exciting video segment, or triggering the playing of the exciting video segment, or displaying a preset introduction picture for describing the exciting video segment.

In the specific implementation process, the displayed picture or video that can characterize the wonderful video segment can be displayed by opening a window separately, or directly displayed in the playback window of the target video, or can be It is superimposed and displayed in the playback window of the target video, which is not limited here. Wherein, the superimposed display may be displayed in a picture-in-picture manner, or may be superimposed and displayed by setting translucence, which is not limited here.

Further, in the embodiment of the present application, considering that both feature information matching and highlight video segment extraction will consume resources, in order to avoid mutual interference and resource preemption during the execution of each task, the feature matching slave target can also be set Determining that the wonderful video segment in the video is implemented at the GCR-Work layer; the described N wonderful video segments are extracted from the target video and implemented at the Media-Worker layer; the video splicing and according to the playing time information, Marking the exciting video segment at the target position corresponding to the playing time information on the playing progress bar of the target video is implemented at the Media-Worker layer.

Specifically, feature matching and/or barrage information analysis is used to determine and extract N exciting video segments from the target video, and the N exciting video segments are spliced into one video to form a spliced video. When receiving When the client sends a request for obtaining the spliced video, send the spliced video to the client for playback, so that the audience does not need to watch the entire target video completely, but only needs to directly watch the spliced video Seeing all the most exciting video clips that he does not want to miss, effectively saves the viewing time of the audience, and enables the audience to obtain all the wonderful video clips in a relatively short period of time.

Based on the same inventive concept, the embodiment of the present invention also provides a device corresponding to the video splicing method in the first embodiment, see the second embodiment.

Embodiment two

This embodiment provides a video splicing device, as shown in Figure 3, the device includes:

Determining unit 301, configured to determine N highlight video segments from the target video by feature matching and/or barrage information analysis, where N is greater than 1;

An extracting unit 302, configured to extract the N highlight video segments from the target video;

A splicing unit 303, configured to splice the N exciting video segments into one video to form a spliced video;

The sending unit 304 is configured to send the spliced video to the client for playing when receiving a request from the client for obtaining the spliced video.

In the embodiment of the present application, the device may be an electronic device such as a smart phone, a desktop computer, a notebook, or a tablet computer, which is not limited here.

In the embodiment of the present application, the device may be an Android system, an IOS system or a Windows system, which is not limited here.

Because the device introduced in Embodiment 2 of the present invention is the device used to implement the method in Embodiment 1 of the present invention, so based on the method described in Embodiment 1 of the present invention, those skilled in the art can understand the specific structure and deformation of the device , so it will not be repeated here. All devices used in the method of Embodiment 1 of the present invention belong to the intended protection scope of the present invention.

Based on the same inventive concept, this application provides an electronic device embodiment corresponding to Embodiment 1. For details, see

Embodiment three.

Embodiment Three

This embodiment provides an electronic device, as shown in FIG. 4 , including a memory 410, a processor 420, and a computer program 411 stored in the memory 410 and operable on the processor 420. When the processor 420 executes the computer program 411, Any implementation manner in Embodiment 1 can be implemented.

Since the electronic device introduced in this embodiment is the device used to implement the method in Embodiment 1 of this application, based on the method described in Embodiment 1 of this application, those skilled in the art can understand the electronic device of this embodiment. Specific implementation methods and various variations thereof, so how the electronic device implements the method in the embodiment of the present application will not be described in detail here. As long as a person skilled in the art implements the equipment used by the method in the embodiment of the present application, it all belongs to the protection scope of the present application.

Based on the same inventive concept, the present application provides a storage medium corresponding to Embodiment 1, see Embodiment 4 for details.

Embodiment four

This embodiment provides a computer-readable storage medium 500 , as shown in FIG. 5 , on which a computer program 511 is stored. When the computer program 511 is executed by a processor, any implementation manner in the first embodiment can be implemented.

The technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

The method, device, device and medium provided in the embodiments of the present application determine and extract N exciting video segments from the target video by using feature matching and/or using barrage information analysis, and splicing the N exciting video segments For one video, form a mosaic video, and when receiving a request from the client for obtaining the mosaic video, send the mosaic video to the client for playback, so that the audience can watch the entire object without complete For the video, you only need to watch the mosaic directly to see all the most exciting video clips that you don't want to miss, which effectively saves the audience's viewing time and enables the audience to get all the wonderful video clips in a short period of time.

Further, according to the attribute information of the target video, it is judged whether the target video needs to adopt the accurate time stamp extraction mode, and when necessary, the video decoding is performed on the target video, and then extracted according to the time stamp information of the decoded target video The wonderful video segment, when not needed, directly searches the video unit whose corresponding time stamp information is closest to the time stamp information of the target frame in the undecoded target video, to extract the wonderful video segment, effectively reducing It shortens the video extraction time that does not require accurate extraction, and also ensures the extraction accuracy of some videos that require accurate extraction.

Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, systems, or computer program products. Accordingly, the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

While preferred embodiments of the invention have been described, additional changes and modifications to these embodiments can be made by those skilled in the art once the basic inventive concept is appreciated. Therefore, it is intended that the appended claims be construed to cover the preferred embodiment as well as all changes and modifications which fall within the scope of the invention.

Apparently, those skilled in the art can make various changes and modifications to the embodiments of the present invention without departing from the spirit and scope of the embodiments of the present invention. In this way, if the modifications and variations of the embodiments of the present invention fall within the scope of the claims of the present invention and equivalent technologies, the present invention also intends to include these modifications and variations.

Claims (10)

1. A method for video stitching, comprising:

determining N wonderful video segments from the target video by adopting feature matching and/or bullet screen information analysis, wherein N is greater than 1;

extracting the N wonderful video segments from the target video;

splicing the N wonderful video segments into one video to form a spliced video;

and when a request for requesting to acquire the spliced video sent by a client is received, sending the spliced video to the client for playing.

2. The method of claim 1, wherein said determining N highlight video segments in the target video comprises:

setting characteristic information according to the video category of the target video; performing feature matching on the target video to determine a target frame matched with the feature information in the target video; determining the N highlight video segments in the target video according to the target frame and a preset highlight video intercepting rule, wherein the highlight video segments comprise the target frame, and the highlight video intercepting rule corresponds to the characteristic information; or,

acquiring a target video and bullet screen information, wherein the bullet screen information comprises bullet screen quantity information of the target video in a historical playing process; and determining the N wonderful video segments of which the bullet screen conditions meet preset requirements in the target video according to the bullet screen information.

3. The method of claim 1, wherein said extracting said N highlight video segments from said target video comprises:

acquiring attribute information of the target video;

judging whether the target video needs to adopt a timestamp accurate extraction mode according to the attribute information;

if so, performing video decoding on the target video; extracting the N wonderful video segments from the decoded target video according to the wonderful video interception rule and the timestamp information of the decoded target video;

if not, searching a video unit with corresponding timestamp information closest to the timestamp information of the target frame in an undecoded target video, wherein the target video comprises N video units, and N is a positive integer greater than 1; and determining and extracting the highlight video segment according to the nearest video unit.

4. The method of claim 1, wherein said splicing said N highlight video segments into one video to form a spliced video comprises:

splicing the N wonderful video segments into a video, and inserting a prompt video before each wonderful video segment, wherein the prompt video is used for describing the wonderful video segment to be played to form a spliced video; or,

splicing the N wonderful video segments into one video, and inserting an interval video between every two wonderful video segments, wherein the interval video is used for representing that the playing of the previous wonderful video segment is finished and the playing of the next wonderful video segment is about to be played to form a spliced video; or,

and splicing the N wonderful video segments into one video, and superposing playing prompt information in the initial segment video of each wonderful video segment, wherein the prompt information is used for describing the wonderful video segments which are being played to form a spliced video.

5. The method of claim 1, wherein said determining N highlight video segments in the target video is performed at a GCR-word layer; the extracting the N highlight video segments from the target video is implemented in a Media-Worker layer.

6. A video stitching device, comprising:

the determining unit is used for determining N wonderful video segments from the target video by adopting feature matching and/or bullet screen information analysis, wherein N is larger than 1;

an extracting unit, configured to extract the N highlight video segments from the target video;

the splicing unit is used for splicing the N wonderful video segments into one video to form a spliced video;

and the sending unit is used for sending the spliced video to the client side for playing when receiving a request which is sent by the client side and used for requesting to acquire the spliced video.

7. The apparatus of claim 6, wherein the splicing unit is further configured to:

splicing the N wonderful video segments into a video, and inserting a prompt video before each wonderful video segment, wherein the prompt video is used for describing the wonderful video segment to be played to form a spliced video; or,

splicing the N wonderful video segments into one video, and inserting an interval video between every two wonderful video segments, wherein the interval video is used for representing that the playing of the previous wonderful video segment is finished and the playing of the next wonderful video segment is about to be played to form a spliced video; or,

and splicing the N wonderful video segments into one video, and superposing playing prompt information in the initial segment video of each wonderful video segment, wherein the prompt information is used for describing the wonderful video segments which are being played to form a spliced video.

8. The apparatus of claim 6, wherein the extraction unit is further to:

acquiring attribute information of the target video;

judging whether the target video needs to adopt a timestamp accurate extraction mode according to the attribute information;

if so, performing video decoding on the target video; extracting the N wonderful video segments from the decoded target video according to the wonderful video interception rule and the timestamp information of the decoded target video;

if not, searching a video unit with corresponding timestamp information closest to the timestamp information of the target frame in an undecoded target video, wherein the target video comprises N video units, and N is a positive integer greater than 1; and determining and extracting the highlight video segment according to the nearest video unit.

9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any of claims 1-6 when executing the program.

10. A computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, is adapted to carry out the method of any one of claims 1 to 6.