A method and device for generating a disassembled video, an electronic device, and a storage medium

Abstract

Embodiments of the present application provide a generation method and device of a split video, electronic equipment and a storage medium, comprising: in response to the acquisition operation of the high-energy word list of the barrage, obtaining the high-energy word list of the barrage; the high-energy word list of the barrage contains multiple high-energy words of the barrage; acquiring the barrage of the video; if the barrage contains high-energy words of the barrage, the barrage is determined as high-energy barrage; acquiring the matching information table corresponding to the high-energy word corresponding to the high-energy barrage; according to the matching information table corresponding to the high-energy barrage, calculating the high-energy index of the barrage per second; according to the high-energy index of the barrage per second, fitting the smooth high-energy barrage curve about the high-energy index of the barrage per second; the video interval of the video whose smooth high-energy barrage curve exceeds the preset split threshold is determined as the high-energy interval of the video; according to the high-energy interval, generating the split video. The split video containing the high-energy interval is obtained through the barrage, the emotional information of the barrage is fully utilized, and the high-energy interval of the video and then the split video can be accurately and quickly obtained.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet, in particular to a generation method of a split video, a generation device of a split video, an electronic device and a computer readable storage medium. BACKGROUND

[0002] At present, most of the split video schemes for video highlights are pure visual schemes, for example, by identifying pre-defined actions in the video to extract the highlights in the video. The limitation of this method is that only pre-defined actions can be identified, and the processing content is limited. In addition, the highlight content in part of the video is the highlight dialogue of the actor. When delivering the highlight dialogue, the actor often does not have special actions. For this case, the current split video scheme based on action recognition cannot handle it. SUMMARY

[0003] The purpose of the embodiment of the present application is to provide a generation method of a split video, so as to realize splitting the highlights in the video by using the bullet screen, and obtaining a plurality of split videos containing the highlights of the video. The specific technical scheme is as follows:

[0004] In the first aspect of the present application, a generation method of a split video is provided, comprising:

[0005] In response to the acquisition operation of the bullet screen high-energy word table, the bullet screen high-energy word table is obtained; wherein the bullet screen high-energy word table contains a plurality of bullet screen high-energy words;

[0006] The bullet screen of the video is acquired;

[0007] If the bullet screen contains the bullet screen high-energy word, the bullet screen is determined as a high-energy bullet screen;

[0008] The matching information table corresponding to the bullet screen high-energy word corresponding to the high-energy bullet screen is acquired;

[0009] According to the matching information table corresponding to the high-energy bullet screen, the bullet screen high-energy index per second is calculated;

[0010] According to the bullet screen high-energy index per second, the smooth high-energy bullet screen curve about the bullet screen high-energy index per second is fitted;

[0011] The video interval of the video where the smooth high-energy bullet screen curve exceeds the preset split threshold is determined as the high-energy interval of the video;

[0012] According to the high-energy interval, a split video is generated.

[0013] In the second aspect of the present application, a generation device of a split video is also provided, comprising:

[0014] The barrage high-energy word table acquisition module is configured to obtain a barrage high-energy word table in response to an acquisition operation of the barrage high-energy word table, wherein the barrage high-energy word table comprises a plurality of barrage high-energy words.

[0015] The video barrage acquisition module is configured to acquire a barrage of a video.

[0016] The high-energy barrage confirmation module is configured to determine the barrage as a high-energy barrage if the barrage comprises the barrage high-energy word.

[0017] The matching information table acquisition module is configured to acquire a matching information table corresponding to the barrage high-energy word corresponding to the high-energy barrage.

[0018] The per-second barrage high-energy index calculation module is configured to calculate a per-second barrage high-energy index according to the matching information table corresponding to the high-energy barrage.

[0019] The smooth high-energy barrage curve fitting module is configured to fit a smooth high-energy barrage curve about the per-second barrage high-energy index according to the per-second barrage high-energy index.

[0020] The high-energy interval determination module is configured to determine a video interval of the video, in which the smooth high-energy barrage curve exceeds a preset segmentation threshold, as a high-energy interval of the video.

[0021] The clip video generation module is configured to generate a clip video according to the high-energy interval.

[0022] In another aspect of the present application, an electronic device is provided, which comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; the memory is configured to store a computer program; and the processor is configured to execute the program stored in the memory to implement the method for generating a clip video.

[0023] In another aspect of the present application, a computer readable storage medium is provided, which stores instructions, and when the instructions are executed on a computer, the computer executes the method for generating a clip video.

[0024] The embodiment of the present application provides a generation method of a split video, which obtains a barrage high-energy word table and a barrage of a video, and can provide data support for further data matching, wherein the barrage high-energy word table comprises a plurality of barrage high-energy words; if the barrage comprises a barrage high-energy word, the barrage is determined as a high-energy barrage; a matching information table corresponding to the barrage high-energy word corresponding to the high-energy barrage is obtained, so as to provide a data basis for further calculation of a barrage high-energy index per second; according to the matching information table corresponding to the high-energy barrage, the barrage high-energy index per second is calculated, the barrage high-energy index per second can reflect the wonderful degree of the high-energy barrage per second in the video barrage, and is beneficial to determining a high-energy interval of the video; according to the barrage high-energy index per second, a smooth high-energy barrage curve about the barrage high-energy index per second is fitted; a video interval of the video, in which the smooth high-energy barrage curve exceeds a preset split threshold, is determined as the high-energy interval of the video, the smooth high-energy barrage curve fitted by the barrage high-energy index per second can directly and quickly determine the high-energy interval of the video, and then a split video is generated according to the high-energy interval. The embodiment of the present application fully utilizes the emotional information in the user barrage by analyzing the barrage information of the video, so that the high-energy interval (that is, the wonderful segment of the video) of the video can be accurately and quickly recognized and split, the speed and accuracy of splitting the high-energy interval of the video are improved, and then a plurality of split videos containing the high-energy interval are obtained. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description.

[0026] Figure 1 A video wonderful segment schematic diagram provided in the embodiment of the present application;

[0027] Figure 2 A step flowchart of the generation method of the split video provided in the embodiment of the present application;

[0028] Figure 3 A matching information table corresponding to the barrage high-energy word provided in the embodiment of the present application;

[0029] Figure 4 A step flowchart of another generation method of the split video provided in the embodiment of the present application;

[0030] Figure 5 A sub-step flowchart of another generation method of the split video provided in the embodiment of the present application;

[0031] Figure 6 A high-energy barrage curve schematic diagram provided in the embodiment of the present application;

[0032] Figure 7A smooth high-energy barrage curve diagram provided in an embodiment of the present application is shown in the figure;

[0033] Figure 8 A high-energy interval interception diagram provided in an embodiment of the present application is shown in the figure;

[0034] Figure 9 A flow diagram of a strip video generation method provided in an embodiment of the present application is shown in the figure;

[0035] Figure 10 A structural block diagram of a strip video generation device in an embodiment of the present application is shown in the figure;

[0036] Figure 11 A structural diagram of an electronic device embodiment in an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present application will be described below with reference to the accompanying drawings.

[0038] Most of the current video highlight content stripping schemes are pure visual schemes, for example, by identifying pre-defined actions in the video to extract the highlights in the video, the limitation of this method is that only pre-defined actions can be identified, and the processing content is limited; in addition, the highlight content in part of the video is the brilliant dialogue of the actors, and the actors often do not have special actions when saying brilliant lines. For this situation, the current action recognition-based stripping scheme cannot handle it.

[0039] Referring to Figure 1 , a video highlight segment diagram provided in an embodiment of the present application is shown in the figure, in which two people in this video segment are only talking and have no special actions, so the stripping scheme based on action recognition cannot find this highlight segment. But from the video segment of Figure 1 we can easily get that the key information such as "classic" and "famous scene" appears in many places in the video barrage, it can be understood that the barrage in the video is the embodiment of the emotional resonance of the user when watching the video, and the content of the barrage can reflect the emotional characteristics of the video content to a certain extent, so it can be judged that this is a highlight segment in the video, therefore, the stripping scheme based on barrage analysis will be adopted in the present scheme embodiment to split it, and the split video segment is taken as a short video for the user to enjoy.

[0040] One of the core ideas of the embodiments of the present application is that by acquiring the barrage high-energy word table and the barrage of the video, data support can be provided for further data matching, wherein the barrage high-energy word table comprises a plurality of barrage high-energy words; if the barrage contains a barrage high-energy word, the barrage is determined as a high-energy barrage; the matching information table corresponding to the barrage high-energy word corresponding to the high-energy barrage is acquired, which provides a data basis for further calculation of the barrage high-energy index per second; according to the matching information table corresponding to the high-energy barrage, the barrage high-energy index per second is calculated, and through the barrage high-energy index per second, the brilliance degree of the high-energy barrage per second in the video barrage can be reflected, which is beneficial to determining the high-energy interval of the video; according to the barrage high-energy index per second, a smooth high-energy barrage curve about the barrage high-energy index per second is fitted; the video interval of the video whose smooth high-energy barrage curve exceeds the preset segmentation threshold is determined as the high-energy interval of the video, and the smooth high-energy barrage curve fitted by the barrage high-energy index per second can directly and quickly determine the high-energy interval of the video; and then the high-energy interval is used to generate a split video. The embodiments of the present application analyze the barrage information of the video, fully utilize the emotional information in the user barrage, and thus can accurately and quickly identify and split the high-energy interval (i.e. the brilliant segment) of the video, improve the speed and accuracy of splitting the high-energy interval of the video, and then obtain a plurality of split videos containing the high-energy interval.

[0041] Specifically, referring to Figure 2 , a step flowchart of a method for generating a split video provided in an embodiment of the present application is shown, which can specifically include the following steps:

[0042] Step 201, in response to the acquisition operation of the barrage high-energy word table, a barrage high-energy word table is obtained; wherein the barrage high-energy word table comprises a plurality of barrage high-energy words;

[0043] Wherein the barrage high-energy word table is composed of a plurality of emotional tags; for the emotional tags, they can reflect the characteristics of the emotional category, and at the same time contain a plurality of barrage high-energy words in the emotional tags. It can be understood that there are a number of barrage high-energy words in each emotional tag that can reflect the characteristics of the emotional category.

[0044] For the barrage high-energy word, it can be acquired by analyzing the characteristics of a number of video segments of different emotional categories and their corresponding barrage. It should be noted that in the embodiments of the present application, the acquisition method of the barrage high-energy word can be a way of artificially analyzing the characteristics of a number of video segments of different emotional categories and their corresponding barrage. The high-energy barrage word is selected by subjective consciousness, which can better reflect the high correlation between the barrage high-energy word and the user's emotion. It can be understood that the person skilled in the art can select the acquisition method of the barrage high-energy word according to the actual situation, and the embodiments of the present application do not limit this.

[0045] It should be noted that the naming of the emotional label and the selection of the barrage high-energy word can be adjusted by the person skilled in the art according to the actual situation, and the embodiment of the present application does not limit this.

[0046] In the embodiment of the present application, the barrage high-energy word table is obtained in response to the acquisition operation of the barrage high-energy word table. Through the acquisition of the barrage high-energy word table, data support can be provided for further data matching. The barrage high-energy word table contains multiple emotional labels, and the emotional labels contain multiple barrage high-energy words. That is, there are several barrage high-energy words in each emotional label that can reflect the characteristics of the emotional category. Therefore, the hit rate of the barrage in the video segment that hits the barrage high-energy word corresponding to the emotional label can be increased, and the accuracy of identifying the highlight segment of the video can be improved.

[0047] Step 202, acquiring the barrage of the video;

[0048] The video is a complete video that has not been split. In the embodiment of the present application, the complete video is mainly split to obtain a split video containing a highlight segment. It can be understood that the barrage of the video is part or all of the barrage in the complete video.

[0049] Step 203, if the barrage contains the barrage high-energy word, determining the barrage as a high-energy barrage;

[0050] In an optional embodiment of the present application, if the barrage contains the barrage high-energy word in step 203, the barrage is determined as a high-energy barrage. Specifically, it can include the following sub-steps:

[0051] Sub-step S11, judging whether the barrage of the video contains the barrage high-energy word in the barrage high-energy word table;

[0052] Sub-step S12, if the barrage contains the barrage high-energy word, determining the barrage as a high-energy barrage.

[0053] The high-energy barrage is the barrage containing the barrage high-energy word. For example, assuming that a video barrage is “he is really too fond of her, it is so sweet”, the barrage can hit the barrage high-energy words “too fond” and “so sweet” under the emotional label “sweet” in the barrage high-energy word table. Therefore, the barrage is marked as a high-energy barrage.

[0054] In a specific implementation, it is judged whether the barrage of the video contains the barrage high-energy word in the barrage high-energy word table. If the barrage contains the barrage high-energy word, the barrage is determined as a high-energy barrage. It can be understood that the barrage of the video and the barrage high-energy word in the barrage high-energy word table can be matched. If the barrage contains the barrage high-energy word, it can be determined that the barrage and the barrage high-energy word have been matched successfully, and then the matched barrage can be marked as a high-energy barrage.

[0055] In the embodiment of the present application, by acquiring the barrage high-energy word table and all barrages in the video, the barrages in the video segment are matched with the barrage high-energy words in the barrage high-energy word table in sequence. Specifically, the barrages of the video can be matched with the barrage high-energy words in the barrage high-energy word table. If a barrage contains a barrage high-energy word, it is considered that the barrage and the barrage high-energy word have been successfully matched. Then, the matched barrage can be confirmed as a high-energy barrage, and the barrage is marked. By marking the barrage as a high-energy barrage, whether the barrage is a high-energy barrage can be quickly determined.

[0056] In step 204, a matching information table corresponding to the barrage high-energy word corresponding to the high-energy barrage is acquired.

[0057] The matching information table can be acquired from the process of marking the barrage in the video as a high-energy barrage. The matching information table can include a high-energy barrage, a time point of the appearance of the high-energy barrage, an identification of a hit barrage high-energy word, an emotional label corresponding to the high-energy barrage, and a barrage high-energy word hit by the high-energy barrage.

[0058] Referring to Figure 3 , a matching information table corresponding to a barrage high-energy word is shown, which is provided in the embodiment of the present application, Figure 3 In the matching information table, the first column is a high-energy barrage, the second column is a time point of the appearance of the high-energy barrage, the third column is an identification of a hit barrage high-energy word, the fourth column is an emotional label corresponding to the high-energy barrage, and the fifth column is a barrage high-energy word hit by the high-energy barrage. The last three rows of data correspond to the case that one barrage corresponds to multiple emotional labels.

[0059] In an example, it is assumed Figure 3 The keyword "want to cry" in the first barrage "All soldiers hold spears and swords against Xiaonanchen Wang Zhou Shengchen, the shadow in the trailer wants to cry" in the matching information table can be matched to the "want to cry" barrage high-energy word in the barrage high-energy word table. The barrage is marked as a high-energy barrage, and the high-energy barrage can be classified into the "sad and cruel" emotional label corresponding to the "want to cry" barrage high-energy word. In addition, for the case that one barrage corresponds to multiple emotional labels, for example, there are "moved" and "sad and cruel" emotional labels, and the two emotional labels contain the "wuwuwu" barrage high-energy word, then Figure 3 The "wuwuwu" contained in the barrage "wuwuwu, good to see, my beautiful Xiaonanchen Wang Zhou Shengchen!" in the matching information table can be matched to the "moved" and "sad and cruel" emotional labels.

[0060] In an optional embodiment of the present application, the step 204 of acquiring the matching information table corresponding to the barrage high-energy word corresponding to the high-energy barrage can specifically include the following substeps:

[0061] In the sub-step S21, the barrage in the video is matched with the barrage high-energy words in the barrage high-energy word table to obtain one or more matching results corresponding to the barrage high-energy words; the matching result is a matching result corresponding to a high-energy barrage that matches successfully.

[0062] In the sub-step S22, the matching results corresponding to each high-energy barrage are counted to obtain a matching information table corresponding to the barrage high-energy words corresponding to the high-energy barrage.

[0063] The matching information table is a result of matching the barrage with the barrage high-energy words in the barrage high-energy word table, that is, the matching results of the barrage high-energy words can be summarized or counted in the form of the matching information table.

[0064] In the specific implementation, the barrage in the video is matched with the barrage high-energy words in the barrage high-energy word table to obtain one or more matching results corresponding to the barrage high-energy words; the matching result is a matching result corresponding to a high-energy barrage that matches successfully, and the matching results corresponding to each high-energy barrage are counted to obtain a matching information table corresponding to the barrage high-energy words corresponding to the high-energy barrage.

[0065] In the embodiment of the application, by obtaining the barrage high-energy word table and all barrages in the video, the barrages in the video segment are matched with the barrage high-energy words in the barrage high-energy word table in sequence, if a barrage high-energy word is contained in a barrage, the barrage is considered to be a high-energy barrage, and the barrage is marked, the matching information table corresponding to the barrage high-energy words corresponding to the high-energy barrage is obtained by marking the barrage as a high-energy barrage, which provides data support for further calculating the barrage high-energy index per second.

[0066] In step 205, the barrage high-energy index per second is calculated according to the matching information table corresponding to the high-energy barrage.

[0067] The barrage high-energy index per second can reflect the wonderful degree of the video segment per unit time, and it should be noted that the barrage high-energy index per second is in direct proportion to the wonderful degree of the video segment, that is, the higher the barrage high-energy index per second, the higher the wonderful degree of the corresponding video segment.

[0068] In the embodiment of the application, the barrage high-energy index per second is calculated according to the matching information table corresponding to the high-energy barrage, wherein the barrage high-energy index per second can reflect the wonderful degree of the high-energy barrage per second in the video barrage, specifically, the higher the barrage high-energy index per second, the higher the wonderful degree of the corresponding video segment, which provides data support for further fitting the smooth high-energy barrage curve about the barrage high-energy index per second, and is beneficial to determining the high-energy interval of the video.

[0069] In step 206, a smooth high-energy bullet curve about the bullet high-energy index per second is fitted according to the bullet high-energy index per second.

[0070] For the smooth high-energy bullet curve, it can be expressed as a value corresponding to the bullet high-energy index per second in a unit time, and it can be understood that the exciting degree of the video segment can be determined according to the value of the bullet high-energy index per second.

[0071] In the embodiment of the present application, the bullet high-energy index per second is calculated through the matching information table corresponding to the high-energy bullet, wherein the bullet high-energy index per second can reflect the exciting degree of the high-energy bullet per second in the video bullet, so that the smooth high-energy bullet curve about the bullet high-energy index per second can be fitted according to the bullet high-energy index per second, and then the high-energy interval of the video can be reflected through the smooth high-energy bullet curve.

[0072] In step 207, the video interval of the video whose smooth high-energy bullet curve exceeds the preset segmentation threshold is determined as the high-energy interval of the video.

[0073] The preset segmentation threshold can be a value set in advance, and the higher the threshold of the preset segmentation threshold, the higher the requirement for the exciting degree of the video. The person skilled in the art can adjust it according to the actual situation, and the present application does not limit it.

[0074] The video interval can be a segment in the video, and the high-energy interval can be an exciting segment in the video segment.

[0075] In an example, assuming that the three bullet high-energy indexes per second corresponding to the smooth high-energy bullet curve at the 340th second, the 341st second and the 342nd second are 2, 3 and 4 respectively, and the preset segmentation threshold is 2.6, then the video interval corresponding to the 341st second and the 342nd second can be determined as the high-energy interval of the video, i.e. the exciting segment in the video. By analogy, assuming that the preset segmentation threshold is set to a value greater than 4, then there is no high-energy interval of the video in the video segment, i.e. there is no exciting segment in the video.

[0076] In the embodiment of the present application, the video interval of the video whose smooth high-energy bullet curve exceeds the preset segmentation threshold is determined as the high-energy interval of the video, so that the exciting segment of the video can be quickly determined.

[0077] In step 208, the high-energy intervals in the video whose time interval is in a preset unit time and adjacent are merged to obtain a merged high-energy interval, and the merged high-energy interval is taken as a split video.

[0078] For the time interval, it can be the time length between two adjacent high-energy intervals; for the preset unit time, those skilled in the art can adjust it according to the actual situation, and the implementation of the present application does not limit it.

[0079] Wherein, for the merged high-energy interval, it can be the high-energy interval obtained by merging two adjacent high-energy intervals; for the clip video, it can be the video segment containing the high-energy interval cut from the long video, that is, the clip video contains the highlight segment of the video.

[0080] In an example, assuming that the preset unit time is 5 seconds and the time interval of two adjacent high-energy intervals is 3 seconds, the two adjacent high-energy intervals can be merged to obtain a merged high-energy interval, that is, two merged highlight segments, and the merged high-energy interval is taken as a clip video. Similarly, assuming that the time interval of two adjacent high-energy intervals is 8 seconds, the two adjacent high-energy intervals cannot be merged.

[0081] In the embodiment of the present application, by obtaining the bullet screen high-energy word table and the bullet screen of the video, data support can be provided for further data matching, wherein the bullet screen high-energy word table contains a plurality of bullet screen high-energy words; if the bullet screen contains a bullet screen high-energy word, the bullet screen is determined as a high-energy bullet screen; the matching information table corresponding to the bullet screen high-energy word corresponding to the high-energy bullet screen is obtained, which provides a data basis for further calculation of the per-second bullet screen high-energy index; according to the matching information table corresponding to the high-energy bullet screen, the per-second bullet screen high-energy index is calculated, and through the per-second bullet screen high-energy index, the highlight degree of the per-second high-energy bullet screen in the video bullet screen can be reflected, which is beneficial to determine the high-energy interval of the video; according to the per-second bullet screen high-energy index, a smooth high-energy bullet screen curve about the per-second bullet screen high-energy index is fitted; the video interval of the video whose smooth high-energy bullet screen curve exceeds the preset segmentation threshold is determined as the high-energy interval of the video, and the smooth high-energy bullet screen curve fitted by the per-second bullet screen high-energy index can directly and quickly determine the high-energy interval of the video, and then generate the clip video according to the high-energy interval. The embodiment of the present application analyzes the bullet screen information of the video, fully utilizes the emotional information in the user's bullet screen, so as to accurately and quickly identify and split the high-energy interval (i.e. the highlight segment) of the video, improve the speed and accuracy of splitting the high-energy interval of the video, and then obtain a plurality of clip videos containing the high-energy interval.

[0082] Referring to Figure 4 , a step flow chart of another method for generating a clip video provided in the embodiment of the present application is shown;

[0083] Step 501, in response to the obtaining operation of the bullet screen high-energy word table, the bullet screen high-energy word table is obtained; wherein the bullet screen high-energy word table contains a plurality of bullet screen high-energy words;

[0084] In the embodiment of the present application, in response to the obtaining operation of the barrage high-energy word table, the barrage high-energy word table is obtained, and through obtaining the barrage high-energy word table and all the barrages in the video, data support can be provided for further data matching, wherein the barrage high-energy word table contains multiple emotional labels, and multiple barrage high-energy words are contained in the emotional labels, that is, in each emotional label, there are several barrage high-energy words that can reflect the characteristics of the emotional category, so as to increase the hit rate of the barrage in the video segment hitting the barrage high-energy words corresponding to the emotional label, and improve the accuracy of identifying the highlight segment of the video. The foregoing embodiment has explained and described the related content in detail, and will not be repeated here.

[0085] Step 502, obtaining the barrage of the video;

[0086] Step 503, if the barrage contains the barrage high-energy word, the barrage is determined as a high-energy barrage;

[0087] In the embodiment of the present application, through obtaining the barrage high-energy word table and all the barrages in the video, the barrage high-energy words in the barrage high-energy word table are used to match all the barrages in the video segment in turn, specifically, the barrage of the video and the barrage high-energy words in the barrage high-energy word table can be matched, if a barrage contains a barrage high-energy word, it is considered that the barrage and the barrage high-energy word have been matched successfully, and then the matched barrage can be confirmed as a high-energy barrage, which can quickly distinguish whether the barrage is a high-energy barrage. The foregoing embodiment has explained and described the related content in detail, and will not be repeated here.

[0088] Step 504, obtaining the matching information table corresponding to the barrage high-energy word corresponding to the high-energy barrage; the matching information table contains the appearance time point of the high-energy barrage, and the appearance time point of the high-energy barrage corresponds to other barrages;

[0089] Wherein, the matching information table is the matching result of the barrage and the barrage high-energy word in the barrage high-energy word table, that is, the matching result of the barrage high-energy word can be summarized or counted through the matching information table; for the appearance time point of the high-energy barrage, it is the time point when the high-energy barrage appears in the video segment; for other barrages, they are other barrages corresponding to the appearance time point of the high-energy barrage, that is, in addition to the barrage marked as a high-energy barrage, the barrage in the video also includes other unmarked barrages.

[0090] In an example, referring to Figure 3At the 340th second, there can be two high-energy bullet screens, "Will the general's bitterness be understood? I feel sorry for Zhou Shengchen" and "All soldiers hold spears and lances against Xiaonancheng, the shadow in the trailer, and want to cry after watching it", but actually, at the 340th second, there can be more high-energy bullet screens or other bullet screens, and it can be understood that the other bullet screens are the other bullet screens that are not marked as high-energy bullet screens, that is, there should be high-energy bullet screens and other bullet screens at the time point of the appearance of the high-energy bullet screens.

[0091] It should be noted that the data in the above embodiments Figure 3 Only the matching information table for the high-energy bullet screen is screened, and it can be understood that the above-mentioned examples are only as an example, and are set to be relatively simple for convenient explanation, and therefore each item of data is set to be relatively simple, and in actual application, the high-energy bullet screen or the other bullet screen corresponding to a time point can be more than the examples listed above, that is, a time point can correspond to high-energy bullet screens and other bullet screens.

[0092] In the embodiment of the application, the matching information table corresponding to the high-energy bullet screen is obtained, the time point of the appearance of the high-energy bullet screen in the matching information table is obtained, and the other bullet screen corresponding to the time point of the appearance of the high-energy bullet screen is obtained, which provides convenience for further statistics of the number of bullet screens.

[0093] In step 505, the high-energy bullet screen index per second is calculated according to the matching information table corresponding to the high-energy bullet screen.

[0094] In the embodiment of the application, the high-energy bullet screen index per second can be calculated through the matching information table corresponding to the high-energy bullet screen, the high-energy bullet screen index per second can reflect the wonderful degree of the high-energy bullet screen per second in the video bullet screen, data support is provided for further fitting of the smooth high-energy bullet screen curve about the high-energy bullet screen index per second, and it is beneficial to determine the high-energy interval of the video.

[0095] In an optional embodiment of the application, referring to Figure 5 The substep flowchart of another method for generating a split video provided in the embodiment of the application is shown, and the step of calculating the high-energy bullet screen index per second in step 505 can specifically include the following substeps.

[0096] In substep S31, the video duration of the video is obtained.

[0097] In substep S32, the number of high-energy bullet screens in the video is counted to obtain a first number of bullet screens.

[0098] In substep S33, the number of high-energy bullet screens corresponding to the time point of the appearance of the high-energy bullet screen is counted to obtain a second number of bullet screens.

[0099] Sub-step S34, counting the second number of bullet screens and the number of other bullet screens corresponding to the time point of appearance of the high-energy bullet screen to obtain a third number of bullet screens;

[0100] Sub-step S35, taking the ratio of the first number of bullet screens to the video duration of the video as a first ratio;

[0101] Sub-step S36, taking the ratio of the second number of bullet screens to the third number of bullet screens as a second ratio;

[0102] Sub-step S37, taking the ratio of the second number of bullet screens to the first ratio as a third ratio;

[0103] Sub-step S38, adding the second ratio and the third ratio to obtain a value as the high-energy index of bullet screens per second.

[0104] Among them, for the video duration, it is the video duration of the entire video.

[0105] For the first number of bullet screens, it is the number of high-energy bullet screens in the video. Specifically, the number of high-energy bullet screens in the video can be counted according to the bullet screen information recorded in the matching information table corresponding to the high-energy bullet screen, which is used to calculate the high-energy index of bullet screens per second.

[0106] For the second number of bullet screens, it is the number of high-energy bullet screens corresponding to the time point of appearance of the high-energy bullet screen in the video. For example, there are 2 high-energy bullet screens corresponding to the 340th second, so the number of high-energy bullet screens corresponding to the 340th second is 2. Specifically, the number of high-energy bullet screens corresponding to the time point of appearance of the high-energy bullet screen in the video can be counted according to the time point of appearance of the high-energy bullet screen recorded in the matching information table corresponding to the high-energy bullet screen, which is used to calculate the high-energy index of bullet screens per second.

[0107] For the third number of bullet screens, it is the number of all bullet screens corresponding to the time point of appearance of the high-energy bullet screen in the video. For example, in the 340th second corresponding to the high-energy bullet screen, the number of high-energy bullet screens is 2 and the number of other bullet screens is 8, so there are a total of 10 bullet screens appearing in the 340th second. Specifically, the number of other bullet screens corresponding to the time point of appearance of the high-energy bullet screen in the video can be counted according to the time point of appearance of the high-energy bullet screen recorded in the matching information table corresponding to the high-energy bullet screen, so as to obtain the number of all bullet screens corresponding to the time point of appearance of the high-energy bullet screen in the video, which is used to calculate the high-energy index of bullet screens per second.

[0108] For the first ratio, it is the ratio of the first number of bullet screens to the video duration, which can represent the average number of high-energy bullet screens per second of the video. It should be noted that since the first ratio can represent the average number of high-energy bullet screens per second of the video, the first ratio can be an integer or a decimal, and therefore, the value of the average number of high-energy bullet screens per second of the video can be adjusted or selected by a person skilled in the art according to the actual situation, and the embodiments of the present application do not limit this.

[0109] For the second ratio, it is the ratio of the second number of bullet screens to the third number of bullet screens. Specifically, the second ratio is the ratio of the number of high-energy bullet screens corresponding to the time point of appearance of the high-energy bullet screens to the number of all bullet screens corresponding to the time point of appearance of the high-energy bullet screens.

[0110] For the third ratio, it is the ratio of the second number of bullet screens to the first ratio. Specifically, the first ratio can represent the average number of high-energy bullet screens per second of the video, and the third ratio is the ratio of the number of high-energy bullet screens corresponding to the time point of appearance of the high-energy bullet screens to the average number of high-energy bullet screens per second of the video.

[0111] Among them, for the high-energy index of bullet screens per second, it can reflect the wonderful degree of the corresponding high-energy bullet screens per second in the video bullet screens, which is beneficial to determine the high-energy interval of the video.

[0112] In a specific implementation, the video duration of the video is obtained, the number of high-energy bullet screens in the video is counted to obtain the first number of bullet screens, that is, the number of all high-energy bullet screens in the video, the number of high-energy bullet screens corresponding to the time point of appearance of the high-energy bullet screens is counted to obtain the second number of bullet screens, that is, the number of high-energy bullet screens corresponding to the unit time, the second number of bullet screens and the number of other bullet screens corresponding to the time point of appearance of the high-energy bullet screens are counted to obtain the third number of bullet screens, that is, the number of all bullet screens corresponding to the unit time, and then the ratio of the first number of bullet screens to the video duration of the video is taken as the first ratio, the ratio of the second number of bullet screens to the third number of bullet screens is taken as the second ratio, and the ratio of the second number of bullet screens to the first ratio is taken as the third ratio, so that the value obtained by adding the second ratio and the third ratio is taken as the high-energy index of bullet screens per second; wherein the high-energy index of bullet screens per second and the number of bullet screens have the following relationship:

[0113]

[0114] : represents the number of high-energy bullet screens corresponding to the t-th second, that is, the second number of bullet screens described above;

[0115] : represents the total number of bullet screens of the t-th second, that is, the third number of bullet screens described above;

[0116] represents the proportion of the number of high-energy bullet screens in the t-th second to the total number of bullet screens in the t-th second, that is, the second ratio described above, and reflects the exciting degree in the time;

[0117] represents the proportion of the number of high-energy bullet screens in the t-th second to the average number of high-energy bullet screens per second of the video, that is, the third ratio described above, and reflects the exciting degree of the content in the t-th second compared with the whole film (the entire video).

[0118] is a proportion factor, used to adjust the weights of the two formulas, which can be set according to actual conditions.

[0119] is a proportion factor, used to adjust the weights of the two formulas, which can be set according to actual conditions.

[0120] In an example, assuming that the number of all high-energy bullet screens in the video is 50, the video duration is 5 seconds, the number of high-energy bullet screens corresponding to the 340th second of the video is 6, the number of other bullet screens corresponding to the 340th second is 4, and the total number of bullet screens corresponding to the 340th second is 10, then the proportion of the number of high-energy bullet screens in the 340th second to the total number of bullet screens in the 340th second can be 6 / 10=0.6, and by analogy, the proportion of the number of high-energy bullet screens in the 340th second to the average number of high-energy bullet screens per second of the video can be 6 / (50 / 5)=0.6. In addition, assuming that =2, =1, then the high-energy bullet index per second is 0.6*2+0.6*1=1.8.

[0121] It should be noted that the examples described above are only examples for convenience of explanation, and therefore the data are set to be relatively simple, and in actual application, a person skilled in the art can adjust the parameters according to actual conditions, and the embodiments of the present application do not limit this.

[0122] In the embodiments of the present application, the video duration of the video is obtained, the number of high-energy bullet screens in the video is counted to obtain a first number of bullet screens according to the obtained matching information table corresponding to the high-energy bullet word, the number of high-energy bullet screens corresponding to the time point of appearance of the high-energy bullet screens is counted to obtain a second number of bullet screens, the second number of bullet screens and the number of other bullet screens corresponding to the time point of appearance of the high-energy bullet screens are counted to obtain a third number of bullet screens, and then the ratio of the first number of bullet screens to the video duration of the video is taken as a first ratio, the ratio of the second number of bullet screens to the third number of bullet screens is taken as a second ratio, and the ratio of the second number of bullet screens to the first ratio is taken as a third ratio, so that the value obtained by adding the second ratio and the third ratio can be taken as a high-energy bullet index per second, and the high-energy bullet index per second is conducive to determining the high-energy interval of the video.

[0123] Step 506: Fit all the high-energy indexes per second of the video into a high-energy barrage curve;

[0124] The high-energy barrage curve can be a curve fitted by the high-energy index of each barrage per second calculated above.

[0125] In a specific implementation, after calculating the high-energy index of each bullet comment per second in the video through the steps of the above embodiments, all the high-energy indexes of bullet comments per second in the video are connected by line segments to obtain a high-energy bullet comment curve graph with respect to the high-energy index of bullet comments per second.

[0126] Reference Figure 6 The diagram illustrates a high-energy barrage curve provided in an embodiment of the present invention. Figure 6 The curve in the graph is composed of the high-energy index of all bullet comments per second in the video. The horizontal axis represents the time point when the high-energy bullet comments appear, and the vertical axis represents the high-energy index of bullet comments per second in the video. As can be seen from the graph, the values ​​are uneven, and the higher the value, the more exciting the corresponding video segment is.

[0127] Step 507: Smooth the high-energy barrage curve to obtain a smoothed high-energy barrage curve;

[0128] The smoothed high-energy bullet screen curve is obtained by smoothing the high-energy bullet screen curve. In this embodiment of the invention, the high-energy bullet screen curve can be smoothed by the moving average method. Specifically, it is necessary to calculate the average value of the current value within a certain window, and then use this average value to replace the current value, so that the average value is used as the smoothed value to depict the curve. For example, Figure 6 High-energy bullet screen clips often have many jagged edges on their curves. Directly using a preset segmentation threshold to segment them can easily break a complete and exciting video clip into several smaller segments, causing disjointed storylines. Therefore, it is advisable to... Figure 6 The high-energy barrage curve is calculated by averaging the high-energy index per second (0-6) corresponding to the barrage per second within a unit time period (e.g., horizontal axis: 0-2500 seconds). This average value is then used as the optimized value to depict a smooth high-energy barrage curve. Figure 7 The diagram illustrates a smooth high-energy barrage curve provided in an embodiment of the present invention, such as... Figure 7 As shown, Figure 7 Compared to Figure 6 The raised peaks become flatter, and the concave parts are also lifted, making the overall curve appear to have a smoother transition and reducing the impact of some extreme values ​​(such as...). Figure 6The "burr" is convex). It should be noted that, for the method of smoothing the high-energy barrage curve into a smooth high-energy barrage curve, the sliding window average method can be used, or other curve smoothing methods such as the quadratic exponential smoothing method and the five-point cubic smoothing method can be used. Those skilled in the art can select according to the actual situation, and the embodiments of the present application do not limit this.

[0129] It can be understood that, since there are relatively many burrs in the high-energy barrage curve, directly using a preset segmentation threshold to segment it can easily segment a complete wonderful video segment into several small segments, causing the plot to be incoherent, so the high-energy barrage curve needs to be smoothed first. The smooth high-energy barrage curve is the same as the high-energy barrage curve, the horizontal axis also represents the appearance time point of the high-energy barrage, and the vertical axis also represents the high-energy index of the barrage per second in the video. The higher the corresponding value, the more wonderful the video segment corresponding to the value.

[0130] In a specific implementation, after calculating each high-energy index of the barrage per second in the video through the steps of the above embodiments, all high-energy indexes of the barrage per second in the video are connected by line segments to obtain a high-energy barrage curve of the high-energy index of the barrage per second, and then the high-energy barrage curve is smoothed to obtain a smooth high-energy barrage curve, which can ensure the coherence of the plot, i.e., the coherence of the video segment.

[0131] Step 508, determining the video interval of the video where the smooth high-energy barrage curve exceeds the preset segmentation threshold as the high-energy interval of the video;

[0132] The high-energy interval is the video interval where the smooth high-energy barrage curve exceeds the preset segmentation threshold, which can be understood as a wonderful segment in the video.

[0133] Referring to Figure 8 , a schematic diagram of intercepting a high-energy interval is shown, the dashed line (ratio) represents a smooth high-energy barrage curve, the horizontal solid line (thres) represents a preset segmentation threshold, the solid line (ratio_cut) represents an actually selected high-energy interval, the horizontal axis represents the appearance time point of the high-energy barrage, and the vertical axis represents the high-energy index of the barrage per second in the video.

[0134] In an example, the part where the high-energy index of the barrage per second is higher than the preset segmentation threshold is considered as a high-energy interval, but in order to ensure the integrity of the video, the solid line (ratio_cut) is actually selected as the initial result of the high-energy interval, as shown in Figure 8 Figure 8 ​As shown, there is a part of the per-second barrage high-energy index higher than the preset segmentation threshold value in about the 120th-200th second, and there is a part of the per-second barrage high-energy index higher than the preset segmentation threshold value in about the 400th-500th second, so the two video intervals corresponding to the 120th-200th second and the 400th-500th second can be taken as the high-energy interval of the video, i.e., the wonderful segment in the video.

[0135] In the embodiment of the present application, after the smooth high-energy barrage curve is obtained, the video interval in which the smooth high-energy barrage curve exceeds the preset segmentation threshold value is determined as the high-energy interval of the video, and the position of the high-energy interval can be intuitively and quickly determined through the smooth high-energy barrage curve.

[0136] In an optional embodiment of the present application, after the video interval of the video in which the smooth high-energy barrage curve exceeds the preset segmentation threshold value is determined as the high-energy interval of the video in the step 508, the method further comprises:

[0137] The starting point of the high-energy interval is updated to the first point at which the per-second barrage high-energy index starts to change from falling to rising, and the ending point of the high-energy interval is updated to the second point at which the per-second barrage high-energy index starts to change from falling to rising.

[0138] In the actual implementation, considering the lag in bullet comments and to ensure complete coverage of the highlights, the starting point of the high-energy interval is updated to the first point where the bullet comment high-energy index per second begins to rise from a decrease, and the ending point is updated to the second point where the bullet comment high-energy index per second begins to rise from a decrease. Specifically, the first point where the bullet comment high-energy index per second begins to rise from a decrease is the first point where the same smooth high-energy bullet comment curve changes from a decrease to a rise, and the second point where the bullet comment high-energy index per second begins to rise is the second point where the same smooth high-energy bullet comment curve changes from a decrease to a rise. Simultaneously, the high-energy interval is the video segment where the smooth high-energy bullet comment curve corresponding to the bullet comment high-energy index per second exceeds a preset segmentation threshold. In other words, the high-energy interval is the most exciting segment within the highlights, and the bullet comment high-energy index per second... The "high-energy index" can be understood as the entire highlight segment. That is, by tracing the starting point of the high-energy interval back to the point where the per-second barrage high-energy index begins to rise from a decrease, we can avoid situations where there is insufficient or disjointed plot. In other words, we are aligning the starting point of a small highlight segment with the starting point of the entire highlight segment. Similarly, updating the ending point of the high-energy interval to the point where the per-second barrage high-energy index begins to rise can be understood as aligning the ending point of a small highlight segment with the ending point of the entire highlight segment. By updating the starting and ending points of the high-energy interval to the starting and ending points of the per-second barrage high-energy index, we avoid situations where there is insufficient or disjointed plot, ensuring the integrity of the high-energy interval.

[0139] like Figure 8 As shown, assuming the selected high-energy interval is the video interval corresponding to the time period from the 120th second to the 200th second, that is, the first bulge in the figure that exceeds the horizontal solid line (thres preset segmentation threshold), but in the actual process of selecting the high-energy interval of the video, considering the lag of the bullet comments, and in order to fully cover the exciting parts, the starting point of the bullet comment high-energy index per second corresponding to the first bulge starting to change from decreasing to increasing is taken as the starting point of the high-energy interval (the first bulge). Similarly, the second point of the bullet comment high-energy index per second corresponding to the first bulge starting to change from decreasing to increasing is taken as the ending point, and thus as the ending point of the high-energy interval (the first bulge).

[0140] Step 509: Merge adjacent high-energy intervals in the video with time intervals within a preset unit time to obtain merged high-energy intervals;

[0141] The time interval can be the length of time between two adjacent high-energy intervals; the preset unit time can be adjusted by those skilled in the art according to the actual situation, and the implementation of the present invention does not limit this.

[0142] The high-energy interval can be obtained by merging two adjacent high-energy intervals, and the cut video can be a video segment containing a high-energy interval cut from a long video, i.e., the cut video contains a highlight segment of the video.

[0143] In an example, assuming that the preset unit time is 5 seconds and the time interval of the two adjacent high-energy intervals is 3 seconds, the two adjacent high-energy intervals can be merged, and thus a merged high-energy interval, i.e., two merged highlight segments, can be obtained.

[0144] In step 510, the start point of the merged high-energy interval is updated to the start time point of the shot corresponding to the merged high-energy interval, and the end point of the merged high-energy interval is updated to the end time point of the shot corresponding to the merged high-energy interval.

[0145] In an optional embodiment of the present application, the method further comprises:

[0146] The video is cut into a plurality of shots, and the start time point and the end time point of each shot in the video are obtained; wherein the shot contains the high-energy interval.

[0147] In order to ensure the integrity of the video, the video file needs to be analyzed first to obtain the start time point and the end time point of each shot in the video. Since the shot cutting algorithm is a mature algorithm, any shot cutting method can be used, and the algorithm details are not described herein.

[0148] In an example, assuming that the start point of the merged high-energy interval is 50 seconds and the end point is 70 seconds, wherein the start time point of the current shot corresponding to the merged high-energy interval is 40 seconds and the end time point is 80 seconds, the start point of the merged high-energy interval, i.e., 50 seconds, is updated to the start time point of the shot corresponding to the merged high-energy interval, i.e., 40 seconds, and the end point of the merged high-energy interval, i.e., 70 seconds, is updated to the end time point of the shot corresponding to the merged high-energy interval, i.e., 80 seconds, so as to ensure the integrity of the video.

[0149] In the embodiment of the present application, the start point of the merged high-energy interval is updated to the start time point of the shot corresponding to the merged high-energy interval, and the end point of the merged high-energy interval is updated to the end time point of the shot corresponding to the merged high-energy interval, so as to ensure the integrity of the cut video.

[0150] Step 511, taking the merged high-energy interval after updating the start point and the end point as a candidate segmented video, and taking the candidate segmented video with a video duration within a preset video duration as the segmented video.

[0151] In a specific implementation, the merged high-energy interval after updating the start point and the end point is taken as a candidate segmented video, and the candidate segmented video with a video duration within a preset video duration is taken as the segmented video, effectively ensuring the visual perception of the segmented video.

[0152] In an example, it is assumed that only the candidate segmented video with a video duration within a preset video duration is output, and the candidate segmented video that does not meet the duration requirement is discarded, where it is assumed that the preset video duration is 20 seconds and the preset video duration is 40 seconds. When the video duration of the candidate segmented video is 36 seconds, the candidate segmented video can be output as the segmented video.

[0153] In order for those skilled in the art to better understand the technical solutions of the embodiments of the present application, the following exemplary description is made through specific examples:

[0154] Referring to Figure 9 , a flowchart of a method for generating a segmented video is shown;

[0155] S41, obtaining a barrage high-energy word table;

[0156] The barrage high-energy word table includes a plurality of barrage high-energy words.

[0157] S42, obtaining the barrage of the video;

[0158] S43, barrage high-energy word matching;

[0159] If the barrage includes a barrage high-energy word, the barrage is determined as a high-energy barrage, and a matching information table corresponding to the barrage high-energy word corresponding to the high-energy barrage is obtained.

[0160] S44, high-energy barrage distribution statistics;

[0161] According to the matching information table corresponding to the high-energy barrage, the number of high-energy barrages in the video, the number of high-energy barrages corresponding to the time point of the appearance of the high-energy barrage, and the number of barrages corresponding to the time point of the appearance of the high-energy barrage are counted, and then the barrage high-energy index per second is calculated.

[0162] S45, smoothing the high-energy barrage curve;

[0163] According to the barrage high-energy index per second, a smooth high-energy barrage curve about the barrage high-energy index per second is fitted.

[0164] S46, intercepting the high-energy interval;​​

[0165] The video interval of the video whose smooth high-energy bullet screen curve exceeds the preset segmentation threshold is determined as a high-energy interval of the video.

[0166] S47, integrity processing of the split video;

[0167] The high-energy intervals in the video that are adjacent and within a preset unit time are merged to obtain a merged high-energy interval, and in order to ensure the integrity of the split video, the video is segmented into multiple shots in advance, the starting point of the merged high-energy interval is updated to the starting time point of the shot corresponding to the merged high-energy interval, and the ending point of the merged high-energy interval is updated to the ending time point of the shot corresponding to the merged high-energy interval, and the merged high-energy interval after updating the starting point and the ending point is taken as a candidate split video.

[0168] S48, time length filtering of the split video;

[0169] The candidate split video whose video time length is within a preset video time length is taken as a split video.

[0170] S49, output result, that is, output the split video.

[0171] In summary, the embodiment of the present application realizes the identification and splitting of the wonderful video segments in the long video by mining the bullet screen information of the video, can effectively solve the problem that the split is performed based on the traditional computer vision method, so that the wonderful dialogue content cannot be identified, and at the same time, the embodiment of the present application can also well identify other wonderful scenes in the video, such as fighting, through the matching of the high-energy words of the bullet screen. Compared with the method of analyzing the wonderful segments in the video based on the computer vision method, the embodiment of the present application has the advantage of short time consumption, for example, for a 45-minute episode with 100,000 bullet screens, the time for generating the split video is only 1-2 minutes, which is several times or even dozens of times faster than the method of splitting based on the computer vision.

[0172] In the embodiment of the present application, by acquiring the barrage high-energy word table and the barrage of the video, data support can be provided for further data matching, wherein the barrage high-energy word table comprises a plurality of barrage high-energy words; if the barrage contains a barrage high-energy word, the barrage is determined as a high-energy barrage; the matching information table corresponding to the barrage high-energy word corresponding to the high-energy barrage is acquired, which provides a data basis for further calculation of the barrage high-energy index per second; according to the matching information table corresponding to the high-energy barrage, the barrage high-energy index per second is calculated, and through the barrage high-energy index per second, the brilliant degree of the high-energy barrage per second in the video barrage can be reflected, which is beneficial to determining the high-energy interval of the video; according to the barrage high-energy index per second, the smooth high-energy barrage curve about the barrage high-energy index per second is fitted; the video interval of the video whose smooth high-energy barrage curve exceeds the preset segmentation threshold is determined as the high-energy interval of the video, and the smooth high-energy barrage curve fitted by the barrage high-energy index per second can directly and quickly determine the high-energy interval of the video, and then generate the split video according to the high-energy interval. The embodiment of the present application analyzes the barrage information of the video, fully utilizes the emotional information in the user barrage, so as to accurately and quickly identify and split the high-energy interval (i.e. the brilliant segment) of the video, improve the speed and accuracy of splitting the high-energy interval of the video, and then obtain a plurality of split videos containing the high-energy interval.

[0173] Meanwhile, the high-energy intervals in the video which are adjacent and have a time interval within a preset unit time are merged to obtain a merged high-energy interval, which is beneficial to ensuring the integrity of the high-energy interval, updating the starting point of the merged high-energy interval to the starting time point of the shot corresponding to the merged high-energy interval, updating the ending point of the merged high-energy interval to the ending time point of the shot corresponding to the merged high-energy interval, and taking the merged high-energy interval after updating the starting point and the ending point as a candidate split video, so as to ensure the integrity of the candidate split video, and taking the candidate split video with a video length within a preset video length as a split video, so as to effectively ensure the visual effect of the split video.

[0174] It should be noted that, for the method embodiment, in order to simply describe, it is expressed as a series of action combinations, but those skilled in the art should know that the embodiment of the present application is not limited by the described action sequence, because according to the embodiment of the present application, certain steps can be performed in other order or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions involved are not necessarily necessary for the embodiment of the present application.

[0175] Reference Figure 10 , a structural block diagram of a split video generation device according to an embodiment of the present application is shown, which can specifically include the following modules:

[0176] The bullet screen high-energy word list acquisition module 1001 is used to obtain the bullet screen high-energy word list in response to the bullet screen high-energy word list acquisition operation; wherein, the bullet screen high-energy word list contains multiple bullet screen high-energy words;

[0177] The video bullet comment acquisition module 1002 is used to acquire the bullet comments of the video;

[0178] The high-energy bullet screen confirmation module 1003 is used to identify the bullet screen as a high-energy bullet screen if the bullet screen contains the high-energy words in the bullet screen.

[0179] The matching information table acquisition module 1004 is used to acquire the matching information table corresponding to the high-energy words in the barrage;

[0180] The high-energy index calculation module 1005 for bullet comments per second is used to calculate the high-energy index for bullet comments per second based on the matching information table corresponding to the high-energy bullet comments.

[0181] The smooth high-energy barrage curve fitting module 1006 is used to fit a smooth high-energy barrage curve about the high-energy index per second based on the barrage high-energy index per second.

[0182] The high-energy interval determination module 1007 is used to determine the video interval of the video where the smooth high-energy barrage curve exceeds a preset segmentation threshold as the high-energy interval of the video.

[0183] The segmented video generation module 1008 is used to generate segmented videos based on the high-energy range.

[0184] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0185] This invention also provides an electronic device, such as... Figure 11 As shown, it includes a processor 1101, a communication interface 1102, a memory 1103, and a communication bus 1104, wherein the processor 1101, the communication interface 1102, and the memory 1103 communicate with each other through the communication bus 1104.

[0186] Memory 1103 is used to store computer programs;

[0187] When processor 1101 executes the program stored in memory 1103, it performs the following steps:

[0188] Obtain a list of high-energy words for bullet comments and the bullet comments of the video; wherein, the list of high-energy words for bullet comments contains multiple high-energy words for bullet comments;

[0189] If the bullet screen contains the high-energy words mentioned above, then the bullet screen is identified as a high-energy bullet screen.

[0190] obtain a match information table corresponding to the high-energy barrage corresponding high-energy word;

[0191] According to the match information table corresponding to the high-energy barrage, calculate the high-energy index of the barrage per second;

[0192] According to the high-energy index of the barrage per second, fit a smooth high-energy barrage curve about the high-energy index of the barrage per second;

[0193] The video interval of the smooth high-energy barrage curve exceeding the preset segmentation threshold is determined as the high-energy interval of the video.

[0194] The high-energy intervals in the video that are adjacent and within a preset unit time are merged to obtain a merged high-energy interval, and the merged high-energy interval is taken as a split video.

[0195] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0196] The communication interface is used for communication between the terminal and other devices.

[0197] The memory can include a Random Access Memory (RAM) and can also include a non-volatile memory, such as at least one disk memory. Optionally, the memory can also be at least one storage device located away from the aforementioned processor.

[0198] The processor mentioned above can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processing (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.

[0199] In yet another embodiment of the present disclosure, there is provided a computer readable storage medium having instructions stored therein, which when executed on a computer, cause the computer to perform the method of generating a strip video according to any one of the above-mentioned embodiments.

[0200] In the above-mentioned embodiments, the implementation can be wholly or partially in software, hardware, firmware, or any combination thereof. When implemented in software, the implementation can be in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed by a computer, the computer program instructions cause the computer to perform the processes or functions described in the embodiments of the present disclosure. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable apparatus. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center through a wired (for example, coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (for example, infrared, wireless, microwave, etc.) manner. The computer readable storage medium can be any available medium accessible by a computer or a data storage device such as a server, data center, etc. that includes one or more available media sets. The available media can be a magnetic medium (for example, floppy disk, hard disk, magnetic tape), an optical medium (for example, DVD), or a semiconductor medium (for example, Solid State Disk (SSD)), etc.

[0201] It should be noted that, in this document, the terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the processes, methods, articles or devices including a series of elements not only include those elements, but also include other elements not explicitly listed or inherent to such processes, methods, articles or devices. Without more limitations, the element defined by the statement "including a" does not exclude the presence of another identical element in the process, method, article or device including the element.

[0202] The various embodiments in the specification are described in a related manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments. In particular, for the system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiments.

[0203] The above only describes the preferred embodiments of the present application, and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for generating segmented video, characterized in that, include: In response to the operation of obtaining the high-energy word list of bullet comments, the high-energy word list of bullet comments is obtained; wherein, the high-energy word list of bullet comments contains multiple high-energy words of bullet comments; Get the video's bullet comments; If the bullet screen contains the high-energy words mentioned above, then the bullet screen is identified as a high-energy bullet screen. Obtain the matching information table corresponding to the high-energy words in the high-energy bullet comments; Calculate the high-energy index of bullet comments per second based on the matching information table corresponding to the high-energy bullet comments; Based on the high-energy index of bullet comments per second, fit a smooth high-energy bullet comment curve with respect to the high-energy index of bullet comments per second; The video range in which the smooth high-energy barrage curve exceeds a preset segmentation threshold is determined as the high-energy range of the video. Based on the high-energy range, generate segmented video; The matching information table contains the appearance times of the high-energy bullet comments, and the appearance times of the high-energy bullet comments correspond to other bullet comments. The step of calculating the bullet comment high-energy index per second based on the matching information table corresponding to the high-energy bullet comments includes: Obtain the video duration; The number of high-energy bullet comments in the video is counted to obtain the first bullet comment count; The number of high-energy bullet comments corresponding to the time points when the high-energy bullet comments appear is counted to obtain the second bullet comment count; The number of the third bullet comments is obtained by counting the number of the second bullet comments and the number of other bullet comments corresponding to the time points when the high-energy bullet comments appeared; The ratio of the number of the first bullet comments to the video duration is taken as the first ratio. The ratio of the number of the second bullet comments to the number of the third bullet comments is used as the second ratio. The ratio of the second number of bullet comments to the first ratio is used as the third ratio. The value obtained by adding the second ratio to the third ratio is used as the high-energy index of the barrage per second.

2. The method according to claim 1, characterized in that, If the bullet comment contains the high-energy words mentioned above, then the bullet comment is identified as a high-energy bullet comment, including: Determine whether the video's bullet comments contain any high-energy words from the bullet comment high-energy word list; If the bullet screen contains the high-energy words mentioned above, then the bullet screen is identified as a high-energy bullet screen.

3. The method according to claim 1, characterized in that, The step of obtaining the matching information table corresponding to the high-energy words in the high-energy bullet comments includes: The video's bullet comments are matched with the bullet comment high-energy words in the bullet comment high-energy word list to obtain one or more matching results corresponding to the bullet comment high-energy words; wherein, the matching result is the matching result corresponding to the successfully matched high-energy bullet comments; The matching results corresponding to each of the high-energy bullet comments are statistically analyzed to obtain the matching information table corresponding to the high-energy words of the bullet comments.

4. The method according to claim 1, characterized in that, The step of fitting a smooth high-energy barrage curve based on the high-energy index per second includes: Fit all the high-energy indexes of the bullet comments per second in the video into a high-energy bullet comment curve; The high-energy barrage curve is smoothed to obtain a smoothed high-energy barrage curve.

5. The method according to claim 1, characterized in that, After determining the video range where the smooth high-energy barrage curve exceeds a preset segmentation threshold as the high-energy range of the video, the method further includes: The starting point of the high-energy interval is updated to the first point where the high-energy index of the barrage per second begins to rise from a decrease, and the ending point of the high-energy interval is updated to the second point where the high-energy index of the barrage per second begins to rise from a decrease.

6. The method according to claim 1, characterized in that, The method further includes: The video is divided into multiple shots, and the start and end time points of each shot in the video are obtained; wherein, each shot includes the high-energy interval.

7. The method according to claim 6, characterized in that, The step of generating segmented videos based on the high-energy range includes: The high-energy intervals in the video that are adjacent and have time intervals within a preset unit time are merged to obtain the merged high-energy interval. The starting point of the merged high-energy interval is updated to the starting time point of the lens corresponding to the merged high-energy interval, and the ending point of the merged high-energy interval is updated to the ending time point of the lens corresponding to the merged high-energy interval. The merged high-energy intervals after updating the start and end points are used as candidate segmented videos, and the candidate segmented videos with a video duration within the preset video duration are used as segmented videos.

8. A device for generating segmented video, characterized in that, include: The bullet screen high-energy word list acquisition module is used to respond to the bullet screen high-energy word list acquisition operation and obtain the bullet screen high-energy word list; wherein, the bullet screen high-energy word list contains multiple bullet screen high-energy words; The video bullet comment acquisition module is used to acquire the bullet comments in the video; A high-energy bullet screen confirmation module is used to identify a bullet screen as a high-energy bullet screen if the bullet screen contains high-energy words. The matching information table acquisition module is used to acquire the matching information table corresponding to the high-energy words in the barrage; The high-energy index calculation module for bullet comments per second is used to calculate the high-energy index for bullet comments per second based on the matching information table corresponding to the high-energy bullet comments. A smooth high-energy barrage curve fitting module is used to fit a smooth high-energy barrage curve about the high-energy index per second based on the barrage high-energy index per second. The high-energy interval determination module is used to determine the video interval of the video where the smooth high-energy barrage curve exceeds a preset segmentation threshold as the high-energy interval of the video; The video segmentation generation module is used to generate a video segmentation video based on the high-energy range; The matching information table contains the appearance time points of the high-energy bullet comments, and the appearance time points of the high-energy bullet comments correspond to other bullet comments. The bullet comment high-energy index calculation module per second is specifically used to obtain the video duration of the video; count the number of high-energy bullet comments in the video to obtain the first bullet comment count; count the number of high-energy bullet comments corresponding to the appearance time points of the high-energy bullet comments to obtain the second bullet comment count; count the second bullet comment count and the number of other bullet comments corresponding to the appearance time points of the high-energy bullet comments to obtain the third bullet comment count; use the ratio of the first bullet comment count to the video duration of the video as the first ratio; use the ratio of the second bullet comment count to the third bullet comment count as the second ratio; use the ratio of the second bullet comment count to the first ratio as the third ratio; and add the second ratio and the third ratio to obtain the value as the bullet comment high-energy index per second.

9. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the steps of the method described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.

Images