A method, device, equipment and medium for producing short video of game play

Abstract

The present application discloses a method, device, equipment and medium for making a short video of play, which relates to the field of video production technology. The method includes: obtaining an initial video captured by a preset camera in a play place; identifying fellow tourists from the initial video, and classifying the identified fellow tourists into one category to obtain a corresponding cluster of fellow tourists; extracting tourist feature information of each fellow tourist from the initial video; determining the type label corresponding to the cluster of fellow tourists based on the tourist feature information of each fellow tourist, and filtering out the corresponding target audio and video template from a preset audio and video template mapping table based on the type label corresponding to the cluster of fellow tourists and the shooting date type corresponding to the initial video; creating a corresponding short video of play based on the target audio and video template and the preferred video clip corresponding to the cluster of fellow tourists in the initial video. Through the above scheme, the efficiency of video production can be improved and the single video style can be avoided when making a short video of play.

Description

Technical Field

[0001] The present invention relates to the field of video production technology, and in particular to a method, device, equipment and medium for producing a short game video. Background Art

[0002] At present, network cameras are installed on amusement equipment or key attractions in theme parks, amusement parks, scenic spots and other amusement places, and play videos are automatically produced through video analysis methods based on the videos taken by the network cameras. Compared with the traditional manual video production of manually taking photos or videos of the play process, this automatic video production system is simpler and faster, which greatly improves the efficiency of play video generation. However, on the one hand, when using the above-mentioned video automatic production system to produce videos, the short videos generated by each person at the same time are of the same style and lack personalized features; on the other hand, replacements are made during different holidays, resulting in low efficiency in the video production process. In summary, how to improve video production efficiency and avoid the problem of single video style when producing short play videos needs to be further resolved. Summary of the invention

[0003] In view of this, the purpose of the present invention is to provide a method, device, equipment and medium for making a short game video, which can improve the efficiency of video production and avoid the problem of a single video style when making a short game video. The specific scheme is as follows:

[0004] In a first aspect, the present application discloses a method for producing a short game video, comprising:

[0005] Obtain the initial video captured by the preset camera in the recreation area;

[0006] Identifying fellow tourists from the initial video, and classifying the identified fellow tourists into one category to obtain a corresponding fellow tourist cluster;

[0007] Extracting tourist feature information of each of the traveling tourists from the initial video;

[0008] Determine the type label corresponding to the cluster of fellow tourists based on the tourist characteristic information of each of the fellow tourists, and select the corresponding target audio and video template from the preset audio and video template mapping table based on the type label corresponding to the cluster of fellow tourists and the shooting date type corresponding to the initial video;

[0009] A corresponding short travel video is created based on the target audio and video template and the preferred video clips in the initial video corresponding to the cluster of fellow tourists.

[0010] Optionally, identifying fellow tourists from the initial video and classifying the identified fellow tourists into one category to obtain a corresponding fellow tourist cluster includes:

[0011] Selecting the preferred video frames corresponding to the tourists in the initial video according to the trained artificial intelligence model;

[0012] Inputting the preferred video frame into the trained deep learning algorithm model to obtain facial features of tourists in the initial video;

[0013] Inputting the facial features into the trained facial feature clustering algorithm model to obtain a target frequency table representing the number of times the tourist in the initial video appears in the preset camera;

[0014] Based on the time information, location information, position information corresponding to the tourists in the initial video and the frequency information in the target frequency table, the traveling companions are identified and the number of traveling companions is recorded, and then the identified traveling companions are classified into one category to obtain a corresponding traveling companion cluster.

[0015] Optionally, determining the type label corresponding to the traveling companion cluster based on the tourist characteristic information of each traveling companion includes:

[0016] The first tourist cluster label and the second tourist cluster label of the traveling tourist cluster are marked according to the characteristic information, and the type label is determined according to the first tourist cluster label and the second tourist cluster label.

[0017] Optionally, marking the first tourist cluster label of the cluster of tourists traveling together according to the characteristic information includes:

[0018] The category of the traveling companion cluster is determined based on the age information and gender information in the characteristic information, and the first tourist cluster label is marked according to the category of the traveling companion cluster.

[0019] Optionally, marking the second tourist cluster label of the traveling tourist cluster according to the characteristic information includes:

[0020] Inputting the preferred video frames corresponding to the tourists in the traveling group cluster into the trained expression recognition algorithm model to obtain the facial expressions corresponding to the tourists in the traveling group cluster, and determining the expression labels corresponding to the traveling group cluster through the facial expressions corresponding to the tourists in the traveling group cluster;

[0021] Inputting the preferred video frames corresponding to the tourists in the traveling companion cluster into the trained clothing analysis algorithm model to obtain clothing types corresponding to the tourists in the traveling companion cluster, and determining clothing labels corresponding to the traveling companion cluster by the clothing types corresponding to the tourists in the traveling companion cluster;

[0022] A second tourist cluster label of the traveling tourist cluster is determined based on the expression label and the clothing label.

[0023] Optionally, before selecting a corresponding target audio and video template from a preset audio and video template mapping table based on the type label corresponding to the cluster of fellow tourists and the shooting date type corresponding to the initial video, the method further includes:

[0024] The video and audio in the preset video template pool and the preset audio template pool are screened and classified according to the date type, the first visitor cluster label and the second visitor cluster label, and a preset audio and video template mapping table is created.

[0025] Optionally, before screening and classifying the videos and audios in the preset video template pool and the preset audio template pool according to the date type, the first visitor cluster label and the second visitor cluster label, and creating a preset audio and video template mapping table, the method further includes:

[0026] Shooting a video of a scenic spot and a video of an amusement item in the amusement place corresponding to the date type, and creating the preset video template pool based on the video of the scenic spot and the video of the amusement item;

[0027] The preset audio template pool is created based on the theme audio corresponding to the date type, the first visitor cluster label, and the audio corresponding to the second visitor cluster label.

[0028] In a second aspect, the present application discloses a device for making a short game video, comprising:

[0029] The video acquisition module is used to obtain the initial video collected by the preset camera in the recreational area;

[0030] A tourist cluster determination module, used for identifying fellow tourists from the initial video and classifying the identified fellow tourists into one category to obtain a corresponding fellow tourist cluster;

[0031] A feature information extraction module, used to extract the tourist feature information of each of the traveling tourists from the initial video;

[0032] An audio and video module screening module, used to determine the type label corresponding to the cluster of fellow tourists based on the tourist characteristic information of each of the fellow tourists, and to screen out the corresponding target audio and video template from a preset audio and video template mapping table based on the type label corresponding to the cluster of fellow tourists and the shooting date type corresponding to the initial video;

[0033] A travel short video production module is used to create a corresponding travel short video based on the target audio and video template and the preferred video clips in the initial video corresponding to the cluster of fellow tourists.

[0034] In a third aspect, the present application discloses an electronic device, comprising: a memory for storing a computer program;

[0035] The processor is used to execute the computer program to implement the steps of the aforementioned method for producing a short video of the game.

[0036] In a fourth aspect, the present application discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the steps of the aforementioned method for producing a short gameplay video are implemented.

[0037] When making a short video of a tour, the present application first obtains an initial video captured by a preset camera in the tour area, identifies fellow tourists from the initial video, and classifies the identified fellow tourists into one category to obtain a corresponding cluster of fellow tourists, then extracts the tourist characteristic information of each of the fellow tourists from the initial video, determines the type label corresponding to the cluster of fellow tourists based on the tourist characteristic information of each of the fellow tourists, and based on the type label corresponding to the cluster of fellow tourists and the shooting date type corresponding to the initial video, selects the corresponding target audio and video template from the preset audio and video template mapping table, and finally creates a corresponding short video of the tour based on the target audio and video template and the preferred video clip corresponding to the cluster of fellow tourists in the initial video. It can be seen that when making a short video of the tour, the present application determines the cluster of fellow tourists by the information obtained from the initial video and assigns the corresponding type label to the cluster of fellow tourists, and selects the corresponding target audio and video template from the preset audio and video template mapping table based on the type label, and then makes a short video of the tour based on the target audio and video template and the initial video. Therefore, when producing short travel videos, the present application uses the tourist information obtained from the initial video to analyze the fellow tourists and assign corresponding type labels to the clusters of fellow tourists. The style features of the clusters of fellow tourists can be distinguished by the type labels corresponding to the clusters of fellow tourists, and the corresponding target audio and video templates can be matched according to the style features corresponding to the type labels, thereby avoiding the problem that the short videos generated by each person are of the same style and lack personalized features when the automatic production system for travel short videos is producing videos; on the other hand, the target audio and video template corresponding to the date is selected according to the shooting date type corresponding to the initial video, thereby avoiding the problem of low efficiency caused by the need to manually replace templates on different shooting dates. In summary, the present application can improve the efficiency of video production and avoid the single video style when producing short travel videos. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0039] Figure 1 A flow chart of a method for making a short video of a game provided in this application;

[0040] Figure 2 A flow chart of a specific method for making a short video of a game provided in this application;

[0041] Figure 3 A schematic diagram of the core network structure of the main network of the expression recognition model provided for this application;

[0042] Figure 4 A schematic diagram of the structure of a short video production device for playing games provided in this application;

[0043] Figure 5 A structural diagram of an electronic device provided for this application. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] In the prior art, when the automatic video production system is producing videos, the short videos generated by each person at the same time are of the same style, lacking personalized features; on the other hand, they are replaced on different holidays, resulting in low efficiency in the video production process. To this end, the present application provides a method for producing a short video of play, which can improve the efficiency of video production and avoid a single video style when producing a short video of play.

[0046] The embodiment of the present invention discloses a method for producing a short video of a game, see Figure 1 Said method comprises:

[0047] Step S11: Acquire the initial video captured by the preset camera in the recreation area.

[0048] In this embodiment, the recreational places include but are not limited to scenic spots, amusement parks, and parks. Accordingly, the preset cameras in the recreational places are preset cameras installed in various areas of the recreational places, and the preset cameras are used to shoot the video of tourists playing in the recreational places to obtain the initial video. By obtaining the video of tourists playing in the recreational places shot by the preset cameras, it is convenient to judge the cluster of tourists traveling together and label the cluster of tourists traveling together through the initial video, and further edit and synthesize the short video of the play through the initial video and the corresponding audio and video templates selected.

[0049] Step S12: identifying fellow tourists from the initial video, and classifying the identified fellow tourists into one category to obtain a corresponding fellow tourist cluster.

[0050] In this embodiment, the method of identifying the traveling companions from the initial video and classifying the identified traveling companions into one category to obtain the corresponding traveling companion clusters includes: selecting the preferred video frames corresponding to the tourists in the initial video according to the trained artificial intelligence model; inputting the preferred video frames into the trained deep learning algorithm model to obtain the facial features of the tourists in the initial video; inputting the facial features into the trained facial feature clustering algorithm model to obtain the target frequency table representing the number of times the tourists in the initial video appear in the preset camera; identifying the traveling companions based on the time information, location information, position information corresponding to the tourists in the initial video and the frequency information in the target frequency table and recording the number of traveling companions, and then classifying the identified traveling companions into one category to obtain the corresponding traveling companion clusters. That is, firstly, the preferred video frame in the initial video is selected by the artificial intelligence model, and the preferred video frame is a video frame that can clearly display the appearance and other features of the tourists in the initial video. It can be understood that each tourist in the recreational place corresponds to a preferred video frame, so as to determine the tourist feature information of the traveling companions through the preferred video frame. After obtaining the preferred video frame, the preferred video frame is input into the trained deep learning algorithm model to obtain the facial features of the tourists corresponding to the preferred video frame, and the facial features of the tourists are used as input, and the tourists are clustered into n categories through the facial feature clustering algorithm, which are represented by C1, C2, ... Cn, where each category represents a tourist, and the number of data in each category is further counted, represented by M1, M2, ... Mn, and the target frequency table of the number of times the tourists appear in the preset camera in the initial video is established through the above data.

[0051] Furthermore, after obtaining the target frequency table, the time information, location information, position information and frequency information in the target frequency table are used to identify fellow tourists and record the number of fellow tourists, and the fellow tourists are determined as corresponding clusters of fellow tourists. Specifically, if there are i preset cameras in a certain recreational place, and s tourists appear in more than j cameras at the same time, j can be adjusted according to the actual situation, but must be greater than i / 2, and the distance between tourists is less than the preset distance, then the above tourists are determined to be fellow tourists, and the fellow tourists are labeled and the number of fellow tourists is recorded. Through the above technical solution, fellow tourists who play together in the tourist place are identified, so that the type labels of fellow tourists can be marked later and videos of fellow tourists playing together can be produced when making short videos of the play.

[0052] Step S13: extracting tourist feature information of each of the traveling tourists from the initial video.

[0053] In this embodiment, the tourist characteristic information includes but is not limited to age information and gender information. That is, by identifying the preferred initial video frame in the initial video, the tourist characteristic information of the traveling companions is obtained, so that the traveling companion clusters are subsequently labeled with type labels through the tourist characteristic information, and then the corresponding target audio and video templates are screened out through the labeling of type labels.

[0054] Step S14: Determine the type label corresponding to the traveling companion cluster based on the tourist characteristic information of each traveling companion, and filter out the corresponding target audio and video template from the preset audio and video template mapping table based on the type label corresponding to the traveling companion cluster and the shooting date type corresponding to the initial video.

[0055] In this embodiment, the type label of the cluster of fellow tourists is determined based on the characteristic information of each fellow tourist in the cluster of fellow tourists extracted from the initial video, and the type label is a label that characterizes the characteristics of the cluster of fellow tourists. Furthermore, the target audio and video templates that meet the shooting date type of the day and are related to the characteristics of the cluster of fellow tourists are screened out from the preset audio and video template mapping table through the type label corresponding to the cluster of fellow tourists and the shooting date type corresponding to the initial video. Through the above technical solution, when making a short video of the tour, the audio and video templates can be screened according to the type of shooting date and the characteristics of the cluster of fellow tourists, thus avoiding the problem of a single style of the short videos of the tour produced in the tour venue.

[0056] Step S15: creating a corresponding short travel video based on the target audio and video template and the preferred video clips in the initial video corresponding to the cluster of fellow tourists.

[0057] In this embodiment, the preferred video clips corresponding to the cluster of fellow tourists in the initial video are selected through the pre-trained video selection model, that is, the clear and beautiful play video clips corresponding to the cluster of fellow tourists, and the preferred video clips are integrated with the target audio and video template to obtain the play short video. Through the above technical solution, when producing the play short video, it is possible to match the appropriate audio and video template by analyzing the cluster of fellow tourists, and select the preferred video clips that meet the wishes of tourists to produce the play short video.

[0058] It can be seen that when this embodiment is producing a short video of play, the cluster of fellow tourists is determined by the information obtained from the initial video and the corresponding type label is assigned to the cluster of fellow tourists, and the corresponding target audio and video template is screened out from the preset audio and video template mapping table based on the type label, and then the short video of play is produced based on the target audio and video template and the initial video. Therefore, when this application is producing a short video of play, on the one hand, the tourist information obtained from the initial video is used to analyze the fellow tourists and assign the corresponding type label to the cluster of fellow tourists, and the style characteristics of the cluster of fellow tourists can be distinguished by the type label corresponding to the cluster of fellow tourists, and the corresponding target audio and video template is matched according to the style characteristics corresponding to the type label, thereby avoiding the problem that the short videos generated by each person are of the same style and lack personalized features when the automatic production system of the short video of play is used for video production; on the other hand, the target audio and video template corresponding to the date of the day is selected according to the shooting date type corresponding to the initial video, thereby avoiding the problem that the template needs to be manually replaced on different shooting dates, resulting in low efficiency. In summary, this application can improve the efficiency of video production and avoid the single video style when producing a short video of play.

[0059] See also Figure 2 As shown, an embodiment of the present invention discloses a method for producing a short game video. Compared with the previous embodiment, this embodiment further illustrates and optimizes the technical solution.

[0060] Step S21: Obtain the initial video captured by the preset camera in the recreation area.

[0061] Step S22: identifying fellow tourists from the initial video, and classifying the identified fellow tourists into one category to obtain a corresponding fellow tourist cluster.

[0062] Step S23: extracting tourist feature information of each of the traveling tourists from the initial video.

[0063] Step S24: Mark the first tourist cluster label and the second tourist cluster label of the traveling tourist cluster according to the characteristic information, and determine the type label according to the first tourist cluster label and the second tourist cluster label, and then filter out the corresponding target audio and video template from the preset audio and video template mapping table based on the type label corresponding to the traveling tourist cluster and the shooting date type corresponding to the initial video.

[0064] In this embodiment, marking the first tourist cluster label of the traveling companion cluster according to the characteristic information includes: judging the category of the traveling companion cluster based on the age information and gender information in the characteristic information, and marking the first tourist cluster label according to the category of the traveling companion cluster. In a specific embodiment, the first tourist cluster label includes but is not limited to family, sister group, mixed group, single female and single male. When there are children in the traveling companion cluster, the first tourist label of the traveling companion cluster is marked as a family label, and multiple females in the traveling companion cluster are marked as a sister group label. If there are both males and females in the traveling companion cluster, it is marked as a mixed group label. The priority order of the first tourist cluster label is family>sister group>mixed group>single female>single male. It can be understood that the first tourist label can be adjusted according to the actual application scenario.

[0065] In the present embodiment, marking the second tourist cluster label of the traveling companion cluster according to the characteristic information includes: inputting the preferred video frames corresponding to the tourists in the traveling companion cluster into the trained expression recognition algorithm model to obtain the facial expressions corresponding to the tourists in the traveling companion cluster, and determining the expression labels corresponding to the traveling companion cluster through the facial expressions corresponding to the tourists in the traveling companion cluster; inputting the preferred video frames corresponding to the tourists in the traveling companion cluster into the trained clothing analysis algorithm model to obtain the clothing types corresponding to the tourists in the traveling companion cluster, and determining the clothing labels corresponding to the traveling companion cluster through the clothing types corresponding to the tourists in the traveling companion cluster; and determining the second tourist cluster label of the traveling companion cluster based on the expression labels and the clothing labels.

[0066] In a specific implementation, the main network of the expression recognition algorithm model is MobilenetV1, and its core network structure is as follows: Figure 3As shown, the preferred video frame image in RGB format is input, the size of the image is 224*224, and the last softmax layer is a 4-classification. The expression features corresponding to the tourists are identified by the trained expression recognition algorithm model, and the expression features include but are not limited to expressionless, smiling, big and small, and sad. Based on the expression features of the tourists in the identified traveling companion cluster, the corresponding expression labels are marked for the traveling companion cluster, and the priority order of expression labels is laughing>smile>expressionless>sad. Further, the preferred video frame corresponding to the tourists in the traveling companion cluster is used as input, and the clothing of the tourists in the traveling companion cluster is identified by the clothing analysis algorithm model. The clothing types include but are not limited to ordinary clothing, JK clothing, lolita clothing, and sports clothing. The clothing type priority order is and the clothing labels corresponding to the traveling companion cluster are determined for the clothing of the tourists in the traveling companion cluster. The clothing label priority order is lolita clothing>JK clothing>sports clothing>ordinary clothing. It can be understood that both the labels and the priority order can be adjusted according to the actual situation. Furthermore, the second tourist cluster label is determined according to the expression label and the clothing label. The second tourist cluster label of laughing + loli clothing is A1, the second tourist cluster label of laughing + JK clothing is A2, and so on. In this embodiment, the second tourist cluster label can be divided into 20 categories.

[0067] In this embodiment, before the corresponding target audio and video template is selected from the preset audio and video template mapping table based on the type label corresponding to the cluster of tourists and the shooting date type corresponding to the initial video, it also includes: establishing a preset video template pool and a preset audio template pool. Specifically, the video of the scenic spots and the video of the amusement items in the recreational place corresponding to the date type are shot, and the preset video template pool is created based on the video of the scenic spots and the video of the amusement items; the preset audio template pool is created based on the theme audio corresponding to the date type, the audio corresponding to the first tourist cluster label and the second tourist cluster label. Specifically, the date type includes but is not limited to holiday theme date type, four seasons theme date type, and solar term theme date type, and the video of the scenic spots and the video of the tourist items in the recreational place corresponding to different date types are shot, and the above videos are made into video templates, and a preset video template pool is created; and a preset audio template pool is created, which includes all the holidays of the whole year and the theme audio related to the corresponding date type, wherein different date type themes cover a variety of categories of audio, including but not limited to happy themes, sad themes, childish themes, sports themes, and two-dimensional themes.

[0068] Further, the video and audio in the preset video template pool and the preset audio template pool are screened and classified according to the date type, the first visitor cluster label and the second visitor cluster label, and a preset audio and video template mapping table is created. The storage form of the preset audio and video template mapping table is:

[0069] {Date type: {First visitor cluster label: {Second visitor cluster label: video1+audio1, video2+audio2, ...}}};

[0070] After the preset template mapping table is established, based on the type label corresponding to the cluster of fellow tourists and the shooting date type corresponding to the initial video, the corresponding target audio and video template is mapped and searched from the preset audio and video template mapping table.

[0071] Step S25: creating a corresponding short travel video based on the target audio and video template and the preferred video clips in the initial video corresponding to the cluster of fellow tourists.

[0072] It can be seen that in this embodiment, the preset algorithm model is used to automatically identify the clusters of tourists traveling together and their types, and then the audio and video templates are matched in a targeted manner to determine the theme style of the short video, so as to achieve the effect of different faces for different people when producing short videos of the tour, thereby improving the satisfaction of tourists.

[0073] See also Figure 4 As shown, the embodiment of the present invention discloses a specific device for making a short game video, including:

[0074] The video acquisition module 11 is used to obtain the initial video captured by the preset camera in the recreational area;

[0075] A tourist cluster determination module 12 is used to identify fellow tourists from the initial video and classify the identified fellow tourists into one category to obtain a corresponding fellow tourist cluster;

[0076] A feature information extraction module 13 is used to extract the tourist feature information of each of the traveling tourists from the initial video;

[0077] The audio and video module screening module 14 is used to determine the type label corresponding to the cluster of fellow tourists based on the tourist characteristic information of each of the fellow tourists, and screen out the corresponding target audio and video template from the preset audio and video template mapping table based on the type label corresponding to the cluster of fellow tourists and the shooting date type corresponding to the initial video;

[0078] The travel short video production module 15 is used to create a corresponding travel short video based on the target audio and video template and the preferred video clips corresponding to the cluster of fellow tourists in the initial video.

[0079] It can be seen that when this embodiment is producing a short video of play, the cluster of fellow tourists is determined by the information obtained from the initial video and the corresponding type label is assigned to the cluster of fellow tourists, and the corresponding target audio and video template is screened out from the preset audio and video template mapping table based on the type label, and then the short video of play is produced based on the target audio and video template and the initial video. Therefore, when this application is producing a short video of play, on the one hand, the tourist information obtained from the initial video is used to analyze the fellow tourists and assign the corresponding type label to the cluster of fellow tourists, and the style characteristics of the cluster of fellow tourists can be distinguished by the type label corresponding to the cluster of fellow tourists, and the corresponding target audio and video template can be matched according to the style characteristics corresponding to the type label, thereby avoiding the problem that the short videos generated by each person are of the same style and lack personalized features when the automatic production system of the short video of play is used for video production; on the other hand, the target audio and video template corresponding to the date of the day is selected according to the shooting date type corresponding to the initial video, thereby avoiding the problem that the template needs to be manually replaced on different shooting days, resulting in low efficiency. In summary, this application can improve the efficiency of video production and avoid the single video style when producing a short video of play.

[0080] In some specific embodiments, the tourist cluster determination module 12 specifically includes:

[0081] A video frame optimization unit, used to optimize the preferred video frames corresponding to the tourists in the initial video according to the trained artificial intelligence model;

[0082] A facial feature acquisition unit, used for inputting the preferred video frame into a trained deep learning algorithm model to obtain facial features of tourists in the initial video;

[0083] A frequency table acquisition unit, used for inputting the facial features into the trained facial feature clustering algorithm model to obtain a target frequency table representing the number of times the tourist in the initial video appears in the preset camera;

[0084] The companion identification unit is used to identify the companion tourists and record the number of companions based on the time information, location information, position information corresponding to the tourists in the initial video and the frequency information in the target frequency table, and then classify the identified companion tourists into one category to obtain a corresponding companion tourist cluster.

[0085] In some specific embodiments, the audio and video module screening module 14 specifically includes:

[0086] A first label marking unit, configured to determine the category of the traveling companion cluster based on the age information and the gender information in the characteristic information, and mark the first tourist cluster label according to the category of the traveling companion cluster;

[0087] An expression label determination unit is used to input the preferred video frames corresponding to the tourists in the traveling companion cluster into the trained expression recognition algorithm model to obtain the facial expressions corresponding to the tourists in the traveling companion cluster, and determine the expression labels corresponding to the traveling companion cluster through the facial expressions corresponding to the tourists in the traveling companion cluster;

[0088] A clothing label determination unit is used to input the preferred video frames corresponding to the tourists in the traveling companion cluster into the trained clothing analysis algorithm model to obtain clothing types corresponding to the tourists in the traveling companion cluster, and determine clothing labels corresponding to the traveling companion cluster according to the clothing types corresponding to the tourists in the traveling companion cluster;

[0089] The first label marking unit is used to determine a second tourist cluster label of the traveling tourist cluster based on the expression label and the clothing label.

[0090] In some specific embodiments, the game short video production device further includes:

[0091] A mapping table creation module is used to screen and classify the video and audio in the preset video template pool and the preset audio template pool according to the date type, the first visitor cluster label and the second visitor cluster label, and create a preset audio and video template mapping table.

[0092] In some specific embodiments, the game short video production device further includes:

[0093] A preset video template pool creation module, used for shooting the video of the scenic spots and the video of the amusement items in the amusement place corresponding to the date type, and creating the preset video template pool based on the video of the scenic spots and the video of the amusement items;

[0094] A preset audio template pool creation module is used to create the preset audio template pool based on the theme audio corresponding to the date type, the first visitor cluster label, and the audio corresponding to the second visitor cluster label.

[0095] Figure 5 An electronic device 20 provided in an embodiment of the present application is shown. The electronic device 20 may further include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the method for making a short video of the game disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment may be specifically an electronic computer.

[0096] In this embodiment, the power supply 23 is used to provide voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0097] In addition, the memory 22, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0098] The operating system 221 is used to manage and control the hardware devices on the electronic device 20, and the computer program 222 may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program that can be used to complete the method for making a short video of the game performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 may further include a computer program that can be used to complete other specific tasks.

[0099] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein the computer program, when executed by a processor, implements the aforementioned disclosed method for making a short video of a game. The specific steps of the method can be referred to the corresponding contents disclosed in the aforementioned embodiments, and will not be repeated here.

[0100] Finally, it should be noted that, in this article, relational terms such as first and second, etc. 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 these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0101] The above is a detailed introduction to a method, device, equipment and medium for producing a short video of gameplay provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for producing a short game video, characterized in that: include: Get the initial video collected by the preset camera in the playground; Identifying fellow tourists from the initial video, and classifying the identified fellow tourists into one category to obtain a corresponding fellow tourist cluster; Extracting tourist feature information of each of the traveling tourists from the initial video; Based on the tourist characteristic information of each of the traveling companions, the type label corresponding to the traveling companion cluster is determined, and based on the type label corresponding to the traveling companion cluster and the shooting date type corresponding to the initial video, the corresponding target audio and video template is screened out from the preset audio and video template mapping table, the audio and video template mapping table includes the mapping relationship between the shooting date type, the first tourist cluster label, the second tourist cluster label, the video in the preset video template pool and the audio in the preset audio template pool, the establishment process of the preset video template pool includes: shooting the scenic spot video and the amusement project video in the amusement place corresponding to the date type, and creating it based on the scenic spot video and the amusement project video, the establishment process of the preset audio template pool includes: creating it based on the theme audio corresponding to the date type, the first tourist cluster label and the audio corresponding to the second tourist cluster label; Wherein, determining the type label corresponding to the cluster of traveling companions based on the tourist characteristic information of each of the traveling companions includes: marking a first tourist cluster label and a second tourist cluster label of the cluster of traveling companions according to the characteristic information, and determining the type label according to the first tourist cluster label and the second tourist cluster label, wherein the category of the cluster of traveling companions is judged based on age information and gender information in the characteristic information, and the first tourist cluster label is marked according to the category of the cluster of traveling companions; and determining the second tourist cluster label of the cluster of traveling companions based on an expression label and a clothing label, wherein the expression label is determined based on the facial expressions corresponding to the tourists in the cluster of traveling companions, and the clothing label is determined based on the clothing types corresponding to the tourists in the cluster of traveling companions; A corresponding short travel video is created based on the target audio and video template and the preferred video clips in the initial video corresponding to the cluster of fellow tourists.

2. The method for producing a short game video according to claim 1, characterized in that: The step of identifying the traveling companions from the initial video and classifying the identified traveling companions into one category to obtain a corresponding traveling companion cluster includes: Selecting the preferred video frames corresponding to the tourists in the initial video according to the trained artificial intelligence model; Inputting the preferred video frame into the trained deep learning algorithm model to obtain facial features of tourists in the initial video; Inputting the facial features into the trained facial feature clustering algorithm model to obtain a target frequency table representing the number of times the tourist in the initial video appears in the preset camera; Based on the time information, location information, position information corresponding to the tourists in the initial video and the frequency information in the target frequency table, the traveling companions are identified and the number of traveling companions is recorded, and then the identified traveling companions are classified into one category to obtain a corresponding traveling companion cluster.

3. The method for producing a short game video according to claim 2, characterized in that: The step of marking the second tourist cluster label of the cluster of tourists traveling together according to the characteristic information comprises: Inputting the preferred video frames corresponding to the tourists in the traveling group cluster into the trained expression recognition algorithm model to obtain the facial expressions corresponding to the tourists in the traveling group cluster, and determining the expression labels corresponding to the traveling group cluster through the facial expressions corresponding to the tourists in the traveling group cluster; Input the preferred video frames corresponding to the tourists in the traveling companion cluster into the trained clothing analysis algorithm model to obtain clothing types corresponding to the tourists in the traveling companion cluster, and determine clothing labels corresponding to the traveling companion cluster according to the clothing types corresponding to the tourists in the traveling companion cluster; A second tourist cluster label of the traveling tourist cluster is determined based on the expression label and the clothing label.

4. A device for making a short video of a game, characterized in that: include: The video acquisition module is used to obtain the initial video collected by the preset camera in the recreational area; A tourist cluster determination module, used for identifying fellow tourists from the initial video and classifying the identified fellow tourists into one category to obtain a corresponding fellow tourist cluster; A feature information extraction module, used to extract the tourist feature information of each of the traveling tourists from the initial video; An audio and video module screening module, used for determining the type label corresponding to the cluster of fellow tourists based on the tourist characteristic information of each of the fellow tourists, and screening out the corresponding target audio and video template from a preset audio and video template mapping table based on the type label corresponding to the cluster of fellow tourists and the shooting date type corresponding to the initial video, wherein the audio and video template mapping table includes a mapping relationship between the shooting date type, the first tourist cluster label, the second tourist cluster label, the video in the preset video template pool, and the audio in the preset audio template pool, wherein the process of establishing the preset video template pool includes: shooting the video of the scenic spots and the video of the amusement items in the amusement place corresponding to the date type, and creating the video based on the video of the scenic spots and the video of the amusement items, and the process of establishing the preset audio template pool includes: creating the video based on the theme audio corresponding to the date type, the audio corresponding to the first tourist cluster label and the second tourist cluster label; The audio and video module screening module includes: a first label marking unit, which is used to determine the category of the traveling companion cluster based on the age information and gender information in the feature information, and mark the first tourist cluster label according to the traveling companion cluster category; a second label marking unit, which is used to determine the second tourist cluster label of the traveling companion cluster based on the expression label and the clothing label, wherein the expression label is determined based on the facial expressions corresponding to the tourists in the traveling companion cluster, and the clothing label is determined based on the clothing types corresponding to the tourists in the traveling companion cluster; A travel short video production module is used to create a corresponding travel short video based on the target audio and video template and the preferred video clips in the initial video corresponding to the cluster of fellow tourists.

5. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor is used to execute the computer program to implement the steps of the method for producing a short gameplay video as described in any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that: Used to store computer programs; wherein, when the computer program is executed by a processor, the steps of the method for producing a short gameplay video as described in any one of claims 1 to 3 are implemented.

Images