Short video generation method, system and terminal

Abstract

The present invention relates to a short video generation method, system, and terminal, and relates to the field of video production technology. The method includes: obtaining image material information and short video demand information uploaded by ordinary users; retrieving demand style information and demand duration value based on the short video demand information; determining material demand characteristics based on the demand style information; identifying the material demand characteristics and image material information to determine selected material information; retrieving style demand material information from a preset video material library based on the demand style information, and combining the selected material information with the style demand material information to form used material information; synthesizing and editing the used material information based on the demand style information and the demand duration value to generate short video output information, and outputting the short video output information to a terminal held by the ordinary user for display. The present invention has the effect of facilitating the production of short videos by ordinary users.

Description

Technical Field

[0001] The present invention relates to the field of video production technology, and in particular to a short video generation method, system and terminal. Background Art

[0002] With the rapid development of mobile internet products and technologies, short video products have become the most advanced and popular ecosystem in the current internet industry. The creation and dissemination of short video content is an important means and method for providing content services to internet users.

[0003] Currently, when producing short videos, producers generally need to formulate a detailed shooting plan based on the script or outline, and use different shooting techniques to create unique visual effects, create an atmosphere suitable for the storyline, collect materials, and then import the captured video and audio materials into specialized video editing software to edit and sort the materials. They can also add video effects and special effects as needed, adjust audio materials, and choose suitable background music and sound effects to enhance the viewing experience. They can then choose to publish and share the short videos on different social media platforms.

[0004] In the process of producing short videos, producers need to import the captured video and audio materials into special video editing software for production. Special video editing software requires producers to master certain editing, audio processing and other technical and artistic skills, which makes it inconvenient for ordinary users to produce short videos. Summary of the Invention

[0005] In order to facilitate ordinary users to produce short videos, the present invention provides a short video generation method, system and terminal.

[0006] In a first aspect, the present invention provides a short video generation method, which adopts the following technical solution:

[0007] A short video generation method, comprising:

[0008] Obtain image material information and short video demand information uploaded by ordinary users;

[0009] Retrieve the required style information and required duration value based on the short video demand information;

[0010] Determine the required material characteristics based on the required style information;

[0011] Identify the material information based on the material requirements and image material information to determine the selected material information;

[0012] Based on the required style information, retrieve the style required material information from the preset video material library, and combine the selected material information with the style required material information to form the used material information;

[0013] The material information is synthesized and edited according to the required style information and the required duration value to generate short video output information, and the short video output information is output to the terminal held by the ordinary user for display.

[0014] Optionally, methods for determining material requirement characteristics include:

[0015] Retrieve style vocabulary based on required style information;

[0016] According to the correspondence between the style vocabulary and the preset synonyms, the synonyms corresponding to the style vocabulary are determined;

[0017] Determine synonymous comprehensive vocabulary based on style vocabulary and synonyms;

[0018] According to the corresponding relationship between the synonymous comprehensive vocabulary and the preset style-related vocabulary, the style-related vocabulary corresponding to the synonymous comprehensive vocabulary is determined;

[0019] Summarize and form style comprehensive vocabulary based on synonymous comprehensive vocabulary and style related vocabulary;

[0020] According to the correspondence between the style comprehensive vocabulary and the preset style material features, the style material features corresponding to the style comprehensive vocabulary are determined, and the style material features are used as the material demand features.

[0021] Optional methods for determining synonymous comprehensive words include:

[0022] Determine whether there is only one style word;

[0023] If yes, then summarize the style words and synonyms and form a synonymous comprehensive word;

[0024] If not, then based on the consistency between the style vocabulary and synonyms, the consistent vocabulary is determined;

[0025] When consistent vocabulary does not exist, style vocabulary and synonyms are aggregated to form a synonymous comprehensive vocabulary;

[0026] When consistent vocabulary exists, the style vocabulary is adjusted based on the consistent vocabulary and used as the style adjustment vocabulary;

[0027] According to the correspondence between the style adjustment vocabulary and the preset synonym adjustment vocabulary, the synonym adjustment vocabulary corresponding to the style adjustment vocabulary is determined;

[0028] Based on the style adjustment vocabulary and synonym adjustment vocabulary, they are summarized and formed into a synonym comprehensive vocabulary.

[0029] Optionally, the method for determining the selected material information includes:

[0030] Based on the material requirement features, identification and matching are performed from the image material information to form matching material information;

[0031] Determine matching remaining information based on image material information and matching material information;

[0032] Retrieving the remaining image features based on the matching remaining information;

[0033] Determine feature association values ​​based on remaining image features and material requirement features;

[0034] Determine whether the feature correlation value is greater than a preset correlation reference value;

[0035] If yes, then selecting the matching residual information corresponding to the remaining image features based on the feature association value to obtain image selection information;

[0036] Based on the combination of matching material information and image selection information and as selected material information;

[0037] If not, the matching material information will be used as the selected material information.

[0038] Optionally, a method for determining a feature association value includes:

[0039] According to the correspondence between the material demand characteristics and the preset demand local characteristics, the demand local characteristics corresponding to the material demand characteristics are determined;

[0040] Determine whether the remaining image features are consistent with the required local features;

[0041] If yes, then determine the feature similarity value based on the required local features;

[0042] According to the correspondence between the feature similarity value and the preset similarity association value, a similarity association value corresponding to the feature similarity value is determined, and the similarity association value is used as the feature association value;

[0043] If not, then determine the demand feature category information corresponding to the material demand feature based on the correspondence between the material demand feature and the preset demand feature category information;

[0044] Determining the residual feature type information corresponding to the residual image feature according to the correspondence between the residual image feature and the preset residual feature type information;

[0045] A category association value is determined according to the required feature category information and the remaining feature category information, and the category association value is used as the feature association value.

[0046] Optionally, methods for determining the category-related value include:

[0047] Determine whether the required feature type information is consistent with the remaining feature type information;

[0048] If yes, the preset category consistent association value is output as the category association value;

[0049] If not, then determining the category initial association value corresponding to the required feature category information and the remaining feature category information according to the correspondence between the required feature category information, the remaining feature category information and the preset category initial association value;

[0050] When the category initial association value is greater than the preset category base association value, the category initial association value is used as the category association value;

[0051] When the category initial association value is not greater than the preset category base association value, determining the demand-related category information corresponding to the demand-related category information according to the correspondence between the demand-feature category information and the preset demand-related category information;

[0052] According to the correspondence between the demand-related category information, the remaining feature category information and the preset category secondary association value, the category secondary association value corresponding to the demand-related category information and the remaining feature category information is determined;

[0053] The sum of the category initial association value and the category secondary association value is calculated and used as the category comprehensive association value, and the category comprehensive association value is used as the category association value.

[0054] Optionally, the method for generating short video output information includes:

[0055] Determining a style duration value corresponding to the required style information according to a correspondence between the required style information and the preset style duration value;

[0056] Calculate the difference between the required duration value and the style duration value and use it as the duration remaining value;

[0057] Determining the material duration value corresponding to the material information according to the correspondence between the material information used and the preset material duration value;

[0058] Determine whether the material duration value is greater than the remaining duration value;

[0059] If yes, calculate the ratio between the material duration value and the remaining duration value and use it as the material duration ratio value;

[0060] Adjusting and synthesizing the material information based on the material duration ratio to generate short video output information;

[0061] If not, calculate the difference between the material duration value and the remaining duration value and use it as the duration missing value;

[0062] Determine remaining material information based on selected material information and image material information;

[0063] Determine the duration and supplement the material information based on the missing duration value and the remaining material information;

[0064] The supplementary material information is combined with the used material information according to the duration and synthesized and edited to generate short video output information.

[0065] Optionally, methods for determining the duration supplementary material information include:

[0066] Determining the remaining material duration value corresponding to the remaining material information according to the correspondence between the remaining material information and the preset remaining material duration value;

[0067] Calculate the ratio between the remaining material duration value and the missing duration value and use it as the remaining material ratio value;

[0068] According to the correspondence between the remaining material ratio value and the preset remaining material ratio influence value, the remaining material ratio influence value corresponding to the remaining material ratio value is determined;

[0069] Retrieving remaining material features based on remaining material information;

[0070] Determine the remaining correlation value based on the remaining material characteristics and material demand characteristics;

[0071] Determining a residual correlation influence value corresponding to the residual correlation value according to a correspondence between the residual correlation value and a preset residual correlation influence value;

[0072] Calculate the sum of the remaining material proportion impact value and the remaining associated impact value and use it as the remaining material comprehensive impact value;

[0073] The remaining materials are sorted from large to small based on their comprehensive impact values, and the remaining material information is selected based on the sorting results and the missing values ​​of duration to form material selection information, and the material selection information is used as the duration supplementary material information.

[0074] In a second aspect, the present invention provides a short video generation system, which adopts the following technical solution:

[0075] A short video generation system, comprising:

[0076] Acquisition module, used to obtain image material information and short video demand information;

[0077] A memory, configured to store a program of the short video generation method according to any one of the first aspects;

[0078] The processor loads and executes the program in the memory.

[0079] In a third aspect, the present invention provides an intelligent terminal, which adopts the following technical solution:

[0080] A smart terminal includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute the short video generation method as described in any one of the first aspects.

[0081] In summary, the present invention includes at least one of the following beneficial technical effects:

[0082] 1. By acquiring image material information and short video demand information and retrieving required style information and required duration value, the selected material information is selected according to the style and the style required material information is retrieved. The used material information is then combined and edited to generate short video output information and output to the terminal held by ordinary users for display. In this way, ordinary users can input materials and automatically generate short videos after the materials are input, and can make subsequent adjustments by ordinary users, thereby facilitating the production of short videos by ordinary users;

[0083] 2. Retrieve style vocabulary from the required style information and search for synonyms. Then, identify the synonymous comprehensive vocabulary and search for style-related vocabulary. Then, aggregate the style comprehensive vocabulary and search for style material characteristics. These style material characteristics are then used as material requirement characteristics, thereby improving the accuracy of the obtained material requirement characteristics.

[0084] 3. By judging whether there is only one style vocabulary, if there is only one, directly aggregate the style vocabulary and synonyms to form a synonymous comprehensive vocabulary. If there is more than one, the consistent vocabulary is determined and aggregated according to the existence to form a synonymous comprehensive vocabulary, thereby improving the accuracy of the obtained synonymous comprehensive vocabulary. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 This is a flow chart of a method for generating a short video according to an embodiment of the present application;

[0086] Figure 2 is a flow chart of a method for determining material requirement characteristics according to an embodiment of the present application;

[0087] Figure 3 is a flow chart of a method for determining a synonymous comprehensive vocabulary according to an embodiment of the present application;

[0088] Figure 4 This is a flow chart of a method for determining selected material information according to an embodiment of the present application;

[0089] Figure 5 is a flow chart of a method for determining a feature association value according to an embodiment of the present application;

[0090] Figure 6is a flow chart of a method for determining a category association value according to an embodiment of the present application;

[0091] Figure 7 This is a flow chart of the method for generating short video output information in an embodiment of the present application. DETAILED DESCRIPTION

[0092] The present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0093] A short video generation method obtains image material information and short video demand information, selects materials based on vocabulary corresponding to the style, adjusts the materials based on the required duration, and then synthesizes and edits them to generate a short video and outputs it. In this way, ordinary users can input materials and the short video is automatically generated after the materials are input, and can be adjusted by ordinary users later, thereby facilitating the production of short videos by ordinary users.

[0094] Reference Figure 1 , an embodiment of the present invention discloses a short video generation method, which includes:

[0095] Step S100: Obtain image material information and short video demand information uploaded by ordinary users.

[0096] Image material information refers to the material information uploaded by ordinary users for short video production, and short video requirement information refers to the parameters such as the style and duration of the short video that ordinary users need to produce. Image material information and short video requirement information are obtained after being uploaded by ordinary users.

[0097] Step S101: Retrieve the required style information and required duration value based on the short video demand information.

[0098] The required style information refers to the style information of the short video that ordinary users need to create, and the required duration value refers to the duration value of the short video that ordinary users need to create. The required style information and required duration value can be retrieved through the short video demand information to facilitate subsequent use.

[0099] Step S102: Determine material requirement characteristics based on the required style information.

[0100] The material requirement characteristics refer to the shape, color, and other characteristics corresponding to the style required by ordinary users. By analyzing the required style information, the material requirement characteristics are determined to facilitate subsequent use. The specific steps for determining the material requirement characteristics are shown in steps S200 to S205.

[0101] Step S103: identifying the material information based on the material requirement characteristics and the image material information to determine the selected material information.

[0102] Among them, the selected material information refers to the material information after selecting the materials uploaded by ordinary users. By identifying the material requirement characteristics with the image material information, and using the materials in the image material information that are consistent with the material requirement characteristics as the selected material information, it is convenient for subsequent use.

[0103] Step S104: retrieving style requirement material information from a preset video material library based on the required style information, and combining the selected material information with the style requirement material information to form used material information.

[0104] The video material library refers to a database storing materials of various styles, which are obtained through pre-input. Style requirement material information refers to the material information corresponding to the required style, and usage material information refers to the material information required to produce a short video. Materials that match the required style information are retrieved from the preset video material library and used as style requirement material information. The selected material information is then combined with the style requirement material information to form usage material information, facilitating subsequent use.

[0105] Step S105: Synthesize and edit the used material information according to the required style information and the required duration value to generate short video output information, and output the short video output information to a terminal held by an ordinary user for display.

[0106] The short video output information refers to the short video information that needs to be output, and the terminal held by the ordinary user refers to the terminal that the ordinary user can receive and display the short video. The terminal held by the ordinary user can be a mobile phone or a computer. By synthesizing and editing the required style information and the required duration value with the used material information, the short video output information is generated and output to the terminal held by the ordinary user for display. The ordinary user can input the material and the short video is automatically generated after the material is generated and can be subsequently adjusted by the ordinary user, thereby facilitating the production of short videos by ordinary users.

[0107] exist Figure 1 In step S102, in order to further ensure the rationality of the material demand characteristics, it is necessary to further analyze and calculate the material demand characteristics. Figure 2 The steps shown are explained in detail.

[0108] Reference Figure 2 ,The method for determining the material requirement characteristics includes the following steps:

[0109] Step S200: Retrieving style vocabulary based on required style information.

[0110] Among them, style vocabulary refers to the vocabulary corresponding to the style produced by ordinary users' needs. The style vocabulary is retrieved through the required style information to facilitate subsequent use.

[0111] Step S201: determining a synonymous word corresponding to a style word according to a correspondence between the style word and preset synonymous words.

[0112] Synonyms are words that have the same meaning as style words. Different style words correspond to different synonyms. Synonyms are retrieved from a database that stores different style words and their corresponding synonyms. This database is retrieved through pre-input. Synonyms are determined through style word queries to facilitate subsequent use.

[0113] Step S202: Determine a synonymous comprehensive vocabulary based on the style vocabulary and the synonym vocabulary.

[0114] The synonymous comprehensive vocabulary refers to the vocabulary corresponding to the synthesis of style vocabulary and synonyms. By analyzing the style vocabulary and synonyms, the synonymous comprehensive vocabulary is determined to facilitate subsequent use. The specific steps for determining the synonymous comprehensive vocabulary refer to steps S300 to S306.

[0115] Step S203: determining the style-related vocabulary corresponding to the synonymous comprehensive vocabulary according to the corresponding relationship between the synonymous comprehensive vocabulary and the preset style-related vocabulary.

[0116] Among them, style-related words refer to words that are mutually related to synonymous comprehensive words. Different synonymous comprehensive words correspond to different style-related words. The style-related words are obtained by querying from a database that stores different synonymous comprehensive words and their corresponding style-related words. The database is obtained after pre-input. The style-related words are determined by querying the synonymous comprehensive words to facilitate subsequent use.

[0117] Step S204: Summarize the synonymous comprehensive words and the style-related words to form a style comprehensive word.

[0118] Among them, the style comprehensive vocabulary refers to the comprehensive vocabulary corresponding to the required style. By summarizing synonymous comprehensive vocabulary and style-related vocabulary, the summarized vocabulary is used as the style comprehensive vocabulary to facilitate subsequent use.

[0119] Step S205: determining the style material features corresponding to the style comprehensive vocabulary according to the corresponding relationship between the style comprehensive vocabulary and the preset style material features, and using the style material features as the material requirement features.

[0120] The style material characteristics refer to the material characteristics of items corresponding to the comprehensive vocabulary of the required style. Different style comprehensive vocabulary corresponds to different style material characteristics. The style material characteristics are obtained by querying from a database that stores different style comprehensive vocabulary and corresponding style material characteristics. This database is obtained through pre-input. The style material characteristics are determined through the style comprehensive vocabulary query and used as the material requirement characteristics, thereby improving the accuracy of the obtained material requirement characteristics.

[0121] exist Figure 2 In step S202, in order to further ensure the rationality of the synonymous comprehensive vocabulary, it is necessary to further analyze and calculate the synonymous comprehensive vocabulary separately, specifically by Figure 3 The steps shown are explained in detail.

[0122] Reference Figure 3 , the method for determining the synonymous comprehensive vocabulary includes the following steps:

[0123] Step S300: Determine whether there is only one style word. If yes, execute step S301; if no, execute step S302.

[0124] Here, by judging whether there is only one style word, it is judged whether the words can be directly merged and summarized.

[0125] Step S301: Summarize style words and synonyms to form a synonymous comprehensive word.

[0126] Among them, when there is only one style word, it means that the words can be directly merged and summarized at this time, so the style word and synonymous words are summarized and formed into a synonymous comprehensive word, thereby improving the accuracy of the obtained synonymous comprehensive word.

[0127] Step S302: determining consistent words based on the consistency between the style words and the synonyms.

[0128] Among them, consistent vocabulary refers to the vocabulary that corresponds to the consistency between style vocabulary and synonyms. When there is more than one style vocabulary, it means that the vocabulary cannot be directly merged and summarized at this time. Consistent vocabulary is formed by analyzing the consistency between style vocabulary and synonyms.

[0129] Step S303: When the consistent vocabulary does not exist, the style vocabulary and synonyms are aggregated to form a synonymous comprehensive vocabulary.

[0130] Among them, when the consistent vocabulary does not exist, it means that there is no need to delete the vocabulary at this time, so the style vocabulary and synonyms are aggregated and formed into a synonymous comprehensive vocabulary, thereby improving the accuracy of the obtained synonymous comprehensive vocabulary.

[0131] Step S304: When a consistent vocabulary exists, the style vocabulary is adjusted based on the consistent vocabulary and used as the style adjustment vocabulary.

[0132] Among them, the style adjustment vocabulary refers to the corresponding vocabulary after the style vocabulary is adjusted. When consistent vocabulary exists, it means that the vocabulary needs to be deleted at this time. Therefore, the consistent vocabulary in the style vocabulary is deleted and adjusted, and the deleted vocabulary is used as the style adjustment vocabulary for convenience in subsequent use.

[0133] Step S305: determining a synonymous adjustment word corresponding to the style adjustment word according to the correspondence between the style adjustment word and the preset synonymous adjustment words.

[0134] Synonymous adjustment words are words with the same meaning as the corresponding style adjustment words. Different style adjustment words correspond to different synonymous adjustment words. Synonymous adjustment words are obtained by querying from a database that stores different style adjustment words and their corresponding synonymous adjustment words. This database is obtained through pre-input. Synonymous adjustment words are determined through style adjustment word query to facilitate subsequent use.

[0135] Step S306: Summarize the style-adjusted words and the synonym-adjusted words to form a synonym-comprehensive word.

[0136] Among them, the accuracy of the acquired synonymous comprehensive vocabulary is improved by aggregating the style adjustment vocabulary and the synonym adjustment vocabulary to form the synonymous comprehensive vocabulary.

[0137] exist Figure 1 In step S103, in order to further ensure the rationality of the selected material information, it is necessary to further analyze and calculate the selected material information. Figure 4 The steps shown are explained in detail.

[0138] Reference Figure 4 , the method for determining the selected material information includes the following steps:

[0139] Step S400: performing identification and matching from image material information based on material requirement features to form matching material information.

[0140] Among them, matching material information refers to the material information corresponding to the successful matching based on the material requirement characteristics. The material corresponding to the characteristics consistent with the material requirement characteristics is identified and matched from the image material information and used as the matching material information for subsequent use.

[0141] Step S401: determining matching remaining information based on image material information and matching material information.

[0142] Among them, the matching residual information refers to the material information remaining after the successful matching in the image material uploaded by ordinary users. By removing the matching material information in the image material information, the remaining material in the image material information is used as the matching residual information for convenience in subsequent use.

[0143] Step S402: Retrieve remaining image features based on the matching remaining information.

[0144] Among them, the remaining image features refer to the features existing in the remaining images in the uploaded image material. By extracting the features of the images in the matching remaining information and using them as the remaining image features, it is convenient for subsequent use.

[0145] Step S403: determining a feature association value based on the remaining image features and the material requirement features.

[0146] The feature association value refers to the association value corresponding to the association between the remaining image features and the material requirement features. By analyzing the remaining image features and the material requirement features, the feature association value is determined to facilitate subsequent use. The specific steps for determining the feature association value are shown in steps S500 to S506.

[0147] Step S404: Determine whether the feature association value is greater than a preset association reference value. If yes, proceed to step S405; if no, proceed to step S407.

[0148] The correlation reference value refers to the minimum value corresponding to when features are correlated. By judging whether the feature correlation value is greater than the preset correlation reference value, it is determined whether there is a correlation between the remaining image features and the material requirement features.

[0149] Step S405: Selecting the remaining matching information corresponding to the remaining image features based on the feature association value to obtain image selection information.

[0150] Image selection information refers to the image information corresponding to the remaining images selected from the uploaded image material. When the feature correlation value is greater than the preset correlation baseline value, it indicates that there is a correlation between the remaining image features and the required material features. Therefore, by selecting the remaining image features corresponding to the feature correlation value greater than the preset correlation baseline value, the matching residual information corresponding to the selected remaining image features is retrieved and used as image selection information, thereby facilitating subsequent use.

[0151] Step S406: combining the matching material information with the image selection information and using the result as the selected material information.

[0152] Herein, by combining the matching material information with the image selection information, and using the combined material as the selected material information, the accuracy of the obtained selected material information is improved.

[0153] Step S407: using the matched material information as the selected material information.

[0154] Among them, when the feature correlation value is not greater than the preset correlation reference value, it means that there is no correlation between the remaining image features and the material requirement features at this time, so the matching material information is used as the selected material information, thereby improving the accuracy of the obtained selected material information.

[0155] exist Figure 4 In step S403, in order to further ensure the rationality of the feature correlation value, it is necessary to further analyze and calculate the feature correlation value. Figure 5 The steps shown are explained in detail.

[0156] Reference Figure 5 , the method for determining the feature association value includes the following steps:

[0157] Step S500: determining a demand local feature corresponding to the material demand feature according to a correspondence between the material demand feature and the preset demand local feature.

[0158] A local requirement feature refers to a material requirement feature that corresponds only to a portion of the overall feature. Different material requirement features correspond to different local requirements features. These features are retrieved by querying a database that stores different material requirement features and their corresponding local requirements features. This database is retrieved after pre-entering. Determining the local requirements features through a material requirement feature query facilitates subsequent use.

[0159] Step S501: Determine whether the remaining image features are consistent with the required local features. If yes, proceed to step S502; if not, proceed to step S504.

[0160] Here, by judging whether the remaining image features are consistent with the required local features, it is judged whether the remaining images in the uploaded image material are directly related.

[0161] Step S502: determining feature similarity values ​​according to required local features.

[0162] The feature similarity value refers to the similarity value corresponding to similar features. Different required local features correspond to different feature similarity values. When the remaining image features match the required local features, it indicates that there is a direct correlation between the remaining images in the uploaded image material. Therefore, the feature similarity value is retrieved using the required local features that match the remaining image features, facilitating subsequent use.

[0163] Step S503: determining a similarity association value corresponding to the feature similarity value according to a correspondence between the feature similarity value and a preset similarity association value, and using the similarity association value as the feature association value.

[0164] The similarity association value refers to the association value corresponding to feature similarity. Different feature similarity values ​​correspond to different similarity association values. The similarity association value is obtained by querying from a database that stores different feature similarity values ​​and their corresponding similarity association values. The database is obtained through pre-input. The similarity association value is determined by querying the feature similarity value and is used as the feature association value, thereby improving the accuracy of the obtained feature association value.

[0165] Step S504: Determine the demand feature category information corresponding to the material demand feature according to the correspondence between the material demand feature and the preset demand feature category information.

[0166] Among them, the demand feature category information refers to the category information to which the material demand feature belongs. Different material demand features correspond to different demand feature category information. The demand feature category information is obtained by querying from a database that stores different material demand features and corresponding demand feature category information. The database is obtained after pre-input.

[0167] When the remaining image features are not consistent with the required local features, it means that there is no direct correlation between the remaining images in the uploaded image material. Therefore, the required feature type information is determined through the material required feature query to facilitate subsequent use.

[0168] Step S505: determining the remaining feature category information corresponding to the remaining image features according to the correspondence between the remaining image features and the preset remaining feature category information.

[0169] The remaining feature category information refers to the category information to which the remaining image features belong. Different remaining image features correspond to different remaining feature category information. The remaining feature category information is obtained by querying from a database that stores different remaining image features and corresponding remaining feature category information. The database is obtained after pre-input. Determining the remaining feature category information through the remaining image feature query facilitates subsequent use.

[0170] Step S506: determining a category association value according to the required feature category information and the remaining feature category information, and using the category association value as the feature association value.

[0171] The category association value refers to the association value corresponding to the presence of correlation between categories. The category association value is determined by analyzing the demand feature category information and the remaining feature category information. This category association value is then used as the feature association value to improve the accuracy of the acquired feature association value. For details on the steps for determining the category association value, see steps S600 to S606.

[0172] exist Figure 5 In step S506, in order to further ensure the rationality of the category association value, it is necessary to further analyze and calculate the category association value. Figure 6 The steps shown are explained in detail.

[0173] Reference Figure 6 , the method for determining the category association value includes the following steps:

[0174] Step S600: Determine whether the required feature type information is consistent with the remaining feature type information. If yes, execute step S601; if not, execute step S602.

[0175] Here, by judging whether the required feature category information is consistent with the remaining feature category information, it is judged whether there is a consistent association between the categories.

[0176] Step S601: Outputting a preset category-consistent association value as the category association value.

[0177] The category consistency association value refers to the association value corresponding to category consistency. This association value is obtained through pre-input. When the required feature category information is consistent with the remaining feature category information, it indicates that a consistent association exists between the categories. Therefore, the preset category consistency association value is output as the category association value, thereby improving the accuracy of the obtained category association value.

[0178] Step S602: determining the category initial association value corresponding to the demand feature category information and the remaining feature category information according to the correspondence between the demand feature category information, the remaining feature category information and the preset category initial association value.

[0179] Among them, the category initial association value refers to the initial association value corresponding to when there is an association between categories. Different demand feature category information and remaining feature category information correspond to different category initial association values. The category initial association value is obtained by querying from a database that stores different demand feature category information and remaining feature category information and the corresponding category initial association values. The database is obtained after pre-input.

[0180] When the required feature category information is inconsistent with the remaining feature category information, it means that there is no consistent association between the categories at this time. Therefore, the initial category association value is determined by querying the required feature category information and the remaining feature category information to facilitate subsequent use.

[0181] Step S603: When the category initial association value is greater than the preset category reference association value, the category initial association value is used as the category association value.

[0182] The category base association value refers to the minimum association value corresponding to an existing category association. This value is obtained by pre-entering the category base association value. When the category initial association value is greater than the preset category base association value, it indicates that the categories are associated. Therefore, the category initial association value is used as the category association value, thereby improving the accuracy of the obtained category association value.

[0183] Step S604: when the category initial association value is not greater than the preset category reference association value, the demand association category information corresponding to the demand feature category information is determined according to the correspondence between the demand feature category information and the preset demand association category information.

[0184] Demand-related category information refers to category information associated with demand feature categories. Different demand feature category information corresponds to different demand-related category information. The demand-related category information is retrieved by querying from a database that stores different demand feature category information and corresponding demand-related category information. This database is retrieved through pre-input. If the category initial association value is not greater than the preset category baseline association value, it indicates that there is no direct association between the categories. Therefore, the demand-related category information is determined through the demand feature category information query to facilitate subsequent use.

[0185] Step S605: Determine the category secondary association value corresponding to the demand-related category information and the remaining feature category information according to the correspondence between the demand-related category information, the remaining feature category information and the preset category secondary association value.

[0186] The category secondary association value refers to the association value corresponding to a category with a secondary association. Different demand-related category information and remaining feature category information correspond to different category secondary association values. Category secondary association values ​​are obtained by querying a database that stores different demand-related category information and remaining feature category information and their corresponding category secondary association values. This database is obtained through pre-input. Determining the category secondary association value through querying the demand-related category information and remaining feature category information facilitates subsequent use.

[0187] Step S606: Calculate the sum of the category initial association value and the category secondary association value and use it as the category comprehensive association value, and use the category comprehensive association value as the category association value.

[0188] Among them, the comprehensive association value of the category refers to the comprehensive association value corresponding to the existence of the category association. By calculating the sum of the category initial association value and the category secondary association value and using it as the category comprehensive association value, and then using the category comprehensive association value as the category association value, the accuracy of the obtained category association value is improved.

[0189] exist Figure 1 In step S105, in order to further ensure the rationality of the short video output information, it is necessary to further analyze and calculate the short video output information. Figure 7 The steps shown are explained in detail.

[0190] Reference Figure 7 , the method for generating short video output information includes the following steps:

[0191] Step S700: determining a style duration value corresponding to the required style information according to a correspondence between the required style information and a preset style duration value.

[0192] The style duration value refers to the baseline duration value corresponding to the style during the short video production process. Different required style information corresponds to different style duration values. The style duration value is obtained by querying from a database that stores different required style information and corresponding style duration values. This database is obtained through pre-input. Determining the style duration value through querying the required style information facilitates subsequent use.

[0193] Step S701: Calculate the difference between the required duration value and the style duration value and use it as the duration remaining value.

[0194] The remaining duration value refers to the remaining duration value corresponding to the required duration minus the style base duration. The difference between the required duration value and the style duration value is calculated and used as the remaining duration value for subsequent use.

[0195] Step S702: determining the material duration value corresponding to the used material information according to the correspondence between the used material information and the preset material duration value.

[0196] The material duration value refers to the duration corresponding to the normal production of the material being used. Different material information corresponds to different material duration values. The material duration value is obtained by querying from a database that stores different material information and corresponding material duration values. This database is pre-entered and then obtained. Determining the material duration value by querying the material information facilitates subsequent use.

[0197] Step S703: Determine whether the material duration value is greater than the remaining duration value. If yes, execute step S704; if not, execute step S706.

[0198] Among them, by judging whether the material duration value is greater than the remaining duration value, it is determined whether the used material can be used to directly produce a short video.

[0199] Step S704: Calculate the ratio between the material duration value and the remaining duration value and use it as the material duration ratio value.

[0200] Among them, the material duration ratio value refers to the ratio between the material duration and the remaining duration. When the material duration value is greater than the remaining duration value, it means that the material used at this time can be directly used to produce a short video. Therefore, the ratio between the material duration value and the remaining duration value is calculated and used as the material duration ratio value for subsequent use.

[0201] Step S705: Based on the material duration ratio value, the material information is adjusted and synthesized and edited to generate short video output information.

[0202] Among them, the duration of the used material information is adjusted according to the material duration ratio value, and the adjusted material is synthesized and edited to generate short video output information, thereby improving the accuracy of the obtained short video output information.

[0203] Step S706: Calculate the difference between the material duration value and the duration remaining value and use it as the duration missing value.

[0204] Among them, the missing value of duration refers to the missing value corresponding to the missing duration. When the material duration value is not greater than the remaining duration value, it means that the material used at this time cannot be used to directly produce a short video. Therefore, the difference between the material duration value and the remaining duration value is calculated and used as the missing value of duration for subsequent use.

[0205] Step S707: Determine remaining material information according to the selected material information and the image material information.

[0206] The remaining material information refers to the remaining material information after the uploaded material is selected. The selected material information in the image material information is deleted and adjusted, and the adjusted image material information is used as the remaining material information for easy subsequent use.

[0207] Step S708: Determine the duration and supplement the material information based on the missing duration value and the remaining material information.

[0208] The duration-supplemented material information refers to the material information corresponding to the material supplemented based on the duration. By analyzing the missing duration values ​​and the remaining material information, the duration-supplemented material information is determined to facilitate subsequent use. The specific steps for determining the duration-supplemented material information are as follows: Steps S800 to S807.

[0209] Step S709: The supplementary material information is combined with the used material information according to the duration and synthesized and edited to generate short video output information.

[0210] Among them, by combining the duration supplementary material information with the usage material information, and then synthesizing and editing the combined materials to generate short video output information, the accuracy of the obtained short video output information is improved.

[0211] exist Figure 7 In step S708 shown, in order to further ensure the rationality of the duration supplementary material information, it is necessary to perform further separate analysis and calculation on the duration supplementary material information, which is specifically described in detail through the following steps.

[0212] The method for determining the duration supplementary material information includes the following steps:

[0213] Step S800: determining the remaining material duration value corresponding to the remaining material information according to the correspondence between the remaining material information and the preset remaining material duration value.

[0214] The remaining material duration value refers to the duration value of the remaining material when producing a short video. Different remaining material information corresponds to different remaining material duration values. The remaining material duration value is obtained by querying from a database that stores different remaining material information and corresponding remaining material duration values. The database is obtained after pre-entry. Determining the remaining material duration value by querying the remaining material information facilitates subsequent use.

[0215] Step S801: Calculate the ratio between the remaining material duration value and the missing duration value and use it as the remaining material ratio value.

[0216] Among them, the remaining material ratio value refers to the ratio between the remaining material duration and the missing duration. By calculating the ratio between the remaining material duration value and the missing duration value and using it as the remaining material ratio value, it is convenient for subsequent use.

[0217] Step S802: determining a remaining material proportion influence value corresponding to the remaining material proportion value according to a correspondence between the remaining material proportion value and a preset remaining material proportion influence value.

[0218] The remaining material ratio impact value refers to the degree of influence of the ratio between the remaining material duration and the missing duration on material selection. Different remaining material ratio values ​​correspond to different remaining material ratio impact values. The remaining material ratio impact value is obtained by querying from a database that stores different remaining material ratio values ​​and corresponding remaining material ratio impact values. The database is obtained through pre-input. Determining the remaining material ratio impact value through the remaining material ratio query facilitates subsequent use.

[0219] Step S803: Retrieve remaining material features based on the remaining material information.

[0220] The remaining material features refer to the features contained in the image corresponding to the remaining material. The remaining material features are retrieved through the remaining material information to facilitate subsequent use.

[0221] Step S804: determining a remaining correlation value according to the remaining material characteristics and the material demand characteristics.

[0222] The remaining correlation value refers to the correlation value corresponding to the correlation between the remaining material characteristics and the material requirement characteristics. By analyzing the remaining material characteristics and the material requirement characteristics, the remaining correlation value is determined to facilitate subsequent use. The steps for determining the remaining correlation value can refer to steps S500 to S506.

[0223] Step S805: determining a residual correlation influence value corresponding to the residual correlation value according to a correspondence between the residual correlation value and a preset residual correlation influence value.

[0224] The residual correlation impact value refers to the degree of influence on material selection based on the residual correlation value. Different residual correlation values ​​correspond to different residual correlation impact values. The residual correlation impact value is obtained by querying from a database that stores different residual correlation values ​​and their corresponding residual correlation impact values. The database is obtained after pre-input. Determining the residual correlation impact value through the residual correlation value query facilitates subsequent use.

[0225] Step S806: Calculate the sum of the remaining material proportion impact value and the remaining associated impact value and use it as the remaining material comprehensive impact value.

[0226] Among them, the comprehensive impact value of the remaining materials refers to the comprehensive impact value generated by selecting the remaining materials. The sum of the remaining material proportion impact value and the remaining associated impact value is calculated and used as the comprehensive impact value of the remaining materials to facilitate subsequent use.

[0227] Step S807: Sort the remaining materials from large to small based on their comprehensive impact values, and select the remaining material information according to the sorting result and the missing value of duration to form material selection information, and use the material selection information as the duration supplementary material information.

[0228] Among them, the material selection information refers to the material information corresponding to the remaining materials after selection. By sorting the comprehensive impact values ​​of the remaining materials from large to small, and selecting them in sequence according to the sorting results, the consistency of the sum of the durations corresponding to the selected remaining material information and the missing duration value is analyzed. When they are consistent, the selected remaining material information is directly used as the material selection information. When they are inconsistent, the selection is adjusted again based on the sorting results until the sum of the durations corresponding to the selected remaining material information is consistent with the missing duration value. The material selection information is then used as the duration supplementary material information, thereby improving the accuracy of the obtained duration supplementary material information.

[0229] Based on the same inventive concept, an embodiment of the present invention provides a short video generation system, including:

[0230] Acquisition module, used to obtain image material information and short video demand information;

[0231] A memory for storing a program for the short video generation method as described above;

[0232] The processor loads and executes the program in the memory.

[0233] Based on the same inventive concept, an embodiment of the present invention provides a smart terminal, including a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute the short video generation method as described above.

[0234] Those skilled in the art will clearly understand that for the sake of convenience and brevity, the division of the above-mentioned functional modules is only used as an example for illustration. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working processes of the above-mentioned systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0235] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A short video generation method, characterized in that: include: Obtain image material information and short video demand information uploaded by ordinary users; Based on the short video demand information, the required style information and required duration value are retrieved. The required style information refers to the style information of the short video that the general user needs to produce; Determine the material requirements based on the required style information. Material requirements refer to the shape and color characteristics corresponding to the style required by ordinary users. Identify the material requirements and image material information to determine the selected material information. The selected material information refers to the material information after selecting the materials uploaded by ordinary users. Based on the required style information, style required material information is retrieved from a preset video material library, and the selected material information is combined with the style required material information to form the used material information. The style required material information refers to the material information corresponding to the required style, and the used material information refers to the material information required to produce the short video; Synthesize and edit the used material information according to the required style information and the required duration value to generate short video output information, and output the short video output information to the terminal held by the general user for display; Methods for determining material demand characteristics include: Retrieve style vocabulary based on required style information; According to the correspondence between the style vocabulary and the preset synonyms, the synonyms corresponding to the style vocabulary are determined; Determine synonymous comprehensive vocabulary based on style vocabulary and synonyms; According to the corresponding relationship between the synonymous comprehensive vocabulary and the preset style-related vocabulary, the style-related vocabulary corresponding to the synonymous comprehensive vocabulary is determined; Summarize and form style comprehensive vocabulary based on synonymous comprehensive vocabulary and style related vocabulary; According to the correspondence between the style comprehensive vocabulary and the preset style material features, the style material features corresponding to the style comprehensive vocabulary are determined, and the style material features are used as the material requirement features; Methods for determining selected material information include: Based on the material requirement features, identification and matching are performed from the image material information to form matching material information; Determine matching remaining information based on image material information and matching material information; Retrieving the remaining image features based on the matching remaining information; Determine feature association values ​​based on remaining image features and material requirement features; Determine whether the feature correlation value is greater than a preset correlation reference value; If yes, then selecting the matching residual information corresponding to the remaining image features based on the feature association value to obtain image selection information; Based on the combination of matching material information and image selection information and as selected material information; If not, the matching material information will be used as the selected material information.

2. The short video generation method according to claim 1, characterized in that Methods for determining synonymous comprehensive words include: Determine whether there is only one style word; If yes, then summarize the style words and synonyms and form a synonymous comprehensive word; If not, then based on the consistency between the style vocabulary and synonyms, the consistent vocabulary is determined; When consistent vocabulary does not exist, style vocabulary and synonyms are aggregated to form a synonymous comprehensive vocabulary; When consistent vocabulary exists, the style vocabulary is adjusted based on the consistent vocabulary and used as the style adjustment vocabulary; According to the correspondence between the style adjustment vocabulary and the preset synonym adjustment vocabulary, the synonym adjustment vocabulary corresponding to the style adjustment vocabulary is determined; Based on the style adjustment vocabulary and synonym adjustment vocabulary, they are summarized and formed into a synonym comprehensive vocabulary.

3. The short video generation method according to claim 1, characterized in that Methods for determining feature association values ​​include: According to the correspondence between the material demand characteristics and the preset demand local characteristics, the demand local characteristics corresponding to the material demand characteristics are determined; Determine whether the remaining image features are consistent with the required local features; If yes, then determine the feature similarity value based on the required local features; According to the correspondence between the feature similarity value and the preset similarity association value, a similarity association value corresponding to the feature similarity value is determined, and the similarity association value is used as the feature association value; If not, then determine the demand feature category information corresponding to the material demand feature based on the correspondence between the material demand feature and the preset demand feature category information; Determining the residual feature type information corresponding to the residual image feature according to the correspondence between the residual image feature and the preset residual feature type information; A category association value is determined according to the required feature category information and the remaining feature category information, and the category association value is used as the feature association value.

4. The short video generation method according to claim 3, characterized in that: Methods for determining category-related values ​​include: Determine whether the required feature type information is consistent with the remaining feature type information; If yes, the preset category consistent association value is output as the category association value; If not, then determining the category initial association value corresponding to the required feature category information and the remaining feature category information according to the correspondence between the required feature category information, the remaining feature category information and the preset category initial association value; When the category initial association value is greater than the preset category base association value, the category initial association value is used as the category association value; When the category initial association value is not greater than the preset category base association value, determining the demand-related category information corresponding to the demand-related category information according to the correspondence between the demand-feature category information and the preset demand-related category information; According to the correspondence between the demand-related category information, the remaining feature category information and the preset category secondary association value, the category secondary association value corresponding to the demand-related category information and the remaining feature category information is determined; The sum of the category initial association value and the category secondary association value is calculated and used as the category comprehensive association value, and the category comprehensive association value is used as the category association value.

5. The short video generation method according to claim 1, characterized in that: The method for generating short video output information includes: Determining a style duration value corresponding to the required style information according to a correspondence between the required style information and the preset style duration value; Calculate the difference between the required duration value and the style duration value and use it as the duration remaining value; Determining the material duration value corresponding to the material information according to the correspondence between the material information used and the preset material duration value; Determine whether the material duration value is greater than the remaining duration value; If yes, calculate the ratio between the material duration value and the remaining duration value and use it as the material duration ratio value; Adjusting and synthesizing the material information based on the material duration ratio to generate short video output information; If not, calculate the difference between the material duration value and the remaining duration value and use it as the duration missing value; Determine remaining material information based on selected material information and image material information; Determine the duration and supplement the material information based on the missing duration value and the remaining material information; The supplementary material information is combined with the used material information according to the duration and synthesized and edited to generate short video output information.

6. The short video generation method according to claim 5, characterized in that: Methods for determining the duration supplementary material information include: Determining the remaining material duration value corresponding to the remaining material information according to the correspondence between the remaining material information and the preset remaining material duration value; Calculate the ratio between the remaining material duration value and the missing duration value and use it as the remaining material ratio value; According to the correspondence between the remaining material ratio value and the preset remaining material ratio influence value, the remaining material ratio influence value corresponding to the remaining material ratio value is determined; Retrieving remaining material features based on remaining material information; Determine the remaining correlation value based on the remaining material characteristics and material demand characteristics; Determining a residual correlation influence value corresponding to the residual correlation value according to a correspondence between the residual correlation value and a preset residual correlation influence value; Calculate the sum of the remaining material proportion impact value and the remaining associated impact value and use it as the remaining material comprehensive impact value; The remaining materials are sorted from large to small based on their comprehensive impact values, and the remaining material information is selected based on the sorting results and the missing values ​​of duration to form material selection information, and the material selection information is used as the duration supplementary material information.

7. A short video generation system, characterized in that: include: Acquisition module, used to obtain image material information and short video demand information; A memory for storing a program of the short video generation method according to any one of claims 1 to 6; The processor loads and executes the program in the memory.

8. An intelligent terminal, characterized in that: The apparatus comprises a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute the short video generation method according to any one of claims 1 to 6.

Images