Video generation method and system based on artificial intelligence

Abstract

The invention discloses a video generation method and system based on artificial intelligence, and belongs to the technical field of video generation methods. A target information module comprises body temperature evaluation information, emotion model information, environment model information, family relationship information and weather information; the target information evaluation module is used for evaluating the target information module; the family relationship information comprises a third-generation co-hall relationship, a second-generation relationship, a single family relationship, an adult solitary relationship, an isolated-nest elderly relationship and an orphan relationship, and the lens sensing control library comprises rightward movement, ascending, descending, leftward movement, strong vibration, departure and approaching. According to the method, overall evaluation can be carried out, the states of family members can be known, the color of the corresponding video and the actual display effect of the video can be linked, and the video can better meet the use effect which a photographed person wants to express. And poor experience caused by influence on the emotion of the video object due to lack of scene in video generation is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of home-related video generation, and in particular to a video generation method and system based on artificial intelligence. Background Art

[0002] Smart Home refers to the use of Internet of Things technology, cloud computing, big data analysis and other technologies to achieve intelligent management of the home environment and improve the comfort and convenience of living. Among the current video generation devices, there is no video generation device or device specifically for the home.

[0003] In the actual use process, the video applications and systems in the related technologies do not consider the selection of accessories when facing family objects, and the adaptability between the various accessories. The effect of applying and coordinating accessories is poor. After the video is shot, it is only independent and scattered video data, and does not comprehensively consider the members of the smart family, or other related parameters of the family members and the correlation between the various parameters, resulting in the user video generation being relatively monotonous and without any correlation. In addition, the appropriate video generation form is not considered for family relationships, resulting in some videos that do not conform to specific family relationships cannot be reasonably displayed, affecting the experience of family members viewing videos. Therefore, a video generation method based on artificial intelligence is proposed. Summary of the invention

[0004] The technical problem to be solved by the present invention is: how to solve the problem that in the existing video applications and video systems, in actual use, the types of parameters of the target need to be obtained in addition to the video itself are not taken into account, and the shooting is too mechanical and without warmth; and the current video acquisition mainly focuses on the adaptability of each parameter, and the application and synergy effects are poor, which has a certain impact on the use of video generation. A video generation method based on artificial intelligence is provided.

[0005] The present invention solves the above technical problems through the following technical solutions, which include: Target information module: The target information module includes body temperature assessment information, emotion model information, environment model information, family relationship information and weather information; Body temperature assessment information: The body temperature monitoring module is used to monitor the body temperature information of the target object, and the body temperature analysis module is used to analyze the target body temperature information to obtain body temperature assessment information; Emotion model information: the emotion acquisition module is used to acquire the target's emotions, and the emotion processing module is used to process the acquired target's emotion information to obtain the emotion model information; Environmental model information: The environmental acquisition module is used to acquire environmental information of the target object, and the environmental processing module is used to process the environmental acquisition of the target object to obtain environmental model information; Family relationship information: The family relationship collection module is used to collect and process the family relationship information of the target object; Weather information: The weather collection and processing module is used to collect and process the weather information of the target object; A lens sensor control library, the lens sensor control library includes moving to the right, rising, falling, moving to the left, strong vibration, moving away and approaching, the moving to the right corresponds to the family portrait feeling in the lens sensor control library, the rising corresponds to the macro display in the lens sensor control library, the falling corresponds to the detail display in the lens sensor control library, the moving to the left corresponds to the documentary in the lens sensor control library, the strong vibration corresponds to the fast switching lens in the lens sensor control library, the moving away corresponds to the gradual zooming out in the lens sensor control library, and the approaching corresponds to the gradual zooming in in the lens sensor control library.

[0006] The lens sensor control library includes moving to the right (family portrait feeling), rising (macro display), falling (detail display), moving to the left (documentary), strong vibration (rapid lens switching), moving away (gradually zooming out), and approaching (gradually zooming in).

[0007] Furthermore, the emotions include happiness, anger, sadness and fear.

[0008] Furthermore, the target information evaluation module sequentially evaluates family relationship information, environmental model information, emotional model information, body temperature evaluation information and weather information. By first evaluating the family relationship information, the relationship between the subjects can be understood first, thereby first laying the tone for the generation of this video, thereby avoiding some video shooting forms that affect the emotions of elderly people living alone after shooting, protecting family members to the greatest extent, and also recording the current situation that needs to be recorded.

[0009] Furthermore, the environmental information includes a sense of daily life, a sense of festival celebrations, a sense of family warmth, a sense of community activities, a sense of natural scenery, a sense of growth records and a sense of outdoor camping.

[0010] Furthermore, the family relationship information includes three-generation living together relationship, two-generation relationship, single-parent family relationship, adult living alone relationship, lonely elderly relationship and orphan relationship.

[0011] Furthermore, the weather information includes sunny days, rainy days, cloudy days and snowy days.

[0012] Furthermore, the weather collection and processing module obtains local weather information through the Internet of Things.

[0013] Furthermore, a video generation method based on artificial intelligence comprises the following steps: S1: Acquire target information, wherein the target information includes body temperature assessment information, emotion model information, environment model information, weather information, and family relationship information; S2: calling a preset lens sensor control library, wherein the lens sensor control library includes shooting techniques and lens states, and the lens states correspond to the shooting techniques; S3: Determine a target designated lens based on the target information and the lens sensor control library; S4: Generate a target video based on the target designated shot; The lens sensor control library includes moving to the right, rising, falling, moving to the left, strong vibration, moving away and approaching, the moving to the right corresponds to the family portrait feeling in the lens sensor control library, the rising corresponds to the macro display in the lens sensor control library, the falling corresponds to the detail display in the lens sensor control library, the moving to the left corresponds to the documentary in the lens sensor control library, the strong vibration corresponds to the fast switching lens in the lens sensor control library, the moving away corresponds to the gradual zooming out in the lens sensor control library, and the approaching corresponds to the gradual zooming in in the lens sensor control library.

[0014] Furthermore, the body temperature assessment information: the body temperature monitoring module is used to monitor the body temperature information of the target object, and the body temperature analysis module is used to analyze the target body temperature information to obtain the body temperature assessment information; Emotion model information: the emotion acquisition module is used to acquire the target's emotions, and the emotion processing module is used to process the acquired target's emotion information to obtain the emotion model information; Environmental model information: The environmental acquisition module is used to acquire environmental information of the target object, and the environmental processing module is used to process the environmental acquisition of the target object to obtain environmental model information; Family relationship information: The family relationship collection module is used to collect and process the family relationship information of the target object; Weather information: The weather collection and processing module is used to collect and process the weather information of the target object.

[0015] Further, the emotions include happiness, anger, sadness and fear; The environmental information includes a sense of daily life, a sense of festival celebrations, a sense of family warmth, a sense of community activities, a sense of natural scenery, a sense of growth records, and a sense of outdoor camping; The family relationship information includes three-generation relationships, two-generation relationships, single-parent family relationships, adult living alone relationships, lonely-nest elderly relationships, and orphan relationships; The weather information includes sunny days, rainy days, cloudy days and snowy days; The weather collection and processing module obtains local weather information through the Internet of Things.

[0016] Compared with the prior art, the present invention has the following advantages: the video generation method based on artificial intelligence first collects the target information modules of family members before performing target information evaluation, determines family relationships, and determines the temperature evaluation information, emotional model information, environmental model information and weather information of each family member, thereby avoiding the situation that a video that does not take family relationships into consideration cannot effectively express the overall state that the family members want to express at this time; at the same time, by performing an overall evaluation on each temperature evaluation information, emotional model information, environmental model information and weather information, the state of the family members can be understood, and the color of the corresponding video and the actual display effect of the video can be linked, thereby increasing the use effect that the video is more in line with what the person being photographed wants to express.

[0017] The target information evaluation module sequentially evaluates family relationship information, environmental model information, emotional model information, body temperature evaluation information and weather information. By reasonably arranging the target information modules, the target information in sequence can make the information of the target information evaluation module confirm the first stage of video generation. By optimizing the lens sensor control library in the first stage of video generation, a video that is more in line with the current family relationship can be generated, avoiding inappropriate video generation, affecting the emotions of the video object, and thus bringing a poor experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the video generation method based on artificial intelligence of the present invention.

[0019] Figure 2 It is a schematic diagram of the lens sensor control library of the present invention. DETAILED DESCRIPTION

[0020] The following is a detailed description of an embodiment of the present invention. This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiment.

[0021] like Figure 1 As shown, a schematic diagram of a video generation method based on artificial intelligence includes obtaining target information, calling a preset lens sensor control library, determining a target designated lens based on the target information and the lens sensor control library, and generating a target video based on the target designated lens.

[0022] The target information module includes body temperature assessment information, emotion model information, environment model information, family relationship information and weather information; Target information evaluation module: used to evaluate the target information module; Body temperature assessment information: The body temperature monitoring module is used to monitor the body temperature information of the target object, and the body temperature analysis module is used to analyze the target body temperature information to obtain body temperature assessment information; Emotion model information: the emotion acquisition module is used to acquire the target's emotions, and the emotion processing module is used to process the acquired target's emotion information to obtain the emotion model information; Environmental model information: The environmental acquisition module is used to acquire environmental information of the target object, and the environmental processing module is used to process the environmental acquisition of the target object to obtain environmental model information; Family relationship information: The family relationship collection module is used to collect and process the family relationship information of the target object; Weather information: The weather collection and processing module is used to collect and process the weather information of the target object; like Figure 2 As shown, Figure 2 The lens sensor control library includes rightward movement, rise, fall, leftward movement, strong vibration, distance and approach. The rightward movement corresponds to the family portrait feeling in the lens sensor control library, the rise corresponds to the macro display in the lens sensor control library, the fall corresponds to the detail display in the lens sensor control library, the leftward movement corresponds to the documentary feeling in the lens sensor control library, the strong vibration corresponds to the fast switching lens in the lens sensor control library, the distance corresponds to the gradual distance in the lens sensor control library, and the approach corresponds to the gradual distance in the lens sensor control library. The lens sensor control library includes rightward movement (family portrait feeling), rise (macro display), fall (detail display), leftward movement (documentary), strong vibration (fast switching lens), distance (gradual distance) and approach (gradual distance).

[0023] In some specific implementations, the emotions include happiness, anger, indifference, sadness, and fear; the system also includes custom settings, including recording duration, video resolution, and speech-to-text conversion.

[0024] In some specific implementations, the environmental information includes a sense of daily life, a sense of holiday celebrations, a sense of family warmth, a sense of community activities, a sense of natural scenery, a sense of growth records, and a sense of outdoor camping.

[0025] In some specific implementations, the family relationship information includes three-generation living together relationship, two-generation relationship, single-parent family relationship, adult living alone relationship, lonely elderly relationship and orphan relationship.

[0026] With respect to the acquisition of family relationship information, it can be manually input by the user before the video is shot, or similarity can be judged based on the environmental information collected before the video is shot. Since the present invention is mainly aimed at family users, they are all family members. In some specific implementations, the user can manually revise. Preferably, the target information evaluation module sequentially evaluates family relationship information, environmental model information, emotional model information, body temperature evaluation information, and weather information.

[0027] For example, if a video captures an old person, a middle-aged person and a child, based on the similarity matching, it is a three-generation family relationship; if a video captures an old person, the preferred video is automatically judged as a lonely old person relationship. If the old person has manually selected the lonely old person relationship in the last video shooting, the family relationship information can also be directly determined as family relationship information in this time.

[0028] In some specific implementations, the weather information includes sunny days, rainy days, cloudy days and snowy days.

[0029] In some specific implementations, the weather collection and processing module obtains local weather information through the Internet of Things, and the weather information can be displayed in real time on the video, or the display can be turned off manually. Preferably, it is displayed in real time to help the photographer recall later and better record the usage scenario at the time.

[0030] A video generation method based on artificial intelligence comprises the following steps: S1: Acquire target information, wherein the target information includes body temperature assessment information, emotion model information, environment model information, weather information, and family relationship information; evaluate the target information, and obtain a target information evaluation module; First, the body temperature assessment information can be manually input by the photographer for each target object, or the video settings can be directly marked with the corresponding body temperature acquisition module. The body temperature assessment information helps to understand whether the subject is in a stable state at this time. For example, in some cases of high body temperature such as fever, if the weather information is normal at 25 degrees Celsius, the body temperature information can help to determine that the subject may be sick at this time. At this time, it is not appropriate to increase the fill light to reflect the glossiness of the subject's face, and only normal display is required.

[0031] The emotional model information can be compared with the objects in the video to identify the emotional information of each object. Emotions are mainly divided into happiness, anger, sadness and fear. The objects in the video can be identified by pre-establishing a virtual scene. In this scene, the emotions of the characters are divided into four categories: happiness, anger, sadness and fear. By comparing the characters in the video, the emotional state of the object can be distinguished. By judging the proportion of the four categories of happiness, anger, sadness and fear among the number of people in the object, the emotion of one of the four categories with the highest proportion can be compared with the emotional model information to set the emotion of the object in the video to that emotion. The emotional model information can also be obtained by using a convolutional neural network (CNN).

[0032] The target information evaluation module evaluates family relationship information, environmental model information, emotional model information, body temperature evaluation information and weather information in turn. The weather information can be manually input by the user, or directly obtained online through the Internet of Things, or directly obtained through video based on the user's current weather scene. In some embodiments, the information can be directly obtained through manual input by the user or automatic acquisition by video, preferably automatic acquisition. After obtaining the above information, a corresponding target information evaluation module is generated; the target information evaluation module gives the tone of the video at this time by judging the family relationship information, environmental model information, emotional model information, body temperature assessment information and weather information. For example, by judging that three generations live together, the emotion is happy, the weather is sunny, and the environmental information is a sense of growth record, then the video target information evaluation module defines the current information as a family portrait recording ritual sense. It is necessary to make the video black and white documentary and the video colorful and vivid, highlight the characters, and weaken the environment as the main input points, and the weather is sunny. The light input can be appropriately adjusted, and AI repair of automatically opening the eyes can be added to avoid multiple shots of a video and increase the convenience of shooting. Otherwise, because of the large number of people in this scene, multiple shots may be taken repeatedly or everyone’s state is not in place during one shot, resulting in poor video quality and failure to meet good expectations.

[0033] Family relationship information includes three-generation relationships, two-generation relationships, single-parent family relationships, adult living alone relationships, lonely elderly relationships, and orphan relationships; the corresponding family relationship information is determined by the video area. Specifically, through the video character relationship based on association rules, through video analysis technology, especially character relationship mining technology, the characters in the video and their relationships can be identified. This application uses face recognition technology to identify individuals in the video, and a method for generating a social relationship graph of video characters based on a graph convolutional network, and infers the relationship between the characters through interactions and conversations. The specific steps are frame-level sub-graph generation, multi-channel time series accumulation module, fragment-level graph convolutional network module, model training and reasoning.

[0034] For example, by judging that the relationship of lonely elderly people is obtained, the mood is flat, the weather is cloudy, the body temperature is normal (about 37°C), and the environmental information is a sense of daily life, the video target information evaluation module defines the current information as the relationship of lonely elderly people. It is necessary to use black and white documentary videos and colorful and vivid videos to highlight the characters and weaken the environment as the main input points. If the weather is cloudy, the light input can be appropriately increased to highlight the facial features of the characters and reduce the bright features of the main characters. The focus is on documentary and the input of bright colors can be reduced to avoid causing emotional fluctuations in the elderly and causing some emotional thoughts, etc., so as to make it more in line with the current usage needs of the elderly, record his current state, facilitate subsequent nostalgia, and minimize emotional fluctuations.

[0035] The target information evaluation module sequentially evaluates family relationship information, environmental model information, emotional model information, body temperature evaluation information and weather information. By reasonably arranging the target information modules, the target information in sequence can make the information of the target information evaluation module confirm the first stage of video generation. By optimizing the lens sensor control library in the first stage of video generation, a video that is more in line with the current family relationship can be generated, avoiding inappropriate video generation, affecting the emotions of the objects in the video, and thus bringing a poor experience.

[0036] Family relationship information includes three-generation relationships, two-generation relationships, single-parent family relationships, adult single-family relationships, lonely elderly relationships, and orphan relationships; the corresponding family relationship information is determined by the video area.

[0037] The environmental model information mainly includes the sense of daily life, the sense of festival celebration, the sense of family warmth, the sense of community activities, the sense of natural scenery, the sense of growth record and the sense of outdoor camping. The environmental model information can be manually input by the user or automatically identified by the video device, but the automatic identification by the video device is preferred, which can increase the user experience and improve the convenience of taking pictures. By judging the objects in the video, for example, the furnishings in the video are living room furniture (sofa, coffee table, TV cabinet, TV, dining table, dining chair, etc.), bedroom furniture (bed, wardrobe, bedside table, etc.), kitchen furniture (cabinet, range hood, stove, etc.), bathroom furniture (washbasin, toilet, bathroom screen), through the above environment, it can be judged that the object of the video at this time is the corresponding sense of daily life. For example, by judging that the above furniture appears in the video object, taking kitchen furniture as an example, when the cabinet appears in the video, the first similarity of the video being the daily life in the kitchen is 0.7, if the range hood appears at the same time, the second similarity of the video being the daily life in the kitchen is 0.8, if the stove appears at the same time, the third similarity of the video being the daily life in the kitchen is 0.9, and the daily life scene is located by collecting environmental information.

[0038] S2: Call the preset lens sensor control library, which includes shooting techniques and lens states, and the lens states correspond to the shooting techniques; specifically, it is controlled through MotionCtrl, which is a model that can independently control the camera motion and object motion in the video. It designs two main modules: CMCM and OMCM. CMCM is a lightweight module composed of several fully connected layers. Its main task is to process and fuse the camera posture sequence (RT) to mobilize the lens sensor control library.

[0039] The lens sensor control library includes moving to the right, rising, falling, moving to the left, strong vibration, moving away and approaching, the moving to the right corresponds to the family portrait feeling in the lens sensor control library, the rising corresponds to the macro display in the lens sensor control library, the falling corresponds to the detail display in the lens sensor control library, the moving to the left corresponds to the documentary in the lens sensor control library, the strong vibration corresponds to the fast switching lens in the lens sensor control library, the moving away corresponds to the gradual zooming out in the lens sensor control library, and the approaching corresponds to the gradual zooming in in the lens sensor control library.

[0040] S3: Based on the target information and the lens sensor control library, the target designated lens is determined, and the target information and the lens sensor control library are linked; a deep learning model: a multi-layer perceptron (MLP), a bidirectional LSTM or a Transformer network can be used to fuse the mapped multimodal features. The model can learn the complex interactive relationship between the video target information and the lens sensor control library, and automatically optimize the feature fusion strategy through end-to-end training to achieve the determination of the target designated lens. Through the joint training method, the model uses the target information and the information of the lens sensor control library to optimize the fusion strategy of the multimodal features when determining the target designated lens. This includes setting a loss function, such as a cross entropy loss plus a consistency loss, to ensure the consistency between the target designated lens and the actual state. After fusion, an additional calibration module is used to perform a final correction on each video clip according to the target designated lens, combined with the contrast between the target information and the lens sensor control library. This step ensures the matching degree of the target video in step S4.

[0041] S4: Generate a target video based on the target designated shot; In some specific embodiments, the body temperature assessment information: the body temperature monitoring module is used to monitor the body temperature information of the target object, and the body temperature analysis module is used to analyze the target body temperature information to obtain the body temperature assessment information; Emotion model information: the emotion acquisition module is used to acquire the target's emotions, and the emotion processing module is used to process the acquired target's emotion information to obtain the emotion model information; Environmental model information: The environmental acquisition module is used to acquire environmental information of the target object, and the environmental processing module is used to process the environmental acquisition of the target object to obtain environmental model information; Family relationship information: The family relationship collection module is used to collect and process the family relationship information of the target object; Weather information: The weather collection and processing module is used to collect and process the weather information of the target object.

[0042] In some specific embodiments, the emotions include happiness, anger, sadness and fear; the environmental information includes a sense of daily life, a sense of holiday celebrations, a sense of family warmth, a sense of community activities, a sense of natural scenery, a sense of growth records and a sense of outdoor camping; the family relationship information includes three-generation relationships, two-generation relationships, single-parent family relationships, adult living alone relationships, lonely elderly people relationships and orphan relationships; the weather information includes sunny days, rainy days, cloudy days and snowy days; the weather collection and processing module obtains local weather information through the Internet of Things.

[0043] The three-generation relationship and the two-generation relationship correspond to rightward movement, upward movement, downward movement, leftward movement, strong vibration, distance and approach in the lens sensor control library; The single-parent family relationship, the adult living alone relationship, the lonely elderly relationship and the orphan relationship correspond to rising, falling, moving to the left, strong vibration, moving away and approaching in the lens sensor control library.

[0044] For example, by judging that the relationship of the elderly living alone is obtained, the mood is flat, the weather is cloudy, the environmental information is a sense of daily life, and the body temperature is normal (about 37°C), the video target information evaluation module defines the current information as the relationship of the elderly living alone. The video needs to be a black-and-white documentary and a colorful and vivid video, highlighting the characters and weakening the environment as the main input points. If the weather is cloudy, the light input can be appropriately increased to highlight the facial features of the characters and reduce the bright features of the main characters. The focus is on documentary and the input of bright colors can be reduced to avoid causing emotional fluctuations in the elderly and causing some emotional thoughts, etc., so as to make it more in line with the current usage needs of the elderly, record his current state, facilitate subsequent nostalgia, and minimize emotional fluctuations. The relationships of single-parent families, adults living alone, lonely elderly people and orphans correspond to rising, falling, moving to the left, strong vibration, moving away and approaching in the lens sensor control library; at this time, the video shooting can choose rising, falling, moving to the left, strong vibration, moving away and approaching in the lens sensor control library, so as to perform secondary processing of the above information in combination with the lens sensor control library, such as selecting moving away, which corresponds to gradually zooming out in the lens sensor control library, so that the task can be highlighted first, and then the surrounding environment of the character can be displayed after zooming out, forming a secondary processing of the video after the description of the character, thereby highlighting the use scenario of the character to the environment, which is more convenient for the use demand of the relationship between lonely elderly people at this time. Through this video, they can first convey their own safety and status to others, and then convey the living environment, which greatly facilitates the use scenario of lonely elderly people, highlights their need to record, and can also It is convenient for users to find the video later. They can quickly understand the information corresponding to the video at that time. The effective record makes the video record more ceremonial. The input of weather and other information can also be displayed in real time in the video. It can also be recorded in some records, so that users can recall the usage scene at that time. It is more convenient and more suitable for current users. In addition, the family portrait relationship in the lens sensor control library is not set for the relationship of lonely elderly people, which reduces the harm or sentimentality to the elderly. The body temperature assessment information helps to understand the emotional fluctuations of the subject at this time. If the normal body temperature is 37℃, it is normal. If the body temperature of the subject is 39℃ at this time, it is necessary to balance the color of the subject's face in the video. It can usually be displayed as a regular facial chromaticity to avoid the blush caused by high fever, so as to better display the video, improve the recording purpose of the video, and increase the use effect of the video. Based on the target specified lens, the target video is generated. (The rising corresponds to the macro display in the lens sensor control library; others include, the falling corresponds to the detail display in the lens sensor control library, the moving to the left corresponds to the documentary in the lens sensor control library, the strong vibration corresponds to the fast switching lens in the lens sensor control library, the moving away corresponds to the gradual zooming out in the lens sensor control library, and the approaching corresponds to the gradual zooming in in the lens sensor control library).

[0045] The video generation method based on artificial intelligence first collects the target information modules of family members before conducting target information evaluation, determines family relationships, and judges the temperature evaluation information, emotion model information, environment model information and weather information of each family member, so as to avoid the situation that a video that does not take family relationships into consideration cannot effectively express the overall state that the family members want to express at this time. At the same time, by conducting an overall evaluation on each temperature evaluation information, emotion model information, environment model information and weather information, the state of the family members can be understood, and the color of the corresponding video and the actual display effect of the video can be linked, thereby increasing the use effect that the video is more in line with what the person being photographed wants to express.

[0046] The target information evaluation module sequentially evaluates family relationship information, environmental model information, emotional model information, body temperature evaluation information and weather information. By reasonably arranging the target information modules, the target information in sequence can make the information of the target information evaluation module confirm the first stage of video generation. By optimizing the lens sensor control library in the first stage of video generation, a video that is more in line with the current family relationship can be generated, avoiding inappropriate video generation, affecting the emotions of the video object, and thus bringing a poor experience.

[0047] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0048] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the above integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium, which can be a personal computer, a server, or a network device, etc. The computer-readable storage medium executes some steps of the method described in each embodiment of the present invention. The storage medium includes: a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a volatile memory or a non-volatile memory. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM) or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM) and direct RAM bus RAM (DR RAM). Various media that can store program code are available.

[0049] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0050] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0051] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A video generation method based on artificial intelligence, characterized in that: The steps include: S1: Acquire target information, the target information including body temperature assessment information, emotion model information, environment model information, weather information and family relationship information; evaluate the target information to obtain a target information evaluation module; S2: calling a preset lens sensor control library, wherein the lens sensor control library includes shooting techniques and lens states, and the lens states correspond to the shooting techniques; S3: Determine a target designated lens based on the target information and the lens sensor control library; S4: generating a target video based on the target designated shot; The lens sensor control library includes moving to the right, rising, falling, moving to the left, strong vibration, moving away and approaching, the moving to the right corresponds to the family portrait feeling in the lens sensor control library, the rising corresponds to the macro display in the lens sensor control library, the falling corresponds to the detail display in the lens sensor control library, the moving to the left corresponds to the documentary in the lens sensor control library, the strong vibration corresponds to the fast switching lens in the lens sensor control library, the moving away corresponds to the gradual zooming out in the lens sensor control library, and the approaching corresponds to the gradual zooming in in the lens sensor control library.

2. The method for generating video based on artificial intelligence according to claim 1, characterized in that: Body temperature assessment information: The body temperature monitoring module is used to monitor the body temperature information of the target object, and the body temperature analysis module is used to analyze the target body temperature information to obtain body temperature assessment information; Emotion model information: the emotion acquisition module is used to acquire the target's emotions, and the emotion processing module is used to process the acquired target's emotion information to obtain the emotion model information; Environmental model information: The environmental acquisition module is used to acquire environmental information of the target object, and the environmental processing module is used to process the environmental acquisition of the target object to obtain environmental model information; Family relationship information: The family relationship collection module is used to collect and process the family relationship information of the target object; Weather information: The weather collection and processing module is used to collect and process the weather information of the target object.

3. The video generation method based on artificial intelligence according to claim 2, characterized in that: The environmental information includes a sense of daily life, a sense of festival celebrations, a sense of family warmth, a sense of community activities, a sense of natural scenery, a sense of growth records, and a sense of outdoor camping; In step S3, the target designated shot is determined by a deep learning model.

4. The video generation method based on artificial intelligence according to claim 2, characterized in that: The target information evaluation module sequentially evaluates family relationship information, environmental model information, emotional model information, body temperature evaluation information and weather information.

5. The video generation method based on artificial intelligence according to claim 1, characterized in that: The generation method also includes custom settings, which include recording duration, video resolution, and speech-to-text conversion.

6. The video generation method based on artificial intelligence according to claim 1, characterized in that: The weather collection and processing module obtains local weather information through the Internet of Things.

7. The video generation method based on artificial intelligence according to claim 1, characterized in that: The weather information includes sunny days, rainy days, cloudy days and snowy days.

8. The method for generating video based on artificial intelligence according to claim 1, characterized in that: The family relationship information includes three-generation living together relationship, two-generation relationship, single-parent family relationship, adult living alone relationship, lonely elderly relationship and orphan relationship.

9. The video generation method based on artificial intelligence according to claim 1, characterized in that: The emotions include happiness, anger, sadness and fear.

10. A video generation system based on artificial intelligence, characterized in that: Applicable to the video generation method as described in any one of claims 1 to 9.

Images