Video generation method and device, equipment, medium and program product

Abstract

The invention discloses a video generation method, device and equipment, a medium and a program product, and is applied to a server, and the method comprises the steps that the server receives a plurality of sub-monitoring videos, sent by camera equipment, of an object of a target type in a preset historical period and video tag information of the sub-monitoring videos, the video tag information comprises content value index data; according to the content value index data, determining the video content value of the sub-monitoring video; determining at least one target sub-monitoring video in the plurality of sub-monitoring videos according to the video content value; and splicing the at least one target sub-monitoring video to obtain a beautiful time video, so that the terminal equipment can play the beautiful time video in a beautiful time page of the monitoring application. According to the embodiment of the invention, the valuable content can be quickly and accurately obtained from mass data, and the personalized emotion recording requirement of the user is met.

Description

Technical Field

[0001] This application belongs to the field of video surveillance technology, and in particular relates to a video generation method, apparatus, device, medium and program product. Background Technology

[0002] With the widespread adoption of surveillance technology, surveillance equipment has been extensively used in homes, businesses, public areas, and other sectors, generating massive amounts of surveillance video data. This video data records activity information in various scenarios, providing users with valuable information for review and viewing.

[0003] However, faced with massive amounts of surveillance video data, existing surveillance video processing methods mostly rely on manual screening. Taking home surveillance as an example, if a user wants to view highlights of the activities of the elderly, children, or pets in the past day, they need to replay the surveillance videos of the past 24 hours one by one. This is time-consuming and laborious, greatly affecting the efficiency and accuracy of users in quickly obtaining valuable content from massive amounts of data, and failing to meet users' personalized emotional recording needs. Summary of the Invention

[0004] This application provides a video generation method, apparatus, device, medium, and program product that can quickly and accurately extract valuable content from massive amounts of data, meeting users' personalized emotional recording needs.

[0005] In a first aspect, embodiments of this application provide a video generation method applied to a server, the method comprising: Receive multiple sub-monitoring videos of a target type object within a preset historical time period, as well as video tag information for the sub-monitoring videos, sent by the camera device. The video tag information includes content value index data. The video content value of the sub-monitoring videos is determined based on the content value index data. Based on the value of the video content, at least one target sub-monitoring video is identified from multiple sub-monitoring videos; At least one target sub-monitoring video is stitched together to obtain a "Beautiful Moments" video, which is then played on the "Beautiful Moments" page of the monitoring application by the terminal device.

[0006] Based on the same inventive concept, in a second aspect, embodiments of this application also provide a video generation method applied to a terminal device, the terminal device having a monitoring application installed, the method comprising: The "Good Times" page of the monitoring application is displayed. The type filtering area on the "Good Times" page includes information of multiple types. Receive type selection instructions input in the type filtering area. The type selection instructions include information about the target type from multiple types. The system sends target type information to the server so that the server can receive multiple sub-monitoring videos of the target type object within a preset historical time period sent by the camera device, as well as video tag information of the sub-monitoring videos. The video tag information includes content value index data. Based on the content value index data, the video content value of the sub-monitoring videos is determined. Based on the video content value, at least one target sub-monitoring video is determined from the multiple sub-monitoring videos. The at least one target sub-monitoring video is stitched together to obtain the "Beautiful Moments" video. Receive the "Beautiful Moments" video sent by the server and play the "Beautiful Moments" video on the "Beautiful Moments" page of the monitoring application.

[0007] In one possible embodiment of the second aspect, receiving a "Good Times" video sent by a server and playing the "Good Times" video on the "Good Times" playback page of a monitoring application includes: Receive videos of wonderful moments sent by the server; The candidate display area on the "Good Times" playback page displays brief information about the "Good Times" video using a preset visual style; Upon receiving the playback command corresponding to the brief information, the "Beautiful Time" video will be played in the video playback area of ​​the "Beautiful Time" playback page.

[0008] Based on the same inventive concept, in a third aspect, embodiments of this application also provide a video generation method applied to a camera device, the method comprising: Obtain information about the target type; Within a preset historical period, sub-monitoring images are acquired at preset time intervals; When the sub-monitoring image contains objects of the target type, image tag information for the sub-monitoring image is generated based on the target type, object characteristics in the sub-monitoring image, image quality of the sub-monitoring image, and image acquisition time of the sub-monitoring image. Based on multiple sub-monitoring images, generate multiple sub-monitoring videos of preset duration; Based on the image tag information of each sub-monitoring image in the sub-monitoring video, generate video tag information for the sub-monitoring video; The system sends multiple sub-monitoring videos of a target type within a preset historical time period, along with video tag information for each sub-monitoring video, to the server. The server receives these sub-monitoring videos and their video tags from the camera device. The video tag information includes content value index data. Based on the content value index data, the system determines the video content value of each sub-monitoring video. Based on the video content value, at least one target sub-monitoring video is identified from among the multiple sub-monitoring videos. The system then stitches together the at least one target sub-monitoring video to obtain a "Beautiful Moments" video, which is played on the "Beautiful Moments" page of the monitoring application by the terminal device.

[0009] Based on the same inventive concept, in a fourth aspect, embodiments of this application also provide a video generation apparatus applied to a server, the apparatus comprising: The receiving module is used to receive multiple sub-monitoring videos of a target type object within a preset historical time period sent by the camera device, as well as the video tag information of the sub-monitoring videos. The video tag information includes content value index data. The determination module is used to determine the video content value of sub-monitoring videos based on content value index data; The determination module is also used to identify at least one target sub-monitoring video among multiple sub-monitoring videos based on the value of the video content. The splicing module is used to splice at least one target sub-monitoring video to obtain a "Beautiful Moments" video, which can be played on the "Beautiful Moments" page of the monitoring application by the terminal device.

[0010] Based on the same inventive concept, in a fifth aspect, embodiments of this application also provide a video generation apparatus, applied to a terminal device, the terminal device having a monitoring application installed, the apparatus comprising: The display module is used to display the "Good Times" page of the monitoring application. The type filtering area on the "Good Times" page includes information of multiple types. The receiving module is used to receive the type selection instruction input in the type filtering area. The type selection instruction includes information about the target type among multiple types. The sending module is used to send target type information to the server, so that the server can receive multiple sub-monitoring videos of the target type object within a preset historical time period sent by the camera device, as well as the video tag information of the sub-monitoring videos. The video tag information includes content value index data. Based on the content value index data, the video content value of the sub-monitoring videos is determined. Based on the video content value, at least one target sub-monitoring video is determined from the multiple sub-monitoring videos. The at least one target sub-monitoring video is stitched together to obtain the "Beautiful Moments" video. The receiving module is also used to receive the "Beautiful Moments" video sent by the server and play the "Beautiful Moments" video on the "Beautiful Moments" page of the monitoring application.

[0011] Based on the same inventive concept, in a sixth aspect, embodiments of this application also provide a video generation apparatus, applied to a camera device, the apparatus comprising: The acquisition module is used to acquire information about the target type. The acquisition module is also used to acquire sub-monitoring images at preset time intervals within a preset historical period. The generation module is used to generate image tag information for sub-monitoring images when the sub-monitoring image contains objects of the target type, based on the target type, object features in the sub-monitoring image, image quality of the sub-monitoring image, and image acquisition time of the sub-monitoring image. The generation module is also used to generate multiple sub-monitoring videos of preset durations based on multiple sub-monitoring images; The generation module is also used to generate video tag information for the sub-monitoring video based on the image tag information of each sub-monitoring image in the sub-monitoring video; The sending module is used to send multiple sub-monitoring videos of a target type object within a preset historical time period, along with video tag information for these sub-monitoring videos, to the server. The server receives these sub-monitoring videos and their video tag information from the camera device. The video tag information includes content value index data. Based on the content value index data, the module determines the video content value of each sub-monitoring video. Based on the video content value, it identifies at least one target sub-monitoring video from among the multiple sub-monitoring videos. The module then stitches together the at least one target sub-monitoring video to obtain a "Beautiful Moments" video, which is then played on the "Beautiful Moments" page of the monitoring application by the terminal device.

[0012] Based on the same inventive concept, in a seventh aspect, embodiments of this application also provide a video generation device, the device including a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the video generation method of any embodiment of the first aspect, the video generation method of any embodiment of the second aspect, or the video generation method of any embodiment of the third aspect.

[0013] Based on the same inventive concept, in an eighth aspect, embodiments of this application also provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the video generation method of any embodiment of the first aspect, the video generation method of any embodiment of the second aspect, or the video generation method of any embodiment of the third aspect.

[0014] Based on the same inventive concept, in a ninth aspect, embodiments of this application also provide a computer program product, wherein instructions in the computer program product, when executed by a processor of a device, enable the device to perform the video generation method of any embodiment of the first aspect, the video generation method of any embodiment of the second aspect, or the video generation method of any embodiment of the third aspect.

[0015] This application discloses a video generation method, apparatus, device, medium, and program product. The method is applied to a server. The server can receive multiple sub-monitoring videos of a specific object (a target type object) within a preset historical time period, sent by a camera device, along with video tag information (including content value index data) of the sub-monitoring videos. Then, based on the content value index data, the value of the video content of the sub-monitoring videos is determined. Next, based on the value of the video content of each sub-monitoring video, at least one target sub-monitoring video with higher video content is selected from the multiple sub-monitoring videos. This at least one target sub-monitoring video is then stitched together to form a "Beautiful Moments" video, thereby automatically generating a "Beautiful Moments" video of the specific object within the preset historical time period. The user can then play this "Beautiful Moments" video on the "Beautiful Moments" page of the monitoring application on the terminal device. This application embodiment can accurately and quickly filter out valuable videos of specific objects (such as family members or pets) that users are interested in from a large amount of monitoring data and stitch them together to form a "Beautiful Moments" video. Users can simply watch wonderful "Beautiful Moments" videos with one click through the "Beautiful Moments" page of the monitoring application on their terminal device. These videos include the kind smiles of the elderly, the happy play of children, and the cute actions of pets. Users do not need to manually filter the videos, which saves time and effort and fully meets the personalized emotional recording needs of users. Attached Figure Description

[0016] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings, in which the same or similar reference numerals denote the same or similar features, and the drawings are not drawn to scale.

[0017] Figure 1 This is a schematic flowchart of a video generation method provided in an embodiment of this application; Figure 2 This is a schematic diagram of a "Good Times" page in the video generation method provided in this application embodiment; Figure 3 This is a schematic diagram of the structure of the video acquisition module of the camera device in the video generation method provided in this application embodiment; Figure 4 This is another flowchart illustrating the video generation method provided in the embodiments of this application; Figure 5 This is another flowchart illustrating the video generation method provided in the embodiments of this application; Figure 6 This is another flowchart illustrating the video generation method provided in the embodiments of this application; Figure 7 This is another flowchart illustrating the video generation method provided in the embodiments of this application; Figure 8 This is a schematic diagram of the server structure in the video generation method provided in this application embodiment; Figure 9 This is a schematic diagram of the camera device in the video generation method provided in this application embodiment; Figure 10 This is a schematic diagram of a video generation apparatus provided in an embodiment of this application; Figure 11 This is a schematic diagram of a video generation apparatus provided in an embodiment of this application; Figure 12 This is a schematic diagram of a video generation apparatus provided in an embodiment of this application; Figure 13 This is a schematic diagram of the structure of a video generation device provided in an embodiment of this application. Detailed Implementation

[0018] The features and exemplary embodiments of various aspects of this application will now be described in detail. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only configured to explain this application and are not configured to limit this application. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples of this application.

[0019] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0020] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0021] Various modifications and variations can be made to this application without departing from its spirit or scope, which will be apparent to those skilled in the art. Therefore, this application is intended to cover modifications and variations falling within the scope of the corresponding claims (the claimed technical solutions) and their equivalents. It should be noted that the implementation methods provided in the embodiments of this application can be combined with each other without contradiction.

[0022] Before describing the technical solutions provided in the embodiments of this application, in order to facilitate understanding of the embodiments of this application, this application first specifically explains the problems existing in the related technologies: With the development of smart home and Internet of Things technologies, my country's security industry market continues to expand. More and more families are adopting video surveillance systems to meet multiple needs such as security and family companionship. Related technologies have launched special applications such as "Baby Time" and "Pet Time" to record precious family moments and meet various personalized needs of users.

[0023] However, faced with massive amounts of surveillance video data, existing surveillance video processing methods mostly rely on manual screening. Taking home surveillance as an example, if a user wants to view highlights of the activities of the elderly, children, or pets in the past day, they need to replay the surveillance videos of the past 24 hours one by one. This is time-consuming and laborious, greatly affecting the efficiency and accuracy of users in quickly obtaining valuable content from massive amounts of data, and failing to meet users' personalized emotional recording needs.

[0024] In addition, some related technologies use a solution of ordinary home smart cameras + cloud-based frame extraction AI analysis + video editing to generate precious moments videos. Home smart cameras generally use a monocular wide-angle fixed-focus lens (such as a 4mm focal length, with a diagonal field of view of about 105°) to maximize panoramic coverage, but they are somewhat inadequate in capturing details. In most scenarios, key targets (family members, pets) are too far from the camera, making them small and lacking clarity, which greatly reduces the actual effect of recording precious family moments. Furthermore, the camera itself lacks AI analysis capabilities, and target recognition mainly relies on cloud-based AI frame extraction analysis. Due to the cost limitations of AI analysis, the frame extraction frequency is low, and the content recognition and classification are not accurate and precise enough, which easily leads to false alarms or missed alarms, ultimately affecting the product's user experience.

[0025] In addition, the video editing and generation stage mainly relies on basic video editing and splicing techniques. It lacks AI intelligent assistance in video clip selection and does not incorporate generative large model technology. The content presentation is still mainly based on the original video material, and the creative dimensions of fun need to be expanded, making it difficult to meet the diverse needs of users.

[0026] Based on this, embodiments of this application provide a video generation method, apparatus, device, medium, and program product that can quickly and accurately extract valuable content from massive amounts of data, meeting users' personalized emotional recording needs.

[0027] The technical solution of this application embodiment is described below using a video generation system as an example. The video generation system may include a terminal device, a camera device, and a server.

[0028] The terminal device is equipped with a monitoring application (App). Users can log in to the monitoring application through the terminal device to watch highlights in the monitoring video. For example, the monitoring application can automatically present a video of a happy moment (e.g., 5-10 minutes long) of an elderly person, child, or pet during a 24-hour period.

[0029] First, the terminal device needs to determine what type of memorable moments the user wants to watch, i.e., determine the target type. For example, the user can specify at least one of the following types as the target type in the terminal device's monitoring application: elderly, child, pet, etc.; or, the default type can be used as the target type.

[0030] Then, the camera captures surveillance images of objects of the target type and generates image tag information corresponding to the surveillance images. For example, the image tag information may include the target type of the object appearing in the surveillance image, the characteristics of the object in the surveillance image (if the target type is a face, the object characteristics may include age, expression, gender, etc.; if the target type is a pet, the object characteristics may include species, breed, color, etc.), the image quality of the surveillance image, and the image acquisition time of the surveillance image.

[0031] Next, the camera equipment can extract multiple sub-video segments of preset duration from the surveillance video. That is, a long surveillance video (e.g., 2 minutes long) showing a target object can be divided into multiple shorter surveillance videos, such as each sub-video lasting 10 seconds. For example, if the target object is a pet, and the surveillance video showing a pet is 2 minutes long, it can be split into multiple 10-second segments. Then, based on the image tag information of the surveillance images, video tag information for the surveillance video can be generated.

[0032] Next, the camera equipment can send each surveillance video and video tag information to the server. The server will then select multiple surveillance videos with content value and use artificial intelligence (AI) to stitch them together into a 5-10 minute "Beautiful Moment" video. The server will then send the "Beautiful Moment" video to the terminal device.

[0033] Therefore, users can play videos of the target type of object's "good moments" throughout the day on the "Good Moments" page of the monitoring application on the terminal device, thus achieving the ability to quickly and accurately extract valuable content from massive amounts of data and meeting users' personalized emotional recording needs.

[0034] The video generation method provided in this application embodiment will be described in detail below with reference to the accompanying drawings.

[0035] Figure 1 This is a schematic diagram of a video generation method provided in an embodiment of this application, such as... Figure 1 As shown, the method may include steps S110 to S240.

[0036] The terminal device determines, through steps S110-S130, which type of beautiful moments video the user wants to watch.

[0037] S110, the terminal device displays the "Good Times" page of the monitoring application. The type filtering area on the "Good Times" page includes information of multiple types. The monitoring application is installed on the terminal device.

[0038] Among them, the types in the type filter area on the "Good Times" page can include human body, human face, pet, etc.

[0039] Specifically, step S110 can be understood as installing a monitoring application APP on the terminal device, then opening the "Good Times" page of the monitoring application, and displaying information on multiple types (such as elderly, children, and pets) that the user can select in the type filtering area of ​​the "Good Times" page.

[0040] In one example, such as Figure 2 As shown, the "Good Times" page of the monitoring application app can include a type filtering area, a candidate display area, a video playback area, and a video operation area. The operation process is as follows: 1) Users can select the target type (such as elderly, children, pets, etc.) in the type filtering area.

[0041] Type Filtering Area: This area can provide users with quick type filtering in the form of drop-down menus, tag lists, etc. After the user selects the target type, the candidate area will display only brief information about the videos of the specified target type.

[0042] 2) The camera equipment and server will automatically generate a video of the beautiful moments of the target type and display it in the display candidate area in the form of AI video cards. The AI ​​video cards can include a thumbnail of the beautiful moments video, video text, target type type tag, time and other brief information.

[0043] Display candidate area (or AI video card list area): Displays brief information about the generated wonderful moments videos in the form of cards. The cards include information such as video thumbnails, text, target type and date.

[0044] 3) When the AI ​​video card is clicked, the corresponding video will be automatically loaded and played in the video playback area. When playing the video, the user can perform operations such as pausing and fast forwarding in the video operation area.

[0045] Video playback area: It can retrieve and play videos of beautiful moments of objects of the target type specified by the user from the cloud or the local terminal device. While playing, it displays information such as video text, type tags, and timestamps. It supports pausing the video in the video operation area, and you can also click on the card to enter a secondary page to expand the details.

[0046] Video operation area: Provides a variety of operations such as sharing, downloading, volume adjustment, and speed adjustment, optimizing the user's video operation interaction experience.

[0047] As a crucial entry point for users to remotely access and control camera equipment, the monitoring application (APP) provides functions such as live video streaming, video playback, two-way intercom, and camera control. In this embodiment, the monitoring application (APP) offers a dedicated AI Time Page (Beautiful Moments Page), providing users with convenient content presentation and interactive operations, offering a seamless one-stop experience.

[0048] It should be noted that, Figure 2 The "Beautiful Moments" page design shown is an exemplary presentation. Any design that adopts a similar layout structure (including functional areas and interaction logic) falls within the scope of protection of this application.

[0049] For example, the "Good Times" page can also be set with personalized templates: preset theme templates (warm daily life / energetic pets) and user style learning mechanism (historical behavior training model).

[0050] S120, the terminal device receives the type selection instruction entered in the type filtering area, the type selection instruction includes information about the target type among multiple types.

[0051] Specifically, step S120 can be understood as follows: after seeing the type filter area on the "Good Times" page displaying information on multiple types (such as elderly, children, and pets) that the user can select, one or more of the multiple types as the target type according to their needs.

[0052] Through steps S110 and S120, the terminal device can determine which type of object's happy moments the user wants to watch, that is, determine the target type. For example, the user can specify at least one of the types such as elderly, children, and pets as the target type in the terminal device's monitoring application; or, the default type can be used as the target type.

[0053] S130, the terminal device sends target type information to the server.

[0054] Specifically, the terminal device can send target type information to the server to inform the server which type of video of a beautiful moment the user wants to watch. Simultaneously, the camera device also needs to know the target type information; for example, the terminal device can send the target type information to the camera device, or the server can send the target type information to the camera device.

[0055] The camera device captures surveillance images of objects of a target type in steps S140-S190, and generates image tag information corresponding to the surveillance images. For example, the image tag information may include the target type of the object appearing in the surveillance image, the characteristics of the object (e.g., if the target type is a face, the characteristics may include age, expression, gender, etc.; if the target type is a pet, the characteristics may include species, breed, color, etc.), the image quality of the surveillance image, and the image acquisition time of the surveillance image. Then, multiple surveillance images are stitched together to form a short surveillance video (i.e., a sub-survey video). For example, if the acquisition time interval between surveillance images is 0.2 seconds, and the duration of each short surveillance video is 10 seconds, then 50 surveillance images can be stitched together to form a short surveillance video, and video tag information for that surveillance video can be generated based on the image tag information of the surveillance images.

[0056] For example, such as Figure 3 As shown, the core components of the camera equipment's mechanical structure include a binocular lens and a gimbal base. The binocular lens consists of a wide-angle lens and a telephoto (fixed-focus or zoom) lens, arranged coaxially side-by-side and sharing the gimbal. In addition, it includes basic components such as a night lighting unit and a microphone. Figure 3 The mechanical structure shown is only for illustrating its structural layout. Any design that conforms to this layout falls within the scope of this invention.

[0057] S140, The camera device acquires information about the target type.

[0058] Specifically, the camera device can obtain target type information from the terminal device or from the server.

[0059] S150, the camera device acquires sub-monitoring images at preset time intervals (e.g., 0.2s) within a preset historical period (e.g., 24 hours a day).

[0060] The preset historical time period is the time interval required to generate a "Beautiful Moment" video. For example, a "Beautiful Moment" video of a target type object can be generated every 24 hours, or every 12 hours. The preset historical time period can be set according to user needs, and this embodiment of the application does not limit this.

[0061] The preset time interval is the time interval for capturing images, which can be once every 0.2 seconds. The preset time interval can be set according to the requirements of video smoothness, but this application embodiment does not limit it.

[0062] Specifically, the camera device can acquire monitoring images at preset time intervals (such as 0.2s) within a preset historical period (such as 24 hours a day). These monitoring images can be called sub-monitoring images.

[0063] S160, when the sub-monitoring image contains an object of the target type, the camera device generates image tag information for the sub-monitoring image based on the target type, object characteristics in the sub-monitoring image, image quality of the sub-monitoring image, and image acquisition time of the sub-monitoring image.

[0064] Specifically, if the camera device finds that the similarity between the type of an object in the sub-monitoring image and the target type is greater than a preset similarity threshold, it considers that the sub-monitoring image contains an object of the target type. Then, it can generate image tag information for the sub-monitoring image. The image tag information for the sub-monitoring image may include: the type of object in the sub-monitoring image (target type), the characteristics of the object in the sub-monitoring image (the setting of the characteristic parameters is determined by the target type. For example, if the target type is a face, the characteristic parameters may include age, expression, gender, etc.; if the target type is a pet, the characteristic parameters may include species, breed, color, etc.), the image quality of the sub-monitoring image, and the image acquisition time of the sub-monitoring image (time stamp when the image was acquired).

[0065] For example, the image quality of a sub-monitoring image can be comprehensively evaluated by type similarity (the similarity between the type of object in the sub-monitoring image and the target type), object feature parameters (such as facial expressions), image sharpness, lighting, occlusion, etc.

[0066] S170, the camera device extracts multiple sub-monitoring videos of preset duration (e.g., 10s) from the monitoring video based on multiple sub-monitoring images.

[0067] Specifically, the camera device can determine the surveillance video to be captured (the surveillance video containing the target type) based on whether an object of the target type appears in the sub-monitoring image. Then, it can capture multiple sub-monitoring videos of preset duration (e.g., 10 seconds) from that surveillance video. That is, a long surveillance video (e.g., 2 minutes long) containing the target type object can be captured as multiple short surveillance videos, such as each sub-monitoring video being 10 seconds long.

[0068] S180, the camera device generates video tag information for the sub-monitoring video based on the image tag information of each sub-monitoring image in the sub-monitoring video.

[0069] Specifically, after the camera equipment generates multiple sub-monitoring videos, each sub-monitoring video needs to be labeled. This can be done by generating video label information for each sub-monitoring video based on the image label information of each sub-monitoring image in the sub-monitoring video.

[0070] S190, the camera device sends to the server multiple sub-monitoring videos of the target type within a preset historical time period, as well as the video tag information of the sub-monitoring videos. Specifically, after generating multiple sub-monitoring videos of the target type within a preset historical time period and tagging the videos, the camera device can send the multiple sub-monitoring videos and their video tag information to the server.

[0071] In one example, the process of identifying target types and preprocessing (generating sub-monitoring videos and video tags) on the camera device side is as follows: 1) Recognize target types in the sub-monitoring images captured by the camera equipment. For example, detect target types (such as human bodies, pets, actions, etc.) in real time based on the YOLOv8 lightweight model.

[0072] For example, the division of labor between camera equipment and cloud server AI functions: the camera equipment performs real-time object recognition of target types (such as using the YOLOv8 lightweight model) and image quality enhancement (noise reduction, beautification), while the cloud server is responsible for filtering out videos with high content value (such as sorting by sentiment value) and generating content creation (background music, copywriting).

[0073] 2) Based on the type of the target being identified, further feature analysis is performed. For example, for human faces, further feature analysis (age, expression, gender) is performed, and for pets, feature analysis (species and actions, such as "dog - running") is performed to generate structured object feature parameters.

[0074] 3) Image quality enhancement. For example, using deep learning-based algorithms to perform image enhancement tasks such as portrait beautification, image pixel removal, and optimization of image color saturation, contrast, and tone.

[0075] 4) Data Encapsulation and Upload. The processed video stream can be compressed into TS format, segmented at fixed intervals (e.g., 10 seconds), and sub-monitoring video files containing timestamps can be generated. Video tags (including timestamps, target type, object characteristic parameters, and image quality evaluation) are added to each sub-monitoring video file before encapsulation, packaging, and uploading to the server. For example, video tag information can be transmitted according to a structured data transmission protocol.

[0076] Through steps S200-S230, the server can select multiple sub-monitoring videos with content value from multiple sub-monitoring videos within a preset historical time period and the video tag information of the sub-monitoring videos. Then, it uses artificial intelligence (AI) to stitch the multiple sub-monitoring videos with content value into a beautiful moment video, and then sends the beautiful moment video to the terminal device.

[0077] S200, the server receives multiple sub-monitoring videos of the target type object within a preset historical time period sent by the camera device, as well as the video tag information of the sub-monitoring videos. The video tag information includes content value index data.

[0078] The video tag information includes content value indicator data, such as timestamps, target types, object characteristic parameters, and picture quality evaluation.

[0079] Specifically, the server can receive a large amount of monitoring data: such as multiple sub-monitoring videos (e.g., 10-second videos) of a target type of object (e.g., a pet) sent by the camera device within a preset historical time period (e.g., 24 hours a day), as well as video tag information for the sub-monitoring videos (containing multiple content value index data, such as timestamps, target type, object characteristic parameters, and image quality evaluation). The server needs to generate "Beautiful Moments" videos within the preset historical time period based on this information.

[0080] S210, the server determines the video content value of the sub-monitoring video based on the content value index data.

[0081] Specifically, the server can determine the value of the sub-monitoring video content based on the various content value indicators (such as timestamps, target types, object characteristic parameters, and image quality evaluation) in the video tag information of the sub-monitoring video.

[0082] For example, the video content value of a sub-surveillance video can be determined by a weighted scoring model that integrates target type (face / pet), emotion value (smile duration), and scene richness (number of objects of the target type).

[0083] S220: The server determines at least one target sub-monitoring video among multiple sub-monitoring videos based on the value of the video content.

[0084] Specifically, the server can select the most valuable sub-monitoring video from multiple sub-monitoring videos based on the value of the video content. For example, the determination rule could be that the score is greater than a certain threshold, or the multiple sub-monitoring videos could be sorted according to the value of the video content, and a certain number of sub-monitoring videos could be selected as target sub-monitoring videos. The number of target sub-monitoring videos can be determined by the duration of the beautiful moments video (such as 5 minutes, 10 minutes, etc.).

[0085] S230: The server stitches together the surveillance videos of at least one target sub-monitor to obtain the "Beautiful Moments" video.

[0086] Specifically, the server can use AI to intelligently stitch together multiple target sub-monitoring videos with high content value to obtain videos of beautiful moments within a preset historical time period, such as videos with a duration of 5 to 10 minutes.

[0087] S240: The terminal device receives the "Beautiful Moments" video sent by the server and plays the "Beautiful Moments" video on the "Beautiful Moments" page of the monitoring application.

[0088] Specifically, users can play videos of objects of a target type during preset historical periods on the "Good Times" page of the monitoring application on the terminal device. This enables users to quickly and accurately extract valuable content from massive amounts of data, satisfying their personalized emotional recording needs.

[0089] According to the video generation method provided in this application embodiment, the terminal device first determines, through steps S110-S130, which type (target type) of object's beautiful moments the user wants to watch. Then, the camera device captures monitoring images of the target type object appearing through steps S140-S190, and generates image tag information corresponding to the monitoring images. For example, the image tag information may include the target type of the object appearing in the monitoring image, the characteristics of the object in the monitoring image (e.g., if the target type is a face, the object characteristics may include age, expression, gender, etc.; if the target type is a pet, the object characteristics may include species, breed, color, etc.), the image quality of the monitoring image, and the image acquisition time of the monitoring image. Then, multiple monitoring images are stitched together into a short-duration monitoring video (i.e., a sub-monitoring video). For example, if the acquisition time interval between monitoring images is 0.2s, and the duration of each short-duration monitoring video is 10s, then 50 monitoring images can be stitched together into a short-duration monitoring video, and video tag information for the monitoring video can be generated based on the image tag information of the monitoring images. Next, through steps S200-S230, the server selects multiple sub-monitoring videos with content value from a preset historical time period and their video tag information. Then, using artificial intelligence (AI), these high-value sub-monitoring videos are stitched together into a short video clip, which the server then sends to the terminal device. Therefore, users can play videos of the target type within a preset historical time period on the "Beautiful Moments" page of the monitoring application on their terminal device. This allows for the rapid and accurate extraction of valuable content from massive amounts of data, satisfying users' personalized emotional recording needs.

[0090] In some embodiments, such as Figure 4 As shown, step S240, where the terminal device receives the "Beautiful Moments" video sent by the server and plays it on the "Beautiful Moments" playback page of the monitoring application, may include steps S241-S243: S241, the terminal device receives the "Beautiful Moments" video sent by the server; S242, the terminal device displays brief information about the "Beautiful Moments" video in a preset visual style in the display candidate area of ​​the "Beautiful Moments" playback page; S243, upon receiving a playback instruction corresponding to the brief information, the terminal device plays the "Beautiful Moments" video in the video playback area of ​​the "Beautiful Moments" playback page. The brief information includes at least one of the following: target type information, video text of the "Beautiful Moments" video, and timestamp information.

[0091] Specifically, after the terminal device receives the AI-processed "Beautiful Moments" video (which can be sent directly, or a brief description of the video can be sent first) from the server, it can display a candidate area on the "Beautiful Moments" playback page with a preset visual style (e.g., Figure 2 The AI ​​video card format displays brief information about the "Beautiful Moments" video. After the user clicks the card, the corresponding "Beautiful Moments" video can be played in the video playback area (it can be played locally on the terminal device or retrieved from the cloud). This embodiment combines the AI ​​production process of video with surveillance video, using a server to intelligently process the video to obtain "Beautiful Moments" videos, which are then displayed on the user interface.

[0092] The presentation of data elements (such as multiple types of information) on the "Good Times" page can include lists, dropdowns, etc., and the card layout style can be single column / double column, etc.

[0093] This application embodiment utilizes the layout structure of the "Beautiful Moments" page in the client APP, combining the spatial arrangement of video playback, operation, type filtering area, and card list. The functional areas in the "Beautiful Moments" page layout are clearly defined, providing users with convenient content presentation and operation interaction.

[0094] In some embodiments, such as Figure 5 As shown, step S210, where the server determines the video content value of the sub-monitoring video based on the content value index data, may include steps S211 to S214: S211, The server determines the type value based on the target type; S212, The server determines the emotional value based on the object characteristics in the sub-monitoring video; S213, The server determines the picture quality value based on the video picture quality of the sub-monitoring video; S214, The server determines the video content value of the sub-monitoring video based on the type value, emotional value, picture quality value, and preset weight coefficients.

[0095] Among them, the type value, emotional value, and picture quality value each correspond to different preset weight coefficients, such as emotional value (smile lasting >3 seconds, weight 40%), type value (human face + pet, weight 30%), and picture quality value (clarity, weight 30%).

[0096] Specifically, the server can determine the value of the sub-monitoring video from multiple dimensions, such as type value, emotional value, and picture quality value, based on the content value index data, and then synthesize them to obtain the video content value of the sub-monitoring video.

[0097] For example, the user can specify the priority of the target type and determine the type value based on the priority; or, the type value can be determined by the similarity between the type of the object in the video and the real type, for example, the similarity between the type of the object in the video and the real type is 90%.

[0098] For example, the server can extract multiple sub-monitoring videos and their video tag information. Then, the data is aggregated by time window (e.g., 24 hours), that is, videos of beautiful moments are generated according to a preset historical time period (e.g., 24 hours). Then, based on the content value index data in the video tag information, the values ​​of multiple dimensions such as type value, emotional value, and picture quality value are determined, and the weighted values ​​are used to obtain the video content value of the sub-monitoring video.

[0099] In this embodiment, the server can comprehensively evaluate the video content value of a sub-monitoring video through multiple dimensions, resulting in a more accurate evaluation of the video content value.

[0100] In some embodiments, such as Figure 6 As shown, step S220, where the server determines at least one target sub-monitoring video among multiple sub-monitoring videos based on the value of the video content, may include steps S221 and S222: S221, the server divides the preset historical time period into multiple sub-historical time periods; S222, the server selects at least one target sub-monitoring video from the multiple sub-monitoring videos based on the value of the video content.

[0101] Specifically, a preset historical time period can be divided into several sub-time periods, such as dividing a day into a morning period, an afternoon period, and an evening period. Then, beautiful video clips (sub-monitoring videos) from the morning period, beautiful video clips (sub-monitoring videos) from the afternoon period, and beautiful video clips (sub-monitoring videos) from the evening period can be filtered out and then spliced ​​together to form a beautiful video of the preset historical time period.

[0102] The embodiments of this application can avoid the generated "wonderful moments" videos covering multiple time periods throughout the day, and avoid the time periods of the "wonderful moments" videos being too concentrated, such as only focusing on a specific time period (e.g., only including video clips in the afternoon), thereby more comprehensively reflecting the wonderful moments of the day.

[0103] In some embodiments, such as Figure 7 As shown, the video tag information also includes the video capture time; in step S230, the server stitches together at least one target sub-monitoring video to obtain the "Beautiful Moments" video, including steps S231~S233: S231, the server sorts at least one target sub-monitoring video according to the time sequence based on the video acquisition time; S232, the server adds a smooth transition video with a preset transition duration between two adjacent target sub-monitoring videos and then splices them to obtain the initial beautiful moments video; S233, the server adds background music and video text to the initial beautiful moments video to obtain the beautiful moments video.

[0104] Specifically, the server can sort at least one target sub-monitoring video according to the time of video capture. Before splicing, a smooth transition video with a preset transition duration is added between two adjacent target sub-monitoring videos to achieve a natural and smooth transition, resulting in the initial beautiful moment video. Then, background music and video text are added to the initial beautiful moment video to obtain the beautiful moment video.

[0105] This application embodiment can use AI processing to smoothly stitch together multiple target sub-monitoring videos and add background music and video text, thereby enhancing the viewing experience and appeal of the "Good Times" videos, increasing their emotional value, and improving the user experience.

[0106] It should be noted that when splicing target sub-monitoring videos, the videos can be spliced ​​according to the time dimension, or according to the established content concept logic (such as "panoramic introduction → close-up focus → emotional climax → ending freeze frame").

[0107] In one example, such as Figure 8 As shown, the core functional modules of the server mainly include a business management module, a signaling module, a media module, a video editing module, and an AI processing module. The server can be a cloud server.

[0108] Business Management Module: Responsible for the server's overall network communication, task scheduling, business control and other functions, it is the control center of the server.

[0109] Signaling module: Responsible for communication with the device, including device access, heartbeat, signaling control, device event reporting, device media control, and all other device-related communication functions. The signaling module will also dynamically issue commands based on business needs to control the camera device to execute object recognition algorithms for specified target types.

[0110] Media module: Responsible for live streaming and forwarding of device video streams, uploading and storing videos, device voice intercom and other media-related functions, and responsible for live streaming and playback of device videos.

[0111] Video editing module: Filters videos according to analysis type, quality and other dimensions on a periodic basis, extracts a certain number of high-quality segments (multiple target sub-monitoring videos) within that period, and forwards them to the AI ​​processing module through the business management module to achieve efficient and accurate video content preprocessing.

[0112] AI processing module: Based on a generative video model, it first analyzes the emotional tone and scene characteristics of the video, intelligently matches upbeat background music, and simultaneously combines it with the on-screen content. Through a natural language processing model, it generates positive and uplifting text, achieving a deep integration of audio, visuals, and text, thereby enhancing the video's appeal and virality.

[0113] The background music can be matched with the BPM (beats per minute) from the music library based on the emotional tone of the video (such as "warm" or "cheerful"). For example, 80-100 BPM light music can be selected for a warm scene. Then, the ambient sound is mixed with the background music (e.g., volume ratio: ambient sound 30%, music 70%).

[0114] The copy can be extracted from the visual keywords (such as "family + interaction + pet") based on a large language model (LLM) such as GPT-4, and positive copy within 20 characters (such as "Happiness is always there, wonderful times with family and pets!") can be generated. Then, narration can be generated based on the copy.

[0115] In this embodiment, the server, through the collaborative operation of the business management module, signaling module, media module, video editing module, and AI processing module, can generate higher-quality videos of precious moments, effectively improving the user experience.

[0116] In one example, such as Figure 9 As shown, the core functional modules of the camera equipment may include a control and management module, a video acquisition module, an AI analysis and processing module, and a video processing module.

[0117] As a core component of the camera equipment, the control and management module is responsible for the overall network communication, task scheduling, hardware driving and control of the camera equipment. It is the control center of the camera equipment and the core processing module of the dynamic target intelligent tracking algorithm on the camera equipment side.

[0118] The video acquisition module is responsible for the real-time acquisition, encoding, and compression of audio and video data. Under the scheduling of the control and management module, the video acquisition module synchronously transmits the video data to the AI ​​analysis module for processing.

[0119] AI Analysis and Processing Module: This module has a pre-built matrix of AI algorithm models. Through the collaboration of multiple AI models, it can achieve the requirements of object recognition and video quality optimization. The control and management module can upload structured data (such as timestamps, object type, object feature parameters, and image quality evaluation) of the target recognition results (image label information) to the server. Simultaneously, it can synchronously transmit the video data processed by the AI ​​analysis and processing module to the video processing module for further processing.

[0120] The video processing module compresses, segments, and packages real-time video stream data into video files, generates index information for these files, and then adds an AI index to the video files based on the results of the AI ​​analysis and processing module before uploading them to the server. For example, image tag information can be used as an AI index.

[0121] This application embodiment integrates an AI analysis module (such as object detection of target types and image quality enhancement) on the camera device side, which can process video streams in real time and upload structured tag data (such as timestamps, feature parameters, image quality, etc.), reduce the cloud frame extraction frequency (from 1 frame / second in related technologies to real-time processing, such as shooting once every 0.2 seconds), reduce the server's computing cost, shorten the target recognition latency from the second level to less than 0.2 seconds, and improve the real-time performance of interaction.

[0122] In one example, the camera device integrates a dynamic target intelligent tracking algorithm, achieving the dual goals of recording the entire scene without blind spots and intelligently capturing key target details.

[0123] The camera equipment uses a binocular lens; for example, the main parameters of the wide-angle lens are: focal length 4mm, horizontal angle of view... 90°, vertical viewing angle The field of view is 46°, and the resolution is 2560×1440 pixels. Key parameters of the telephoto lens: focal length 8mm (a zoom lens, such as 6mm-12mm, is preferable; the specific choice depends on the actual usage scenario), horizontal angle of view... 48°, vertical viewing angle The angle is 32° and the resolution is 2560×1440 pixels.

[0124] The dynamic target intelligent tracking algorithm mainly includes four stages: 1) Real-time detection of target objects.

[0125] 1.1) Detection of objects of the target type.

[0126] For example, the algorithm module built into the camera device is based on a deep learning-based object detection algorithm, which operates according to a fixed period. To detect whether a target object exists in a video frame, specifically we can... Set to 0.2 seconds. Since the AI ​​analysis and processing module processes each frame of video in real time, this algorithm execution module only needs to execute according to a fixed period. Simply extract the structured data from the analysis results.

[0127] If the target type object appears in the video, the subsequent steps are executed; otherwise, the camera returns to the preset position (the preset position is the default state of the camera, usually when recording with a wide-angle lens and the gimbal angle is adjusted to the preset setting), and the algorithm module enters a waiting state until the next loop is started.

[0128] In one example, the camera device employs a dual-lens design to enhance image quality and coverage. Specifically, it uses a coaxial, parallel arrangement of a wide-angle and a telephoto lens. The wide-angle lens provides panoramic coverage (90° horizontal field of view), while the telephoto lens captures details (8mm focal length, 2560×1440 resolution). This addresses the limitation of monocular cameras in some technologies where panoramic coverage and detail capture are mutually exclusive. For example, in a home setting, it can capture both a panoramic view of the living room and detailed movements of a baby crib in the distance. The coaxial, parallel mechanical arrangement of the wide-angle (panoramic coverage) and telephoto (detail capture) lenses allows for coordinated switching via a shared gimbal.

[0129] 1.2) Calculate the target position. Once the target object appears in the video frame, calculate its specific coordinates and size.

[0130] Target coordinates and size: Coordinates of the top left corner of the target ,width ,high The coordinates of the target's center point: For example, the image coordinate system has its origin at the top left corner of the image. The positive X-axis direction is to the right, and the positive Y-axis direction is downward; Image dimensions: width W=2560, height H=1440; Image center coordinates: .

[0131] 2) Switch lenses according to the size of the target.

[0132] 2.1) Calculate the resolution of the target object * .

[0133] 2.2) Determine and execute the camera switch. If If so, then the wide-angle lens will be activated. If the telephoto lens is used, then it can be switched to a telephoto lens. If the telephoto lens is a zoom lens, the focal length can be adjusted according to the size of the target to maintain it at a suitable ratio.

[0134] For example, when switching between wide-angle and telephoto lenses (such as adding a 0.5-second fade-in effect during the transition from panorama to close-up), the 180-degree axis rule should be followed to avoid visual abrupt changes.

[0135] For example, the binocular lens switching standard: To ensure the accuracy of AI model recognition, the target object needs to achieve a certain pixel density in the image. Taking a human face as an example, according to the GA / T standard, the interpupillary distance (pupillary distance) should be no less than 60 pixels, and based on industry experience, it should not be less than 40 pixels. For general target objects (taking animals as an example), if species or movement needs to be identified, it is recommended that the resolution be no less than 100×100 pixels, and no less than 50×50 pixels.

[0136] In a home setting, the main target objects are people, faces, and pets. For different targets, we need to set different switching criteria and preset ranges to avoid repeated camera switching caused by a single preset.

[0137] The threshold is the resolution of the target object; when the target size is larger than... When the target size is smaller than a certain value, switch to wide-angle lens; when the target size is smaller than a certain value, switch to wide-angle lens. When needed, switch to wide-angle lens.

[0138] To ensure good recognition accuracy, Set to 1000×1000 Set to 200×200; the specific standard can be adjusted based on practical results.

[0139] For example, a dynamic lens switching logic based on target size (resolution threshold) (e.g., target >1000×1000 pixels switches to wide-angle, <200×200 pixels switches to telephoto).

[0140] 3) Calculate the target offset.

[0141] 3.1) Calculate the offset and the center point of the target object. Image center coordinates Offsets on the X and Y axes; Offset on the X-axis Offset on the Y-axis A negative offset indicates a leftward / upward offset; a positive offset indicates a rightward / downward offset.

[0142] 3.2) Calculate the offset rate.

[0143] Offset on the X-axis Offset on the Y-axis In addition, the system has a preset offset rate threshold. Rat, with a reference value set to 0.3, means that if the offset exceeds 0.3 times half the image width, the gimbal angle needs to be adjusted.

[0144] 3.3) Determine whether adjustments are needed. Then the X-axis needs to be adjusted in angle. If | If so, the Y-axis angle needs to be adjusted.

[0145] 4) Calculate the gimbal adjustment angle.

[0146] A gimbal angle adjustment mechanism using a PID control algorithm (configuration of proportional gain coefficient, integral, and derivative parameters) is employed to improve angle rotation stability.

[0147] 4.1) Calculate the gimbal angle adjustment rate using the following formula: in: This refers to the gimbal rotation rate; This is the proportional gain coefficient, with a reference value of 2; This is the integral gain coefficient, with a reference value of 0.05; This is the differential gain coefficient, with a reference value of 0.1; The offset of the target in the image.

[0148] According to the above formula, the gimbal angle rotation rate on the X-axis is: ; The gimbal rotation rate on the Y-axis is: ; 4.2) Calculate the angle adjustment value using the following formula: If the current lens is a wide-angle lens, then the gimbal rotation angle along the X-axis is: / 2; otherwise, the gimbal rotation angle along the X-axis: / 2. If the current lens is a wide-angle lens, then the gimbal rotation angle along the Y-axis is: / 2; otherwise, the gimbal rotation angle on the Y-axis: Off / 2. A negative angle indicates a leftward / upward shift; a positive shift indicates a rightward / downward shift.

[0149] 4.3) Calculate the target angle of the gimbal: ; ; After obtaining the angles of the gimbal on the X and Y axes, set the latest angle for the gimbal to keep the target in the center of the image.

[0150] The PTZ rotation mechanism based on the PID control algorithm automatically adjusts the angle when the target offset rate exceeds 30%, ensuring that the target type of object is always centered in the frame. Combined with the server's generative video model, special effects and subtitles are automatically added during the video splicing stage. Compared with the splicing process of related technologies, the creative dimension of fun is increased by more than 50%, meeting users' personalized sharing needs.

[0151] 5) Control the gimbal to adjust the angle so that the target is kept in the center of the screen.

[0152] The entire algorithm employs a periodic loop mechanism, following a fixed period. Then, perform the above steps in sequence.

[0153] In this embodiment, the camera device integrates a dynamic target intelligent tracking algorithm. By analyzing the size and orientation of the target in the video in real time, it dynamically switches the lens and gimbal angles, achieving the dual goals of recording the entire scene without blind spots and intelligently capturing key target details.

[0154] It should be noted that the intelligent video surveillance method using a binocular gimbal architecture with a rear wide-angle lens and a telephoto (fixed focus or zoom) lens in this application addresses the pain point of monocular cameras in complex indoor scenes where it is difficult to balance panoramic coverage and detail capture. By integrating a dynamic target intelligent tracking algorithm, it achieves the dual goals of recording the entire scene without blind spots and intelligently capturing key target details. Through image quality enhancement technology, intelligent video editing, and generative video models, it realizes dynamic recording of beautiful moments. It should also be noted that in this application embodiment, the camera device is responsible for the identification and preprocessing of target types; the service device is responsible for video filtering and generative content creation. The entire algorithm execution cycle, video time span, processing frequency, and AI video type (such as parent-child / pet themes) are dynamically controlled by business rules (such as target type) to achieve scene-based intelligent creation.

[0155] Based on the same inventive concept, embodiments of this application also provide a video generation apparatus, applied to a server, such as... Figure 10 As shown, the device 1000 may include a receiving module 1010, a determining module 1020, and a splicing simulation module 1030: The receiving module 1010 is used to receive multiple sub-monitoring videos of a target type object within a preset historical time period sent by the camera device, as well as video tag information of the sub-monitoring videos. The video tag information includes content value index data. The determination module 1020 is used to determine the video content value of the sub-monitoring video based on the content value index data; the determination module 1020 is also used to determine at least one target sub-monitoring video among multiple sub-monitoring videos based on the video content value; the splicing module 1030 is used to splice at least one target sub-monitoring video to obtain a "Beautiful Moments" video, which is used by the terminal device to play the "Beautiful Moments" video on the "Beautiful Moments" page of the monitoring application.

[0156] In some embodiments, the determining module is used to determine the video content value of the sub-monitoring video based on content value index data. Specifically, it can be used to: determine the type value based on the target type; determine the emotional value based on the object characteristics in the sub-monitoring video; determine the picture quality value based on the video picture quality of the sub-monitoring video; and generate the video content value of the sub-monitoring video based on the type value, emotional value, picture quality value, and preset weight coefficients.

[0157] In some embodiments, the determining module is used to determine at least one target sub-monitoring video among multiple sub-monitoring videos based on the value of the video content. Specifically, it can be used to: divide a preset historical time period into multiple sub-historical time periods; and filter out at least one target sub-monitoring video in each sub-historical time period from the multiple sub-monitoring videos based on the value of the video content.

[0158] In some embodiments, the video tag information also includes the video capture time; the splicing module is used to splice at least one target sub-monitoring video to obtain a beautiful moment video, specifically it can be used to: sort at least one target sub-monitoring video in chronological order based on the video capture time; add a smooth transition video with a preset transition duration between two adjacent target sub-monitoring videos and then splice them to obtain an initial beautiful moment video; add background music and video text to the initial beautiful moment video to obtain a beautiful moment video.

[0159] Based on the same inventive concept, embodiments of this application also provide a video generation apparatus, applied to a terminal device, the terminal device having a monitoring application installed, such as... Figure 11 As shown, the device 1100 may include a display module 1110, a receiving module 1120, and a sending module 1130: the display module 1110 is used to display the "Good Times" page of the monitoring application, and the type filtering area in the "Good Times" page includes information on multiple types; the receiving module 1120 is used to receive the type selection instruction input in the type filtering area, and the type selection instruction includes information on the target type among multiple types; the sending module 1130 is used to send the target type information to the server so that the server can receive multiple sub-monitoring videos of the target type object sent by the camera device within a preset historical time period, as well as the video tag information of the sub-monitoring videos, the video tag information including content value index data; determine the video content value of the sub-monitoring videos according to the content value index data; determine at least one target sub-monitoring video among the multiple sub-monitoring videos according to the video content value; and stitch the at least one target sub-monitoring video together to obtain the "Good Times" video; the receiving module 1120 is also used to receive the "Good Times" video sent by the server and play the "Good Times" video on the "Good Times" page of the monitoring application.

[0160] Based on the same inventive concept, embodiments of this application also provide a video generation apparatus, applied to camera equipment, such as... Figure 12 As shown, the device 1200 may include an acquisition module 1210, a generation module 1220, and a sending module 1230: the acquisition module 1210 is used to acquire target type information; the acquisition module 1210 is also used to acquire sub-monitoring images at preset time intervals within a preset historical period; the generation module 1220 is used to generate image tag information for sub-monitoring images based on the target type, object features in the sub-monitoring images, image quality of the sub-monitoring images, and image acquisition time of the sub-monitoring images when the sub-monitoring images include objects of the target type; the generation module 1220 is also used to extract multiple sub-monitoring videos of preset durations from the monitoring video based on multiple sub-monitoring images; the generation module 1220 is also used to extract sub-monitoring videos of various sub-monitoring images from the monitoring video. The image tag information is used to generate video tag information for sub-monitoring videos; the sending module 1230 is used to send multiple sub-monitoring videos of the target type object within a preset historical time period and the video tag information of the sub-monitoring videos to the server, so that the server can receive multiple sub-monitoring videos of the target type object within a preset historical time period and the video tag information of the sub-monitoring videos sent by the camera device. The video tag information includes content value index data; the video content value of the sub-monitoring videos is determined according to the content value index data; at least one target sub-monitoring video is determined among the multiple sub-monitoring videos according to the video content value; the at least one target sub-monitoring video is stitched together to obtain a "Beautiful Moments" video, which is used by the terminal device to play the "Beautiful Moments" video on the "Beautiful Moments" page of the monitoring application.

[0161] The various modules in the video generation apparatus provided in this application embodiment can achieve... Figures 1 to 7 The functions of each step in the provided video generation method, and the corresponding technical effects they achieve, will not be elaborated here for the sake of brevity.

[0162] Figure 13 A schematic diagram of the hardware structure of the video generation device provided in an embodiment of this application is shown.

[0163] The video generating device may include a processor 1301 and a memory 1302 storing computer program instructions.

[0164] Specifically, the processor 1301 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0165] Memory 1302 may include a mass storage device for data or instructions. For example, and not limitingly, memory 1302 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where suitable, memory 1302 may include removable or non-removable (or fixed) media. Where suitable, memory 1302 may be internal or external to the video generation device. In a particular embodiment, memory 1302 is a non-volatile solid-state memory. Processor 1301 implements any of the video generation methods described above by reading and executing computer program instructions stored in memory 1302.

[0166] In one example, the video generating device may further include a communication interface 1303 and a bus 1304. Wherein, as Figure 13 As shown, the processor 1301, memory 1302, and communication interface 1303 are connected via bus 1304 and communicate with each other. Communication interface 1303 is mainly used to realize communication between various modules, devices, units, and / or equipment in this embodiment. Bus 1304 includes hardware, software, or both, coupling the components of the video generation device together. This device can execute the video generation method in this embodiment based on the various units / components in the video generation device, thereby achieving a combination... Figures 1 to 6 The method for generating the video described.

[0167] Furthermore, in conjunction with the video generation methods described in the above embodiments, this application embodiment can provide a computer storage medium for implementation. This computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the video generation methods described in the above embodiments.

[0168] This application also provides a computer program product, wherein the instructions in the computer program product, when executed by the processor of an electronic device, cause the electronic device to perform various processes implementing any of the above-described video generation method embodiments.

[0169] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0170] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0171] The above are merely specific embodiments of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method of generating a video, characterized by, Applied to a server, the method comprises: receiving a plurality of sub-monitoring videos of objects of a target type sent by a camera device within a preset historical period and video tag information of the sub-monitoring videos, the video tag information comprising content value index data; determining video content value of the sub-monitoring videos according to the content value index data; determining at least one target sub-monitoring video from the plurality of sub-monitoring videos according to the video content value; splicing at least one target sub-monitoring video to obtain a good time video for playing the good time video in a good time page of a monitoring application on a terminal device.

2. The method of claim 1, wherein, The method comprises: determining type value according to the target type; determining emotional value according to object features in the sub-monitoring videos; determining picture quality value according to video picture quality of the sub-monitoring videos; determining video content value of the sub-monitoring videos based on the type value, the emotional value, the picture quality value and a preset weight coefficient.

3. The method of claim 1, wherein, The method comprises: dividing the preset historical period into a plurality of sub-historical periods; screening at least one target sub-monitoring video within each sub-historical period from the plurality of sub-monitoring videos according to the video content value.

4. The method of claim 1, wherein, The video tag information further comprises video collection time; the method of splicing at least one target sub-monitoring video to obtain a good time video comprises: sorting at least one target sub-monitoring video in chronological order based on the video collection time; splicing after adding a smooth transition video of a preset transition duration between adjacent two target sub-monitoring videos to obtain an initial good time video; adding background music and video script to the initial good time video to obtain a good time video.

5. A method of generating a video, characterized by, Applied to a terminal device, the terminal device is installed with a monitoring application, and the method comprises: displaying a good time page of the monitoring application, the type filtering area in the good time page comprising information of a plurality of types; receiving a type selection instruction input in the type filtering area, the type selection instruction comprising information of a target type from the plurality of types; sending the information of the target type to a server, so that the server receives a plurality of sub-monitoring videos of objects of the target type sent by a camera device within a preset historical period and video tag information of the sub-monitoring videos, the video tag information comprising content value index data; determines video content value of the sub-monitoring videos according to the content value index data; determines at least one target sub-monitoring video from the plurality of sub-monitoring videos according to the video content value; and splices at least one target sub-monitoring video to obtain a good time video; receiving the good time video sent by the server and playing the good time video in the good time page of the monitoring application.

6. The method of claim 5, wherein, The receiving the good time video sent by the server and playing the good time video in a good time playing page of the monitoring application comprises: receiving the good time video sent by the server; displaying the brief information of the good time video in a display candidate area in the good time playing page in a preset visual style; in the case of receiving a playing instruction corresponding to the brief information, playing the good time video in a video playing area in the good time playing page.

7. A method of generating a video, characterized by, The method applied to a camera device comprises: obtaining information of a target type; obtaining sub-monitoring images in a preset time interval within a preset historical period; in the case of including an object of the target type in the sub-monitoring image, generating image label information of the sub-monitoring image based on the target type, object features in the sub-monitoring image, image picture quality of the sub-monitoring image and image acquisition time of the sub-monitoring image; based on a plurality of the sub-monitoring images, intercepting a plurality of sub-monitoring videos of a preset time length in a monitoring video; generating video label information of the sub-monitoring video based on the image label information of each of the sub-monitoring images in the sub-monitoring video; sending a plurality of the sub-monitoring videos of the object of the target type within the preset historical period and the video label information of the sub-monitoring video to a server, so that the server receives a plurality of sub-monitoring videos of an object of a target type within a preset historical period and video label information of the sub-monitoring video sent by a camera device, the video label information comprises content value index data; according to the content value index data, determining a video content value of the sub-monitoring video; according to the video content value, determining at least one target sub-monitoring video in a plurality of the sub-monitoring videos; splicing at least one target sub-monitoring video to obtain a good time video, so that a terminal device plays the good time video in a good time page of a monitoring application.

8. An apparatus for generating a video, the apparatus comprising: The device applied to a server comprises: a receiving module, configured to receive a plurality of sub-monitoring videos of an object of a target type within a preset historical period and video label information of the sub-monitoring video sent by a camera device, the video label information comprising content value index data; a determining module, configured to determine a video content value of the sub-monitoring video according to the content value index data; the determining module is further configured to determine at least one target sub-monitoring video in a plurality of the sub-monitoring videos according to the video content value; a splicing module, configured to splice at least one target sub-monitoring video to obtain a good time video, so that a terminal device plays the good time video in a good time page of a monitoring application.

9. An apparatus for generating a video, the apparatus comprising: The device applied to a terminal device, on which a monitoring application is installed, comprises: a display module, configured to display a good time page of the monitoring application, a type filtering area in the good time page comprising information of a plurality of types; The receiving module is configured to receive a type selection instruction input in the type filtering area, and the type selection instruction comprises information of a target type in the types. The sending module is configured to send the information of the target type to a server, so that the server receives a plurality of sub-monitoring videos of objects of the target type sent by the camera device within a preset historical period and video tag information of the sub-monitoring videos, the video tag information comprises content value index data, determines a video content value of the sub-monitoring videos according to the content value index data, determines at least one target sub-monitoring video from the plurality of sub-monitoring videos according to the video content value, and splices the at least one target sub-monitoring video to obtain a good time video. The receiving module is further configured to receive the good time video sent by the server and play the good time video in a good time page of the monitoring application.

10. An apparatus for generating a video, the apparatus comprising: The device is applied to a camera device and comprises: The obtaining module is configured to obtain information of a target type. The obtaining module is further configured to obtain sub-monitoring images at preset time intervals within a preset historical period. The generating module is configured to, in a case where the sub-monitoring images comprise objects of the target type, generate image tag information of the sub-monitoring images based on the target type, object features in the sub-monitoring images, image picture quality of the sub-monitoring images, and image acquisition time of the sub-monitoring images. The generating module is further configured to intercept a plurality of sub-monitoring videos of a plurality of preset time lengths based on the plurality of sub-monitoring images. The generating module is further configured to generate video tag information of the sub-monitoring videos based on the image tag information of each of the sub-monitoring images in the sub-monitoring videos. The sending module is configured to send a plurality of sub-monitoring videos of objects of the target type within a preset historical period and the video tag information of the sub-monitoring videos to a server, so that the server receives a plurality of sub-monitoring videos of objects of the target type sent by the camera device within a preset historical period and video tag information of the sub-monitoring videos, the video tag information comprises content value index data, determines a video content value of the sub-monitoring videos according to the content value index data, determines at least one target sub-monitoring video from the plurality of sub-monitoring videos according to the video content value, and splices the at least one target sub-monitoring video to obtain a good time video, so that a terminal device plays the good time video in a good time page of a monitoring application.

11. A video generation device, comprising: The device comprises a processor and a memory storing computer program instructions, and the processor implements the video generation method in any one of claims 1 to 4, the video generation method in any one of claims 5 to 6, or the video generation method in claim 7 when executing the computer program instructions.

12. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer program instructions, which, when executed by a processor, implement the video generation method according to any one of claims 1 to 4, the video generation method according to any one of claims 5 to 6, or the video generation method according to claim 7.

13. A computer program product, characterised in that, The instructions in the computer program product are executed by a processor of a device, so that the device can implement the video generation method according to any one of claims 1 to 4, the video generation method according to any one of claims 5 to 6, or the video generation method according to claim 7.

Images