Creation method of memory video of electronic equipment and related device
By aggregating, filtering, and providing knowledge guidance through photos taken by children, the problem of meaningless images in recall videos in existing technologies has been solved, resulting in highly relevant children's recall videos and enhancing the recall and learning experience.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENZHEN LUKA DR TECHNOLOGY CO LTD
- Filing Date
- 2025-12-09
- Publication Date
- 2026-05-01
AI Technical Summary
Existing technologies tend to include unclear, repetitive, and meaningless images in children's memory videos, resulting in fragmented videos that fail to effectively evoke fond memories.
By aggregating, filtering, defining themes, and enhancing the images within the album group with knowledge-guided information, duplicate and low-quality images are removed, images relevant to the theme are retained, and children's knowledge-guided information is added to the images to create exclusive memory videos for children.
It effectively reduces the appearance of meaningless pictures, improves the relevance of recall videos and the effect of knowledge transmission, and can better evoke children's memories and learning interest.
Smart Images
Figure CN121967818A_ABST
Abstract
Description
Methods and related devices for creating memory videos of electronic devices Technical Field
[0001] This application relates to the field of artificial intelligence technology, and in particular to a method and related apparatus for creating memory videos of electronic devices. Background Technology
[0002] The phone's photo album allows users to select multiple pictures and generate a slideshow-like video based on the order in which the pictures are selected.
[0003] Applying the technology used to generate slideshow-like videos to learning machines presents the following problems: Since the photos in the learning machine's album are primarily self-taken by children, these photos often contain unclear or meaningless images (e.g., a completely black photo taken directly at an object, or numerous photos of the ground with no meaningful features), or a large number of photos piling up without a clear focus (e.g., taking many photos of leaves at the zoo). Directly including these unclear, repetitive, and randomly taken images as normal scene material in the memory video would cause the algorithm to incorrectly integrate memories of play, holidays, etc., into a fragmented and invalid collection of segments. The resulting video would fail to evoke positive memories and would lose the core meaning of a memory video. Summary of the Invention
[0004] In view of this, this application provides a method for creating a memory video of an electronic device, which provides a way to construct a memory video based on pictures in the electronic device's photo album, and the constructed memory video can effectively evoke the user's memories.
[0005] In a first aspect, this application provides a method for creating a memory video of an electronic device, the method comprising the following steps:
[0006] The images in the album group used to create exclusive learning memory videos for children are aggregated to obtain at least one duplicate image group, and the similarity between the images in the duplicate image group is greater than or equal to a preset similarity.
[0007] Remove images from each of the duplicate image groups whose image quality meets the removal criteria from the album group;
[0008] Based on the location attributes of the shooting location and the time period attributes of the shooting time of each of the remaining pictures in the album group, the theme of the children's exclusive learning memory video is determined;
[0009] Remove images from the remaining images in the album group that have a relevance to the theme that is less than or equal to a preset relevance.
[0010] Based on a preset children's knowledge base, at least some of the remaining pictures in the album group are subjected to children's knowledge-guided enhancement processing to obtain the target album group;
[0011] Based on the target album group and the theme, create the children's exclusive learning memory video.
[0012] Optionally, the remaining at least some of the images in the album group are subjected to child-knowledge-guided enhancement processing based on a preset children's knowledge base to obtain the target album group, including:
[0013] Obtain knowledge-guided scene elements from at least a portion of the target images in the album group, wherein the scene elements include text and / or landscape in the target images;
[0014] Obtain knowledge guidance information corresponding to the scene elements from the children's knowledge base;
[0015] In the associated regions of the scene elements in each of the target images, the knowledge guidance information is added.
[0016] Optionally, the scene elements include traditional Chinese characters, and the knowledge guidance information corresponding to the traditional Chinese characters includes the simplified Chinese characters corresponding to the traditional Chinese characters. Adding the knowledge guidance information to the associated region of the area where the scene elements are located in each of the target images includes:
[0017] The region where the traditional Chinese character is located in the target image is defined as the associated region.
[0018] The traditional Chinese character is deleted from the associated area, and the simplified Chinese character is added to the associated area.
[0019] Optionally, the scene elements include landscapes, the knowledge guidance information includes ancient poems corresponding to the landscapes, and the knowledge guidance information is added to the associated regions of the areas where the scene elements are located in each of the target images, including:
[0020] Identify all target images containing the landscape in the album group;
[0021] The sentences of the ancient poems are divided to determine the target sentences to be added to the target images that at least partially contain the landscape.
[0022] Add the corresponding target sentence to the associated region in each target image containing the landscape.
[0023] Optionally, removing images from each of the duplicate image groups that meet the removal criteria from the album group includes:
[0024] Perform the following steps for each of the repeating image groups:
[0025] At least some of the images in the repeating image group are used as reference images;
[0026] Obtain quality index data for each of the reference images, wherein the quality index data includes data values for at least one of the following parameters: integrity, clarity, and brightness of the target object in the image;
[0027] Determine the differences between the data values of the same parameter in each of the aforementioned quality indicator data;
[0028] The parameter whose difference is greater than or equal to the preset difference is used as the non-reusable parameter of the repeating image group;
[0029] Obtain the data value of each image in the repeating image group with respect to the non-reusable parameter;
[0030] Images whose data values of the non-reusable parameters meet the removal conditions are removed from the album group.
[0031] Optionally, before removing images from the remaining images in the album group that have a relevance to the theme less than or equal to a preset relevance, the method further includes:
[0032] Obtain the image feature data of each remaining image in the album group;
[0033] Determine the target keyword group corresponding to each of the image feature data, wherein the target keyword group contains several target keywords;
[0034] Each of the target keyword groups is matched with the topic to obtain the relevance of each image to the topic.
[0035] Optionally, determining the theme of the children's special learning memory video based on the location attributes of the shooting locations and the time period attributes of the shooting time of each of the remaining pictures in the album group includes:
[0036] Target recognition is performed on at least a portion of the remaining images in the album group to determine images that include the target object;
[0037] Perform facial expression recognition on the image containing the target object to obtain the emotion tag of the target object;
[0038] The topic is obtained by processing the location attribute, the time period attribute, and the emotion tag using a preset text generation model.
[0039] Secondly, this application provides an apparatus for creating a memory video of an electronic device, the apparatus comprising:
[0040] The image aggregation module is used to aggregate images in an album group used to create exclusive learning memory videos for children, and obtain at least one duplicate image group, wherein the similarity between the images in the duplicate image group is greater than or equal to a preset similarity.
[0041] The first filtering module is used to remove images from each of the duplicate image groups whose image quality meets the removal criteria from the album group;
[0042] The theme determination module is used to determine the theme of the children's exclusive learning memory video based on the location attributes of the shooting location and the time period attributes of the shooting time of each of the remaining pictures in the album group;
[0043] The second filtering module is used to remove images from the remaining images in the album group whose relevance to the theme is less than or equal to a preset relevance.
[0044] The information enhancement module is used to perform child-knowledge-guided enhancement processing on at least some of the remaining pictures in the album group based on a preset children's knowledge base, so as to obtain the target album group;
[0045] The generation module is used to create the children's exclusive learning memory video based on the target album group and the theme.
[0046] Thirdly, this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the method for creating a memory video of the electronic device provided in the embodiments of the present invention.
[0047] Fourthly, this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the method for creating a memory video of an electronic device provided in the embodiments of the present invention.
[0048] In this embodiment, by removing images from the album group that meet the rejection criteria, a large number of duplicate and low-quality images can be reduced. Then, the remaining images in the album group are filtered according to the theme of the children's learning-themed memory video, ensuring that the retained images are highly relevant to the theme. This greatly reduces the possibility of randomly captured, meaningless, and cluttered images being included as normal scene material in the children's learning-themed memory video. Furthermore, this solution further applies strong knowledge-guided processing to at least some of the remaining images in the album, establishing a connection between knowledge and memory in the children's learning-themed memory video created based on these images and the theme. This not only effectively evokes children's memories but also achieves a better knowledge transfer effect. Attached Figure Description
[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0050] Figure 1 is a flowchart of a method for creating a memory video of an electronic device according to an embodiment of this application;
[0051] Figure 2 is an interactive schematic diagram of the display interface of the electronic device provided in an embodiment of this application;
[0052] Figure 3 is a second interactive schematic diagram of the display interface of the electronic device provided in an embodiment of this application;
[0053] Figure 4 is a schematic diagram of the interaction of the electronic device display interface provided in the embodiment of this application;
[0054] Figure 5 is a schematic diagram of the interaction of the electronic device display interface provided in the embodiment of this application;
[0055] Figure 6 is a schematic diagram of the interaction of the electronic device display interface provided in the embodiment of this application;
[0056] Figure 7 is a schematic diagram of the interaction of the electronic device display interface provided in the embodiment of this application;
[0057] Figure 8 is a schematic diagram of the interaction of the electronic device display interface provided in the embodiment of this application;
[0058] Figure 9 is a schematic diagram of the structure of a device for creating a memory video of an electronic device according to an embodiment of this application;
[0059] Figure 10 is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation
[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0061] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.
[0062] In this document, references to "embodiment" or "implementation" mean that a particular feature, structure, or characteristic described in connection with an embodiment or implementation may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.
[0063] Please refer to Figure 1. Figure 1 is a flowchart of a method for creating a memory video of an electronic device according to an embodiment of this application. The method for creating a memory video of an electronic device includes the following steps:
[0064] 101. Aggregate the images in the album group used to create exclusive learning memory videos for children to obtain at least one duplicate image group.
[0065] In this embodiment of the invention, the aforementioned electronic device may be a learning device or other terminal device (such as a mobile phone). The learning device refers to a portable hardware device that integrates core functions such as image acquisition, intelligent recognition, learning content generation, and interactive feedback, and can be applied to scenarios such as learning assistance and interactive activities.
[0066] The method for creating memory videos on electronic devices described above can be applied to a server, with a communication connection between the server and the electronic device. In another application scenario, the method for creating memory videos on electronic devices can also be applied to the electronic device itself.
[0067] Images captured by electronic devices can be placed in albums, which may include one or more album groups. For example, images taken at the same location and within the same time period can be placed in the same album group. Exemplarily, the process can be triggered based on user-preset needs, real-time operational needs, or automatically created needs in the background. Then, the images in the album are grouped according to their shooting time and location, resulting in several album groups. One of these album groups is then used to create a video documenting the child's learning memories, ensuring that all images in the album group belong to the same shooting location and were taken within the same time period.
[0068] The aggregation method can be based on clustering the similarity between images. That is, images with a similarity greater than or equal to a preset similarity are grouped into a duplicate image group, and the similarity between images within different duplicate image groups is less than the preset similarity. The similarity between images within a duplicate image group is greater than or equal to the preset similarity.
[0069] 102. Remove images from each duplicate image group that meet the removal criteria from the album group.
[0070] The rejection criteria may include, but are not limited to, image quality being lower than a preset image quality or image quality ranking not being among the top preset numbers in a duplicate image group. For example, rejection criteria could be that the image clarity is lower than a preset clarity, or the brightness is lower than a preset brightness. Alternatively, the image quality of each image in a duplicate image group can be sorted from high to low, and images with lower image quality can be deleted.
[0071] In this case, the images within the same duplicate image group might have been captured due to accidental clicks causing the image to be taken in a different order. Therefore, this solution can remove these duplicate images from the album group when creating a memory video. Methods for removing images from the album group include, but are not limited to, removing the image from the album group and keeping it in the album, or deleting it directly.
[0072] 103. Based on the location attributes of the shooting locations and the time period attributes of the shooting time of each remaining picture in the album group, determine the theme of the children's exclusive learning memory video.
[0073] The remaining images in the album group refer to the images that are still retained in the album group after the processing described in step 102 above. The detailed information for each image in the album can include the shooting time and location. The shooting time period can be considered as the time period in which each image was taken, such as a weekend or summer vacation.
[0074] Location attributes can be considered as the functional positioning of the shooting location, while time period attributes can be attributes such as the time range and seasonal characteristics of the shooting time period. For example, the shooting location can be a zoo, and the location attribute of the zoo can be science education or family-friendly. The shooting time period is summer vacation, and the time period attribute of the shooting time period is leisure or companionship.
[0075] The theme of a children's personalized learning memory video can be considered its core idea and emotional essence. The methods for determining the theme of a children's personalized learning memory video based on location and time period attributes can be: determining keywords for the theme based on location and time period attributes, then combining these keywords to obtain the theme; or directly using the theme to construct a network model to process the location and time period attributes to obtain the theme.
[0076] For example, if the shooting location is a zoo and the shooting time is during summer vacation, the theme could be "A joyful encounter with animals that summer".
[0077] 104. Remove images from the remaining images in the album group whose relevance to the theme is less than or equal to the preset relevance.
[0078] The relevance between an image and a topic can be defined as the degree of matching between features extracted from the image and the topic. Preset relevance levels can be set according to requirements.
[0079] Removing an image from an album group can be done in ways including, but not limited to, removing the image from the album group while keeping it in the album, or deleting it directly.
[0080] 105. Based on a preset children's knowledge base, perform children's knowledge-guided enhancement processing on at least some of the remaining pictures in the album group to obtain the target album group.
[0081] The children's knowledge base can include knowledge enhancement information corresponding to several scene elements. These scene elements can be keywords or scenes, etc.
[0082] It can detect each image in an album group. If the corresponding keywords or scenes are detected in each image in the album group, then the images are enhanced with children's knowledge guidance based on the corresponding knowledge enhancement information. The album group after the children's knowledge guidance enhancement is the target album group.
[0083] 106. Based on the target album group and theme, create exclusive memory videos for children's learning experiences.
[0084] Obtain the preset basic configuration parameters for the memory video. The basic configuration parameters for the memory video may include, but are not limited to, at least one of the following: duration, range of the number of images, interval duration, etc.
[0085] Optionally, after obtaining the target image group, further image filtering can be performed based on the number of images contained in the target image group. For example, if the number of images in the target image group exceeds the upper limit of the image number range in the preset basic configuration parameters for the memory video, the images in the target image group are filtered so that the images ultimately retained fall within that image number range.
[0086] Based on the basic configuration parameters of the memory video, the target album group, and the theme, the constituent elements of the children's exclusive learning memory video are determined, and then the video is generated according to these elements. These constituent elements may include at least some images from the target album group, as well as at least one of the following: background music, image color tone, image layout, and transitions.
[0087] In this embodiment of the invention, by removing images from the album group whose image quality meets the rejection criteria, a large number of duplicate and low-quality images can be reduced. Then, the remaining images in the album group are filtered according to the theme of the children's learning-specific memory video, so that the retained images have a high degree of relevance to the theme of the children's learning-specific memory video. This greatly reduces the possibility of randomly shot meaningless and messy scenes being included as normal scene material in the children's learning-specific memory video. In addition, this solution further performs strong children's knowledge-guided processing on at least some of the remaining images in the album, so that the children's learning-specific memory video created based on these images and the theme establishes a connection between knowledge and memory. This not only effectively evokes children's memories but also achieves a better knowledge transfer effect.
[0088] It is understood that in the specific implementation of this application, data related to target objects, pictures in the album, voice description information, question data, knowledge data, answer data, etc. are involved. When the embodiments in this application are applied to specific products or technologies, user permission or consent is required. Furthermore, the collection, use and processing of related data, as well as the training, deployment and invocation of algorithm models, must comply with the relevant laws, regulations and standards of the relevant countries and regions.
[0089] Optionally, the above 105 may include the following steps: obtaining scene elements available for knowledge guidance from at least some of the target images in the album group, the scene elements including text and / or landscape in the target images; obtaining knowledge guidance information corresponding to the scene elements from the children's knowledge base; and adding knowledge guidance information to the associated areas of the areas where the scene elements are located in each target image.
[0090] The script can be Chinese characters or minority languages, etc. The knowledge guidance information for the script includes, but is not limited to: other ways of writing the script, pinyin, translations in the preset languages, definitions, and pictorial representations.
[0091] A landscape can be a specific background object or a scene composed of one or more background objects. Knowledge-guided information about a landscape includes, but is not limited to: ancient poems, historical origins, related allusions, cultural significance, and descriptive information about the landscape.
[0092] The area where the scene element is located can be the image area where the scene element is located in the picture. The associated area can be a preset setting, or it can be determined from the picture based on the knowledge guidance information and the area where the scene element is located.
[0093] For example, a scene element in an image may include Chinese characters, and the corresponding knowledge guidance information may include pinyin, with the associated area being above the area where the Chinese characters are located. Alternatively, a scene element in an image may include a waterfall, and the corresponding knowledge guidance information may be ancient poems describing waterfalls, with the associated area being one side of the area where the waterfall is located.
[0094] In the above scheme, by determining the corresponding knowledge guidance information based on scene elements in at least some of the pictures and displaying the knowledge guidance information in the relevant area on the picture, the memory of the scene elements and knowledge guidance information can be deepened, so that when children see the same knowledge guidance information later, they can think of the scene elements and better recall their memories.
[0095] Optionally, the scene elements include traditional Chinese characters, and the knowledge guidance information corresponding to the traditional Chinese characters includes the simplified Chinese characters corresponding to the traditional Chinese characters. The way to add knowledge guidance information to the associated areas of the scene elements in each target image can be: taking the area where the traditional Chinese characters are located in the target image as the associated area; deleting the traditional Chinese characters from the associated area and adding simplified Chinese characters to the associated area.
[0096] By deleting traditional Chinese characters from the associated area and adding simplified Chinese characters to the associated area, the effect of switching traditional Chinese characters in the target image to simplified Chinese characters is achieved.
[0097] For example, if the target image before the knowledge guidance information is added to the recall video, the display effect of the target image on the electronic device when playing the recall video is shown in Figure 2. The word "station" displayed in the target image is difficult for children to recognize. If the target image after the knowledge guidance information is added to the recall video, the display effect of the recall video on the electronic device is shown in Figure 3. Obviously, the traditional Chinese characters in the target image have been replaced with simplified Chinese characters, which reduces the difficulty of children's literacy.
[0098] In the above solution, converting traditional Chinese characters to simplified Chinese characters can reduce the difficulty of literacy for children. If children are distracted because they cannot recognize traditional Chinese characters, it will disrupt their viewing rhythm. Simplified Chinese characters can help them focus their attention on the picture and the memory itself.
[0099] Optionally, the scene elements include landscapes, and the knowledge guidance information includes ancient poems corresponding to the landscapes. The way to add knowledge guidance information to the associated areas of the scene elements in each target image can be: determining all target images containing landscapes in the album group; dividing the sentences of the ancient poems and determining the target sentences to be added in at least some of the target images containing landscapes; and adding the corresponding target sentences to the associated areas of each target image containing landscapes.
[0100] Classical Chinese poetry can be used to describe the form or grandeur of a landscape. One way to divide classical Chinese poetry is by using semantic completeness as the dividing criterion, resulting in multiple sentences.
[0101] Considering that the frame switching in the memory video is relatively fast, adding all the sentences of the ancient poem to a single image might cause children to pause the playback to finish watching, or they might not be able to finish watching the poem without pausing. Therefore, this solution chooses to add different sentences of the ancient poem to different target images containing the landscape. The resulting memory video contains target images of different target sentences from the same ancient poem, arranged in the reading order of each target sentence.
[0102] For example, if the number of target images containing the landscape is the same as the number of sentences in the classical Chinese poem, then each target image containing the landscape needs to have one sentence from the classical Chinese poem added to it. If the number of target images containing the landscape is greater than the number of sentences in the classical Chinese poem, then a corresponding number of target images are selected from each target image according to the number of sentences in the classical Chinese poem, and one sentence from the classical Chinese poem is added to each of the selected target images.
[0103] If the number of target images containing the landscape is less than the number of lines in the classical poem, frame interpolation can be performed on the target images containing the landscape to make the number of target images containing the landscape consistent with the number of lines in the classical poem. Each target image containing the landscape will then have one line from the classical poem added to it. Alternatively, if the number of target images containing the landscape is less than the number of lines in the classical poem, at least two lines from the classical poem can be added to some of the target images, ensuring that each target image contains a different line from the classical poem.
[0104] For example, if the target image before the added knowledge guidance information is added to the recall video, the display effect of one of the target images on the electronic device when playing the recall video is shown in Figure 4. The landscape in the target image is the waterfall on the left. The guidance information corresponding to the waterfall is the ancient poem "Looking at the Waterfall at Mount Lu," which includes four sentences. These four sentences can be added to four images respectively. After adding the target image with the added knowledge guidance information to the recall video, the display effect of the four target images with different sentences from "Looking at the Waterfall at Mount Lu" added to the recall video on the electronic device when playing the recall video is shown in Figures 5 to 8.
[0105] In the above scheme, by adding the corresponding ancient poems to the target images containing specific landscapes, not only can children's memory of the landscapes be deepened, but their learning interest can also be stimulated and their initiative in exploring knowledge can be improved. Moreover, the resulting memory video contains target images of different sentences from the same ancient poem arranged in the reading order of each sentence, with each image containing only a small amount of content, making the viewing process smoother for children.
[0106] Optionally, 102 may include the following steps: For each duplicate image group, perform the following steps: use at least some images in the duplicate image group as reference images; obtain quality index data for each reference image, the quality index data including data values of at least one of the following parameters: integrity, clarity, and brightness of the target object in the image; determine the differences between the data values of the same parameter in each quality index data; use parameters with differences greater than or equal to preset differences as non-reusable parameters for the duplicate image group; obtain data values of each image in the duplicate image group regarding the non-reusable parameters; remove images whose non-reusable parameter data values meet the removal criteria from the album group.
[0107] Multiple images can be sampled from a group of repeating images and used as reference images.
[0108] The target object includes objects that are related to the user, such as the user themselves or objects that the user is interested in. The completeness of the target object characterizes whether the target object is fully rendered in the target image. Differences between the data values of the same parameter in various quality metrics can include differences in completeness, sharpness, and / or brightness.
[0109] Non-reusable parameters refer to parameters that differ significantly between images in a repeating image set. For example, if the images in a repeating image set have small differences in integrity and sharpness but large differences in brightness, then the brightness difference is used as the non-reusable parameter for that repeating image set. Then, the brightness data value for each image in the repeating image set is obtained.
[0110] The exclusion criteria here can be that the data value of the non-reusable parameter does not reach the preset threshold or the data value of the non-reusable parameter does not reach the first preset number of duplicate images in the group.
[0111] In the above scheme, considering that there are many quality problems in the pictures taken in children's scenes, it is necessary to determine whether the pictures need to be deleted based on multiple indicators (completeness, clarity, brightness, etc.). If every parameter of each picture is calculated, the amount of computation is large. Because the similarity between the pictures in the duplicate group is high, some indicator data may be highly similar, while some indicators may differ slightly. Therefore, this scheme samples the indicators of a few pictures and determines which pictures need to be deleted based on the indicators with larger differences, thereby reducing the amount of computation.
[0112] Optionally, before removing images from the remaining images in the album group whose relevance to the theme is less than or equal to a preset relevance, the method further includes: obtaining image feature data for each remaining image in the album group; determining the target keyword group corresponding to each image feature data, wherein the target keyword group contains several target keywords; and matching each target keyword group with the theme to obtain the relevance of each image to the theme.
[0113] Image feature data may include, but is not limited to, pose data, behavior data, and scene feature data of the target object. The target object here includes objects associated with the user, such as the user themselves or objects of interest to the user.
[0114] One way to determine the target keyword groups corresponding to each image feature data is to input each image feature data into a keyword extraction model to obtain the corresponding keyword groups. For example, in the obtained keyword groups, the keyword corresponding to posture data could be "squatting," the keyword corresponding to behavior data could be "feeding fish," and the scene feature data could be "fish restaurant." Each target keyword group is then matched with a topic, and the matching degree is used as the correlation between the corresponding image and the topic.
[0115] In the above solution, by extracting image feature data for each target image in the target image group and calculating the association between keyword groups and the theme of each image feature data, and presenting highly related images in the memory video, the user's memory of the core scene can be quickly awakened, irrelevant images can be avoided to distract attention, and the emotional resonance of the video can be stronger.
[0116] Optionally, the method of determining the theme of the children's learning-themed memory video based on the location attributes of the shooting locations and the time period attributes of the shooting time of the remaining pictures in the album group can be as follows: perform target recognition on at least some of the remaining pictures in the album group to determine the pictures containing the target object; perform facial expression recognition on the pictures containing the target object to obtain the emotional label of the target object; and use a preset text generation model to process the location attributes, time period attributes, and emotional labels to obtain the theme.
[0117] The target object includes objects that are related to the user, such as the user themselves or objects that the user is interested in. Emotion tags can be used to characterize the emotions of the target object; for example, emotion tags could be surprise, joy, etc.
[0118] The preset text generation model can be a pre-trained neural network model. The input of the preset text generation model is location attribute, time period attribute and emotion label, and the output is the topic of the recall video.
[0119] In the above approach, emotion is the core link of memory. The memory video created by setting the theme using emotion tags can quickly evoke children's association with the emotions at that time, making it easier to trigger emotional feelings and making the memory video more memorable.
[0120] Please refer to Figure 9, which is a schematic diagram of the structure of a device for creating a memory video of an electronic device according to an embodiment of this application. The device for creating a memory video of an electronic device includes:
[0121] Image aggregation module 201 is used to aggregate images in an album group for creating exclusive learning memory videos for children to obtain at least one duplicate image group, wherein the similarity between the images in the duplicate image group is greater than or equal to a preset similarity.
[0122] The first filtering module 202 is used to remove images from each of the duplicate image groups whose image quality meets the removal criteria from the album group;
[0123] Theme determination module 203 is used to determine the theme of the children's exclusive learning memory video based on the location attributes of the shooting location and the time period attributes of the shooting time of each of the remaining pictures in the album group;
[0124] The second filtering module 204 is used to remove images from the remaining images in the album group whose relevance to the theme is less than or equal to a preset relevance.
[0125] The information enhancement module 205 is used to perform child knowledge-guided enhancement processing on at least some of the remaining pictures in the album group based on a preset children's knowledge base to obtain the target album group;
[0126] The generation module 206 is used to create the children's exclusive learning memory video based on the target album group and the theme.
[0127] Optionally, the information enhancement module 205 performs child-knowledge-guided enhancement processing on at least some of the remaining images in the album group based on a preset children's knowledge base to obtain the operational aspects of the target album group. Specifically, the information enhancement module 205 is used for:
[0128] Obtain knowledge-guided scene elements from at least a portion of the target images in the album group, wherein the scene elements include text and / or landscape in the target images;
[0129] Obtain knowledge guidance information corresponding to the scene elements from the children's knowledge base;
[0130] In the associated regions of the scene elements in each of the target images, the knowledge guidance information is added.
[0131] Optionally, the scene elements include traditional Chinese characters, and the knowledge guidance information corresponding to the traditional Chinese characters includes the simplified Chinese characters corresponding to the traditional Chinese characters. The information enhancement module 205 is specifically used for: adding the knowledge guidance information to the associated area of the scene element in each of the target images; and for other operational aspects.
[0132] The region where the traditional Chinese character is located in the target image is defined as the associated region.
[0133] The traditional Chinese character is deleted from the associated area, and the simplified Chinese character is added to the associated area.
[0134] Optionally, the scene elements include landscapes, the knowledge guidance information includes ancient poems corresponding to the landscapes, the associated regions of the scene elements in each of the target images, and the operational aspects of adding the knowledge guidance information. The information enhancement module 205 is specifically used for:
[0135] Identify all target images containing the landscape in the album group;
[0136] The sentences of the ancient poems are divided to determine the target sentences to be added to the target images that at least partially contain the landscape.
[0137] Add the corresponding target sentence to the associated region in each target image containing the landscape.
[0138] Optionally, in the operation of removing images from each of the duplicate image groups whose image quality meets the removal criteria from the album group, the first filtering module 202 is specifically used for:
[0139] Perform the following steps for each of the repeating image groups:
[0140] At least some of the images in the repeating image group are used as reference images;
[0141] Obtain quality index data for each of the reference images, wherein the quality index data includes data values for at least one of the following parameters: integrity, clarity, and brightness of the target object in the image;
[0142] Determine the differences between the data values of the same parameter in each of the aforementioned quality indicator data;
[0143] The parameter whose difference is greater than or equal to the preset difference is used as the non-reusable parameter of the repeating image group;
[0144] Obtain the data value of each image in the repeating image group with respect to the non-reusable parameter;
[0145] Images whose data values of the non-reusable parameters meet the removal conditions are removed from the album group.
[0146] Optionally, before removing images from the remaining images in the album group that have a relevance to the theme less than or equal to a preset relevance, the second filtering module 204 is further configured to:
[0147] Obtain the image feature data of each remaining image in the album group;
[0148] Determine the target keyword group corresponding to each of the image feature data, wherein the target keyword group contains several target keywords;
[0149] Each of the target keyword groups is matched with the topic to obtain the relevance of each image to the topic.
[0150] Optionally, in the operation of determining the theme of the children's exclusive learning memory video based on the location attributes of the shooting locations and the time period attributes of the shooting time of each of the remaining pictures in the album group, the theme determination module 203 is specifically used for:
[0151] Target recognition is performed on at least a portion of the remaining images in the album group to determine images that include the target object;
[0152] Perform facial expression recognition on the image containing the target object to obtain the emotion tag of the target object;
[0153] The topic is obtained by processing the location attribute, the time period attribute, and the emotion tag using a preset text generation model.
[0154] Please refer to Figure 10, which is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. This application also provides an electronic device, including: a memory 302, a processor 301, and a computer program stored in the memory 302 and executable on the processor 301. When the processor 301 executes the computer program, it implements the steps in the method for creating a memory video of the electronic device provided in this embodiment of the application.
[0155] When processor 301 runs the computer program for creating the memory video of an electronic device stored in memory 302, it specifically performs the following steps:
[0156] The images in the album group used to create exclusive learning memory videos for children are aggregated to obtain at least one duplicate image group, and the similarity between the images in the duplicate image group is greater than or equal to a preset similarity.
[0157] Remove images from each of the duplicate image groups whose image quality meets the removal criteria from the album group;
[0158] Based on the location attributes of the shooting location and the time period attributes of the shooting time of each of the remaining pictures in the album group, the theme of the children's exclusive learning memory video is determined;
[0159] Remove images from the remaining images in the album group that have a relevance to the theme that is less than or equal to a preset relevance.
[0160] Based on a preset children's knowledge base, at least some of the remaining pictures in the album group are subjected to children's knowledge-guided enhancement processing to obtain the target album group;
[0161] Based on the target album group and the theme, create the children's exclusive learning memory video.
[0162] Optionally, the processor 301 performs child-knowledge-guided enhancement processing on at least a portion of the remaining images in the album group based on a preset children's knowledge base to obtain a target album group, including:
[0163] Obtain knowledge-guided scene elements from at least a portion of the target images in the album group, wherein the scene elements include text and / or landscape in the target images;
[0164] Obtain knowledge guidance information corresponding to the scene elements from the children's knowledge base;
[0165] In the associated regions of the scene elements in each of the target images, the knowledge guidance information is added.
[0166] Optionally, the scene elements include traditional Chinese characters, and the knowledge guidance information corresponding to the traditional Chinese characters includes the simplified Chinese characters corresponding to the traditional Chinese characters; the processor 301 executes the association region of the area where the scene elements are located in each of the target images, adding the knowledge guidance information, including:
[0167] The region where the traditional Chinese character is located in the target image is defined as the associated region.
[0168] The traditional Chinese character is deleted from the associated area, and the simplified Chinese character is added to the associated area.
[0169] Optionally, the scene elements include landscapes, and the knowledge guidance information includes ancient poems corresponding to the landscapes. The processor 301 executes the association region of the area where the scene elements are located in each of the target images, and adds the knowledge guidance information, including:
[0170] Identify all target images containing the landscape in the album group;
[0171] The sentences of the ancient poems are divided to determine the target sentences to be added to the target images that at least partially contain the landscape.
[0172] Add the corresponding target sentence to the associated region in each target image containing the landscape.
[0173] Optionally, the step of removing images from the album group whose image quality meets the removal criteria in each of the duplicate image groups, performed by the processor 301, includes:
[0174] Perform the following steps for each of the repeating image groups:
[0175] At least some of the images in the repeating image group are used as reference images;
[0176] Obtain quality index data for each of the reference images, wherein the quality index data includes data values for at least one of the following parameters: integrity, clarity, and brightness of the target object in the image;
[0177] Determine the differences between the data values of the same parameter in each of the aforementioned quality indicator data;
[0178] The parameter whose difference is greater than or equal to the preset difference is used as the non-reusable parameter of the repeating image group;
[0179] Obtain the data value of each image in the repeating image group with respect to the non-reusable parameter;
[0180] Images whose data values of the non-reusable parameters meet the removal conditions are removed from the album group.
[0181] Optionally, before removing images from the remaining images in the album group that have a relevance to the theme less than or equal to a preset relevance, the processor 301 further performs the following steps:
[0182] Obtain the image feature data of each remaining image in the album group;
[0183] Determine the target keyword group corresponding to each of the image feature data, wherein the target keyword group contains several target keywords;
[0184] Each of the target keyword groups is matched with the topic to obtain the relevance of each image to the topic.
[0185] Optionally, the step of determining the theme of the children's special learning memory video based on the location attributes of the shooting locations and the time period attributes of the shooting time of each of the remaining pictures in the album group, executed by the processor 301, includes:
[0186] Target recognition is performed on at least a portion of the remaining images in the album group to determine images that include the target object;
[0187] Perform facial expression recognition on the image containing the target object to obtain the emotion tag of the target object;
[0188] The topic is obtained by processing the location attribute, the time period attribute, and the emotion tag using a preset text generation model.
[0189] In addition, this embodiment of the invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by the processor 301, it implements the various processes of the method for creating a memory video of an electronic device provided in this embodiment of the invention and can achieve the same technical effect. To avoid repetition, it will not be described again here.
[0190] Those skilled in the art will understand that implementing all or part of the processes in the above embodiments can be accomplished by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.
[0191] In this application, the terms "embodiment" and "implementation" mean that a specific feature, structure, or characteristic described in connection with an embodiment can be included in at least one embodiment of this application. The appearance of these phrases in various locations throughout the specification does not necessarily refer to the same embodiment, nor are they independent or alternative embodiments mutually exclusive with other embodiments. Those skilled in the art will understand, explicitly and implicitly, that the embodiments described in this application can be combined with other embodiments. Furthermore, it should be understood that the features, structures, or characteristics described in the various embodiments of this application can be arbitrarily combined to form another embodiment that does not depart from the spirit and scope of the technical solution of this application, provided there is no contradiction between them.
[0192] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit it. Although this application has been described in detail with reference to the above preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of this application should not depart from the spirit and scope of the technical solutions of this application.
Claims
1. A method for creating a memory video of an electronic device, characterized in that, The method includes the following steps: aggregating images within an album group used to create a children's learning-themed memory video to obtain at least one duplicate image group, wherein the similarity between images within each duplicate image group is greater than or equal to a preset similarity; removing images from each duplicate image group whose image quality meets the removal criteria from the album group; determining the theme of the children's learning-themed memory video based on the location attributes of the shooting location and the time period attributes of the shooting time of each remaining image in the album group; removing images from the remaining images in the album group whose relevance to the theme is less than or equal to a preset relevance; performing children's knowledge-guided enhancement processing on at least some of the remaining images in the album group based on a preset children's knowledge base to obtain a target album group; and creating the children's learning-themed memory video based on the target album group and the theme.
2. The method as described in claim 1, characterized in that, The process of performing child-guided enhancement processing on at least a portion of the remaining images in the album group based on a preset children's knowledge base to obtain a target album group includes: obtaining scene elements available for knowledge guidance from at least a portion of the target images in the album group, wherein the scene elements include text and / or landscapes in the target images; obtaining knowledge guidance information corresponding to the scene elements from the children's knowledge base; and adding the knowledge guidance information to the associated area of the area where the scene elements are located in each target image.
3. The method as described in claim 2, characterized in that, The scene elements include traditional Chinese characters, and the knowledge guidance information corresponding to the traditional Chinese characters includes the simplified Chinese characters corresponding to the traditional Chinese characters. The step of adding the knowledge guidance information to the associated region of the area where the scene elements are located in each of the target images includes: taking the area where the traditional Chinese characters are located in the target image as the associated region; deleting the traditional Chinese characters from the associated region and adding the simplified Chinese characters to the associated region.
4. The method as described in claim 2, characterized in that, The scene elements include landscapes, and the knowledge guidance information includes ancient poems corresponding to the landscapes. Adding the knowledge guidance information to the associated areas of the scene elements in each of the target images includes: determining all target images containing the landscapes in the album group; dividing the sentences of the ancient poems and determining the target sentences to be added in at least some of the target images containing the landscapes; and adding the corresponding target sentences to the associated areas of each target image containing the landscapes.
5. The method according to any one of claims 1 to 4, characterized in that, The step of removing images from each of the duplicate image groups that meet the removal criteria from the album group includes: performing the following steps on each of the duplicate image groups: using at least some images in the duplicate image group as reference images; obtaining quality index data for each of the reference images, wherein the quality index data includes data values of at least one of the following parameters: integrity, clarity, and brightness of the target object in the image; determining the differences between the data values of the same parameter in each of the quality index data; using parameters whose differences are greater than or equal to a preset difference as non-reusable parameters of the duplicate image group; obtaining data values of each image in the duplicate image group with respect to the non-reusable parameters; and removing images whose data values of the non-reusable parameters meet the removal criteria from the album group.
6. The method according to any one of claims 1 to 4, characterized in that, Before removing images from the remaining images in the album group whose relevance to the theme is less than or equal to a preset relevance, the method further includes: obtaining image feature data of each remaining image in the album group; determining target keyword groups corresponding to each image feature data, wherein the target keyword groups contain several target keywords; and matching each target keyword group with the theme to obtain the relevance of each image to the theme.
7. The method according to any one of claims 1 to 3, characterized in that, The step of determining the theme of the children's learning-themed memory video based on the location attributes of the shooting locations and the time period attributes of the shooting time of each of the remaining images in the album group includes: performing target recognition on at least some of the remaining images in the album group to determine images containing target objects; performing facial expression recognition on images containing target objects to obtain the emotion tags of the target objects; and processing the location attributes, the time period attributes, and the emotion tags using a preset text generation model to obtain the theme.
8. An apparatus for creating a memory video of an electronic device, characterized in that, The device for creating a memory video for the electronic device includes: an image aggregation module for aggregating images in an album group used to create a children's learning-themed memory video, obtaining at least one duplicate image group, wherein the similarity between the images in the duplicate image group is greater than or equal to a preset similarity; a first filtering module for removing images from the album group whose image quality meets the removal criteria; a theme determination module for determining the theme of the children's learning-themed memory video based on the location attributes of the shooting location and the time period attributes of the shooting time of the remaining images in the album group; a second filtering module for removing images from the album group whose relevance to the theme is less than or equal to a preset relevance; an information enhancement module for performing children's knowledge-guided enhancement processing on at least some of the remaining images in the album group based on a preset children's knowledge base, obtaining a target album group; and a generation module for creating the children's learning-themed memory video based on the target album group and the theme.
9. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps in the method for creating a memory video of an electronic device as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps in the method for creating a memory video of an electronic device as described in any one of claims 1 to 7.