Infant feeding information intelligent recognition method, device and equipment and storage medium

Abstract

This invention belongs to the field of image processing technology and solves the technical problem of food waste and poor parenting results caused by the inability to accurately grasp infants' feeding information and the practice of raising infants based on general parenting concepts. It provides an intelligent recognition method, device, equipment, and storage medium for infants' feeding information. The method includes: acquiring feeding videos of infants' feeding scenes; analyzing the infants' feeding actions and scene factors influencing these actions; analyzing feeding actions including feeding with utensils and the infant's micro-expressions; and analyzing scene factors including scene changes and changes in target objects within the scene, thereby determining the infant's feeding information corresponding to the current feeding video; and determining the infant's satisfaction with this feeding. Through the analysis of infants' feeding videos, the state of the infant during each feeding can be accurately grasped, allowing for targeted matching of feeding times and types of food, saving resources and improving parenting effectiveness.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method, apparatus, device, and storage medium for intelligent recognition of infant feeding information. Background Technology

[0002] There are significant individual differences in the growth of infants and young children. However, because infants and young children cannot accurately express their needs, young parents, lacking parenting experience and sufficient time to participate in their infants' development, can only make dietary arrangements based on the overall development of the infants and young children in stages. They cannot make scientific arrangements for the infants' and young children's food based on their actual eating information, resulting in a waste of resources and failing to achieve the expected nurturing effect. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a method, apparatus, device and storage medium for intelligent identification of infant feeding information, in order to solve the technical problem of food waste and poor training effect caused by the inability to accurately grasp the infant feeding information and to raise infants based on general parenting concepts.

[0004] The technical solution adopted in this invention is:

[0005] This invention provides an intelligent recognition method for infant feeding information, the method comprising:

[0006] S1: Obtain feeding videos of infants and toddlers;

[0007] S2: Perform target analysis on each frame of the feeding video to obtain the feeding parameters of the infant's current feeding scene;

[0008] S3: Based on the infant's action parameters and scene factors in the feeding parameters, obtain the infant's feeding information for this feeding.

[0009] Preferably, S1 includes:

[0010] S11: Obtain first and second state information of the eating utensils relative to the holding part;

[0011] S12: Obtain the eating video based on the start time corresponding to the first state information and the end time corresponding to the second state information;

[0012] The first state information indicates that the eating utensils are in a grasped state, and the second state information indicates that the eating utensils are released from the grasped state.

[0013] Preferably, S12 includes:

[0014] S121: Obtain the preset duration of videos depicting infants and toddlers eating;

[0015] S122: Based on the state information of the feeding tool relative to the holding part, the basic video of the infant feeding scene is segmented to obtain each first video in the holding state and each second video in the non-holding state.

[0016] S123: Based on the scene information of each second video and the preset duration, each second video is shortened to obtain a second target video that corresponds one-to-one with each second video;

[0017] S124: The first video and the second target video are spliced ​​together according to the shooting time sequence to obtain the eating video.

[0018] Preferably, S2 includes:

[0019] S21: Obtain the frequency difference of the eating actions segmented from the video of the eating process;

[0020] S22: Divide the eating video into multiple eating sub-videos based on the frequency difference;

[0021] S23: Analyze each of the feeding sub-videos to obtain the action parameters of the feeding parameters that correspond one-to-one with each of the feeding sub-videos.

[0022] Preferably, S23 includes:

[0023] S231: Obtain the distance value and distance threshold of the eating utensils relative to the infant;

[0024] S232: Compare each distance value with the distance threshold to obtain each eating action corresponding to the action information;

[0025] The image of one feeding action includes each frame of the video stream corresponding to two consecutive distance values ​​equal to the distance threshold.

[0026] Preferably, S231 includes:

[0027] S2311: Obtain the reference position and the position of the tableware used to represent the infant eating in each frame of the feeding video;

[0028] S2312: Based on the Euclidean distance between each of the reference positions and the tableware positions, obtain the distance gradients corresponding to the time dimension;

[0029] The distance gradients are used to characterize the distance values.

[0030] Preferably, S3 includes:

[0031] S31: Obtain the pre-defined eating weights for each eating scenario;

[0032] S32: Use the feeding weights to weight the action parameters corresponding to the feeding scenario to obtain the feeding score of the infant for this feeding.

[0033] S33: Obtain the eating information by comparing the eating score with the preset eating score table.

[0034] The present invention also provides an intelligent recognition device for infant feeding information, comprising:

[0035] Data acquisition module: used to acquire videos of infants and young children eating.

[0036] Data analysis module: used to perform target analysis on each frame of the feeding video to obtain the feeding parameters of the infant's current feeding scene;

[0037] Data processing module: used to obtain the infant's feeding information for this feeding based on the infant's action parameters and scene factors mentioned in the feeding parameters.

[0038] The present invention also provides an electronic device, comprising: at least one processor, at least one memory, and computer program instructions stored in the memory, wherein the computer program instructions, when executed by the processor, implement the method described in any of the preceding embodiments.

[0039] The present invention also provides a storage medium having stored thereon computer program instructions that, when executed by a processor, implement the method described in any of the preceding claims.

[0040] In summary, the beneficial effects of the present invention are as follows:

[0041] This invention provides an intelligent identification method, device, equipment, and storage medium for infant feeding information. It acquires feeding videos of infants during feeding, analyzes the feeding actions and scene factors influencing these actions, including feeding movements with utensils and the infant's micro-expressions, and scene factors such as scene changes and changes in target objects within the scene. This analysis determines the infant's feeding information corresponding to the current video and assesses the infant's satisfaction with the feeding session. By analyzing these feeding videos, the system can accurately grasp the infant's state during each feeding, allowing for targeted adjustments to feeding times and types of food, saving resources and improving parenting effectiveness. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, and these are all within the protection scope of the present invention.

[0043] Figure 1 This is a flowchart illustrating the intelligent recognition method for infant feeding information in Embodiment 1 of the present invention;

[0044] Figure 2 This is a schematic diagram of the process for obtaining eating videos in Embodiment 1 of the present invention;

[0045] Figure 3 This is a schematic diagram of the process for obtaining action parameters in Embodiment 1 of the present invention;

[0046] Figure 4 This is a schematic diagram of the process for obtaining feeding information in Embodiment 1 of the present invention;

[0047] Figure 5 This is a schematic diagram of the intelligent recognition device for infant feeding information in Embodiment 2 of the present invention;

[0048] Figure 6 This is a schematic diagram of the electronic device in Embodiment 3 of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. In the description of the present invention, it should be understood that the terms "center," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. 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. Unless otherwise specified, an element defined by the phrase "comprising..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. Where there is no conflict, the various features of this invention and its embodiments can be combined with each other, all of which are within the scope of protection of this invention.

[0050] Example 1

[0051] Please see Figure 1 , Figure 1 This is a flowchart illustrating the intelligent recognition method for infant feeding information in Embodiment 1 of the present invention; the method includes:

[0052] S1: Obtain feeding videos of infants and toddlers;

[0053] S2: Perform target analysis on each frame of the feeding video to obtain the feeding parameters of the infant's current feeding scene;

[0054] S3: Based on the infant's action parameters and scene factors in the feeding parameters, obtain the infant's feeding information for this feeding.

[0055] Specifically, AI-powered smart monitoring devices or other electronic devices with image acquisition capabilities are used to collect videos of infants and toddlers eating. The feeding status and effectiveness of each feeding session are determined based on the image information of each frame in the video. The feeding parameters in the videos are categorized into infant-specific information and environmental information. Infant-specific information primarily refers to the infant's movement parameters, while environmental information primarily refers to environmental factors. Based on the infant's movement parameters and environmental factors in the videos, corresponding feeding information is output for each video. Environmental factors include: the feeding scene (McDonald's, playground, restaurant), other targets in the environment besides the target infant, and the main target helping the infant eat. Infant-specific movement parameters include: micro-expressions in the facial area (overall or localized areas), hand movements, feeding rate, and feeding frequency. Feeding information includes: preference, neutral, and dislike.

[0056] The intelligent recognition method for infant feeding information in this embodiment acquires videos of infants' feeding scenes, analyzes the infants' feeding actions and the scene factors influencing these actions. Feeding actions include the feeding motions with utensils and the infant's micro-expressions; scene factors include scene changes and changes in the target object within the scene. This analysis determines the infant's feeding information corresponding to the current video and assesses the infant's satisfaction with the feeding session. By analyzing these videos, the state of the infant during each feeding can be accurately grasped, allowing for targeted adjustments to feeding times and types of food, saving resources and improving parenting effectiveness.

[0057] In one embodiment, please refer to Figure 2 S1 includes:

[0058] S11: Obtain first and second state information of the eating utensils relative to the holding part;

[0059] S12: Obtain the eating video based on the start time corresponding to the first state information and the end time corresponding to the second state information;

[0060] The first state information indicates that the eating utensils are in a grasped state, and the second state information indicates that the eating utensils are released from the grasped state.

[0061] Specifically, the collected videos of infants and toddlers are analyzed to identify feeding videos. The relative positions of the infant's hands and the feeding utensils are analyzed to determine the feeding process. When the infant grasps the feeding utensils, it is considered that the infant has started eating. When the infant releases the feeding utensils, it is considered that the feeding session has ended. The video data from when the infant grasps the feeding utensils to when the infant releases the feeding utensils is recorded as the feeding video. This avoids having too many other useless videos, which would waste computer processing power and reduce the accuracy of the detection.

[0062] It should be noted that the target audience for operating the feeding utensils can be set according to the infant's age. For example, if the infant cannot use the feeding utensils, the operation information of the target audience feeding the infant can be collected. When the infant has just learned to use the feeding utensils, the operation of the feeding utensils will be divided into two stages: the infant operation stage and the infant feeding operation stage. No specific limitations are made here.

[0063] It should be noted that: eating utensils must include at least one of the following: spoon, chopsticks, bowl, plate, etc.; the holding state must include at least one of the following: the state in which an infant or toddler holds the utensils, or the state in which a guardian holds the utensils.

[0064] In one embodiment, S12 includes:

[0065] S121: Obtain the preset duration of videos depicting infants and toddlers eating;

[0066] S122: Based on the state information of the feeding tool relative to the holding part, the basic video of the infant feeding scene is segmented to obtain each first video in the holding state and each second video in the non-holding state.

[0067] S123: Based on the scene information of each second video and the preset duration, each second video is shortened to obtain a second target video that corresponds one-to-one with each second video;

[0068] S124: The first video and the second target video are spliced ​​together according to the shooting time sequence to obtain the eating video.

[0069] Specifically, during infant feeding, the part of the infant holding the eating utensils often separates from the utensils for various reasons. Therefore, an infant feeding video is not a continuous video showing the utensils being held in the hand, but rather a continuous video consisting of multiple feeding segments interspersed with multiple non-feeding segments. This includes, but is not limited to, the following situations: the infant drinking water, playing, or being distracted during feeding; or when the infant is eating with other members and being fed by a caregiver, with the caregiver feeding the infant intermittently. Scenarios corresponding to the separation of the feeding utensils from the holding part are not listed here. The second video corresponding to the separation of the holding part from the feeding utensils includes videos related to the infant feeding and videos unrelated to the feeding. The second video can be edited and optimized according to the preset duration of the feeding video to improve the reliability of the feeding video.

[0070] In one embodiment, please refer to Figure 3 S2 includes:

[0071] S21: Obtain the frequency difference of the eating actions segmented from the video of the eating process;

[0072] S22: Divide the eating video into multiple eating sub-videos based on the frequency difference;

[0073] S23: Analyze each of the feeding sub-videos to obtain the action parameters of the feeding parameters that correspond one-to-one with each of the feeding sub-videos.

[0074] Specifically, an infant's feeding state is related to many factors, and there are significant differences between different feeding states. Feeding frequency is the most obvious indicator of different feeding states. Based on the different feeding frequencies of infants, feeding videos are divided into multiple feeding sub-videos, including but not limited to: early feeding stage, middle feeding stage, and late feeding stage. In the early feeding stage, the infant is very hungry and feeds frequently, such as taking large amounts of food per bite and having frequent feeding needs. In the middle feeding stage, the infant's activity level is low, hunger is not obvious, the amount of food per bite is moderate, and the frequency of feeding needs is low. In the late feeding stage, the infant is not hungry and / or does not like the food, showing resistance to eating, and the feeding frequency is extremely low. The feeding videos are segmented for detection to obtain motion parameters corresponding to each feeding stage.

[0075] In one embodiment, S23 includes:

[0076] S231: Obtain the distance value and distance threshold of the eating utensils relative to the infant;

[0077] S232: Compare each distance value with the distance threshold to obtain each eating action corresponding to the action information;

[0078] The image of one feeding action includes each frame of the video stream corresponding to two consecutive distance values ​​equal to the distance threshold.

[0079] Specifically, the system obtains the real-time positional relationship between the infant and the feeding utensils, determining the distance between them. When the first distance value equals a distance threshold, the feeding utensils enter the infant's mouth area, signifying the start of a feeding session. When the second distance value equals the distance threshold, the feeding utensils leave the infant's mouth area, indicating the end of the feeding session. It should be noted that the distance threshold is a range value, which can reduce false recognition caused by the feeding utensils moving around the infant's mouth. Alternatively, a valid duration can be set for two consecutive distance values ​​equaling the distance threshold. The two distance values ​​equaling the distance threshold are only valid if the interval is greater than or equal to the valid duration. For example, if the valid duration is 0.1 seconds, and the first distance value equals the distance threshold at 10 seconds, the second at 10.05 seconds, and the third at 10.15 seconds, then the two consecutive values ​​are at 10 seconds and 10.15 seconds respectively.

[0080] In one embodiment, S231 includes:

[0081] S2311: Obtain the reference position and the position of the tableware used to represent the infant eating in each frame of the feeding video;

[0082] S2312: Based on the Euclidean distance between each of the reference positions and the tableware positions, obtain the distance gradients corresponding to the time dimension;

[0083] The distance gradients are used to characterize the distance values.

[0084] Specifically, feature extraction is performed on each frame of the image to obtain reference positions representing the feeding sites of infants and the positions of feeding utensils. The reference positions include, but are not limited to, the head region or the mouth region; (using the formula:) Calculate the Euclidean distance between the reference position and the feeding tool position in each frame of the image; center1 is the reference position Roi. head The center point of (x1, y1, w1, h1), center2 is the position of the eating utensils Roi. food The center point of (x2, y2, w2, h2) is distance1, which is the reference position and the Euclidean distance between the tableware position and the reference position. The distance gradient is calculated according to the time corresponding to the time sequence of each frame image. The distance gradient is a series of periodic values ​​from negative to positive. One cycle from positive to negative is recorded as one feeding. The feeding rate and feeding frequency of infants and young children are determined according to the duration of one feeding.

[0085] It should be noted that using a distance filter to filter the distance gradient and eliminate abrupt changes in the distance gradient can eliminate sudden changes in Euclidean distance caused by shaking or waving during a feeding cycle, thereby improving the accuracy of detection.

[0086] In one embodiment, S2 includes:

[0087] S24: Obtain the eating duration threshold corresponding to the difference between two consecutive eating durations;

[0088] S24: Divide each of the eating sub-videos into multiple target videos according to the eating duration threshold;

[0089] S25: Analyze each key image frame of each target video to obtain the scene factor of each target video.

[0090] In one embodiment, S25 includes:

[0091] S251: Obtain the initial scene information of the infant feeding scene and the target scene information of each of the target videos;

[0092] S252: Compare each of the target scene information with the initial scene information, and use the difference between the target scene information and the initial scene information as the scene factor.

[0093] Specifically, the duration of each feeding session is determined by the distance between the feeding utensils and the infant's mouth. When the time difference between two consecutive feeding sessions exceeds a feeding duration threshold, the infant is considered to have received significant additional stimulation, such as changes in the infant's hunger state, changes in the person feeding the infant, changes in the surrounding environment (people and scenes), or drinking water. Each feeding sub-video is divided into multiple target videos. Key image frames of each target video are analyzed to determine the scene factors. The first frame of each target video is preferred, but multiple frames can be selected at preset intervals. For example, in the initial stage of an infant's feeding scene, the parents are not present. If the parents join the feeding scene during the process, the infant eats faster. Analysis of the scene information reveals that the parents' participation in the feeding session constitutes the scene factor for that target video.

[0094] In one embodiment, please refer to Figure 4 S3 includes:

[0095] S31: Obtain the pre-defined eating weights for each eating scenario;

[0096] S32: Use the feeding weights to weight the action parameters corresponding to the feeding scenario to obtain the feeding score of the infant for this feeding.

[0097] S33: Obtain the eating information by comparing the eating score with the preset eating score table.

[0098] Specifically, as infants' feeding time increases, their feeding frequency gradually decreases. It's impossible to determine an infant's mood during feeding or the reason for their corresponding emotion (e.g., liking, disliking, or indifference) simply by observing their facial expressions and feeding actions. By weighting each feeding action with contextual factors, we can obtain feeding scores for different scenarios, thus determining the infant's feeding effectiveness. For example, if an infant uses utensils to feed themselves, their feeding frequency may decrease after a period of time. At this point, the infant's parents should intervene. In the initial stages of a feeding scenario, the infant's feeding frequency increases significantly when parents enter the scenario. This frequency then quickly decreases back to normal. Therefore, parents have a positive influence on the infant's feeding. Thus, the feeding scenario involving parents can be given bonus points in the scoring. Different scenario factors have different effects on the infant's feeding score. For example, the presence of toys or snacks in the scenario will result in a negative score. The weighting effects of each scenario factor are not listed here. The infant feeding score obtained through this method is highly accurate and can effectively monitor and regulate the infant's eating habits.

[0099] In one embodiment, S32 includes:

[0100] S321: Obtain the action information, facial expression information, confidence level corresponding to each facial expression, and total number of feedings corresponding to the action parameters of a single feeding;

[0101] S322: Based on the action information, the facial expression information, and the confidence level, obtain the basic score value for a single meal;

[0102] S323: The basic score values ​​are weighted according to the weights to obtain the eating score.

[0103] Specifically, by acquiring facial expressions of infants and young children, and combining this with the eating rate and duration of each feeding session, a confidence level for the facial expressions is determined. For example, a fast eating rate indicates that the infant is hungry; if the expression after eating is one of enjoyment, a lower confidence level is assigned to that expression, indicating that the infant likes the food and that a hunger factor exists. Conversely, a slow eating rate indicates that the infant is not hungry; if the infant still finishes eating and expresses enjoyment, a higher confidence level is assigned. This confidence level is determined by the formula: y1 = w1 * f1(grad dist1 )+w2*f2(Class emot1 The base score for a single meal is calculated as ) + b. The results show the infant's condition at each feeding, including Class emot1 The most frequent micro-expression category, grad dist1 The feeding rate is represented by 1, where 1 represents liking, 0 represents neutral, and -1 represents disliking.

[0104] The intelligent recognition method for infant feeding information in this embodiment acquires videos of infants' feeding scenes, analyzes the infants' feeding actions and the scene factors influencing these actions. Feeding actions include the feeding motions with utensils and the infant's micro-expressions; scene factors include scene changes and changes in the target object within the scene. This analysis determines the infant's feeding information corresponding to the current video and assesses the infant's satisfaction with the feeding session. By analyzing these videos, the state of the infant during each feeding can be accurately grasped, allowing for targeted adjustments to feeding times and types of food, saving resources and improving parenting effectiveness.

[0105] Example 2

[0106] This invention also provides an intelligent recognition device for infant feeding information, such as... Figure 5 As shown, it includes:

[0107] Data acquisition module: used to acquire videos of infants and young children eating.

[0108] Data analysis module: used to perform target analysis on each frame of the feeding video to obtain the feeding parameters of the infant's current feeding scene;

[0109] Data processing module: used to obtain the infant's feeding information for this feeding based on the infant's action parameters and scene factors mentioned in the feeding parameters.

[0110] The intelligent infant feeding information recognition device of this embodiment acquires feeding videos of infants during feeding, analyzes the feeding actions and scene factors influencing these actions, including feeding actions with utensils and the infant's micro-expressions, and scene factors such as scene changes and changes in target objects within the scene, thereby determining the infant's feeding information corresponding to the current feeding video and assessing the infant's satisfaction with the feeding session. By analyzing infant feeding videos, the device can accurately grasp the infant's state during each feeding, allowing for targeted adjustments to feeding times and types of food, saving resources and improving parenting effectiveness.

[0111] In one embodiment, the data acquisition module includes:

[0112] Status information unit: Acquires the first and second status information of the feeding utensils relative to the infant;

[0113] Eating video unit: The eating video is obtained based on the start time corresponding to the first state information and the end time corresponding to the second state information;

[0114] The first state information indicates that the eating utensils are in a grasped state, and the second state information indicates that the eating utensils are released from the grasped state.

[0115] In one embodiment, the data analysis module includes facial expression information as the proportion of each facial expression to the total number of facial expressions, and the facial expressions include at least one of the following: crying, happy, and calm; the action information includes at least one of the following: eating rate and eating frequency.

[0116] In one embodiment, the data analysis module includes:

[0117] Location information unit: Acquires the head position information of the infant and the position information of the eating utensils;

[0118] Euclidean distance unit: based on the head position information and the tableware position information, using the formula The Euclidean distance between the infant's head and the feeding utensils was obtained;

[0119] Distance gradient unit: Based on the Euclidean distance, outputs the distance gradient corresponding to the time dimension;

[0120] Action information unit: Outputs the action information based on the distance gradient.

[0121] In one embodiment, the data analysis module includes:

[0122] Distance information unit: acquires the distance value and distance threshold of the eating utensils relative to the infant;

[0123] Feeding frequency unit: By comparing the distance value with the distance threshold, the total number of feedings of the infant is obtained;

[0124] Action information unit: Outputs the action information based on the total number of times the food was eaten and the duration of the eating video.

[0125] In one embodiment, the data analysis module includes:

[0126] Expression acquisition unit: Acquires the number of times the infants and toddlers eat in the video information and the expressions before and after each individual feeding;

[0127] Expression Count Unit: Based on the number of meals consumed, output the total number of expressions in the expression information;

[0128] Expression information unit: Outputs the expression information based on the total number of expressions and the expressions after each individual meal.

[0129] In one embodiment, the data processing module includes:

[0130] Information collection unit: Acquires action information, facial expression information, confidence level for each facial expression, and total number of feedings corresponding to a single feeding session;

[0131] Single feeding unit: Based on the action information, facial expression information and the confidence level of a single feeding, the single feeding information of a single feeding is obtained;

[0132] Feeding information unit: Outputs the feeding information based on each individual feeding information and the total number of feedings;

[0133] The eating information includes at least one of the following: like, neutral, and dislike.

[0134] The intelligent recognition device for infant feeding information in this embodiment acquires a video stream including the target area of ​​the infant and standard sitting posture parameters; analyzes the human figure parameters and facial parameters of the infant in each frame of the video stream; compares the human figure parameters and facial parameters with the standard sitting posture parameters to determine whether the infant's sitting posture is abnormal within the corresponding time period of the video stream; by analyzing the human figure parameters and facial parameters of the infant within a time period, the situation where the infant's normal activities are misjudged as abnormal sitting posture is eliminated, thus improving the accuracy of sitting posture detection.

[0135] Example 3

[0136] Embodiment 3 of the present invention discloses an electronic device, such as Figure 6 As shown, it includes at least one processor, at least one memory, and computer program instructions stored in the memory.

[0137] Specifically, the processor 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 embodiments of the present invention.

[0138] The memory may include a large-capacity storage device for data or instructions. For example, and not limitingly, the memory may include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory may include removable or non-removable (or fixed) media. Where appropriate, the memory may be internal or external to a data processing device. In a particular embodiment, the memory is a non-volatile solid-state memory. In a particular embodiment, the memory includes a read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0139] The processor reads and executes computer program instructions stored in the memory to implement any of the intelligent recognition methods for infant feeding information in Embodiment 1 above.

[0140] In one example, the electronic device may also include a communication interface and a bus. The processor, memory, and communication interface are connected via the bus and communicate with each other.

[0141] The communication interface is mainly used to enable communication between various modules, devices, units and / or equipment in the embodiments of the present invention.

[0142] A bus, including hardware, software, or both, couples components of an electronic device together. For example, and not limitingly, a bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, a bus may include one or more buses. While specific buses are described and illustrated in embodiments of the invention, the invention contemplates any suitable bus or interconnect.

[0143] In summary, the embodiments of the present invention provide a method, apparatus, device, and storage medium for intelligent identification of infant feeding information.

[0144] The intelligent recognition method, apparatus, device, and storage medium for infant feeding information in this embodiment acquire a video stream including the target area of ​​the infant and standard sitting posture parameters; analyze the human figure parameters and facial parameters of the infant in each frame of the video stream; compare the human figure parameters and facial parameters with the standard sitting posture parameters to determine whether the infant's sitting posture is abnormal within the corresponding time period of the video stream; by analyzing the human figure parameters and facial parameters of the infant within a time period, the situation where the infant's normal activities are misjudged as abnormal sitting posture is eliminated, thereby improving the accuracy of sitting posture detection.

[0145] It should be clarified that the present invention 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 the present invention 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 the present invention.

[0146] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the required tasks. The programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligently recognizing infant feeding information, characterized in that, The method includes: S1: Obtain feeding videos of infants and toddlers; S2: Perform target analysis on each frame of the feeding video to obtain the feeding parameters of the infant's current feeding scene; S3: Based on the infant's action parameters and scene factors of the feeding parameters, obtain the infant's feeding information for this feeding, and then determine the infant's satisfaction with this feeding. The scene factors include scene changes or changes in target objects in the scene. S2 includes: S21: Obtain the frequency difference of the eating actions segmented from the eating video; S22: Divide the eating video into multiple eating sub-videos based on the frequency difference; S23: Analyze each of the feeding sub-videos to obtain the infant movement parameters that correspond one-to-one with the feeding parameters of each feeding sub-video; S24: Obtain the eating duration threshold corresponding to the difference between two consecutive eating durations; S25: Divide each of the eating sub-videos into multiple target videos according to the eating duration threshold; S26: Analyze each key image frame of each target video to obtain the scene factor of each target video.

2. The intelligent recognition method for infant feeding information according to claim 1, characterized in that, S1 includes: S11: Obtain first and second state information of the eating utensils relative to the holding part; S12: Obtain the eating video based on the start time corresponding to the first state information and the end time corresponding to the second state information; The first state information indicates that the eating utensils are in a grasped state, and the second state information indicates that the eating utensils are released from the grasped state.

3. The intelligent recognition method for infant feeding information according to claim 2, characterized in that, S12 includes: S121: Obtain the preset duration of videos depicting infants and toddlers eating; S122: Based on the state information of the eating utensils relative to the holding part, the basic video of the infant feeding scene is segmented to obtain each first video in the holding state and each second video in the non-holding state. S123: Based on the scene information of each second video and the preset duration, each second video is shortened to obtain a second target video that corresponds one-to-one with each second video; S124: The first video and the second target video are spliced ​​together according to the shooting time sequence to obtain the eating video.

4. The intelligent recognition method for infant feeding information according to claim 1, characterized in that, S23 includes: S231: Obtain the distance value and distance threshold of the eating utensils relative to the infant; S232: Compare each of the distance values ​​with the distance threshold to obtain each feeding action corresponding to the infant's action parameters of the feeding parameters; The image of one feeding action includes each frame of the video stream corresponding to two consecutive distance values ​​equal to the distance threshold.

5. The intelligent recognition method for infant feeding information according to claim 4, characterized in that, S231 includes: S2311: Obtain the reference position and the position of the tableware used to represent the infant eating in each frame of the feeding video; S2312: Based on the Euclidean distance between each of the reference positions and the tableware positions, obtain the distance gradients corresponding to the time dimension.

6. The intelligent recognition method for infant feeding information according to any one of claims 1 to 5, characterized in that, S3 includes: The feeding information is obtained by comparing the infant's feeding score for this feeding with the preset feeding score sheet.

7. An intelligent recognition device for infant feeding information, characterized in that, include: Data acquisition module: used to acquire videos of infants and young children eating. Data analysis module: used to perform target analysis on each frame of the feeding video to obtain the feeding parameters of the infant's current feeding scene; Data processing module: used to obtain the infant's feeding information for this feeding based on the infant's action parameters and scene factors, and then determine the infant's satisfaction with this feeding. The scene factors include scene changes or changes in target objects in the scene. The step of performing target analysis on each frame of the feeding video to obtain feeding parameters for the infant's current feeding scene includes: obtaining the frequency difference of feeding actions segmented from the feeding video; dividing the feeding video into multiple feeding sub-videos based on the frequency difference; analyzing each feeding sub-video to obtain infant action parameters corresponding to the feeding parameters for each feeding sub-video; obtaining a feeding duration threshold corresponding to the difference in feeding duration between two adjacent feeding sessions; dividing each feeding sub-video into multiple target videos based on the feeding duration threshold; and analyzing each key image frame of each target video to obtain scene factors for each target video.

8. An electronic device, characterized in that, include: At least one processor, at least one memory, and computer program instructions stored in the memory, which, when executed by the processor, implement the method as described in any one of claims 1-6.

9. A storage medium storing computer program instructions thereon, characterized in that, The method as described in any one of claims 1-6 is implemented when the computer program instructions are executed by the processor.

Images