Intelligent alarm method and device for animal direct broadcasting room, storage medium and program product

Abstract

The embodiment of the invention discloses an intelligent alarm method and device for an animal direct broadcasting room, a storage medium and a program product. The method comprises the following steps: acquiring an image and food delivery information of an animal direct broadcasting room; performing fine adjustment on the image by adopting a pre-training model; determining the current food shortage grade of the animal based on the fine-tuned image and the food delivery information; and obtaining alarm data of the animal direct broadcasting room, and determining whether to give an alarm according to the alarm data and the current grain shortage level. According to the method, the food supply condition of the animal direct broadcasting room can be automatically monitored, and when the animal food shortage problem occurs, the alarm can be automatically and timely given, so that the alarm accuracy is improved, and the conditions of wrong alarm and missing alarm are reduced.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to an intelligent alarm method, device, storage medium, and program product for an animal live broadcast room. Background Art

[0002] In modern society, live broadcast platforms have become an important channel for people to obtain information and entertainment. As a special branch of this field, animal live broadcast rooms have attracted a large number of viewers. However, managers of animal live broadcast rooms face many challenges, the most critical of which is to ensure the food supply of animals.

[0003] Traditional methods of warning food shortages rely mainly on manual monitoring and regular inspections, but this requires staff to monitor cameras for long periods of time or conduct regular on-site inspections, which not only consumes a lot of human resources, but may also cause human negligence or fatigue, leading to missed reports. In addition, manual detection is difficult to achieve real-time monitoring and timely warnings, which may result in food shortages not being discovered in a timely manner. Summary of the invention

[0004] In view of this, the embodiments of the present disclosure provide an intelligent alarm method, device, storage medium, and program product for an animal live broadcast room, which can automatically monitor the food supply situation in the animal live broadcast room, and when the animal is short of food, it can automatically and promptly issue an alarm, thereby improving the accuracy of the alarm.

[0005] In the first aspect, the embodiment of the present disclosure provides an intelligent alarm method for an animal live broadcast room, which adopts the following technical solutions:

[0006] Obtain images and food delivery information from animal live broadcast rooms;

[0007] Fine-tuning the image using a pre-trained model;

[0008] Determine the current food shortage level of the animal based on the fine-tuned image and food delivery information;

[0009] Obtain the alarm data from the animal live broadcast room, and determine whether to issue an alarm based on the alarm data and the current food shortage level.

[0010] Optionally, the food delivery information refers to data recorded each time the animal is fed, and the food delivery information includes at least one of food delivery status, food delivery time, type of food actually delivered, and food delivery amount; wherein the food delivery status is used to indicate whether the food delivery fails or succeeds.

[0011] Optionally, the animal live broadcast room intelligent alarm method also includes:

[0012] Build a visual language model architecture;

[0013] Collect historical images;

[0014] Setting a configuration file, wherein the configuration file is used to specify parameters and paths for fine-tuning the historical image;

[0015] Setting a pre-training script, wherein the pre-training script is used to specify parameters and paths for training the visual language model architecture;

[0016] Based on the historical images, the configuration file and the pre-training script, the visual language model architecture is trained to obtain the pre-training model.

[0017] Optionally, determining the current food shortage level of the animal based on the fine-tuned image and food delivery information includes:

[0018] Obtaining a prompt word, wherein the prompt word includes an execution requirement for the first grading model and a specified result return structure;

[0019] Using a basic model to convert the prompt words into request examples;

[0020] The request example, the fine-tuned image, and the food delivery information are input into the first grading model to obtain the current food shortage level.

[0021] Optionally, determining the current food shortage level of the animal based on the fine-tuned image and food delivery information includes:

[0022] Based on the food delivery information, obtaining the delivery amount of each type of food actually delivered on that day;

[0023] Based on the fine-tuned image and the food delivery information, obtaining the remaining amount of each type of food actually delivered on that day;

[0024] Based on the delivered amount and the remaining amount, obtaining the consumed amount of each type of food actually delivered on that day;

[0025] Determine whether the animal live broadcast room adopts a diet;

[0026] If a recipe is used, then based on the recipe, the required food types and the required consumption of the required food types in different time periods of the animal are obtained;

[0027] If no recipe is used, then based on the fine-tuned image, the required food types of the animal and the required consumption of the required food types in different time periods are obtained;

[0028] Traverse each type of food required, and record the type of food required being traversed as the first type;

[0029] Recording the type of food actually delivered that successfully matches the first type as the second type;

[0030] When there is no type of food actually delivered that successfully matches the first type, it is assumed that the second type exists, and the consumed amount of the second type is recorded as 0;

[0031] Matching the current time with multiple time periods corresponding to the first category to determine the current time period;

[0032] Obtaining a first score based on the consumed amount of the second category, the required consumption in the current time period, and the weight value of the first category;

[0033] After the traversal of the multiple required food types is completed, a second score is obtained based on the multiple first scores obtained;

[0034] The current food shortage level is determined according to the second score.

[0035] Optionally, determining the current food shortage level according to the second score includes:

[0036] Preset multiple numerical intervals, each numerical interval corresponds to a number;

[0037] If the second score is less than or equal to 0, the current food shortage level is determined to be 0;

[0038] If the second score is greater than 0, the current food shortage level is equal to the number corresponding to the numerical range to which the second score belongs.

[0039] Optionally, determining whether to issue an alarm according to the alarm data and the current food shortage level includes:

[0040] The alarm data includes at least one of the alarm upper limit of the animal live broadcast room, the alarm time interval, and the alarm mode corresponding to each food shortage level;

[0041] If the current food shortage level is 0, no alarm is required;

[0042] If the current food shortage level is greater than 0, the number of alarms in the animal live broadcast room on that day is counted;

[0043] If the number of alarms is 1, an alarm is issued according to the alarm method corresponding to the current food shortage level;

[0044] If the number of alarms is greater than 1, the time interval from the last alarm is obtained;

[0045] If the number of alarms does not exceed the alarm upper limit, and the time interval reaches the alarm time interval threshold, an alarm is issued in accordance with the alarm method corresponding to the current food shortage level.

[0046] In a second aspect, the disclosed embodiment also provides an animal live broadcast room intelligent alarm system, which adopts the following technical solutions:

[0047] The acquisition module is used to obtain images and food delivery information of the animal live broadcast room;

[0048] A fine-tuning module, used to fine-tune the image using a pre-trained model;

[0049] A determination module, for determining the current food shortage level of the animal based on the fine-tuned image and food delivery information;

[0050] The alarm module is used to obtain the alarm data of the animal live broadcast room and determine whether to issue an alarm based on the alarm data and the current food shortage level.

[0051] In a third aspect, the embodiments of the present disclosure further provide a computer device, which adopts the following technical solution:

[0052] The computer device comprises:

[0053] at least one processor; and,

[0054] a memory communicatively connected to the at least one processor; wherein,

[0055] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any of the above-mentioned intelligent alarm methods for animal live broadcast rooms.

[0056] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute any of the above-mentioned intelligent alarm methods for animal live broadcast rooms.

[0057] In a fifth aspect, an embodiment of the present disclosure further provides a computer program product, including a computer program / instruction, which implements the steps of any of the above methods when executed by a processor.

[0058] The animal live broadcast room intelligent alarm method provided by the disclosed embodiment ensures that the problem of animal food shortage can be discovered in time by acquiring the image and food delivery information of the animal live broadcast room in real time. Fine-tuning the image through the pre-training model facilitates the system to recognize the image more accurately, greatly improving the accuracy and efficiency of monitoring. Based on the fine-tuned image and food delivery information, the current food shortage level of the animal can be accurately determined, and combined with the alarm data of the animal live broadcast room, it is determined whether to issue an alarm, which means that the system can issue an early warning in time when the food shortage problem occurs, so that the staff can take measures quickly to prevent the animals from being harmed due to food shortage, which helps the staff to reasonably adjust the food delivery and maintain the health of the animals. Moreover, this kind of monitoring through automation and intelligence can reduce the workload of manual monitoring, reduce labor costs, improve work efficiency, and reduce false alarms or missed reports caused by human negligence.

[0059] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the following preferred embodiments are specifically cited and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0061] Figure 1 A flow chart of an intelligent alarm method for an animal live broadcast room provided in an embodiment of the present disclosure;

[0062] Figure 2 A schematic diagram of a process for constructing a pre-training model provided in an embodiment of the present disclosure;

[0063] Figure 3 A schematic diagram of a flow chart of a method for determining a current food shortage level provided in an embodiment of the present disclosure;

[0064] Figure 4 Another schematic diagram of a process for determining a current food shortage level provided by an embodiment of the present disclosure;

[0065] Figure 5 A principle block diagram of the animal live broadcast intelligent alarm system provided by the embodiment of the present disclosure;

[0066] Figure 6 A schematic diagram of the structure of a computer device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0067] The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0068] It should be clear that the following embodiments of the present disclosure are described by specific specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other in the absence of conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.

[0069] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described herein may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this device and / or practice this method.

[0070] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present disclosure. The drawings only show components related to the present disclosure rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0071] Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the aspects described may be practiced without these specific details.

[0072] Reference Figure 1 The present disclosure provides an intelligent alarm method for an animal live broadcast room, comprising the following steps:

[0073] S1: Obtain images and food delivery information from the animal live broadcast room;

[0074] S2: Fine-tune the image using the pre-trained model;

[0075] S3: Determine the current food shortage level of the animal based on the fine-tuned image and food delivery information;

[0076] S4: Get the alarm data of the animal live broadcast room, according to the alarm data and the current food shortage level YY+242180P

[0077] Determine whether to issue an alarm.

[0078] Among them, the animal live broadcast room is a new form of entertainment and education that uses the Internet platform to display the life, behavior, interaction and other content of animals through real-time live broadcast. It can be used in zoos, science and technology museums, farms, breeding farms and other scenes where it is difficult to monitor in real time manually.

[0079] The animal live broadcast room intelligent alarm method disclosed in the present invention can ensure that the problem of animal food shortage can be discovered in time by obtaining the image and food delivery information of the animal live broadcast room in real time. The image is fine-tuned by the pre-training model, which facilitates the system to recognize the image more accurately, greatly improving the accuracy and efficiency of monitoring. Based on the fine-tuned image and food delivery information, the current food shortage level of the animal can be accurately determined, and combined with the alarm data of the animal live broadcast room, it is determined whether to issue an alarm, which means that the system can issue an early warning in time when the food shortage problem occurs, so that the staff can take quick measures to prevent the animals from being harmed due to food shortage, which helps the staff to reasonably adjust the food delivery and maintain the health of the animals. Moreover, this kind of monitoring through automation and intelligence can reduce the workload of manual monitoring, reduce labor costs, improve work efficiency, and reduce false alarms or missed reports caused by human negligence.

[0080] In summary, this method helps improve the quality of life of animals by ensuring that the food supply of animals is paid attention to and managed in a timely manner through automatic alarms. And by judging the current level of food shortage, it can help staff make more scientific and reasonable decisions, so as to better maintain and care for animals. This method has important application value in the fields of animal protection, pet live broadcasting, zoo management, etc., and can effectively improve the level of animal welfare and safety management.

[0081] In S1, the food delivery information corresponding to several animal live broadcast rooms is obtained. The food delivery information refers to the data recorded each time the animal is fed. The food delivery information includes at least one of the food delivery status, food delivery time, the type of food actually delivered, and the amount of food delivered. The animals can be fed with a feeding machine, and the food delivery information of the feeding machine can be obtained by connecting the feeding machine through an interface. Manual feeding can also be used, and the food delivery information is obtained by manually recording the food delivery information or by analyzing the video stream of the animal live broadcast room. Among them, the food delivery status is used to indicate whether the food delivery fails or succeeds, and the food delivery amount refers to the weight of the food delivered each time.

[0082] A specific implementation method of "obtaining images from animal live broadcast rooms" is: obtain the identifiers of several animal live broadcast rooms, query the corresponding live broadcast time according to the identifiers, and if the corresponding live broadcast time is found, use multiple cameras to collect the video stream of the animal live broadcast room from multiple directions (that is, multiple angles) at regular intervals during the live broadcast time, and use the ffmpeg tool (a multimedia processing tool) to capture key frames from the video stream and generate a set of images in jepg format. For example, use 4 cameras, collect the video stream of the animal live broadcast room from 4 different directions every one hour, the collection time is 2 minutes, and capture several images from the video stream.

[0083] The live broadcast time of some animal live broadcast rooms is pre-set. Through the above method, images of the animal live broadcast rooms can be collected in time according to the live broadcast time, providing effective data for subsequent alarm monitoring processes.

[0084] If the corresponding live broadcast time is not found, the live broadcast status of the animal live broadcast room is detected according to the identification of the animal live broadcast room. The live broadcast status is used to indicate whether it is being broadcast. If it is detected that the animal live broadcast room is being broadcast, Kafka (Apache Kafka, a distributed stream processing platform) is used to collect the corresponding food delivery information in real time. According to the food delivery status and food delivery time included in the food delivery information, the number of consecutive food delivery failures on the day is determined. If the number of consecutive times is greater than the preset number of failures, multiple cameras are triggered to collect the video stream of the animal live broadcast room from multiple directions (that is, multiple angles), and the ffmpeg tool is used to capture key frames from the video stream to generate a set of images in jepg format. If the number of consecutive times is not greater than the preset number of failures, there is no need to obtain images from the animal live broadcast room.

[0085] The live broadcast time of some animal live broadcast rooms is irregular and cannot be known in advance. Through the above method, it is possible to timely detect whether the animal live broadcast room is broadcasting live, and detect the corresponding food delivery information during the live broadcast stage to determine the consecutive number of food delivery failures on the day, and determine whether to trigger the subsequent process based on the consecutive number of failures.

[0086] In S2, the pre-trained model is used to fine-tune the image. The fine-tuning includes denoising and enhancement. Denoising refers to the use of image processing technology to remove noise in the image and improve the image quality; enhancement refers to the use of image enhancement technology to highlight the characteristics of the animal, such as body shape, color, etc., in order to better analyze the status.

[0087] For pre-trained models, refer to Figure 2 The flowchart of the pre-training model construction method shown includes the following steps:

[0088] S21: Build a visual language model architecture;

[0089] S22: Collect historical images;

[0090] S23: Set up configuration files and pre-training scripts;

[0091] S24: Based on historical images, configuration files, and pre-trained scripts, train the visual language model architecture to obtain a pre-trained model.

[0092] In S21, InternVL-Chat-V1.2-Plus is adopted as the visual language model architecture.

[0093] In S22, a large number of historical images of animal live broadcasts are collected and annotated for subsequent use as training samples for the visual language model architecture.

[0094] In S23, the configuration file is a pre-defined SFT (supervised fine-tuning) configuration file, which is used to specify the parameters and paths for fine-tuning historical images. An example of the configuration file YY+242180P is as follows:

[0095] {

[0096] "sharegpt4v_instruct_gpt4-vision_cap100k":{

[0097] "root":"playground / data / ",

[0098] "annotation":

[0099] "playground / sharegpt4v_instruct_gpt4-vision_cap100k.jsonl",

[0100] "data_augment":true,

[0101] "repeat_time":1,

[0102] "length":102025

[0103] }

[0104] }

[0105] Explanation of the above parameters:

[0106] ①sharegpt4v_instruct_gpt4-vision_cap100k: The identifier or name of the dataset (a collection of historical images), usually used to refer to this specific dataset in the program;

[0107] ②root: the root directory of the image data after denoising;

[0108] ③annotation: The location of the annotation file, which usually contains metadata or other relevant information about the historical image, such as object labels in the image;

[0109] ④data_augment: controls whether to perform augmentation processing on historical images. Augmentation processing is a technique to improve the generalization ability of machine learning models by applying a series of transformations to increase the diversity of training samples. Here, true means augmentation processing;

[0110] ⑤repeat_time: The number of times the dataset is reused during training. If set to a value greater than 1, the dataset will be used multiple times to increase the number of training samples.

[0111] ⑥length: indicates the number of data items in the dataset, that is, the number of historical images.

[0112] The pre-training script is a pre-configured LoRA (Low-Rank Adaptation) adapter pre-training script. The pre-training script is used to specify the parameters and paths for training the visual language model architecture. The configuration process uses the LoRA adapter technology. An example of a pre-training script is as follows:

[0113] CUDA_VISIBLE_DEVICES=0,1sh shell / hermes2_yi34b / internvl_chat_v1_2_hermes2_yi34b_448_res_finetun e_continue_lora.sh

[0114] In S24, the configuration file and the pre-training script are deployed and executed, the visual language model architecture is trained using historical images, the visual language model architecture is adjusted, and a pre-training model is obtained.

[0115] In S3, refer to Figure 3 A flowchart of a method for determining a current food shortage level is shown, wherein a specific implementation method of “determining a current food shortage level of an animal based on a fine-tuned image and food delivery information” includes:

[0116] S301: Obtain prompt words;

[0117] S302: converting prompt words into request examples using a basic model;

[0118] S303: Input the request example, the fine-tuned image, and the food delivery information into the first grading model to obtain the current food shortage level.

[0119] In S301, the prompt words include execution requirements for the first classification model and a specified result return structure. The execution requirement refers to requiring the first classification model to obtain a specified result based on the fine-tuned image and food delivery information, and the specified result return structure refers to the structure of the specified result required to be output by the first classification model.

[0120] YY+242180P

[0121] In S302, the basic model uses InternViT-6B-448px-V1-5, which is a natural language model. It performs natural language processing on the prompt words, performs natural language understanding, and generates structured text that is easy for the first classification model to understand. The structured text is called a request example.

[0122] Specifically, the InternGL-InternVL model consisting of the pre-trained model and the basic model is deployed locally. The algorithm service is deployed based on the pre-trained model and the basic model to provide an interface to the outside world. The status of the animal live broadcast room and food delivery information are obtained through the interface, such as whether it is in the live broadcast time period, whether the food delivery is successful, and other parameters. These back-end parameters are integrated and transmitted to the corresponding model, which processes them to obtain fine-tuned images and request examples.

[0123] For example, if the prompt word requires the first classification model to output the result of whether there is food in the animal's food bowl based on the image of a certain animal live broadcast room, the request example is as follows:

[0124] {"prompt":"Please judge whether there is food in the feeding equipment according to the images collected by the cameras in different directions in the SXC10c2 live broadcast room. Please return in the following json format: {\"have_food\":\"Whether there is food\", / / Return yes or no\"reason\":\"Reason and basis for judgment\"}","pictures":["http:pics1":"http:pics2"...]}

[0125] For another example, if the prompt word is to require the first grading model to determine the current food shortage level based on the status and image of the animal live broadcast room and the food delivery information of the feeder, the request example is as follows:

[0126] system_prompt: You are an excellent live broadcast room dispatcher, usually responsible for evaluating the current food shortage level based on the relevant information of the animal live broadcast room.

[0127] user_prompt:

[0128] The current status of the animal live broadcast room {stusio_id} is as follows:

[0129] YY+242180P

[0130] Whether it is during the live broadcast time period: {online}

[0131] Whether feeding status: {can_feed}

[0132] Please judge the current food shortage level of the current animal live broadcast room based on the status of the animal live broadcast room, the images sent back by multiple cameras in the animal live broadcast room, the food delivery information sent back by the corresponding feeder, and your experience. Please return it in the specified json format:

[0133] {"stusio_id":"Animal Live Room ID","warning_level":"Current Food Shortage Level"\\int type, the current food shortage level is 0-10, as the number increases, the current food shortage level is higher, the number 0 means no food shortage, the number 1-10 means food shortage}

[0134] Example results:

[0135] {"stusio_id":"studio123456","warning_level":5}

[0136] This result example indicates that the animal live broadcast room ID, that is, the animal live broadcast room identified as studio123456, currently has a food shortage level of 5.

[0137] In S303, the first classification model can select ChatGPT (Chat Generative Pre-trained Transformer, an artificial intelligence model based on the Transformer architecture), input the fine-tuned image, request example and food delivery information into the first classification model, and the first classification model outputs the current food shortage level.

[0138] Reference Figure 4 Another flowchart of the method for determining the current food shortage level is shown. Another specific implementation of “determining the current food shortage level of the animal based on the fine-tuned image and food delivery information” includes:

[0139] S311: Based on the food delivery information, the delivery amount of each type of food actually delivered on that day is obtained;

[0140] S312: Based on the fine-tuned image and food delivery information, obtain the actual YY+242180P for each day

[0141] The amount of food left over;

[0142] S313: Based on the delivered amount and the remaining amount, the consumed amount of each type of food actually delivered on the day is obtained;

[0143] S314: Determine whether the animal live broadcast room uses a recipe; if so, execute S315; if not, execute S316;

[0144] S315: Obtain the required food types and the required consumption of the required food types in different time periods of the animal based on the recipe, and execute S317;

[0145] S316: based on the fine-tuned image, obtaining the required food types of the animal and the required consumption of the required food types in different time periods, and executing S317;

[0146] S317: traverse each required food type, and record the required food type being traversed as the first type;

[0147] S318: Match the first category with the multiple categories of food actually delivered; if the match is successful, execute S319; if the match fails, execute S320;

[0148] S319: Record the type of food actually delivered that successfully matches the first type as the second type, and execute S321;

[0149] S320: Assume that the second category exists, and record the consumed amount of the second category as 0;

[0150] S321: Match the current time with multiple time periods corresponding to the first category to determine the current time period;

[0151] S322: Obtain a first score based on the consumed amount of the second category, the required consumption in the current time period, and the weight value of the first category;

[0152] S323: after the traversal of the multiple required food types is completed, a second score is obtained based on the multiple first scores obtained;

[0153] YY+242180P

[0154] S324: Determine the current food shortage level according to the second score.

[0155] In S311, based on the food delivery information including the food delivery status, food delivery time, the type of food actually delivered and the food delivery amount, the delivered amount of each type of food actually delivered on that day can be calculated.

[0156] In S312, the fine-tuned image and food delivery information are input into a preset object detection model, and the object detection model analyzes the fine-tuned image in combination with the type of food actually delivered contained in the food delivery information, and identifies and quantifies the remaining amount of food in the feeding utensil. The object detection model is trained using a deep learning model (YOLO, SSD or Mask R-CNN).

[0157] In S313, the amount of food actually put in on that day is subtracted from the corresponding remaining amount to obtain the amount of food actually put in on that day. Specifically, the calculation formula for the amount of food actually put in on that day is as follows:

[0158] Ci=Ai-Ri;

[0159] Among them, Ai is the amount of each type of food i actually put in that day; Ri is the remaining amount of each type of food i actually put in that day; i is the type number of the food actually put in.

[0160] In S314 and S315, the recipe includes several types of food required by the animal live broadcast room for the animals being broadcast live and the required consumption of the several types of food in different time periods.

[0161] In S316, the fine-tuned image is input into a preset image recognition model. The image recognition model determines the type, physiological characteristics (female or male), age stage and body size of the animal by analyzing the fine-tuned image, and predicts the nutritional needs of the animal through these characteristics, and further predicts the required food types and the required consumption of the required food types in different time periods. The image recognition model adopts a convolutional neural network (CNN) YY+242180P

[0162] Trained.

[0163] In S315 and S316, the required consumption of each required food j in time period t is set as Dj,t, where j is the type number of the required food.

[0164] In S317-S322, the required food type being traversed is recorded as the first type. By matching the first type with multiple types of food actually delivered, it can be determined whether the corresponding food is actually delivered. If so, the second type can be determined. If not, it is assumed that the second type exists, but the consumed amount of the second type is recorded as 0. Then, by matching the current time with multiple time periods corresponding to the first type, the time period that successfully matches the current time is recorded as the current time period, and the required consumption corresponding to the consumed amount of the second type is determined. The required consumption of the current time period is subtracted from the consumed amount of the second type to obtain the difference, which may be a positive number or a negative number. The difference is multiplied by the weight value of the first type to obtain the first score.

[0165] Specifically, set the set of multiple types of food actually delivered as I, i∈I, and the set of multiple types of required food as J, j∈J. The required food type being traversed is recorded as the first type j'. If there is a matching type of food actually delivered, it is recorded as the second type i', and its consumed amount is Ci'; if there is no matching type of food actually delivered, it is assumed that the second type i' exists, and its consumed amount is Ci'=0. The current time period is recorded as t', the required consumption of the first type j' in the current time period t' is recorded as Dj',t', and the first score of the first type j' is recorded as Fj', Fj'=Wj'(Dj',t'-Ci'), where Wj' is the weight value of the first type j'.

[0166] It is worth noting that animals have higher physiological requirements for certain foods, but animals can take no or small amounts of other foods. Therefore, the weight value of the first category is a positive number, and the higher the importance of the first category to the animal, the higher its weight value. For example, for carnivorous animals, the weight value of meat food is higher, and the weight values ​​of other non-essential foods are lower.

[0167] In S323, after the traversal of multiple required food types is completed, multiple first scores are obtained, and the multiple first scores are added together to obtain a second score. Specifically, the second score is F, F=∑ j′∈J Fj′.

[0168] In S324, multiple numerical intervals are preset, each numerical interval corresponds to a number. If the second score is less than or equal to 0, the current food shortage level is determined to be 0; if the second score is greater than 0, the current food shortage level is equal to the number corresponding to the numerical interval to which the second score belongs.

[0169] Since some animal studios will feed animals according to a preset recipe, the above two different situations of whether the animal studio adopts a recipe are used to obtain the types of food required by the animals and the required consumption of the required types of food in different time periods, and then calculate the first score of each required type of food, and then obtain the comprehensive score, that is, the second score. The second score can be used to evaluate the current food shortage level of the animal studio.

[0170] In S4, Redis is deployed. Redis is an open source memory database and cache system, which is used to store the alarm data of each animal live broadcast room. According to the identifier of the animal live broadcast room, the alarm data of the current animal live broadcast room is obtained. The alarm data includes the alarm upper limit of the animal live broadcast room, the alarm time interval threshold, and at least one of the alarm methods corresponding to each food shortage level.

[0171] If the current food shortage level is 0, it means that the animals in the animal room have sufficient food and no alarm is needed; if the current food shortage level is greater than 0, the number of alarms in the animal room on that day is counted. If the number of alarms is 1, an alarm is issued according to the alarm method corresponding to the current food shortage level; if the number of alarms is greater than 1, the time interval from the last alarm is obtained; if the number of alarms does not exceed the alarm upper limit, and the time interval reaches the alarm time interval threshold, an alarm is issued according to the alarm method corresponding to the current food shortage level. If the number of alarms exceeds the alarm upper limit, or the time interval does not reach the alarm time interval threshold, no alarm is issued.

[0172] Furthermore, if the current food shortage level is 1-5, it corresponds to a basic alarm, and the alarm method is to notify the manager of the animal live broadcast room by SMS or WeChat; if the current food shortage level is 6-10, it corresponds to an upgraded alarm, and the alarm method is to notify the manager of the animal live broadcast room by phone. When two or more food shortage levels are obtained within the preset time range, the highest of these food shortage levels is used as the current food shortage level, and the remaining food shortage levels are ignored. Regardless of the alarm method, the current food shortage level, comprehensive score, the amount of each type of food actually put in that day, the remaining amount and the consumed amount, as well as the types of food required by the animals and the required consumption of the required food types in different time periods can also be sent to the manager so that the manager can take measures based on this information.

[0173] Based on the above alarm method, the number of alarms can be balanced, which can not only play the role of the alarm, but also avoid excessive alarm frequency, so as not to cause trouble to the relevant personnel in the animal live broadcast room and affect the practicality and efficiency of the alarm system.

[0174] Reference Figure 5 The present disclosure provides an animal live broadcast room intelligent alarm system, comprising:

[0175] An acquisition module 101 is used to acquire images and food delivery information of the animal live broadcast room;

[0176] A fine-tuning module 102, used to fine-tune the image using a pre-trained model;

[0177] A determination module 103, for determining the current food shortage level of the animal based on the fine-tuned image and food delivery information;

[0178] The alarm module 104 is used to obtain the alarm data of the animal live broadcast room and determine whether to issue an alarm based on the alarm data and the current food shortage level.

[0179] The various variations and specific examples of the above-mentioned intelligent alarm method for animal live broadcast rooms are also applicable to the intelligent alarm system for animal live broadcast rooms provided in the present invention. Through the above-mentioned detailed description of the intelligent alarm method for animal live broadcast rooms, those skilled in the art can clearly know the implementation method of the intelligent alarm system for animal live broadcast rooms. For the sake of brevity of the specification, it will not be described in detail here.

[0180] The computer device according to the embodiment of the present disclosure includes a memory and a processor. The memory is used to store non-temporary computer-readable instructions. Specifically, the memory may include one or more computer program products, and the computer program product may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may, for example, include a read-only memory (ROM), a hard disk, a flash memory, etc.

[0181] The processor may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the computer device to perform desired functions. In one embodiment of the present disclosure, the processor is used to run the computer-readable instructions stored in the memory, so that the computer device executes all or part of the steps of the animal live broadcast intelligent alarm method of each embodiment of the present disclosure.

[0182] Those skilled in the art should be able to understand that in order to solve the technical problem of how to obtain a good user experience, the present embodiment may also include well-known structures such as a communication bus and an interface, and these well-known structures should also be included in the protection scope of the present disclosure.

[0183] like Figure 6 A schematic diagram of the structure of a computer device provided in an embodiment of the present disclosure is shown, which is a schematic diagram of the structure of a computer device suitable for implementing the embodiment of the present disclosure. Figure 6The computer device shown is only an example and should not be construed as limiting the functionality and use of the embodiments of the present disclosure.

[0184] No limitations are imposed on the scope.

[0185] like Figure 6 As shown, the computer device may include a processor (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) or a program loaded from a storage device into a random access memory (RAM). In the RAM, various programs and data required for the operation of the computer device are also stored. The processor, ROM, and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0186] Typically, the following devices can be connected to the I / O interface: input devices such as sensors or visual information acquisition devices; output devices such as display screens; storage devices such as tapes, hard disks, etc.; and communication devices. The communication device can allow the computer device to communicate with other devices (such as edge computing devices) wirelessly or by wire to exchange data. Figure 6 A computer device having various devices is shown, but it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed instead.

[0187] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by the processor, all or part of the steps of the animal live studio intelligent alarm method of the embodiment of the present disclosure are executed.

[0188] For detailed description of this embodiment, reference may be made to the corresponding descriptions in the aforementioned embodiments, which will not be repeated here.

[0189] According to the computer-readable storage medium of the embodiment of the present disclosure, non-transitory computer-readable instructions are stored thereon. When the non-transitory computer-readable instructions are executed by the processor, all or part of the steps of the animal live broadcast room intelligent alarm method of each embodiment of the present disclosure are executed.

[0190] The above-mentioned computer-readable storage media include, but are not limited to: optical storage media (e.g., CD-ROM and DVD), magneto-optical storage media (e.g., MO), magnetic storage media (e.g., magnetic tape or mobile hard disk), media with built-in rewritable non-volatile memory (e.g., memory card) and media with built-in ROM (e.g., ROM box).

[0191] For detailed description of this embodiment, reference may be made to the corresponding descriptions in the aforementioned embodiments, which will not be repeated here.

[0192] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, and are not limitations. The above details do not limit the present disclosure to the necessity of adopting the above specific details to be implemented.

[0193] In the present disclosure, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. The block diagrams of the devices, devices, equipment, and systems involved in the present disclosure are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagram. As will be appreciated by those skilled in the art, these devices, devices, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open words, referring to "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the words "and / or" and can be used interchangeably with them, unless the context clearly indicates otherwise. The words "such as" used here refer to the phrase "such as but not limited to", and can be used interchangeably with them.

[0194] Additionally, as used herein, "or" used in a list of items beginning with "at least one" indicates a separate list, so that, for example, a list of "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not mean that the example described is preferred or better than other examples.

[0195] It should also be noted that in the system and method of the present disclosure, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present disclosure.

[0196] Various changes, substitutions, and modifications of the techniques described herein may be made without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of the present disclosure is not limited to the specific aspects of the processes, machines, manufactures, compositions of events, means, methods, and actions described above. Currently existing or later to be developed processes, machines, manufactures, compositions of events, means, methods, or actions that perform substantially the same functions or achieve substantially the same results as the corresponding aspects described herein may be utilized. Thus, the appended claims include such processes, machines, manufactures, compositions of events, means, methods, or actions within their scope.

[0197] YY+242180P

[0198] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0199] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present disclosure to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. An intelligent alarm method for animal live broadcast room, characterized in that: include: Obtain images and food delivery information from animal live broadcast rooms; Fine-tuning the image using a pre-trained model; Determine the current food shortage level of the animal based on the fine-tuned image and food delivery information; Obtain the alarm data from the animal live broadcast room, and determine whether to issue an alarm based on the alarm data and the current food shortage level.

2. The animal live broadcast intelligent alarm method according to claim 1, characterized in that: The food delivery information refers to the data recorded each time the animal is fed, and the food delivery information includes at least one of the food delivery status, food delivery time, the type of food actually delivered, and the amount of food delivered; wherein the food delivery status is used to indicate whether the food delivery fails or succeeds.

3. The animal live broadcast intelligent alarm method according to claim 1, characterized in that: Also includes: Build a visual language model architecture; Collect historical images; Setting a configuration file, wherein the configuration file is used to specify parameters and paths for fine-tuning the historical image; Setting a pre-training script, wherein the pre-training script is used to specify parameters and paths for training the visual language model architecture; Based on the historical images, the configuration file and the pre-training script, the visual language model architecture is trained to obtain the pre-training model.

4. The animal live broadcast intelligent alarm method according to claim 2, characterized in that: The method of determining the current food shortage level of the animal based on the fine-tuned image and food delivery information includes: Obtaining a prompt word, wherein the prompt word includes an execution requirement for the first grading model and a specified result return structure; Using a basic model to convert the prompt words into request examples; The request example, the fine-tuned image, and the food delivery information are input into the first grading model to obtain the current food shortage level.

5. The animal live broadcast intelligent alarm method according to claim 2, characterized in that: The method of determining the current food shortage level of the animal based on the fine-tuned image and food delivery information includes: Based on the food delivery information, obtaining the delivery amount of each type of food actually delivered on that day; Based on the fine-tuned image and the food delivery information, obtaining the remaining amount of each type of food actually delivered on that day; Based on the delivered amount and the remaining amount, obtaining the consumed amount of each type of food actually delivered on that day; Determine whether the animal live broadcast room adopts a diet; If a recipe is used, then based on the recipe, the required food types and the required consumption of the required food types in different time periods of the animal are obtained; If no recipe is used, then based on the fine-tuned image, the required food types of the animal and the required consumption of the required food types in different time periods are obtained; Traverse each type of food required, and record the type of food required being traversed as the first type; Recording the type of food actually delivered that successfully matches the first type as the second type; When there is no type of food actually delivered that successfully matches the first type, it is assumed that the second type exists, and the consumed amount of the second type is recorded as 0; Matching the current time with multiple time periods corresponding to the first category to determine the current time period; Obtaining a first score based on the consumed amount of the second category, the required consumption in the current time period, and the weight value of the first category; After the traversal of the multiple required food types is completed, a second score is obtained based on the multiple first scores obtained; The current food shortage level is determined according to the second score.

6. The animal live broadcast intelligent alarm method according to claim 5, characterized in that: Determining the current food shortage level according to the second score includes: Preset multiple numerical intervals, each numerical interval corresponds to a number; If the second score is less than or equal to 0, the current food shortage level is determined to be 0; If the second score is greater than 0, the current food shortage level is equal to the number corresponding to the numerical range to which the second score belongs.

7. The animal live broadcast intelligent alarm method according to claim 1, characterized in that: The determining whether to issue an alarm according to the alarm data and the current food shortage level includes: The alarm data includes at least one of the alarm upper limit of the animal live broadcast room, the alarm time interval, and the alarm mode corresponding to each food shortage level; YY+242180P If the current food shortage level is 0, no alarm is required; If the current food shortage level is greater than 0, the number of alarms in the animal live broadcast room on that day is counted; If the number of alarms is 1, an alarm is issued according to the alarm method corresponding to the current food shortage level; If the number of alarms is greater than 1, the time interval from the last alarm is obtained; If the number of alarms does not exceed the alarm upper limit, and the time interval reaches the alarm time interval threshold, an alarm is issued in accordance with the alarm method corresponding to the current food shortage level.

8. A computer device, characterized in that: The computer device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the animal live broadcast room intelligent alarm method described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, which are used to enable a computer to execute the animal live broadcast room intelligent alarm method described in any one of claims 1-7.

10. A computer program product comprising computer instructions, characterized in that When the computer instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images