Individual analysis device and method using a dual processing platform with artificial intelligence, and smart farm monitoring system using the same.

Abstract

This invention provides an individual analysis device and method using an artificial intelligence-based dual-processing platform, as well as a smart farm monitoring system using the same. [Solution] The method in the smart farm monitoring system 10 includes the steps of: collecting target data, including video data and sensing data linked to the individual being monitored, using an edge device 200; performing an initial analysis using the edge device, inputting the target data into a first model of a pre-trained artificial intelligence platform to output first analysis data linked to the individual's behavioral patterns; and transmitting the first analysis data and target data to the main server when the first analysis data satisfies pre-set detailed analysis conditions.

Description

Technical Field

[0001] This application relates to an individual analysis apparatus and method for a dual processing platform using artificial intelligence, and a smart farm monitoring system using the same. For example, this application relates to a dual processing system of an artificial intelligence platform for monitoring and analyzing livestock behavior in a smart farm environment.

Background Art

[0002] While the integration of agriculture and ICT technologies is accelerating, the paradigm shift from traditional agricultural and livestock production methods to smart farms has been in full swing. In particular, major advanced countries in smart farms around the world are seriously attempting to connect ICT technologies such as the Internet of Things, nanotechnology, big data, cloud, robots, and drones to the agricultural and livestock industries. In some countries, while utilizing various systems for calculating yields, diagnosing pests and diseases, measuring soil moisture, measuring indicator status, diagnosing harvest times, and monitoring farming conditions, they are in the process of realizing Farm Intelligence.

[0003] On the other hand, in the case of smart farms in the dairy farming field, due to the continuous development of breeding technology, the size of ruminants has increased, and the breeding environment has also changed from existing simple pillar facilities to a large dry space with compost laid. In addition, due to the integration of industries, large-scale breeding in the livestock industry has become widespread. The ultimate goal in modern large-scale ruminant breeding systems is to improve the meat quality of beef cattle, the milk production and quality of dairy cows, and at the same time ensure the health and calmness of livestock. To achieve such a goal, research on the physiological behavior of ruminants, environmental optimization, and management are essential.

[0004] Over the past few decades, research into ruminant behavior and the environmental impact on ruminants has become an essential part of animal husbandry and veterinary medicine. To achieve the aforementioned goals, it has become even more important to consider ruminant behavior and their living environment in addition to conventional physiological research. Ruminant behavior includes responses to other ruminants and other objects, and this is related to their overall responses and adaptations to diverse internal and external conditions. Systematic observation of ruminant behavior has revealed that they primarily exhibit feeding, resting, social, group, estrus, reproductive, and offspring-related behaviors. Considering these ruminant behaviors plays a crucial role in improving productivity, one of the aforementioned goals. Furthermore, various indicators of the living environment, such as cleanliness, temperature and humidity, and feeding density per unit area, have emerged as important factors to consider in livestock productivity. For example, the design of feed and drinking water areas affects the health and physiological state of ruminants.

[0005] Furthermore, while much research has been conducted on the physiological state of ruminants and the regulation of their breeding environment, and regulations have been applied to the breeding environment of ruminants, as the scale of breeding expands, more manpower is required to observe all ruminants. Although sensor products play a somewhat supplementary role, there is the inconvenience of having to attach various sensors to each animal, and this may increase the stress on ruminants. Moreover, as the scale of breeding expands, the economic burden on farmers also increases, which is a limitation. [Prior art documents] [Patent Documents]

[0006] [Patent Document 1] Korean Registered Patent Publication No. 10-1329022 [Overview of the project] [Problems that the invention aims to solve]

[0007] This application aims to solve the problems of the prior art described above and to provide an individual analysis device and method using artificial intelligence with a dual processing base that can efficiently monitor and analyze the behavior of individuals in a smart farm environment, and a smart farm monitoring system using the same. However, the technical problems that the embodiments of this application aim to solve are not limited to those described above, and other technical problems may exist. [Means for solving the problem]

[0008] As a technical means for achieving the above technical challenges, an individual analysis method for a dual processing platform using artificial intelligence according to one embodiment of the present invention may include the steps of: collecting target data, including video data and sensing data, linked to an individual being monitored, using an edge device; performing an initial analysis using the edge device by inputting the target data into a pre-trained first model of an artificial intelligence platform and outputting first analysis data linked to the individual's behavioral pattern; and transmitting the first analysis data and the target data to a main server when the first analysis data satisfies pre-set detailed analysis conditions.

[0009] Furthermore, the first analysis data may include information on the presence or absence of the individual and information on its location, as reflected in the video data. Furthermore, the first analysis data may include first classification information for the behavioral type of the individual. Furthermore, the detailed analysis conditions may include conditions that the first classification information corresponds to a pre-defined abnormal or significant act. Furthermore, the second analysis data may include a second classification of the behavioral types of the individuals. Furthermore, the number of classes in the second classification information can be larger than the number of classes in the first classification information. Furthermore, the first model can be a lightweight model constructed taking into account the resource information of the edge device. Furthermore, the second model may include multiple models pre-built based on each of several different network structures. Furthermore, the second analysis data can be determined by considering the degree of agreement between the output data of each of the multiple models.

[0010] On the other hand, the individual analysis method for a dual processing platform using artificial intelligence according to one embodiment of the present invention includes the steps of: receiving the first analysis data and the target data from the edge device as a result of an initial analysis performed using an edge device, in which target data including video data and sensing data linked to the individual being monitored is input into a first model of an artificial intelligence platform that has been trained in advance, and outputting first analysis data linked to the behavioral pattern of the individual, and the first analysis data satisfies pre-set detailed analysis conditions; and performing a detailed analysis in which at least one of the first analysis data and the target data is input into a second model of an artificial intelligence platform that has been trained in advance, in which second analysis data linked to the behavioral pattern of the individual is derived. Furthermore, in the step of performing the detailed analysis, the second analysis data can be determined by considering the degree of agreement between the output data of each of the multiple models.

[0011] On the other hand, a smart farm monitoring system according to one embodiment of the present invention may include an edge device that collects target data, including video data and sensing data, linked to an individual being monitored, performs an initial analysis by inputting the target data into a first model of a pre-trained artificial intelligence platform to output first analysis data linked to the behavioral patterns of the individual, and transmits the first analysis data and the target data to a main server when the first analysis data satisfies pre-set detailed analysis conditions, and the main server that performs a detailed analysis by inputting at least one of the first analysis data and the target data into a second model of a pre-trained artificial intelligence platform to derive second analysis data linked to the behavioral patterns of the individual.

[0012] The means of solving the problems described above are merely illustrative and should not be construed as intended to limit this application. In addition to the exemplary embodiments described above, additional embodiments may exist in the drawings and the detailed description of the invention. [Effects of the Invention]

[0013] According to the aforementioned solution to the problem of the present invention, it is possible to provide an individual analysis device and method using artificial intelligence with a dual processing base that can efficiently monitor and analyze the behavior of individuals in a smart farm environment, and a smart farm monitoring system using the same.

[0014] According to the aforementioned solution to the problem of the present invention, a dual data processing pipeline is used in which basic data analysis is performed at the edge and important data is selectively transmitted to the server side for advanced analysis. This maximizes the efficiency of data processing and minimizes server load and data transmission costs, even in environments with resource constraints.

[0015] According to the aforementioned solution to the problem of the present invention, a smart farm monitoring system can be constructed to accommodate diverse types of livestock and diverse rearing environments, and it can assist in promptly taking appropriate action by continuously sensing the health status, stress levels, etc., of each individual being monitored.

[0016] According to the aforementioned solution to the problems of the present invention, by implementing a smart farm monitoring system through a modular design for farms of various sizes and different types of smart farms, integration between systems is facilitated, and administrators can customize and individually configure the elements necessary for each smart farm for each individual system. However, the effects obtained with this application are not limited to those described above, and other effects may exist. [Brief explanation of the drawing]

[0017] [Figure 1]It is a schematic configuration diagram of a smart farm monitoring system according to an embodiment of the present application. [Figure 2] It is a conceptual diagram showing the operation flow of a smart farm monitoring system according to an embodiment of the present application. [Figure 3] It is a conceptual diagram for explaining an individual identification method of an artificial intelligence base using video data linked with an individual to be monitored. [Figure 4] It is a schematic configuration diagram of the main server of a smart farm monitoring system according to an embodiment of the present application. [Figure 5] It is a schematic configuration diagram of an edge device of a smart farm monitoring system according to an embodiment of the present application. [Figure 6] It is an operation flowchart for an individual analysis method of a dual processing base using artificial intelligence according to an embodiment of the present application performed using an edge device. [Figure 7] It is an operation flowchart for an individual analysis method of a dual processing base using artificial intelligence according to an embodiment of the present application performed using a main server. [Figure 8] It is a detailed operation flowchart for a detailed analysis process of deriving second analysis data linked with an individual's behavior pattern.

Embodiments for Carrying out the Invention

[0018] Hereinafter, embodiments of the present application will be described in detail so that those having ordinary knowledge in the technical field to which the present application belongs can easily implement it with reference to the attached drawings. However, the present application can be embodied in various different forms and is not limited to the embodiments described here. And, in order to clearly explain the present application in the drawings, parts not related to the explanation are omitted, and similar parts throughout the specification are denoted with similar drawing reference numerals. Throughout the present specification, if a part is said to be "connected" to another part, this includes not only the case where it is "directly connected", but also the case where it is "electrically connected" or "indirectly connected" with other elements interposed therebetween. Throughout this specification, if a member is described as being located "on top of," "at the top," "at the top end," "below," "at the bottom," or "at the bottom end" of another member, this includes not only cases where one member is in contact with another member, but also cases where another member exists between the two members. In the entirety of this specification, when a part "includes" a certain component, this means, unless otherwise stated, that it may include other components rather than excluding them.

[0019] This application relates to an individual analysis device and method using an artificial intelligence-based dual-processing platform, and a smart farm monitoring system using the same. Figure 1 is a schematic diagram of a smart farm monitoring system according to one embodiment of the present invention.

[0020] Referring to Figure 1, the smart farm monitoring system 10 according to one embodiment of the present invention may include a main server 100, an edge device 200, and a user terminal 300. Also referring to Figure 1, the smart farm monitoring system 10 may include a measurement module 21 on the edge device 200 side that measures and transmits various sensing data to the target space to be monitored (e.g., farm, barn, smart farm, etc.) and a camera module 22 that captures and transmits video data to the target space.

[0021] The measurement module 21, camera module 22, main server 100, edge device 200, and user terminal 300 can communicate with each other via the network 20. The network 20 refers to a connected structure that enables information exchange between each node, such as terminals and servers. Examples of such a network 20 include, but are not limited to, the 3GPP (3rd Generation Partnership Project) network, LTE (Long Term Evolution) network, 5G network, WiMAX (World Interoperability for Microwave Access) network, Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), Wi-Fi network, Bluetooth (registered trademark) network, satellite broadcasting network, analog broadcasting network, and DMB (Digital Multimedia Broadcasting) network.

[0022] The user terminal 300 can be any type of wireless communication device, such as a smartphone, smartpad, tablet PC, PCS (Personal Communication System), GSM (Global System for Mobile communication), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), or Wibro (Wireless Broadband Internet) terminal.

[0023] For reference, in the description of the embodiments of this application, the user terminal 300 may be a terminal owned by the manager of a target space (e.g., a farm, barn, smart farm, etc.) managed through the smart farm monitoring system 10 disclosed herein. The data processing pipeline and system architecture of the smart farm monitoring system 10 disclosed in this application will be described below with reference to Figure 2. Figure 2 is a conceptual diagram showing the operation flow of a smart farm monitoring system according to one embodiment of the present invention.

[0024] Referring to Figure 2, a smart farm monitoring system 10 according to one embodiment of the present invention may include edge devices 200 arranged in accordance with each of the target spaces to be monitored, and a main server 100 that acquires data collected from the edge devices 200 in each target space, analyzes the acquired data, and provides analysis results.

[0025] Specifically, the edge device 200 collects target data, including video data and sensing data, linked to the individual being monitored. It can then perform an initial analysis by inputting the collected target data into a pre-trained first model of an artificial intelligence platform, outputting first analysis data linked to the individual's behavioral patterns. Furthermore, if the first analysis data satisfies pre-set detailed analysis conditions, the edge device 200 can transmit the first analysis data and target data to the main server 100.

[0026] In other words, the edge device 200 disclosed in this application collects target data through a measurement module 21 and a camera module 22 installed in the target space to monitor the behavior, location, health status, etc., of individuals that are present in the target space and are to be monitored (for example, livestock that live in the target space such as lactating cows (milking cows), calving cows, growing cows, dairy cows, livestock, calves, pigs, horses, chickens, sheep, etc., and crops cultivated in the target space). The edge device 200 can then apply initial data processing and perform a basic analysis (initial analysis) that takes into account the resource information of the edge device 200 (for example, computing power, etc.) using a pre-trained first model (lightweight model).

[0027] For reference, initial data processing of target data can broadly include preprocessing tasks such as noise removal, data normalization, and initial data classification.

[0028] Furthermore, the edge device 200 can operate to selectively transmit target data collected and / or first analysis data initially analyzed by the first model to the main server 100, based on the results of the analysis of the first model (lightweight model) to distinguish between common and abnormal situations occurring in the target space, or to identify important situations. In order to select the data to be transmitted to the main server 100, predetermined criteria for selecting important data can be applied, and the selected data can be optimized and compressed.

[0029] Furthermore, the main server 100 can input at least one of the first analysis data and target data received from the edge device 200 into a pre-trained second model of the artificial intelligence platform to perform a detailed analysis that derives second analysis data linked to the individual's behavioral patterns.

[0030] In other words, the main server 100 disclosed in this application can perform the following functions: receiving data from the edge device 200; performing advanced analysis (detailed analysis) of the artificial intelligence infrastructure, which is difficult to perform at the edge device 200, using a pre-trained second model, when considering the resources of the edge device 200, on the received data; and providing the results of the advanced analysis (detailed analysis) to the user terminal 300.

[0031] On the other hand, if the smart farm monitoring system 10 is configured to analyze and manage data collected through edge devices 200 deployed for each of multiple target spaces (for example, smart farms such as "Farm A" and "Farm B" in Figure 2), as illustrated in Figure 2, the user terminal 300 may, but is not limited to, be owned by each administrator corresponding to each target space.

[0032] The following describes in detail the functions and operations of the main server 100 and edge device 200 of the smart farm monitoring system 10 disclosed in this application. First, the selective data transmission process through data collection, initial analysis (basic analysis), selection and optimization of important data using the edge device 200 will be described. Later, the process of performing detailed analysis (advanced analysis) on the main server 100 using the data received from the edge device 200 and providing the analysis results to the user terminal 300 will be described.

[0033] First, the edge device 200 can collect target data, including video data and sensing data, in conjunction with the individual being monitored.

[0034] Furthermore, the edge device 200 can perform initial analysis by inputting target data into a pre-trained first model of an artificial intelligence platform and outputting first analysis data linked to the individual's behavioral patterns.

[0035] On the other hand, the edge device 200 disclosed in this application can be initially analyzed using a first model, which is a lightweight model constructed considering the resource information of the edge device 200. In connection with this, the first model can be an edge target model to which a lightweight design has been applied so that it can operate efficiently even on edge devices 200 that fall under the category of equipment with limited computing power, taking into consideration the processor performance, memory capacity, processing speed, etc. of the edge device 200.

[0036] For example, the first model may have a relatively smaller size of input data that can be processed compared to the second model used to perform detailed analysis (advanced analysis) on the main server 100, or it may include multiple layers, consisting of fewer layers than the second model, or it may be designed not to include relatively complex functions, or the number of iterations of a given operation may be set relatively low, but it is not limited to these.

[0037] Furthermore, according to one embodiment of the present invention, the first model can be a model in which the confidence value, which represents the reliability of the class predicted by the model, is set to be relatively smaller than that of the second model. In this regard, the higher the confidence value of the model, the more accurate the prediction result for that class can be. However, in the case of the edge device 200, considering the resources, it is more important to detect the occurrence of specific important situations (e.g., specific types of behavior of individuals) without missing them than to provide highly accurate analysis results. Therefore, a confidence value can be applied to the first model that strictly prevents non-diagnosis (missing important situations / incidents), even if the probability of misdiagnosis (incorrect class classification) is relatively higher.

[0038] In other words, according to one embodiment of the present invention, the confidence value (e.g., first threshold) applied to the first model (lightweight model) for initial analysis can be set to a lower value than the confidence value (e.g., second threshold) applied to the second model (server-side model) for detailed analysis. Specifically, the edge device 200 can derive information about the presence or absence of individuals and their location as first analysis data, which are reflected in the video data.

[0039] In connection with this, Figure 3 is a conceptual diagram illustrating an artificial intelligence-based individual identification method that uses video data linked to the individual being monitored.

[0040] Referring to Figure 3, the edge device 200 uses a first type of video analysis model to recognize individuals present in target image 1, which represents a predetermined frame included in the video data, and identifies a first boundary region B-1 containing the recognized individuals, thereby deriving information on the presence or absence of individuals and their appearance location.

[0041] For example, the edge device 200 can use a first model of video analysis model types to identify bounding boxes corresponding to individuals appearing in target image 1, identify key points indicating specific body parts of individuals, or perform segmentation to classify (understand) the object types of each part that make up target image 1, but is not limited to these operations.

[0042] Furthermore, referring to Figure 3, the edge device 200 can recognize pre-configured identification targets (for example, ear tags, sensors, tags, or recognition labels attached to the ear area of ​​an individual) and identify the second boundary region B-2, which includes the recognized identification target, within the first boundary region B-1.

[0043] On the other hand, according to one embodiment of the present invention, the edge device 200 can derive shape information of objects in the target image 1 identified using a first model, which is an artificial intelligence-based video analysis model, as first analysis data. In this regard, the shape information of an object may specifically include at least one of the object's hue information and partition pattern information. The main server 100, having acquired such shape information of an object as first analysis data, can operate to identify (identify) individuals appearing in the first analysis data or video data linked to the first analysis data using a database (not shown) that matches individual identification information corresponding to each of a plurality of individuals active in the target space with tag identification information assigned for each type of shape information to each of a plurality of objects generated so that their shape information is mutually distinguishable.

[0044] As another example, the edge device 200 can derive first classification information for an individual's behavior type as first analysis data. Furthermore, when the first analysis data satisfies pre-set detailed analysis conditions, the edge device 200 can transmit the first analysis data and target data to the main server 100.

[0045] In this regard, according to one embodiment of the present invention, if the first classification information corresponds to a pre-set abnormal or important action, the edge device 200 can determine that the detailed analysis conditions have been met and transmit the first analysis data and target data to the main server 100.

[0046] Furthermore, the aforementioned abnormal or significant behaviors can be defined to include, for example, if the individual is a dairy cow type of livestock, abnormal behaviors such as riding (reproductive behavior), ingestion (feeding, watering), disease, or stress, but are not limited to these, and can be defined in a variety of ways depending on the type of individual located in the target space. Next, I will explain main server 100.

[0047] The main server 100 inputs target data, including video data and sensing data linked to the individual being monitored, into a pre-trained first model of the artificial intelligence platform. Using the edge device 200, it performs an initial analysis to output first analysis data linked to the individual's behavioral patterns. As a result, the first analysis data, which satisfies pre-set detailed analysis conditions, can be received from the edge device 200 as well as the derived first analysis data and target data.

[0048] Furthermore, the main server 100 can perform a detailed analysis by inputting at least one of the first analysis data and target data into a second model of a pre-trained artificial intelligence platform to derive second analysis data linked to the individual's behavioral patterns.

[0049] Specifically, the main server 100 can operate to determine second analysis data by considering the degree of agreement of the output data of each of the multiple models, using a second model that includes multiple models pre-built based on each of multiple different network structures. On the other hand, multiple models having different network structures can mean, for example, one of the multiple models is designed based on a Convolutional Neural Network (CNN), and another of the multiple models is designed based on a Transformer architecture, but is not limited to this, and a variety of artificial intelligence algorithms or architectures, such as deep learning networks, machine learning algorithms, and guided / unguided learning algorithms that are already publicly known or can be developed in the future, can be applied to each of the multiple models that can be included in the second model disclosed herein.

[0050] In this regard, according to one embodiment of the present invention, the main server 100 can generate a dataset including at least a portion of the target data and first analysis data received from the edge device 200 as the data to be analyzed by the second model.

[0051] Furthermore, the main server 100 can individually input the same dataset generated for each of the multiple models included in the second model, and obtain the prediction results for each of the multiple models.

[0052] Furthermore, the main server 100 can verify whether the prediction results derived through each of the multiple models are mutually consistent. Specifically, if all the prediction results derived through each of the multiple models are consistent, the main server 100 can confirm the prediction results commonly derived through all the models as the second analysis data for the second model. Conversely, if at least some of the prediction results derived through the multiple models are not mutually consistent, the main server 100 can remove the existing prediction results derived through each of the multiple models and perform a detailed analysis again using the newly transmitted data from the edge device 200 (in other words, the target data and the first analysis data).

[0053] On the other hand, in relation to such a multi-model approach, the smart farm monitoring system 10 disclosed in this application, with respect to the second model which includes multiple models, is configured so that each of the multiple models independently performs analysis using the same dataset. The results are considered reliable only if all of the multiple models predict the same class. Conversely, if only one of the models predicts a particular class, the result is considered an error. Considering that in commercial environments such as smart farms, it is important to minimize erroneous results and ultimately provide accurate results with high reliability, the second model, unlike the first model, adopts an approach that uses various models rather than relying on a single model to enhance the reliability of the results. Furthermore, the main server 100 can transmit the second analysis data derived through the detailed analysis via the main server 100 to the user terminal 300.

[0054] Figure 4 is a schematic diagram of the main server of a smart farm monitoring system according to one embodiment of the present invention. Referring to Figure 4, the main server 100 may include a data receiving unit 110, a detailed analysis unit 120, and an analysis information provision unit 130.

[0055] The data receiving unit 110 inputs target data, including video data and sensing data linked to the individual being monitored, into a pre-trained first model of the artificial intelligence platform and performs an initial analysis using the edge device 200 to output first analysis data linked to the individual's behavioral patterns. As a result, the first analysis data can be received from the edge device 200, which is derived when the first analysis data satisfies pre-set detailed analysis conditions.

[0056] The detailed analysis unit 120 can perform a detailed analysis by inputting at least one of the first analysis data and target data into a second model of a pre-trained artificial intelligence platform to derive second analysis data linked to the individual's behavioral patterns.

[0057] Specifically, the detailed analysis unit 120 can operate to determine second analysis data by using a second model that includes multiple models pre-built based on each of several different network structures, taking into account the degree of agreement of the output data of each of the multiple models.

[0058] In this regard, according to one embodiment of the present invention, the detailed analysis unit 120 can generate a dataset including at least a portion of the target data and first analysis data received from the edge device 200 as the data to be analyzed for the second model.

[0059] Furthermore, the detailed analysis unit 120 can individually input the same dataset generated for each of the multiple models included in the second model, and obtain the prediction results for each of the multiple models.

[0060] Furthermore, the detailed analysis unit 120 can verify whether the prediction results derived through each of the multiple models are mutually consistent. Specifically, if all the prediction results derived through each of the multiple models are consistent, the detailed analysis unit 120 can determine the prediction result commonly derived through the multiple models as the second analysis data for the second model. Conversely, if at least some of the prediction results derived through the multiple models are not mutually consistent, the detailed analysis unit 120 can remove the existing prediction results derived through each of the multiple models and perform detailed analysis again using the newly transmitted data from the edge device 200 (in other words, the target data and the first analysis data). The analysis information provision unit 130 can transmit the second analysis data derived through detailed analysis via the main server 100 to the user terminal 300.

[0061] Figure 5 is a schematic diagram of the edge device configuration of a smart farm monitoring system according to one embodiment of the present invention. Referring to Figure 5, the edge device 200 can include a data collection unit 210 and an initial analysis unit 220.

[0062] The collection unit 210 can collect target data, including video data and sensing data, in conjunction with the individual being monitored, using the edge device 200.

[0063] The initial analysis unit 220 can perform initial analysis using the edge device 200 by inputting target data into a pre-trained first model of an artificial intelligence platform and outputting first analysis data linked to the individual's behavioral patterns.

[0064] For example, the initial analysis unit 220 can perform an initial analysis using a first model, which is a lightweight model constructed taking into account the resource information of the edge device 200.

[0065] Specifically, the initial analysis unit 220 can derive information on the presence or absence of individuals and their appearance location, as reflected in the video data, as first analysis data. As another example, the initial analysis unit 220 can derive first classification information for the behavioral type of individuals as first analysis data.

[0066] When the first analysis data satisfies the pre-set detailed analysis conditions, the communication unit 230 can transmit the first analysis data and target data to the main server 100.

[0067] Specifically, according to one embodiment of the present invention, if the first classification information corresponds to a pre-set abnormal act or important act, the communication unit 230 can determine that the detailed analysis conditions have been met and transmit the first analysis data and target data to the main server 100.

[0068] Below, we will briefly explain the operation flow of this invention based on the details described above. Figure 6 is an operation flowchart for an individual analysis method using artificial intelligence on a dual processing platform according to one embodiment of the present invention, which is performed using an edge device.

[0069] The individual analysis method for the dual processing platform using artificial intelligence, as shown in Figure 6, can be performed by the edge device 200 described above. Therefore, even if some details are omitted below, the explanation of the edge device 200 can be applied equally to the explanation of the individual analysis method for the dual processing platform using artificial intelligence.

[0070] Referring to Figure 6, in step S11, the collection unit 210 can collect target data, including video data and sensing data, in conjunction with the individual being monitored, using the edge device 200.

[0071] Next, in step S12, the initial analysis unit 220 can perform an initial analysis using the edge device 200 by inputting target data into a pre-trained first model of the artificial intelligence infrastructure and outputting first analysis data linked to the individual's behavioral patterns.

[0072] For example, in step S12, the initial analysis unit 220 can perform an initial analysis using a first model, which is a lightweight model constructed considering the resource information of the edge device 200.

[0073] Specifically, in step S12, the initial analysis unit 220 can derive information on the presence or absence of individuals and their appearance location, as reflected in the video data, as first analysis data. As another example, in step S12, the initial analysis unit 220 can derive first classification information for the behavioral types of individuals as first analysis data.

[0074] Next, in step S13, the communication unit 230 can transmit the first analysis data and target data to the main server 100 once the first analysis data satisfies the pre-set detailed analysis conditions.

[0075] Specifically, according to one embodiment of the present invention, in step S13, if the first classification information corresponds to a pre-set abnormal act or important act, the communication unit 230 can determine that the detailed analysis conditions have been met and transmit the first analysis data and target data to the main server 100.

[0076] In the above description, steps S11 to S13 can be further divided into additional steps or combined into even fewer steps, as exemplified by the present invention. Furthermore, some steps may be omitted as necessary, and the procedures between steps may be changed.

[0077] Figure 7 is an operation flowchart for an individual analysis method using artificial intelligence on a dual processing platform according to one embodiment of the present invention, performed using the main server. The individual analysis method of the dual processing platform using artificial intelligence shown in Figure 7 can be performed by the main server 100 described above. Therefore, even if the details are omitted below, the explanation of the main server 100 can be applied equally to the explanation of the individual analysis method of the dual processing platform using artificial intelligence.

[0078] Referring to Figure 7, in step S21, the data receiving unit 110 inputs target data, including video data and sensing data linked to the individual being monitored, into a pre-trained first model of the artificial intelligence platform and performs an initial analysis using the edge device 200 to output first analysis data linked to the individual's behavioral pattern. As a result, the first analysis data, which satisfies the pre-set detailed analysis conditions, can be received from the edge device 200 as a result of the first analysis data and target data derived from this analysis.

[0079] Next, in step S22, the detailed analysis unit 120 can perform a detailed analysis by inputting at least one of the first analysis data and target data into a second model of a pre-trained artificial intelligence platform to derive second analysis data linked to the individual's behavioral pattern.

[0080] Specifically, in step S22, the detailed analysis unit 120 can operate to determine second analysis data by using a second model that includes multiple models pre-built based on each of multiple different network structures, taking into account the degree of agreement of the output data of each of the multiple models.

[0081] Furthermore, according to one embodiment of the present invention, in step S22, the detailed analysis unit 120 derives second analysis data including second classification information for the behavior type of an individual, and the number of classes in the derived second classification information can be set to be larger than the number of classes in the first classification information described above.

[0082] In the above description, steps S21 and S22 may be further divided into additional steps or combined into even fewer steps, as embodied in the present invention. Furthermore, some steps may be omitted as necessary, and the procedures between steps may be changed.

[0083] Figure 8 is a detailed operation flowchart for the detailed analysis process that derives second analysis data linked to individual behavior patterns. The detailed analysis process for deriving the second analysis data linked to the individual behavior patterns shown in Figure 8 can be performed by the main server 100 described above. Therefore, even if the details are omitted below, the explanation of the main server 100 can be applied equally to the explanation of the individual analysis method of the dual processing platform using artificial intelligence.

[0084] Referring to Figure 8, in step S221, the detailed analysis unit 120 can generate a dataset that includes at least a portion of the target data and first analysis data received from the edge device 200 as the data to be analyzed for the second model.

[0085] Next, in step S222, the detailed analysis unit 120 can individually input the same dataset generated through step S221 into each of the multiple models included in the second model. Next, in step S223, the detailed analysis unit 120 can obtain the prediction results for each of the multiple models. Next, in step S224, the detailed analysis unit 120 can verify whether the prediction results derived through each of the multiple models are mutually consistent.

[0086] If, as a result of the judgment in step S224, the prediction results derived through each of the multiple models all match, then in step S225, the detailed analysis unit 120 can confirm the prediction results commonly derived through the multiple models as the second analysis data for the second model.

[0087] In contrast, if, as a result of the judgment in step S224, at least some of the prediction results derived through multiple models do not match each other, in step S226, the detailed analysis unit 120 may remove the existing prediction results derived through each of the multiple models and perform a detailed analysis again using the newly transmitted data from the edge device 200 (in other words, the target data and the first analysis data).

[0088] In the above description, steps S221 and S226 may be further divided into additional steps or combined into fewer steps, as embodied in the present invention. Furthermore, some steps may be omitted as necessary, and the procedures between steps may be changed.

[0089] A method for individual analysis of a dual-processing base using artificial intelligence according to one embodiment of the present invention can be embodied in the form of program instructions that can be performed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., individually or in combination. The program instructions recorded on the medium may be specifically designed and configured for the present invention or may be publicly known and usable by those skilled in the computer software art. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include not only machine code, such as that produced by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

[0090] Furthermore, the individual analysis method of the dual processing platform using artificial intelligence described above can also be implemented in the form of a computer program or application executed by a computer stored on a recording medium.

[0091] The description of the present application provided above is illustrative, and a person with ordinary skill in the art to which the present application belongs will understand that it can be easily modified into other specific forms without altering the technical idea or essential features of the present application. Accordingly, the embodiments described above are illustrative in all respects and not limiting. For example, each component described as a single type may be implemented in a dispersed manner, and similarly, components described as dispersed may be implemented in a combined form. The scope of this application is expressed by the claims, which are described below rather than in the detailed description above, and the meaning and scope of the claims, and all modified or altered forms derived from the concept of equivalents thereof, are included in the scope of this application. [Explanation of Symbols]

[0092] 10: Smart Farm Monitoring System 100: Main Server 110: Data receiving unit 120: Detailed analysis department 130:Analysis information provision department 200: Edge devices 210: Collection Department 220: Initial analysis department 230: Communications Department 21: Measurement module 22: Camera Module 300: User terminal 20: Network

Claims

1. In a dual-processing platform for individual analysis using artificial intelligence, The process involves collecting target data, including video data and sensing data, linked to the individual being monitored, using edge devices, and The process includes a step of performing an initial analysis using the edge device, in which the target data is input into a first model of a pre-trained artificial intelligence platform to output first analysis data linked to the behavioral patterns of the individual, and When the first analysis data satisfies the pre-set detailed analysis conditions, the first analysis data and the target data are transmitted to the main server. Includes, The aforementioned main server is This method involves inputting at least one of the first analysis data and the target data into a second model of a pre-trained artificial intelligence platform to perform a detailed analysis that derives second analysis data linked to the behavioral patterns of the individual. An analytical method characterized by the following.

2. The first analysis data mentioned above is: This includes information on the presence or absence of the individual and information on its location, as reflected in the aforementioned video data. The analytical method according to claim 1.

3. The first analysis data mentioned above is: This includes first classification information for the behavioral type of the individual, The detailed analysis conditions mentioned above are: The aforementioned first classification information includes conditions that correspond to pre-defined abnormal or important actions. The analytical method according to claim 1.

4. The second analysis data mentioned above is: This includes a second classification information for the behavioral types of the individual, wherein the number of classes in the second classification information is greater than the number of classes in the first classification information. The analytical method according to claim 3.

5. The first model described above is a lightweight model constructed taking into account the resource information of the edge device. The analytical method according to claim 1.

6. The second model includes multiple models pre-built based on each of several different network structures, The second analysis data mentioned above is: This is determined by considering the degree of agreement in the output data of each of the aforementioned multiple models. The analytical method according to claim 1.

7. In a dual-processing platform for individual analysis using artificial intelligence, An initial analysis is performed using an edge device, in which target data including video data and sensing data linked to the individual being monitored is input into a first model of a pre-trained artificial intelligence platform, and first analysis data linked to the behavioral patterns of the individual is output. As a result, when the first analysis data satisfies the pre-set detailed analysis conditions, the first analysis data and the target data are received from the edge device. The process includes a step of conducting a detailed analysis in which at least one of the first analysis data and the target data is input into a second model of a pre-trained artificial intelligence platform to derive second analysis data linked to the behavioral patterns of the individual. An analytical method characterized by the following.

8. The second model includes multiple models pre-built based on each of several different network structures, The step of conducting the aforementioned detailed analysis is: The second analysis data is determined by considering the degree of agreement of the output data of each of the aforementioned multiple models. The analytical method according to claim 7.

9. The first analysis data mentioned above is: This includes information on the presence or absence of the individual and information on its location, as reflected in the aforementioned video data. The analytical method according to claim 7.

10. The first analysis data mentioned above is: This includes first classification information for the behavioral type of the individual, The detailed analysis conditions mentioned above are: The aforementioned first classification information includes conditions that correspond to pre-defined abnormal or important actions. The analytical method according to claim 7.

11. The second analysis data mentioned above is: This includes a second classification information for the behavioral types of the individual, wherein the number of classes in the second classification information is greater than the number of classes in the first classification information. The analytical method according to claim 10.

12. In a smart farm monitoring system, An edge device collects target data, including video data and sensing data, linked to the individual being monitored, inputs the target data into a first model of a pre-trained artificial intelligence platform to perform an initial analysis that outputs first analysis data linked to the individual's behavioral patterns, and when the first analysis data satisfies pre-set detailed analysis conditions, transmits the first analysis data and the target data to the main server. The system includes a main server that inputs at least one of the first analysis data and the target data into a second model of a pre-trained artificial intelligence platform to perform a detailed analysis that derives second analysis data linked to the behavioral patterns of the individual. A monitoring system characterized by the following features.

Images