Poultry health monitoring system and method

Abstract

A poultry health monitoring system includes a cloud module, a learning and correction module, and a monitoring module. The learning and correction module senses the weight of at least one poultry being carried and a first poultry image, analyzes the number of poultry in the first poultry image, and generates poultry image features and an image-weight relationship. The monitoring module generates a second poultry image including at least one poultry. The cloud module obtains unit weight and activity values ​​for each poultry based on the second poultry image. The present invention further includes a poultry health monitoring method.

Description

Technical Field

[0001] The present invention relates to a health monitoring system and method thereof, and more particularly to a poultry health monitoring system and method thereof that can automatically calibrate and monitor poultry to save manpower. Background Art

[0002] Chickens, a type of poultry, and pigs, a type of livestock, have long been a major source of protein in our diets. They are not only highly nutritious but also a key ingredient in many processed foods. In recent years, chicken has become a key agricultural commodity. Chicken health is closely linked to their dietary habits. Currently, dietary behavior monitoring in poultry facilities is primarily done manually. However, given the large number of chickens housed in poultry facilities, traditional management methods are time-consuming, labor-intensive, and reliant on owner experience, which can lead to challenges in cost control and quality control.

[0003] However, during poultry farming, factors such as the number, range, and response speed of monitoring equipment often limit the number of devices, resulting in unrecoverable losses. Heat stress, particularly in tropical and subtropical regions due to the hot summer climate, has become one of the most challenging issues for poultry farmers. Heat stress reduces poultry growth, negatively impacts egg quality, and can even be linked to sudden, massive mortality. Early farming experience has shown that heat stress is crucial for stable poultry growth and egg quality. Heat stress is typically assessed using the temperature-humidity index (THI), which measures both temperature and humidity. However, the THI is an indirect indicator, and heat stress criteria can vary depending on the chicken breed and the birds' diet and water supply. This can easily lead to inaccurate assessments of heat stress or growth status, resulting in losses in poultry farming.

[0004] Furthermore, poultry farmers are well aware that the health of their poultry is closely related to their individual weight and activity. If poultry are underweight, or are not given enough space or time to move around during their breeding and growth, their health will be seriously affected. Traditionally, addressing this issue requires significant manpower to measure the weight of each poultry, conduct on-site observations and assessments of their health, and manually record their activity levels and activity time. This often consumes significant manpower and time, and processing speeds are often slow to improve, resulting in difficulties in reducing poultry farmers' maintenance costs and ineffective monitoring.

[0005] Therefore, how to design a poultry health monitoring system and method thereof to solve the aforementioned technical problems has become an important topic studied by the inventors of this case. Summary of the Invention

[0006] The purpose of the present invention is to provide a poultry health monitoring system that can solve the technical problems of the existing technology that the breeding and maintenance costs are difficult to reduce and the monitoring efficiency is not obvious, so as to achieve the goals of low maintenance cost, rapid response and full-time monitoring.

[0007] To achieve the aforementioned objectives, the poultry health monitoring system proposed in the present invention includes: a cloud module, a computing core, a learning and correction module, and a monitoring module. The cloud module is configured to store at least one poultry image feature and an image-weight relationship equation corresponding to each poultry image feature. The computing core is coupled to the cloud module and receives the weight value and the first poultry image to analyze the number of the at least one poultry in the first poultry image and generate the at least one poultry image feature and the image-weight relationship equation, wherein the image-weight relationship equation includes the relative relationship between the image feature value and the weight. The learning and correction module is coupled to the computing core and the cloud module and includes a load-bearing structure and a first camera. The load-bearing structure is configured to sense the weight of the at least one poultry carried by the load-bearing structure. The first camera is disposed within the load-bearing structure and is configured to generate a first poultry image of the at least one poultry carried by the load-bearing structure. The monitoring module is coupled to the cloud module and includes a second camera. The second camera is configured to generate a second poultry image including the at least one poultry. The cloud module obtains a unit weight of each poultry according to the second poultry image, the at least one poultry image feature, and the image-weight relationship.

[0008] Furthermore, the poultry health monitoring system described in the present invention further includes an early warning analysis module, which is coupled to the cloud module, and the early warning analysis module outputs at least one of a statistical report and a warning message based on at least one of the unit weight, the activity value and the uniformity.

[0009] Furthermore, the poultry health monitoring system of the present invention further includes a mobile communication platform, which is wirelessly coupled to the early warning analysis module and receives at least one of the statistical report and the warning information.

[0010] Furthermore, in the poultry health monitoring system of the present invention, the mobile communication platform includes one of a workstation, a server, a desktop computer, a laptop computer, a tablet computer, a personal digital assistant or a smart phone.

[0011] Furthermore, in the poultry health monitoring system described in the present invention, the computing core is a deep learning architecture that utilizes an object detection algorithm tool as the computing core to identify the target object. The object detection algorithm tool is a deep learning or image processing method. The cloud module uses the at least one convolutional layer and the at least one pooling layer to compare the second poultry image to determine whether it meets the various poultry image features, and then obtains the unit weight, activity, and uniformity values.

[0012] Furthermore, in the poultry health monitoring system of the present invention, the cloud module includes a server and a cloud database; wherein the server is used to obtain at least one of the unit weight and the activity value; the cloud database is coupled to the server and is used to store at least one of the poultry image feature, the image-weight relationship, the unit weight, and the activity value.

[0013] Furthermore, in the poultry health monitoring system described in the present invention, the server is coupled to the cloud database via one of narrowband internet of things (NB-Iot), LoRaWAN, LTE and Wi-Fi.

[0014] Furthermore, in the poultry health monitoring system of the present invention, the load-bearing structure includes a load-bearing platform and an intermediate platform, wherein the load-bearing platform is used to support at least one poultry, the intermediate platform is disposed on the load-bearing platform, and the first camera is disposed below the intermediate platform.

[0015] Furthermore, in the poultry health monitoring system of the present invention, the load-bearing platform is coupled to the intermediate platform via at least two columns.

[0016] When using the poultry health monitoring system and method described herein, the cloud module of the poultry health monitoring system may pre-store a weight determination database. Furthermore, the learning and correction module first performs a machine learning (ML) program based on an artificial intelligence (AI) model. The learning and correction module first uses the load-bearing structure to sense the weight of at least one poultry carried by the load-bearing structure and uses the first camera to generate the first poultry image of the at least one poultry carried by the load-bearing structure. Finally, the learning and correction module uses the computing core to receive the weight value and the first poultry image, analyze the number of the at least one poultry in the first poultry image, and generate the at least one poultry image features and the image-weight relationship equation for storage in the cloud module, thereby completing the machine learning program. The cloud module stores the at least one poultry image features and the image-weight relationship equation. The monitoring module generates the second poultry image corresponding to the at least one poultry using the second camera, which may be performed sequentially or simultaneously with the aforementioned steps. Finally, the cloud module can obtain the unit weight of each of the poultry based on the second poultry image, the at least one poultry image feature, and the image-weight relationship. Alternatively, the cloud module can obtain the activity value of each of the poultry based on the second poultry image and the at least one poultry image feature. Furthermore, the learning and correction module can continuously repeat the machine learning action over time to continuously correct the at least one poultry image feature and the image-weight relationship stored in the cloud module, so as to make the poultry health monitoring system of the present invention more sensitive and accurate. Since the aforementioned learning, monitoring and correction actions do not require extra human intervention and can operate unattended full-time, it not only saves labor costs, but is also not restricted by time, making poultry breeding and maintenance more efficient.

[0017] To this end, the poultry health monitoring system described in the present invention can solve the technical problems of the existing technology that the breeding and maintenance costs are difficult to reduce and the monitoring efficiency is not obvious, and achieve the goals of low maintenance costs, rapid response and full-time monitoring.

[0018] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments, but this does not limit the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the architecture of a first embodiment of a poultry health monitoring system according to the present invention;

[0020] Figure 2 A schematic diagram of the configuration of the first embodiment of the poultry health monitoring system of the present invention;

[0021] Figure 3 is a schematic diagram of the architecture of a second embodiment of the poultry health monitoring system of the present invention; and

[0022] Figure 4 The figure is a flow chart of the poultry health monitoring method of the present invention.

[0023] Wherein, the reference numerals:

[0024] 10: Cloud Module

[0025] 11: Server

[0026] 12: Cloud Database

[0027] 20: Learning Correction Module

[0028] 21: Load-bearing structure

[0029] 22: First Camera

[0030] 23: Computing core

[0031] 30: Monitoring module

[0032] 31: Second Camera

[0033] 40: Early warning analysis module

[0034] 50: Mobile Communication Platform

[0035] 100: Chicken

[0036] 210:Weight sensor

[0037] 211: Load-bearing platform

[0038] 212: Middle Platform

[0039] S1~S5: Steps DETAILED DESCRIPTION

[0040] The following describes the implementation of the present invention using specific embodiments. Those skilled in the art will readily understand the other advantages and benefits of the present invention from the information provided herein. The present invention may also be implemented or applied through various other specific embodiments, and the details in this specification may be modified and altered based on different perspectives and applications without departing from the spirit of the present invention.

[0041] It should be noted that the structures, proportions, sizes, number of components, etc. shown in the drawings of this specification are only used to match the contents provided in the specification for people familiar with this technology to understand and read, and are not used to limit the conditions under which the present invention can be implemented. Therefore, they have no substantive technical significance. Any modification of the structure, change in the proportional relationship or adjustment of the size should fall within the scope of the technical content provided by the present invention without affecting the efficacy and purpose that can be achieved by the present invention.

[0042] The technical content and detailed description of the present invention are described as follows with reference to the accompanying drawings.

[0043] See also Figures 1 to 2 As shown, Figure 1 Schematic diagram of the architecture of a first embodiment of a poultry health monitoring system according to the present invention; Figure 2 This is a schematic diagram of the configuration of the first embodiment of the poultry health monitoring system of the present invention. In the first embodiment of the present invention, the poultry health monitoring system includes a cloud module 10, a computing core 23, a learning and correction module 20, and a monitoring module 30. The cloud module 10 is configured to store at least one poultry image feature and an image-weight relationship equation corresponding to each poultry image feature. In the first embodiment of the present invention, the cloud module 10 includes a server 11 and a cloud database 12. The server 11 is configured to obtain at least one of the unit weight of each poultry (i.e., the individual weight of any chicken 100) and the activity value. Furthermore, the cloud database 12 is coupled to the server 11 and configured to store at least one of the at least one poultry image feature, the image-weight relationship equation, the unit weight, and the activity value. Furthermore, the server 11 is coupled to the cloud database 12 via one of narrowband internet of things (NB-IoT), LoRaWAN, LTE, and Wi-Fi. The computing core 23 is coupled to the cloud module 10.

[0044] The learning and correction module 20 is coupled to the computing core 23 and the cloud module 10, and includes a load-bearing structure 21 and a first camera 22. The load-bearing structure 21 is used to sense at least one poultry (such as Figure 2The first camera 22 is configured in the load-bearing structure 21 and is used to generate a first poultry image of the at least one poultry supported by the load-bearing structure 21. The computing core 23 is configured in the load-bearing structure 21, and the computing core 23 receives the weight value and the first poultry image to analyze the number of the at least one poultry in the first poultry image and generate the at least one poultry image feature and the image-weight relationship. In the first embodiment of the present invention, the load-bearing structure 21 includes a weight sensor 210, a load-bearing platform 211 and an intermediate platform 212. The load-bearing platform 211 is used to support at least one poultry, the intermediate platform 212 is configured on the load-bearing platform 211, and the first camera 22 is configured below the intermediate platform 212 to shoot at least one chicken 100 from above. The load-bearing platform 211 is coupled to the intermediate platform 212 via at least two columns 213, so that the load-bearing platform 211 and the intermediate platform 212 are linked together. In the first embodiment of the present invention, the computing core 23 is a deep learning architecture that utilizes an object detection algorithm tool as the computing core 23 to identify the target object. The object detection algorithm tool is a deep learning or image processing method, such as a mask region-based convolutional neural network (mask R-CNN) comprising at least one convolution layer and at least one pooling layer. The cloud module 10 uses at least one convolution layer and at least one pooling layer to compare whether the second poultry image generated by the second camera 31 of the monitoring module 30 meets the characteristics of each poultry image, and then obtains the unit weight (i.e., the individual weight of any chicken 100) or the activity or uniformity value according to the image-weight relationship. As Figure 2 As shown, the first camera 22 can capture a top-view image of at least one chicken 100 (e.g., two chickens as shown) and transmit the image data to the computing core 23, allowing the computing core 23 to calculate the number of chickens on the load-bearing platform 211. The load-bearing platform 211 transmits the total weight of the chickens on it to the computing core 23, which then calculates the average weight of the chickens on the load-bearing platform 211. According to one embodiment of the present invention, if the computing core 23 determines that there is only one chicken on the load-bearing platform 211, it can establish a relationship between the chicken's top-view image characteristics (e.g., body length, top-view area, etc.) and weight, i.e., an image-weight formula. However, this embodiment is not intended to limit the scope of this invention. For example, the computing core 23 can also calculate the average top-view image characteristics (e.g., average body length) and average weight of the chickens on the load-bearing platform 211, and can also establish an image-weight formula.

[0045] According to one embodiment of the present invention, the computing core 23 can also determine the activity of chickens by comparing a single chicken image with the time period. For example, if the computing core 23 determines that a chicken image is stationary or has not moved within a predetermined range (e.g., a movement distance of no more than 1 meter) within a predetermined period (e.g., 10 minutes) during the day, the chicken can be determined to be inactive. The computing core 23 can integrate the image data from the first camera 22 and the second camera 31 and determine the proportion of chickens inactive. If the proportion of chickens inactive exceeds a threshold, it indicates that the chickens in the farm may have an infectious disease, and the computing core 23 can issue a warning notification.

[0046] Furthermore, the confusion matrix evaluation results of the learning correction module 20 of the present invention for the machine learning of red-feathered native chickens are as follows:

[0047]

[0048] Here, true positives (TP) refer to the number of birds that were counted manually as "yes" and counted by deep learning; false positives (FP) refer to the number of birds that were counted manually as "no" and counted by deep learning; false negatives (FN) refer to the number of birds that were counted manually as "yes" and counted by deep learning; and true negatives (TN) refer to the number of birds that were counted manually as "no" and counted by deep learning. The above results were obtained using both machine learning and manual counting methods over a three-month rearing cycle. This demonstrates that the machine learning counting method of the present invention can effectively replace manual counting to assess average chicken weight. The multiple average weight data collected daily were further converted into daily standard deviations. In both the red-feathered rooster and red-feathered hens experimental results, the daily standard deviation of average weight increased significantly in the later stages of rearing. This is presumably because adult birds are more likely to fight than young birds. Consequently, weaker birds are unable to compete with stronger birds for food during feeding, leading to significant size differences between birds. From this, we can see that the standard deviation of average weight helps to monitor the overall health of the chickens. If the standard deviation of average weight becomes larger and larger, you can consider zoning to stabilize the average weight of the chickens.

[0049] The monitoring module 30 is coupled to the cloud module 10, and the monitoring module 30 includes a second camera 31. Furthermore, the monitoring module 30 can use a previously established image-weight relationship to quickly determine whether the weight of the chicken is abnormal. The second camera 31 is used to generate a second poultry image including at least one poultry (which can be any other chicken 100 supported outside the load-bearing structure 21). Alternatively, the cloud module 10 obtains the activity value of each poultry based on the second poultry image and the features of the at least one poultry image. Furthermore, the activity value can be the activity distance, activity frequency, and rest period of the individual chicken 100 in the second poultry image according to the judgment conditions of the cloud module 10, which serve as the basis for the cloud module 10 to judge the activity value. For example, a short activity distance and a low activity frequency are judged as poor activity. A threshold can be set for group classification.

[0050] See also Figure 3 FIG2 shows a schematic diagram of the architecture of a second embodiment of a poultry health monitoring system according to the present invention. The second embodiment is substantially similar to the first embodiment, but further includes a warning analysis module 40 and a mobile communication platform 50. The warning analysis module 40 is coupled to the cloud module 10 and outputs at least one of a statistical report and a warning message based on at least one of unit weight and activity level. The mobile communication platform 50 wirelessly couples to the warning analysis module 40 and receives at least one of the statistical report and the warning message, allowing poultry farmers to predict the health status or growth trends of the chickens 100 in advance, allowing them to take proactive measures or preventative measures to reduce the risks and costs of poultry farming. In the second embodiment, the mobile communication platform 50 comprises a workstation, a server, a desktop computer, a laptop computer, a tablet computer, a personal digital assistant, or a smartphone. However, the present invention is not limited thereto.

[0051] See also Figure 4The figure shows a flow chart of the poultry health monitoring method of the present invention. The learning and calibration module 20 first performs a machine learning (ML) program based on an artificial intelligence (AI) model. The learning and calibration module 20 first senses the weight of at least one poultry supported by the load-bearing structure 21 and generates a first poultry image of the at least one poultry supported by the load-bearing structure 21 using the first camera 22. Finally, the learning and calibration module receives the weight value and the first poultry image using the computing core 23, analyzes the number of the at least one poultry in the first poultry image, and generates the at least one poultry image feature and the image-weight relationship equation for storage in the cloud module 10 (step S1), thereby completing the machine learning program. The cloud module 10 then stores the at least one poultry image feature and the image-weight relationship equation (step S2). Sequentially or concurrently with the aforementioned steps, the monitoring module 30 generates a second poultry image corresponding to the at least one poultry using the second camera 31 (step S3). Finally, the cloud module 10 can obtain the unit weight of each of the poultry based on the second poultry image, the at least one poultry image feature, and the image-weight relationship (step S4). Alternatively, the cloud module 10 can obtain the activity value of each of the poultry based on the second poultry image and the at least one poultry image feature (step S5). Furthermore, the learning and correction module 20 can continuously repeat the machine learning action over time to continuously correct the at least one poultry image feature and the image-weight relationship stored in the cloud module 10, so that the poultry health monitoring system of the present invention is more sensitive and accurate. Since the aforementioned learning, monitoring and correction actions do not require unnecessary human intervention and can operate unattended full-time, it not only saves labor costs, but is also not restricted by time, making poultry breeding and maintenance more efficient.

[0052] To this end, the poultry health monitoring system described in the present invention can solve the technical problems of the existing technology that the breeding and maintenance costs are difficult to reduce and the monitoring efficiency is not obvious, and achieve the goals of low maintenance costs, rapid response and full-time monitoring.

[0053] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims attached to the present invention.

Claims

1. A poultry health monitoring system, characterized in that: include: a cloud module for storing at least one poultry image feature and an image weight relationship equation corresponding to each poultry image feature; a computing core coupled to the cloud module, the computing core receiving a weight value and a first poultry image, analyzing the quantity of at least one first poultry in the first poultry image, and generating a feature of the at least one poultry image and an image-weight relationship equation, wherein the image-weight relationship equation includes a relative relationship between the image feature value and the weight; a learning and calibration module coupled to the computing core and the cloud module, and comprising a load-bearing structure and a first camera; wherein the load-bearing structure is used to sense the weight of at least one first poultry supported by the load-bearing structure among a plurality of poultry; the load-bearing structure includes an intermediate platform, and the first camera is disposed below the intermediate platform and is used to generate the first poultry image of the at least one first poultry supported by the load-bearing structure; and a monitoring module coupled to the cloud module and comprising a second camera; wherein the second camera is configured to generate a second poultry image of at least one second poultry among the plurality of poultry that is not supported by the load-bearing structure; The cloud module uses at least one convolutional layer and at least one pooling layer to compare the second poultry image generated by the second camera to see whether it meets the characteristics of each poultry image, and then obtains a unit weight of one of the multiple poultry according to the image-weight relationship.

2. The poultry health monitoring system according to claim 1, wherein: The system further includes a warning analysis module coupled to the cloud module, and the warning analysis module outputs at least one of a statistical report and a warning message based on at least one of the unit weight, activity value and uniformity.

3. The poultry health monitoring system according to claim 2, wherein: The system further includes a mobile communication platform which is wirelessly coupled to the early warning analysis module and receives at least one of the statistical report and the warning information.

4. The poultry health monitoring system according to claim 3, wherein: The mobile communication platform includes one of a workstation, a server, a desktop computer, a laptop computer, a tablet computer, a personal digital assistant, or a smart phone.

5. The poultry health monitoring system according to claim 1, wherein: The computing core utilizes an object detection algorithm tool as a deep learning architecture for identifying a target object. The object detection algorithm tool is a deep learning or image processing method.

6. The poultry health monitoring system according to claim 1, wherein: The cloud module includes a server and a cloud database; wherein the server is used to obtain at least one of the unit weight and the activity value; the cloud database is coupled to the server and is used to store at least one of the image feature of the at least one poultry, the image-weight relationship, the unit weight, and the activity value.

7. The poultry health monitoring system according to claim 6, wherein: The server is coupled to the cloud database via one of narrowband internet of things (NB-Iot), LoRaWAN, LTE and Wi-Fi.

8. The poultry health monitoring system according to claim 1, wherein: The load-bearing structure further includes a load-bearing platform; wherein the load-bearing platform is used to bear the at least one first poultry, and the intermediate platform is configured on the load-bearing platform.

9. The poultry health monitoring system according to claim 8, wherein: The load-bearing platform is coupled to the middle platform via at least two columns.

Images