WO2024001791A1 - 穿戴式传感器与红外相机协同的畜禽体温监测系统及方法 - Google Patents

穿戴式传感器与红外相机协同的畜禽体温监测系统及方法 Download PDF

Info

Publication number
WO2024001791A1
WO2024001791A1 PCT/CN2023/100248 CN2023100248W WO2024001791A1 WO 2024001791 A1 WO2024001791 A1 WO 2024001791A1 CN 2023100248 W CN2023100248 W CN 2023100248W WO 2024001791 A1 WO2024001791 A1 WO 2024001791A1
Authority
WO
WIPO (PCT)
Prior art keywords
livestock
poultry
body temperature
temperature
wearable
Prior art date
Application number
PCT/CN2023/100248
Other languages
English (en)
French (fr)
Inventor
泮进明
王首驿
贺鹏光
林宏建
崔笛
傅迎春
应义斌
Original Assignee
浙江大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 浙江大学 filed Critical 浙江大学
Publication of WO2024001791A1 publication Critical patent/WO2024001791A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/38—Services specially adapted for particular environments, situations or purposes for collecting sensor information
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/695—Control of camera direction for changing a field of view, e.g. pan, tilt or based on tracking of objects
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N5/00—Details of television systems
    • H04N5/30—Transforming light or analogous information into electric information
    • H04N5/33—Transforming infrared radiation
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
    • Y02A40/70—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in livestock or poultry

Definitions

  • the invention relates to the field of body temperature monitoring of livestock and poultry, specifically to multi-source body temperature monitoring based on the continuous monitoring of individual sentinels with wearable temperature sensors and wide-area flexible inspection of infrared cameras, as well as a monitoring system for assessing the health status of body temperature of livestock and poultry and providing abnormal early warning. and methods.
  • Livestock and poultry are endothermic animals. As an important indicator of the physiological functions of livestock and poultry, body temperature can reflect the health status of livestock and poultry to a large extent. In the actual breeding and production process, individual abnormal states can be sensed and discovered earlier and measures can be taken in a timely manner. The more action you take, the more you can reduce production losses.
  • the purpose of the present invention is to provide a livestock and poultry body temperature monitoring system and method that coordinates wearable temperature sensor monitoring and infrared camera inspection, and continuously monitors sentinel individual wearable temperature sensor data with high precision. Integrated with infrared camera wide-area flexible inspection multi-source temperature data, and introducing time series features to build a body temperature change model, it can realize livestock and poultry body temperature monitoring, health status feedback and abnormal warning.
  • a livestock and poultry body temperature monitoring system using wearable sensors and infrared cameras using wearable sensors and infrared cameras:
  • the livestock and poultry house is equipped with different breeding area blocks in its own three-dimensional space. There are individual livestock and poultry in each breeding area block. All livestock and poultry individuals in each breeding area block are randomly equipped with wearable temperature sensors in a fixed proportion to form a sentinel. For individual livestock and poultry, the infrared camera acquires group infrared images based on the breeding area block.
  • the communication terminal is installed in the livestock and poultry house. The wearable temperature sensor and the infrared camera are connected to the local server through the communication terminal. The local server communicates with the client through the cloud server. terminal communication connection.
  • the infrared camera performs inspection along the preset infrared camera inspection route between the breeding area blocks in the livestock and poultry house.
  • the livestock and poultry house is divided into a specific number of breeding area blocks and numbered according to the three-dimensional spatial structure. Individuals in the breeding area blocks are randomly selected according to a specific proportion as sentinels, and wearable temperature sensors are worn to monitor the body temperature and health status of the groups in the area.
  • the wearable temperature sensor has a battery voltage acquisition circuit and a wireless communication module for battery replacement prompts, and the collected data is transmitted to the communication terminal through a wireless data transmission protocol.
  • Livestock and poultry body temperature monitoring method using wearable sensors and infrared cameras includes the following steps:
  • the continuous body temperature data of an appropriate number of sentinel livestock and poultry individuals in the breeding area are obtained through wearable temperature sensors, and the infrared thermal imaging images of all livestock and poultry individuals in each breeding area block and the location of the breeding area block are collected through infrared camera inspection;
  • the communication terminal transmits the continuous body temperature data and infrared thermal imaging data collected in step S1 to the local server.
  • the local server performs filtering of the data and preprocessing of image segmentation to obtain the body temperature value.
  • the preprocessed body temperature value is used to obtain each animal.
  • the body temperature monitoring curve of individual birds changes over time;
  • step S2 uses the body temperature monitoring curve that changes over time obtained in step S2 to construct a growth stage-temperature database, and obtain the normal body temperature change zone (body temperature time series change model) through growth stage-temperature database processing;
  • the temperature values collected in real time are combined with the normal body temperature change band to determine whether the body temperature status is abnormal by using time sequence matching and outlier discrimination.
  • the breeding area block where the livestock and poultry with abnormal body temperature status are located is determined as the abnormal area, and the infrared camera is mobilized to move to the abnormal area. Re-inspection to achieve complete monitoring.
  • the step S2 is specifically:
  • the deep learning method is used to perform instance segmentation on the infrared thermal imaging images of each breeding area block to identify each livestock and poultry individual.
  • Each livestock and poultry individual in the breeding area block is renumbered and extracted.
  • the key pixels in the image where each individual livestock and poultry is located and its temperature information are used as the body temperature value of the individual livestock and poultry;
  • T s and T c directly monitored by the wearable temperature sensor and the infrared camera are processed according to the following formulas to obtain the core body temperature values T sc and T cc respectively.
  • a s and b s are the weight and correction constant of the wearable temperature sensor respectively
  • a c and b c are the weight and correction constant of the infrared camera respectively.
  • the specific step S3 is: in the case of different varieties, different growth stages, and different monitoring methods, obtain body temperature values through wearable temperature sensors and infrared cameras, and establish normal body temperatures at different growth stages under the same variety and various monitoring methods. Change zone.
  • the body temperatures of the corresponding time series are simultaneously established using machine learning methods, which is different from the setting method of a single threshold to construct a normal body temperature change zone.
  • the monitoring method mentioned refers to the position where the wearable temperature sensor is worn.
  • step S3 there are two main steps for abnormality determination:
  • Step 1 Compare the temperature value collected in real time with the normal body temperature change band to determine whether the temperature value falls within the normal body temperature change band:
  • step 2 If it does not fall, the body temperature of the livestock and poultry corresponding to the temperature value is abnormal, and proceed to step 2;
  • Step 2 Compare the temperature value with the range [ ⁇ -3 ⁇ , ⁇ +3 ⁇ ] determined by the following formula:
  • ⁇ is the average temperature of the group
  • ⁇ is the standard deviation of the group temperature
  • T cci is the body temperature value of the group monitored by the wearable temperature sensor and infrared camera
  • n is the number of the group
  • the described group is divided into two situations: all livestock and poultry individuals in a single breeding area block, and all livestock and poultry individuals in all breeding area blocks.
  • the temperature values of the livestock and poultry individuals and the temperature values in a single breeding area block are respectively Compare and judge between groups composed of all livestock and poultry individuals, and between the temperature values of individual livestock and poultry and between groups composed of all livestock and poultry individuals in all breeding area blocks:
  • the body temperature of the livestock and poultry corresponding to the temperature value is not abnormal
  • the body temperature of the livestock and poultry corresponding to the temperature value is abnormal
  • Step 3 Select the livestock and poultry with abnormal body temperature, mobilize the infrared camera to move to the breeding area where the livestock and poultry with abnormal body temperature are located, re-inspect the livestock and poultry, and further monitor the abnormality.
  • This method and system uses two methods to monitor and evaluate the temperature status of individuals and groups in livestock and poultry houses, preprocesses temperature data collected from multiple sources, and can stably characterize and monitor its core temperature;
  • This method optimizes the method of body temperature monitoring and abnormality determination, introduces time series features, and constructs a body temperature time series change model based on variety-growth stage-monitoring method, distinguishes the single threshold judgment method, and more accurately identifies abnormal individuals;
  • This method optimizes the body temperature monitoring and abnormality determination methods, and proposes an outlier abnormality determination method based on big data, which can effectively identify possible abnormal individuals;
  • This method and system adopts solutions for breeding area blocks and sentinel individuals, and evaluates individual and group status by monitoring sentinel status.
  • the division of breeding area blocks facilitates data analysis, processing and rapid positioning;
  • Real-time body temperature monitoring can provide timely understanding of livestock and poultry abnormalities and early prevention to avoid large-scale losses.
  • Figure 1 is a method step diagram according to an embodiment of the present invention.
  • Figure 2 is a schematic structural diagram of the system according to the embodiment of the present invention.
  • Figure 3 is a schematic diagram of the comprehensive body temperature assessment workflow according to the embodiment of the present invention.
  • Figure 4 is a comparison diagram between the embodiment of the present invention and the traditional body temperature monitoring method.
  • the system includes a wearable temperature sensor 6, an infrared camera 3, a communication terminal 7, a local server 8, a cloud server 9 and a client 10;
  • the livestock and poultry house 1 is provided with different breeding area blocks 4 in its own three-dimensional space. There are livestock and poultry individuals 5 in each breeding area block 4. All livestock and poultry individuals 5 in each breeding area block 4 are randomly worn in a fixed proportion.
  • the temperature sensor 6 forms a sentinel livestock and poultry individual.
  • the infrared camera 3 acquires group infrared images based on the breeding area block.
  • the communication terminal 7 is installed in the livestock and poultry house 1.
  • the wearable temperature sensor 6 and the infrared camera 3 both pass the communication terminal 7 It communicates with the local server 8, and the local server 8 communicates with the client 10 via the cloud server 9.
  • the infrared camera 3 performs inspection along the preset infrared camera inspection route 2 between the breeding area blocks 4 in the livestock and poultry house 1 .
  • the local server 8 or the cloud server 9 preprocesses the data and then inputs the normal body temperature change zone to determine the status of the livestock and poultry body temperatures, and quickly locate the livestock and poultry body temperatures.
  • the infrared camera 3 will photograph and review the breeding area blocks with abnormal body temperatures of livestock and poultry to complete monitoring.
  • the body temperature health status of individuals and groups can also be comprehensively assessed based on the temperature data, and the results and warning information are transmitted to the user through the client 10 .
  • the livestock and poultry house is divided into a specific number of breeding area blocks according to the three-dimensional spatial structure.
  • Each breeding area block corresponds to a specific number.
  • a specific number of livestock and poultry individuals in each breeding area block are defined as sentinel livestock and poultry individuals.
  • Sentinel livestock Individual poultry wear wearable temperature sensors to continuously monitor individual body temperature changes, directly and indirectly reflect the group status, and play a role in body temperature monitoring and early warning in the area.
  • the infrared inspection camera moves in the livestock and poultry house along the preset slide rails or tracks in the livestock and poultry house.
  • Each breeding area block takes more than one thermal image to collect the body temperature data of the livestock and poultry.
  • the picture information also includes the positioning of the breeding area block. Data and timing information.
  • the wearable temperature sensor 5 has a battery voltage acquisition circuit and a wireless communication module for battery replacement prompts.
  • the collected data is transmitted to the communication terminal 7 through a wireless data transmission protocol.
  • the wearable temperature sensor 5 is coin-shaped and adopts a modular design. It can be stably worn with customized accessories such as silicone straps and silicone earrings on livestock and poultry at locations where it is easy to wear and can stably represent the body temperature of the individual, including but not limited to Chicken underwing, pig's ear, etc.
  • the wearable temperature sensor 5 continuously collects data.
  • the sampling frequency can be set according to actual needs.
  • the collected information includes timing characteristics and location information.
  • the data collected by the infrared camera 3 includes regional information and is consistent with the timing characteristics of the data collected by the wearable temperature sensor 5. Corresponding.
  • the wearable temperature sensor monitors 24 hours a day at a certain sampling frequency, and the sampling data also includes timing characteristics and location information; the wearable temperature sensor with accessories is worn on the corresponding livestock and poultry at a position that can be used to stably characterize the body temperature of the individual, including but not Limited to chicken underwings, pig ears and other parts;
  • Preset tracks for patrol infrared cameras in livestock and poultry houses and set patrol infrared cameras according to the areas divided in the previous stage.
  • Machine inspection procedures stay in the corresponding area during the inspection, and take more than one infrared thermal image that meets the requirements in the area, and the shooting angle should minimize occlusion;
  • the wearable temperature sensor and infrared camera transmit the collected data to the communication terminal through the wireless data transmission protocol.
  • the data contains timing characteristics and location information.
  • the communication terminal uploads the data to the server at a certain time frequency; the collected data can be selected locally
  • the deployed server can be used for processing, or it can be directly uploaded to the cloud server for processing, which can be flexibly arranged according to the actual situation.
  • local servers and cloud servers can be flexibly arranged to facilitate localized processing and offline processing, and solutions can be flexibly selected to improve computing efficiency and reduce costs.
  • the local server 8 has sufficient computing power, the collected data can be processed directly offline. Otherwise, the data can also be processed directly on the cloud server 9 .
  • the system can continuously optimize the body temperature model based on the collected data during actual application to improve the stability of the system.
  • the implementation process of the present invention is as follows:
  • Continuous high-precision body temperature data of an appropriate number of sentinel livestock and poultry individuals in the breeding area block 4 are respectively obtained through the wearable temperature sensor 6, and the infrared thermal imaging images of all livestock and poultry individuals in each breeding area block 4 are collected through the infrared camera 3 inspection and The location of the breeding area block 4, thus covering more infrared thermal imaging data and positioning data of livestock and poultry animals;
  • the communication terminal 7 transmits the continuous body temperature data and infrared thermal imaging data collected in step S1 to the local server 8.
  • the local server 8 performs filtering of the data and preprocessing of image segmentation to obtain the body temperature value.
  • the image segmentation is specifically to segment the livestock and poultry.
  • the preprocessed body temperature value is used to obtain the body temperature monitoring curve of each livestock and poultry individual that changes over time;
  • step S2 uses the body temperature monitoring curve that changes over time obtained in step S2 to construct a growth stage-temperature database, and obtain the normal body temperature change zone through growth stage-temperature database processing;
  • the temperature values collected in real time are combined with the normal body temperature change band to determine whether the body temperature status is abnormal by using time sequence matching and outlier discrimination, and determine the breeding area block 4 where the livestock and poultry with abnormal body temperature status are located as In the abnormal area, the infrared camera 3 is mobilized to move to the abnormal area for re-inspection to achieve complete monitoring.
  • temperature data can be integrated to provide assessment and feedback on the group body temperature and health status of each breeding area of the livestock and poultry house.
  • Step S2 is specifically: compare the continuous temperature data collected by the wearable temperature sensor with the fluctuation interval threshold to determine abnormal values, eliminate individual abnormal values but retain continuous abnormal values, and retain each temperature value obtained as The body temperature value of individual livestock and poultry; retaining continuous abnormal values means that the abnormal values appear continuously for 5 times or more than 15 minutes.
  • the deep learning method is used to perform instance segmentation on the infrared thermal imaging images of each breeding area block to identify each livestock and poultry individual.
  • Each livestock and poultry individual in the breeding area block is renumbered and extracted.
  • the key pixels in the image where each individual livestock and poultry is located and its temperature information are used as the body temperature value of the individual livestock and poultry;
  • T s and T c directly monitored by the wearable temperature sensor and the infrared camera are processed according to the following formulas to obtain the core body temperature values T sc and T cc respectively.
  • a s and b s are the weight and correction constant of the wearable temperature sensor respectively
  • a c and b c are the weight and correction constant of the infrared camera respectively.
  • the values of a and b are preset according to the monitored object and monitoring method.
  • Step S3 is specifically: in the case of different varieties, different growth stages, and different monitoring methods, obtain the body temperature value through the collection of the wearable temperature sensor 6 and the infrared camera 3, and establish the different growth conditions of the same variety under various monitoring methods through statistical processing.
  • the normal body temperature change band includes the upper curve and the lower curve, which is used as the body temperature time series change model.
  • Step 1 Compare the temperature value collected in real time with the normal body temperature change band to determine whether the temperature value falls within the normal body temperature change band:
  • step 2 If it does not fall, the body temperature of the livestock and poultry corresponding to the temperature value is abnormal, and proceed to step 2;
  • Step 2 Compare the temperature value with the range [ ⁇ -3 ⁇ , ⁇ +3 ⁇ ] determined by the following formula to determine outliers:
  • ⁇ is the average temperature of the group
  • ⁇ is the standard deviation of the group temperature
  • T cci is the body temperature value of the group monitored by the wearable temperature sensor and infrared camera
  • n is the number of the group
  • the group is divided into two situations: all livestock and poultry individuals in a single breeding area block, and all livestock and poultry individuals in all breeding area blocks. They are respectively based on the temperature value of the livestock and poultry individuals and are composed of all livestock and poultry individuals in a single breeding area block. Between groups, the temperature value of individual livestock and poultry and the group composed of all individual livestock and poultry in all breeding area blocks are compared and judged according to the above range [ ⁇ -3 ⁇ , ⁇ +3 ⁇ ]:
  • the body temperature of the livestock and poultry corresponding to the temperature value is not abnormal
  • the body temperature of the livestock and poultry corresponding to the temperature value is abnormal
  • this method can determine abnormal individuals at least 10 hours in advance.
  • Step 3 Select the livestock and poultry with abnormal body temperature status, mobilize the infrared camera to move to the breeding area block 4 where the livestock and poultry with abnormal body temperature status are located, re-inspect the livestock and poultry, and further monitor the abnormal situation.
  • the present invention sets up breeding area blocks and sentinel livestock and poultry individuals, uses two methods to monitor and evaluate the body temperature status of individuals and groups in livestock and poultry houses, can stably characterize and monitor their core temperature, and helps to quickly and stably position. and analyze the group status; the methods of body temperature monitoring and abnormality determination were optimized, time series features were introduced, and a growth stage-temperature body temperature time series change model was constructed.

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Multimedia (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Radiation Pyrometers (AREA)

Abstract

本公开涉及一种穿戴式传感器与红外相机协同的畜禽体温监测系统及方法。畜禽舍在自身的立体空间内设置不同养殖区域块，每个养殖区域块内有畜禽个体，按固定比例随机佩戴上穿戴式温度传感器形成哨兵畜禽个体，红外相机以养殖区域块获取群体红外图像，穿戴式温度传感器和红外相机均通过通讯终端和本地服务器通讯连接，本地服务器经云服务器和客户端通讯连接；方法包括数据获取、数据传输与预处理、数据库构建与模型训练、状态判定与数据复核等步骤。本公开综合穿戴式温度传感器个体连续精确监测与红外相机巡检广域大范围灵活监测的优势，实现无人化养殖场景下较低成本畜禽舍内群体的体温监测与健康评估。

Description

穿戴式传感器与红外相机协同的畜禽体温监测系统及方法 技术领域
本发明涉及畜禽体温监控领域，具体涉及基于穿戴式温度传感器哨兵个体连续监测与红外相机广域灵活巡检相结合的多源体温监控以及用以评估畜禽体温健康状况与异常预警的监测系统及方法。
背景技术
畜禽属于恒温动物，体温作为表征畜禽生理机能的重要指标，在很大程度上可以反映畜禽的健康状况，在实际养殖生产过程中越能更早地感知和发现个体的异常状态并及时采取行动，越能减少生产损失。
传统养殖场针对畜禽个体体温的监测常常采用人工随机巡检的方式，测量肛门的温度来表征畜禽的核心体温，耗费人力且容易造成畜禽的应激反应，不能够实现连续监测。目前，出现了一些基于温度传感器和红外相机的体温监测方式，这些方法都能较好地检测个体的体温，但是对基于温度数据的监测都只采用单一阈值的方式进行判定，这种方法较为粗糙，忽略了由于生长阶段及作息节律的带来的体温变化，没有考虑体温变化的时序特征，存在判别结果单一，鲁棒性差等问题，不利于精细化养殖的要求(例如对于日间体温高的动物，夜晚仍然保持日间的温度实际上已经处于异常，但单一阈值的方式无法在这一维度上进行分类)。此外，在无人化规模化养殖的背景下，采用单一温度传感器监测方式存在脱落、失灵等影响数据稳定性的状况，而采用红外相机监测单一应用成本较高，巡检的方式又不能实现连续监测，获得较好的时序特征。
发明内容
为了解决背景技术中存在的问题，本发明的目的在于提供了一种穿戴式温度传感器监测与红外相机巡检协同的畜禽体温监测系统及方法，将哨兵个体穿戴式温度传感器数据连续高精度监测与红外相机广域灵活巡检多源温度数据融合，引入时序特征构建体温变化模型，能够实现畜禽体温监控、健康状况反馈与异常预警。
本发明解决其技术问题所采用的技术方案是：
一、一种穿戴式传感器与红外相机协同的畜禽体温监测系统：
包括穿戴式温度传感器、红外相机、通讯终端、本地服务器、云服务器和 客户端；
畜禽舍在自身的立体空间内设置不同养殖区域块，每个养殖区域块内有畜禽个体，每个养殖区域块内的所有畜禽个体按固定比例随机佩戴上穿戴式温度传感器，形成哨兵畜禽个体，红外相机以养殖区域块为单位获取群体红外图像，通讯终端安装在畜禽舍内，穿戴式温度传感器和红外相机均通过通讯终端和本地服务器通讯连接，本地服务器经云服务器和客户端通讯连接。
所述的红外相机在畜禽舍内在养殖区域块之间沿预设的红外相机巡检路线进行巡检。
将畜禽舍根据立体空间结构划分成特定数量的养殖区域块并编号，养殖区域块内按特定比例随机选取个体作为哨兵，佩戴穿戴式温度传感器监控该区域内群体的体温健康状态。
所述的穿戴式温度传感器有用于电池更换提示的电池电压采集电路及无线通讯模块，采集的数据通过无线数据传输协议将数据传输到通讯终端。
二、穿戴式传感器与红外相机协同的畜禽体温监测方法，方法包括以下步骤：
S1：数据获取
分别通过穿戴式温度传感器获取养殖区域块合适数量的哨兵畜禽个体的连续体温数据，通过红外相机巡检采集各个养殖区域块内的所有畜禽个体的红外热成像图像及养殖区域块的位置；
S2：数据传输与预处理
通讯终端分别将步骤S1采集的连续体温数据和红外热成像数据传输到本地服务器，本地服务器对数据分别进行筛选、图像分割的预处理获得体温值，以预处理后的体温值处理得到每个畜禽个体的随时序变化的体温监测曲线；
S3：数据库构建与模型训练
在不同品种、不同生长阶段、不同监测方式下利用步骤S2获得的随时序变化的体温监测曲线构建生长阶段-温度数据库，通过生长阶段-温度数据库处理获得正常体温变化带(体温时序变化模型)；
S4：状态判定与数据复核
对实时采集的温度值结合正常体温变化带采用时序匹配、离群判别的方式确定体温状态是否异常，并确定体温状态异常的畜禽所在的养殖区域块作为异常区域，调动红外相机移动到异常区域复检，实现完整的监测。
所述步骤S2具体为：
将穿戴式温度传感器采集的连续温度数据和波动区间阈值进行比较判定出 异常值，对个体出现的异常值予以剔除但是保留连续出现的异常值，保留获得的每个温度值作为畜禽个体的体温值；
将红外相机采集的红外热成像图像，对每一养殖区域块的红外热成像图像采用深度学习方法进行实例分割识别出每个畜禽个体，对养殖区域块内每个畜禽个体重新编号，提取每个畜禽个体所在图像中的关键像素点及其温度信息，作为畜禽个体的体温值；
穿戴式温度传感器与红外相机直接监测到的体温值Ts和Tc分别按照以下公式处理获得核心体温值Tsc、Tcc，具体公式如下：
Tsc＝asTs+bs
Tcc＝acTc+bc
其中，as、bs分别为穿戴式温度传感器的权重和修正常数，ac、bc分别为红外相机的权重和修正常数。
所述步骤S3具体为：在不同品种、不同生长阶段、不同监测方式的情况下，通过穿戴式温度传感器和红外相机采集获得体温值，建立同一品种、各种监测方式下不同生长阶段的正常体温变化带。
在监测时，将对应时序的体温同时利用机器学习方法分别建立对应，区别于单一阈值的设定方式构建正常体温变化带。
所述的监测方式是指穿戴式温度传感器佩戴的位置。
所述步骤S3中，异常判定主要有两步骤：
步骤一，将实时采集的温度值和正常体温变化带进行比较，判断温度值是否落入正常体温变化带：
若未落入，则温度值对应的畜禽的体温状态异常，进行步骤二；
若落入，则温度值对应的畜禽的体温状态不异常，进行步骤三；
步骤二，将温度值和以下公式确定的范围[μ-3σ，μ+3σ]进行比较：

其中，μ为群体温度平均值，σ为群体温度标准差，Tcci为群体通过穿戴式温度传感器与红外相机监测到的体温值，n为群体的数量；
所述的群体分为单个养殖区域块内的所有畜禽个体、所有养殖区域块内的所有畜禽个体的两种情况，分别针对畜禽个体的温度值和由单个养殖区域块内 的所有畜禽个体构成的群体之间、畜禽个体的温度值和由所有养殖区域块内的所有畜禽个体构成的群体之间进行比较判断：
若温度值在范围[μ-3σ，μ+3σ]内，则温度值对应的畜禽的体温状态不异常；
若温度值不在范围[μ-3σ，μ+3σ]内，则温度值对应的畜禽的体温状态异常；
步骤三：选取处体温状态异常的畜禽，调动红外相机移动到体温状态异常的畜禽所在的养殖区域块对该畜禽进行复检，对异常情况进行进一步监测。
与现有发明相比，本发明具有的有益效果是：
1.该方法与系统采用两种方式监测和评估畜禽舍个体和群体的温度状况，对多源采集到的温度数据进行预处理，能够稳定地表征和监测其核心温度；
2.该方法优化了体温监测和异常判定的方法，引入时序特征，构建了基于品种-生长阶段-监测方式的体温时序变化模型，区别单一阈值判定方式，更准确甄别异常个体；
3.该方法优化了体温监测和异常判定的方法，提出一种基于大数据的离群点异常判定方法，能够有效甄别可能存在的异常个体；
4.该方法与系统采用养殖区域块和哨兵个体解决方案，通过监测哨兵状态评估个体和群体状态，养殖区域块的划分有助于数据分析处理和快速定位；
5.通过该方法与系统，穿戴式传感器和红外相机协同的多源数据采集大大减少了系统误判的可能性，能够获取更加稳定准确的体温数据；
6.实时体温监测可以及时了解畜禽异常，提早预防避免大规模损失发生。
附图说明
图1是本发明实施例的方法步骤图。
图2是本发明实施例的系统结构示意图。
图中：1-畜禽舍、2-红外相机巡检路线、3-红外相机、4-养殖区域块、5-畜禽个体、6-穿戴式温度传感器、7-通讯终端、8-本地服务器、9-云服务器、10-客户端。
图3是本发明实施例的体温综合评估工作流示意图。
图4是本发明实施例与传统体温监测方法对比图。
具体实施方式
下面结合附图和实施例，对本发明的具体实施方式作进一步详细描述。以下实施例用于说明本发明，但不用来限制本发明的范围。
如图2所示，系统包括穿戴式温度传感器6、红外相机3、通讯终端7、本地服务器8、云服务器9和客户端10；
畜禽舍1在自身的立体空间内设置不同养殖区域块4，每个养殖区域块4内有畜禽个体5，每个养殖区域块4内的所有畜禽个体5按固定比例随机佩戴上穿戴式温度传感器6，形成哨兵畜禽个体，红外相机3以养殖区域块为单位获取群体红外图像，通讯终端7安装在畜禽舍1内，穿戴式温度传感器6和红外相机3均通过通讯终端7和本地服务器8通讯连接，本地服务器8经云服务器9和客户端10通讯连接。
红外相机3在畜禽舍1内在养殖区域块4之间沿预设的红外相机巡检路线2进行巡检。
将穿戴式温度传感器6和红外相机3采集的数据上传到本地服务器8，本地服务器8或云服务器9对数据进行预处理后输入正常体温变化带进行畜禽体温的状态判定，快速定位畜禽体温异常的养殖区域块，再由红外相机3对畜禽体温异常的养殖区域块进行拍摄复核，完成监测。
复核后还可以综合根据温度数据评估个体与群体体温健康状况，结果与预警信息通过客户端10传输到用户手中。
将畜禽舍根据立体空间结构划分成特定数量的养殖区域块，每一养殖区域块对应特定的编号，每一养殖区域块内的特定数量的畜禽个体被定义为哨兵畜禽个体，哨兵畜禽个体佩戴穿戴式温度传感器连续监测个体体温变化情况，直接和间接反映群体状况，起到该区域内体温监测预警的作用。
红外巡检相机沿畜禽舍内预置的滑轨或轨迹在畜禽舍移动，每块养殖区域块拍摄一张以上热像图以采集畜禽的体温数据，图片信息还包含养殖区域块定位数据与时序信息。
穿戴式温度传感器5有用于电池更换提示的电池电压采集电路及无线通讯模块，采集的数据通过无线数据传输协议将数据传输到通讯终端7。
穿戴式温度传感器5的形状为硬币形，采用模块化设计，可以搭配硅胶表带、硅胶耳钉等定制化配件稳定佩戴于畜禽方便穿戴并可以稳定表征该个体体温的位置，包括但不限于鸡的翼下、猪的耳朵等。
穿戴式温度传感器5连续采集数据，采样频率可根据实际需要进行设定，采集的信息包含时序特征和位置信息，红外相机3采集的数据包含区域信息并且与穿戴式温度传感器5采集数据的时序特征相对应。
穿戴式温度传感器按照一定的采样频率24h不间断监测，采样数据还包含时序特征和位置信息；穿戴式温度传感器加装配件佩戴在对应畜禽可以用以稳定表征该个体体温的位置，包括但不限于鸡的翼下、猪的耳朵等部位；
在畜禽舍内为巡检红外相机预设轨道，按前期划分的区域设定巡检红外相 机巡检程序，巡检时在相应区域停留，在该区域拍摄1张以上符合要求的红外热图像，拍摄角度尽量减少遮挡；
穿戴式温度传感器和红外相机将采集的数据通过无线数据传输协议将数据传输到通讯终端，数据包含时序特征和位置信息，通讯终端以一定的时间频率将数据上传到服务器；采集的数据可以选择本地部署的服务器进行处理，也可以直接上传云服务器处理，可以根据实际情况灵活布置。
系统在实际部署时，可以灵活布置本地服务器和云服务器以方便实现本地化处理及离线处理，灵活选择提高运算效率和降低成本的方案。在本地服务器8算力充足的情况下，可直接离线对采集的数据进行处理，否则数据同样可以直接在云服务器9上处理。
系统实际应用时，根据温度传感器与红外相机与多源的温度数据，经过预处理后，基于两种异常判定方法追踪异常个体，利用红外相机数据核验，综合专家系统对异常对象进行进一步诊断并反馈给用户；
系统可以在实际应用过程中不断根据采集到的数据优化体温模型，以提高系统使用的稳定性。
本发明的实施过程如下：
S1：数据获取
分别通过穿戴式温度传感器6获取养殖区域块4合适数量的哨兵畜禽个体的连续高精度体温数据，通过红外相机3巡检采集各个养殖区域块4内的所有畜禽个体的红外热成像图像及养殖区域块4的位置，从而覆盖更多畜禽动物的红外热成像数据及其定位数据；
S2：数据传输与预处理
通讯终端7分别将步骤S1采集的连续体温数据和红外热成像数据传输到本地服务器8，本地服务器8对数据分别进行筛选、图像分割的预处理获得体温值，图像分割具体是分割出畜禽的头部，以预处理后的体温值处理得到每个畜禽个体的随时序变化的体温监测曲线；
S3：数据库构建与模型训练
在不同品种、不同生长阶段、不同监测方式下利用步骤S2获得的随时序变化的体温监测曲线构建生长阶段-温度数据库，通过生长阶段-温度数据库处理获得正常体温变化带；
S4：状态判定与数据复核
对实时采集的温度值结合正常体温变化带采用时序匹配、离群判别的方式确定体温状态是否异常，并确定体温状态异常的畜禽所在的养殖区域块4作为 异常区域，调动红外相机3移动到异常区域复检，实现完整的监测。
最后还可以综合温度数据，对畜禽舍每一块养殖区域块的群体体温健康状态给予评估与反馈。
步骤S2具体为：将穿戴式温度传感器采集的连续温度数据和波动区间阈值进行比较判定出异常值，对个体出现的异常值予以剔除但是保留连续出现的异常值，保留获得的每个温度值作为畜禽个体的体温值；保留连续出现的异常值是指异常值连续出现5次或15min以上。
将红外相机采集的红外热成像图像，对每一养殖区域块的红外热成像图像采用深度学习方法进行实例分割识别出每个畜禽个体，对养殖区域块内每个畜禽个体重新编号，提取每个畜禽个体所在图像中的关键像素点及其温度信息，作为畜禽个体的体温值；
穿戴式温度传感器与红外相机直接监测到的体温值Ts和Tc分别按照以下公式处理获得核心体温值Tsc、Tcc，具体公式如下：
Tsc＝asTs+bs
Tcc＝acTc+bc
其中，as、bs分别为穿戴式温度传感器的权重和修正常数，ac、bc分别为红外相机的权重和修正常数，a和b的值具体根据监测的对象与监测的方式预先设定。
步骤S3具体为：在不同品种、不同生长阶段、不同监测方式的情况下，通过穿戴式温度传感器6和红外相机3采集获得体温值，通过统计学处理建立同一品种、各种监测方式下不同生长阶段的正常体温变化带，正常体温变化带包含上曲线与下曲线，作为体温时序变化模型。
具体是：
步骤一，将实时采集的温度值和正常体温变化带进行比较，判断温度值是否落入正常体温变化带：
若未落入，则温度值对应的畜禽的体温状态异常，进行步骤二；
若落入，则温度值对应的畜禽的体温状态不异常，进行步骤三；
步骤二，将温度值和以下公式确定的范围[μ-3σ，μ+3σ]进行比较，实现离群点判定：

其中，μ为群体温度平均值，σ为群体温度标准差，Tcci为群体通过穿戴式温度传感器与红外相机监测到的体温值，n为群体的数量；
群体分为单个养殖区域块内的所有畜禽个体、所有养殖区域块内的所有畜禽个体的两种情况，分别针对畜禽个体的温度值和由单个养殖区域块内的所有畜禽个体构成的群体之间、畜禽个体的温度值和由所有养殖区域块内的所有畜禽个体构成的群体之间按照上述范围[μ-3σ，μ+3σ]进行比较判断：
若温度值在范围[μ-3σ，μ+3σ]内，则温度值对应的畜禽的体温状态不异常；
若温度值不在范围[μ-3σ，μ+3σ]内，则温度值对应的畜禽的体温状态异常；
如图4中的实施例所示，对比设定固定阈值的传统体温监测方法，本方法可以至少提前10h判定出异常个体。
步骤三：选取处体温状态异常的畜禽，调动红外相机移动到体温状态异常的畜禽所在的养殖区域块4对该畜禽进行复检，对异常情况进行进一步监测。
由此可见，本发明设置了养殖区域块和哨兵畜禽个体，采用两种方式监测和评估畜禽舍个体和群体的体温状态，能够稳定地表征和监测其核心温度，有助于快速稳定定位和分析群体状况；优化了体温监测和异常判定的方法，引入时序特征，构建了生长阶段-温度的体温时序变化模型，同时提出一种基于大数据的离群点异常判定方法，能够有效锁定可能存在的异常个体；多源数据采集及红外相机复核能够获取更加稳定准确的体温数据，大大减少了系统误判的可能性；实时体温监测可以及时了解畜禽异常，提早预防避免大规模损失发生。

Claims (7)

  1. 一种穿戴式传感器与红外相机协同的畜禽体温监测系统，其特征在于：
    包括穿戴式温度传感器(6)、红外相机(3)、通讯终端(7)、本地服务器(8)、云服务器(9)和客户端(10)；
    畜禽舍(1)在自身的立体空间内设置不同养殖区域块(4)，每个养殖区域块(4)内有畜禽个体(5)，每个养殖区域块(4)内的所有畜禽个体(5)按固定比例随机佩戴上穿戴式温度传感器(6)，形成哨兵畜禽个体，红外相机(3)以养殖区域块为单位获取群体红外图像，通讯终端(7)安装在畜禽舍(1)内，穿戴式温度传感器(6)和红外相机(3)均通过通讯终端(7)和本地服务器(8)通讯连接，本地服务器(8)经云服务器(9)和客户端(10)通讯连接。
  2. 根据权利要求1所述的一种穿戴式传感器与红外相机协同的畜禽体温监测系统，其特征在于：所述的红外相机(3)在畜禽舍(1)内在养殖区域块(4)之间沿预设的红外相机巡检路线(2)进行巡检。
  3. 根据权利要求1所述的一种穿戴式传感器与红外相机协同的畜禽体温监测系统，其特征在于：所述的穿戴式温度传感器(6)有用于电池更换提示的电池电压采集电路及无线通讯模块，采集的数据通过无线数据传输协议将数据传输到通讯终端7。
  4. 应用于权利要求1-3任一所述畜禽体温监测系统的穿戴式传感器与红外相机协同的畜禽体温监测方法，其特征在于方法包括以下步骤：
    S1：数据获取
    分别通过穿戴式温度传感器(6)获取养殖区域块(4)合适数量的哨兵畜禽个体的连续体温数据，通过红外相机(3)巡检采集各个养殖区域块(4)内的所有畜禽个体的红外热成像图像及养殖区域块(4)的位置；
    S2：数据传输与预处理
    通讯终端(7)分别将步骤S1采集的连续体温数据和红外热成像数据传输到本地服务器(8)，本地服务器(8)对数据分别进行筛选、图像分割的预处理获得体温值，以预处理后的体温值处理得到每个畜禽个体的随时序变化的体温监测曲线；
    S3：数据库构建与模型训练
    在不同品种、不同生长阶段和不同监测方式下利用步骤S2获得的随时序变化的体温监测曲线构建生长阶段-温度数据库，通过生长阶段-温度数据库处理获 得正常体温变化带；
    S4：状态判定与数据复核
    对实时采集的温度值结合正常体温变化带采用时序匹配和离群判别的方式确定体温状态是否异常，并确定体温状态异常的畜禽所在的养殖区域块(4)作为异常区域，调动红外相机(3)移动到异常区域复检，实现完整的监测。
  5. 根据权利要求4所述的穿戴式传感器与红外相机协同的畜禽体温监测方法，其特征在于：所述步骤S2具体为：
    将穿戴式温度传感器采集的连续温度数据和波动区间阈值进行比较判定出异常值，对个体出现的异常值予以剔除但是保留连续出现的异常值，保留获得的每个温度值作为畜禽个体的体温值；
    将红外相机采集的红外热成像图像，对每一养殖区域块的红外热成像图像采用深度学习方法进行实例分割识别出每个畜禽个体，对养殖区域块内每个畜禽个体重新编号，提取每个畜禽个体所在图像中的关键像素点及其温度信息，作为畜禽个体的体温值；
    穿戴式温度传感器与红外相机直接监测到的体温值Ts和Tc分别按照以下公式处理获得核心体温值Tsc、Tcc，具体公式如下：
    Tsc＝asTs+bs
    Tcc＝acTc+bc
    其中，as、bs分别为穿戴式温度传感器的权重和修正常数，ac、bc分别为红外相机的权重和修正常数。
  6. 根据权利要求4所述的穿戴式传感器与红外相机协同的畜禽体温监测方法，其特征在于：所述步骤S3具体为：在不同品种、不同生长阶段和不同监测方式的情况下，通过穿戴式温度传感器(6)和红外相机(3)采集获得体温值，建立同一品种和各种监测方式下不同生长阶段的正常体温变化带。
  7. 根据权利要求4所述的穿戴式传感器与红外相机协同的畜禽体温监测方法，其特征在于：所述步骤S4中，异常判定主要有两步骤：
    步骤一，将实时采集的温度值和正常体温变化带进行比较，判断温度值是否落入正常体温变化带：
    若未落入，则温度值对应的畜禽的体温状态异常，进行步骤二；
    若落入，则温度值对应的畜禽的体温状态不异常，进行步骤三；
    步骤二，将温度值和以下公式确定的范围[μ-3σ，μ+3σ]进行比较：

    其中，μ为群体温度平均值，σ为群体温度标准差，Tcci为群体通过穿戴式温度传感器与红外相机监测到的体温值，n为群体的数量；
    所述的群体分为单个养殖区域块内的所有畜禽个体、所有养殖区域块内的所有畜禽个体的两种情况，分别针对畜禽个体的温度值和由单个养殖区域块内的所有畜禽个体构成的群体之间、畜禽个体的温度值和由所有养殖区域块内的所有畜禽个体构成的群体之间进行比较判断：
    若温度值在范围[μ-3σ，μ+3σ]内，则温度值对应的畜禽的体温状态不异常；
    若温度值不在范围[μ-3σ，μ+3σ]内，则温度值对应的畜禽的体温状态异常；
    步骤三：选取体温状态异常的畜禽，调动红外相机移动到体温状态异常的畜禽所在的养殖区域块(4)对该畜禽进行复检，对异常情况进行进一步监测。
PCT/CN2023/100248 2022-07-01 2023-06-14 穿戴式传感器与红外相机协同的畜禽体温监测系统及方法 WO2024001791A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210774043.X 2022-07-01
CN202210774043.XA CN114980011B (zh) 2022-07-01 2022-07-01 穿戴式传感器与红外相机协同的畜禽体温监测方法

Publications (1)

Publication Number Publication Date
WO2024001791A1 true WO2024001791A1 (zh) 2024-01-04

Family

ID=82967808

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/100248 WO2024001791A1 (zh) 2022-07-01 2023-06-14 穿戴式传感器与红外相机协同的畜禽体温监测系统及方法

Country Status (2)

Country Link
CN (1) CN114980011B (zh)
WO (1) WO2024001791A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117686096A (zh) * 2024-01-30 2024-03-12 大连云智信科技发展有限公司 一种基于目标检测的畜禽动物体温检测方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114980011B (zh) * 2022-07-01 2023-04-07 浙江大学 穿戴式传感器与红外相机协同的畜禽体温监测方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100198024A1 (en) * 2009-02-03 2010-08-05 Ron Elazari-Volcani Vitality meter for health monitoring of anonymous animals in livestock groups
CN106974625A (zh) * 2017-03-27 2017-07-25 惠州市物联微电子有限公司 基于智能养殖场动物体温监测系统、异常体温识别方法和异常体温区域定位方法
JP2018074910A (ja) * 2016-11-07 2018-05-17 株式会社Protec 生体モニタシステム及び生体モニタ方法
CN111626985A (zh) * 2020-04-20 2020-09-04 北京农业信息技术研究中心 基于图像融合的家禽体温检测方法及禽舍巡检系统
CN112945395A (zh) * 2021-03-17 2021-06-11 西藏新好科技有限公司 一种基于目标检测的畜禽动物体温评估方法
CN114980011A (zh) * 2022-07-01 2022-08-30 浙江大学 穿戴式传感器与红外相机协同的畜禽体温监测系统及方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106020066A (zh) * 2016-07-25 2016-10-12 深圳市菲明格科技有限公司 动物体温集中监测装置及其系统
CN113923619A (zh) * 2021-10-15 2022-01-11 新希望六和股份有限公司 一种精确测温控制系统
CN114601446A (zh) * 2022-03-16 2022-06-10 内江职业技术学院 一种动物疫病的检测报警装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100198024A1 (en) * 2009-02-03 2010-08-05 Ron Elazari-Volcani Vitality meter for health monitoring of anonymous animals in livestock groups
JP2018074910A (ja) * 2016-11-07 2018-05-17 株式会社Protec 生体モニタシステム及び生体モニタ方法
CN106974625A (zh) * 2017-03-27 2017-07-25 惠州市物联微电子有限公司 基于智能养殖场动物体温监测系统、异常体温识别方法和异常体温区域定位方法
CN111626985A (zh) * 2020-04-20 2020-09-04 北京农业信息技术研究中心 基于图像融合的家禽体温检测方法及禽舍巡检系统
CN112945395A (zh) * 2021-03-17 2021-06-11 西藏新好科技有限公司 一种基于目标检测的畜禽动物体温评估方法
CN114980011A (zh) * 2022-07-01 2022-08-30 浙江大学 穿戴式传感器与红外相机协同的畜禽体温监测系统及方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
QIANGHOU LUAN, HUANG XIN; HE MINNA; LIN GUIGANG; WU JIANHUI: "Design and implementation of a wireless wearable dynamic temperature monitoring system", JOURNAL OF BIOMEDICAL ENGINEERING RESEARCH, vol. 37, no. 2, 15 June 2018 (2018-06-15), pages 215 - 218, XP093124097 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117686096A (zh) * 2024-01-30 2024-03-12 大连云智信科技发展有限公司 一种基于目标检测的畜禽动物体温检测方法
CN117686096B (zh) * 2024-01-30 2024-04-16 大连云智信科技发展有限公司 一种基于目标检测的畜禽动物体温检测方法

Also Published As

Publication number Publication date
CN114980011B (zh) 2023-04-07
CN114980011A (zh) 2022-08-30

Similar Documents

Publication Publication Date Title
WO2024001791A1 (zh) 穿戴式传感器与红外相机协同的畜禽体温监测系统及方法
CN108717678B (zh) 一种智慧养老系统
Wu et al. Using a CNN-LSTM for basic behaviors detection of a single dairy cow in a complex environment
CN107506602A (zh) 一种大数据健康预测系统
CN110333692B (zh) 基于热红外的猪只发热自动监测诊断系统
CN107943078A (zh) 多旋翼双机系统无人机巡检故障诊断系统及方法
CN109784200B (zh) 基于双目视觉的奶牛行为图像获取与体况智能监测系统
CN114037552B (zh) 一种肉鸭生理生长信息巡检方法及系统
CN116824626A (zh) 一种动物异常状态人工智能识别的方法
CN111177468A (zh) 一种基于机器视觉的实验室人员不安全行为安全检查方法
CN114842957A (zh) 一种基于情绪识别的老年痴呆症辅助诊断系统及其方法
CN109949272A (zh) 识别皮肤病种类采集人体皮肤图片的数据采集方法及系统
CN117319809A (zh) 一种监控视野的智能调节方法
CN111678558A (zh) 一种应用于充油设备的红外热成像油位、温度监测系统及方法
CN117082206A (zh) 一种基于mr+4g技术的光伏电站可穿戴式智能巡检装置
Wu et al. Body condition score for dairy cows method based on vision transformer
CN111210913A (zh) 一种基于人体关键节点识别的智能睡眠监护方法
CN113850967B (zh) 基于煤炭装车系统的铁路货车运行安全与故障预警系统
CN115868947A (zh) 一种奶牛健康预警系统和预警方法
CN109376635B (zh) 一种护理质量考核系统及安全事件上报方法
CN114022831A (zh) 一种基于双目视觉的牲畜体况监测方法及系统
CN115137353A (zh) 动物体态监测装置及方法
CN111643061A (zh) 一种基于人工智能的相貌空间集成系统
CN111554398A (zh) 一种基于5g的远程评估生命体征方法及系统
CN118318753B (zh) 一种基于人工智能的畜禽养殖场管控系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23829973

Country of ref document: EP

Kind code of ref document: A1