WO2020119659A1 - 一种智能养猪群养测重方法、装置、电子设备及存储介质 - Google Patents

一种智能养猪群养测重方法、装置、电子设备及存储介质 Download PDF

Info

Publication number
WO2020119659A1
WO2020119659A1 PCT/CN2019/124174 CN2019124174W WO2020119659A1 WO 2020119659 A1 WO2020119659 A1 WO 2020119659A1 CN 2019124174 W CN2019124174 W CN 2019124174W WO 2020119659 A1 WO2020119659 A1 WO 2020119659A1
Authority
WO
WIPO (PCT)
Prior art keywords
group
pig
image information
information
weight
Prior art date
Application number
PCT/CN2019/124174
Other languages
English (en)
French (fr)
Inventor
鞠铁柱
陈春雨
张兴福
何恒翔
Original Assignee
北京小龙潜行科技有限公司
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 北京小龙潜行科技有限公司 filed Critical 北京小龙潜行科技有限公司
Priority to AU2019399813A priority Critical patent/AU2019399813A1/en
Publication of WO2020119659A1 publication Critical patent/WO2020119659A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01GWEIGHING
    • G01G17/00Apparatus for or methods of weighing material of special form or property
    • G01G17/08Apparatus for or methods of weighing material of special form or property for weighing livestock
    • AHUMAN NECESSITIES
    • A01AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
    • A01KANIMAL HUSBANDRY; AVICULTURE; APICULTURE; PISCICULTURE; FISHING; REARING OR BREEDING ANIMALS, NOT OTHERWISE PROVIDED FOR; NEW BREEDS OF ANIMALS
    • A01K29/00Other apparatus for animal husbandry
    • A01K29/005Monitoring or measuring activity, e.g. detecting heat or mating

Definitions

  • the present application relates to the field of computer technology, and in particular to a method, device, electronic device, and storage medium for intelligent pig-raising group weight measurement.
  • the present application provides a method, device, electronic equipment and storage medium for intelligent pig breeding group weight measurement.
  • the present application provides an intelligent pig-raising group weight measurement method, including:
  • the preset weight measurement model is obtained after training based on the image sample input data of the pigs in the group pen and the weight sample result data.
  • the method further includes:
  • the SO includes:
  • training sample data includes image data and weight data of pigs in a designated group pen in a pig farm to be subjected to intelligent group rearing weight measurement;
  • the S01 includes:
  • the S02 includes:
  • the image information in the data pair of the collected pig weight information and image information is used as input, and the weight information in the corresponding data pair is used as output, and the initial weight measurement model is trained based on the method of machine learning to obtain the preset Weight measurement model.
  • the data is stored in pairs, including:
  • Multi-gesture data collection of single pigs in group pens and, comprehensive location data collection of pigs in group pens;
  • the multi-gesture data collection of single pigs in group pens includes:
  • the comprehensive location data collection of pigs in group pens includes:
  • the image collection equipment and ear tag recognition equipment are set at the preset positions around the pen; the fence is divided into two In one area, the trough is placed in one area, the sink is placed in another area, and two electronic scales are placed at the middle fence. Both electronic scales are provided with one-way doors. One of the scales can only enter the drinking area from the drinking area. In the food area, another scale can only enter the drinking area from the feeding area. The pigs walk back and forth on the two scales when drinking water for feeding. Each pig in the group pens wears ears for identification information Mark.
  • the S1 includes:
  • the method further includes:
  • the present application also provides an intelligent swine raising and weight measurement device, including:
  • the first obtaining module is configured to obtain the image information of the group pens of the pig farm to be weighed by intelligent group raising;
  • the second obtaining module is configured to obtain the image information of each pig in the group pen according to the image information of the group pen;
  • the weight measurement module is configured to obtain the weight information of each pig in the group pen according to the image information of each pig in the group pen using a preset weight measurement model;
  • the preset weight measurement model is obtained after training based on the image sample input data of the pigs in the group pen and the weight sample result data.
  • the present application also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor.
  • the processor executes the program as described in the first aspect Describe the steps of the intelligent pig-raising group weight measurement method.
  • the present application further provides a computer-readable storage medium on which a computer program is stored, which when executed by a processor implements the steps of the intelligent pig breeding group weight measurement method as described in the first aspect.
  • the smart pig breeding group weight measurement method provided in this application first obtains the image information of the group pen of the pig farm to be subjected to intelligent group breeding weight measurement, and then obtains the group according to the image information of the group pen The image information of each pig in the pen, and finally according to the image information of each pig in the pen, using a preset weight measurement model to obtain the weight information of each pig in the pen; wherein, The preset weight measurement model is obtained after training based on the image sample input data of the pigs in the group pen and the weight sample result data.
  • the smart pig breeding method for weight measurement provided by this application overcomes the various disadvantages of using the scale to measure weight in the existing methods, especially the stress response caused by driving the weight measurement, and this application solves the traditional The pain points in the pig process achieved stress-free weight measurement.
  • FIG. 1 is a flowchart of a method for intelligent pig breeding group weight measurement provided by an embodiment of the present application
  • FIG. 2 is a schematic diagram of a single pig multi-posture data collection method in a model training stage provided by an embodiment of the present application;
  • FIG. 3 is a schematic diagram of a method for collecting pig comprehensive position data during a model training stage provided by an embodiment of the present application
  • FIG. 4 is a schematic diagram of the setting of a fixed image acquisition device located above a food trough at a model application stage provided by an embodiment of the present application;
  • FIG. 5 is a schematic diagram of the setting of a fixed image acquisition device located above a water tank in a model application stage provided by an embodiment of the present application;
  • FIG. 6 is a schematic diagram of the setting of a slide rail image acquisition device located above the group feeding stage in the model application stage provided by an embodiment of the present application;
  • FIG. 7 is a schematic structural diagram of an intelligent pig group breeding weight measurement device provided by another embodiment of the present application.
  • FIG. 8 is a schematic structural diagram of an electronic device provided by another embodiment of the present application.
  • An embodiment of the present application provides an intelligent pig breeding group weight measurement method. Referring to FIG. 1, the method includes the following steps:
  • Step 101 Acquire image information of a group pen of a pig farm to be weighed by intelligent group breeding.
  • the method of predicting the weight of pigs in the group pen is based on the image of the pigs in the group pen, it is necessary to first obtain the image information of the group pen, where the image information of the group pen contains the group Image information of pigs in the pen.
  • Step 102 Acquire image information of each pig in the group pen according to the image information of the group pen.
  • the image information of the group pen is processed, such as the image segmentation algorithm, etc., to obtain the image information of each pig in the group pen.
  • Step 103 Obtain the weight information of each pig in the group pen according to the image information of each pig in the group pen using a preset weight measurement model;
  • the preset weight measurement model is obtained after training based on the image sample input data of the pigs in the group pen and the weight sample result data.
  • the image information of each pig in the group pen is input into the pre-trained weight measurement model to obtain the weight information of each pig in the group pen.
  • the smart pig group breeding weight measurement method provided in this embodiment first obtains the image information of the group pen of the pig farm to be subjected to the intelligent group breeding weight measurement, and then obtains the image information of the group pen The image information of each pig in the group pen, and finally according to the image information of each pig in the group pen, adopt the preset weight measurement model to obtain the weight information of each pig in the group pen;
  • the preset weight measurement model is obtained after training based on image sample input data and weight sample result data of pigs in group pens.
  • the smart pig breeding method for weight measurement provided in this embodiment can obtain the pig weight information based on the pig image information, thereby overcoming the various disadvantages of using the scale to measure the weight in the existing method (time-consuming and laborious, Low efficiency, inaccurate measurement, stress response, etc.), especially the stress response caused by driving weight measurement, it can be seen that this embodiment solves the pain points in the traditional pig breeding process, and achieves stress-free weight measurement. Higher market significance and practical significance.
  • the method further includes:
  • Step 100 Establish the preset weight measurement model.
  • the intelligent pig breeding group weight measurement method further includes a process of establishing a weight measurement model.
  • the step 100 includes:
  • Step 100a Prepare training sample data; the training sample data includes image data and weight data of pigs in a designated group pen in a pig farm to be subjected to intelligent group rearing weight measurement;
  • Step 100b Based on the prepared training sample data, train the initial weight measurement model based on the machine learning method to obtain the preset weight measurement model.
  • the training sample data includes image data and weight data from pigs in a designated group pen in the same pig farm, the two types of data Need to appear in pairs, that is, an image corresponds to the weight information of the pig when the image was collected, and then based on the prepared training sample data, the initial weight measurement model is trained by machine learning until the accuracy meets the preset After the request, the preset weight measurement model is obtained.
  • CNN or RNN can be used for model learning and training.
  • step 100a may be specifically implemented in the following manner:
  • step 100b may be implemented in the following manner:
  • the image information in the data pair of the collected pig weight information and image information is used as input, and the weight information in the corresponding data pair is used as output, and the initial weight measurement model is trained based on the method of machine learning to obtain the preset Weight measurement model.
  • the weight information and image information of each pig in a designated group pen in the pig farm to be subjected to intelligent group breeding weight measurement are periodically collected , And the weight information and image information collected each time are combined into data and stored in pairs, including:
  • Multi-gesture data collection of single pigs in group pens and, comprehensive location data collection of pigs in group pens;
  • the multi-gesture data collection of single pigs in group pens includes:
  • the comprehensive location data collection of pigs in group pens includes:
  • the image collection equipment and ear tag recognition equipment are set at the preset positions around the pen; the fence is divided into two In one area, the trough is placed in one area, the sink is placed in another area, and two electronic scales are placed at the middle fence. Both electronic scales are provided with one-way doors. One of the scales can only enter the drinking area from the drinking area. In the food area, another scale can only enter the drinking area from the feeding area. The pigs walk back and forth on the two scales when drinking water for feeding. Each pig in the group pens wears ears for identification information Mark.
  • the above step 100 actually belongs to the model training stage.
  • the main purpose is to collect data for model training.
  • data collection includes two aspects, one is single-pig multi-gesture data collection, and the other is comprehensive position data collection.
  • the purpose is to collect as many images as possible of a single pig in a free environment in a group feeding environment; and for the comprehensive position data collection mentioned on the other hand, The purpose is to collect images of pigs in actual group breeding application scenarios; the combination of the two data collections can collect data that is richer and closer to the actual scene, and thus can help build more accurate models.
  • the main purpose is to collect each posture video and image data of the pig as much as possible in free movement, including gesture image data of walking in free movement as much as possible, gesture image data when drinking water And posture image data during feeding.
  • an area is isolated in the group pen, a certain number of square scales as shown in Figure 2 are placed in the area, each scale is surrounded by a fence, a trough is installed on one side of the scale, and the other side
  • the sink is installed so that the pigs can walk back and forth while drinking and feeding, which is convenient for collecting more abundant image data.
  • a pig is placed in each scale, and the video and image information of the pig is collected in 24 hours.
  • Equipped with a data acquisition software system the entire acquisition process is fully automated without human intervention.
  • the data acquisition equipment located above the center of the scale provides part of the modeling data for the fixed image acquisition equipment solution and the slide rail image acquisition equipment solution. It should be noted that, in FIG. 2, the longitudinally telescopic data collection device array located above the gutter side of the scale can telescopically acquire the image above the gutter side.
  • the main purpose is to collect various posture video and image data in the actual group feeding scene.
  • a suitable number of group feeding fences are needed, and the group feeding fences must be modified to separate the fences between the group feeding fences.
  • For two areas place the trough in one area, the sink in another area, and place two electronic scales (such as the two upper and lower electronic scales in Figure 3) at the middle fence.
  • One of the two electronic scales is equipped with a one-way door, that is, one of the scales can only enter the drinking area from the drinking area, and the other scale can only enter the drinking area from the feeding area. .
  • the foregoing step 101 may be implemented as follows:
  • a fixed image acquisition device when it is necessary to collect image information of pigs in a single group pen, a fixed image acquisition device is used; when it is necessary to collect image information of pigs in multiple group pens, a slide rail image acquisition device is used; wherein, The slide rail is installed above the multiple group pens of the pig farm, and the image collection device slides on the slide rail, and each time it slides to the position above the corresponding group pen feeder or water trough, the image of the pig in the corresponding group pen is collected Information; a charging bin is installed at one or both ends of the slide rail for charging the image acquisition device; when the image acquisition device slides on the slide rail, return to the end installed on the slide rail or Charging in the charging compartment at both ends.
  • step 101 actually belongs to the model application phase.
  • the model application phase is entered, that is, step 101 actually belongs to the model application phase.
  • a fixed image acquisition device can be installed above each or each group of sinks or food troughs (as shown in Figure 4 And the collector in Figure 5) to collect image information during pig drinking or feeding. It should be noted that when it is necessary to collect image information of pigs in multiple group pens, a collector can be installed on the slide rail as shown in FIG. 6, and the collector moves on the slide rail to collect different Images of pigs in group pens.
  • a charging bin is installed at one end or both ends of the slide rail ( Figure 6 takes the charging bins at both ends as an example) for charging the image acquisition device, when the image acquisition device is in the slide After the sliding on the rail is completed, return to the charging bin installed at one or both ends of the rail to charge for the next sliding and image acquisition.
  • the data collection at the water tank or the food tank is mainly based on the user scenario.
  • the posture of the pig when eating is better than the posture of drinking water. Therefore, for rectangular food troughs, measurement can be made at the location of the food trough, but some pig farms use round food troughs. In this case, data collection needs to be performed at the water trough location.
  • the image acquisition scene for pigs in a large-scale group pen or image acquisition scene for pigs in multiple group pens it is recommended to use a slide-type image acquisition device.
  • the stage of training the model can be understood as the customization stage.
  • the weight information and the image information need to be collected to form a data pair for training the model
  • the end of the model training stage that is, after the customization stage
  • Entered the model application stage that is, entered the product promotion stage
  • the implementation of a scalable smart group breeding weight measurement system for specific users the important weight measurement algorithm model in the system comes from the data obtained by the customized system And the model obtained after training and optimization.
  • the application stage of the model there is no need to install scales for the group pens, only need to install fixed collection equipment above the sink or food trough of the group pens (that is, the “collectors” in FIGS. 4 and 5), or in the group.
  • the rail image acquisition device that is, the "collector” in Figure 6
  • the track image acquisition equipment conducts daily inspections according to the work strategy set by the customer on the cloud platform, that is, slide along the slide rail to collect image data, and transmit the data that meets the requirements to the cloud, and the track image acquisition device slides along the slide rail. After returning to the charging bin for charging, all the qualified data collected along the way have been transferred to the cloud.
  • the cloud performs image segmentation of the pigs, and then inputs the segmented pig images into the model to obtain each pig.
  • weight information is returned to the clients such as laptops, desktops, mobile phones, etc., customers can obtain the weight information of each pig, and can also obtain the bar weight information, which can be used for data analysis and evaluation for users. Pig growth, feeding levels, and recommendations for subsequent feeding.
  • the tracker in FIG. 4 is used to track each pig in the pen in real time. Its function is to record the movement trajectory of each pig, and it can also assist individual identification for identification.
  • the tracker uses an artificial intelligence algorithm to obtain a tracking model through data training, enabling real-time tracking of all pigs within a specific range.
  • the individual identification device in FIG. 4 is used to identify individual pigs, to prevent the pig from repeatedly eating and causing the system to repeatedly calculate the body weight, thereby causing a deviation in the overall bar average weight.
  • the method further includes:
  • Step 104 Obtain the bar weight information based on the weight information of each pig in the group pen and the number of pigs in the group bar; where, when obtaining the bar weight information, the duplicates are removed according to the identity information of the pigs Pig weight information.
  • each pig in the group pen has an ear tag indicating identity information
  • the duplicate pig weight information is removed according to the pig's identity information, so that The weight information is more accurate.
  • the device includes: a first acquisition module 21, a second acquisition module 22, and a weight measurement module 23. among them:
  • the first obtaining module 21 is configured to obtain image information of a group pen of a pig farm to be subjected to intelligent group rearing weight measurement;
  • the second obtaining module 22 is configured to obtain the image information of each pig in the group pen according to the image information of the group pen;
  • the weight measuring module 23 is configured to obtain the weight information of each pig in the group pen according to the image information of each pig in the group pen using a preset weight measurement model;
  • the preset weight measurement model is obtained after training based on the image sample input data of the pigs in the group pen and the weight sample result data.
  • the smart swine herd breeding weight measurement device provided in this embodiment can be used to perform the smart swine herd breeding weight measurement method described in the above embodiment, and its working principle and technical effect are similar.
  • the smart swine herd breeding weight measurement device provided in this embodiment can be used to perform the smart swine herd breeding weight measurement method described in the above embodiment, and its working principle and technical effect are similar.
  • the electronic device specifically includes the following content: a processor 801, a memory 802, a communication interface 803, and a bus 804;
  • the processor 801, the memory 802, and the communication interface 803 complete communication with each other through the bus 804; the communication interface 803 is configured to implement communication between various modeling software and related equipment such as intelligent manufacturing equipment module libraries Information transfer;
  • the processor 801 is configured to call a computer program in the memory 802.
  • the processor executes the computer program, all the steps in the first embodiment described above are implemented. For example, when the processor executes the computer program Implement the following steps:
  • Step 101 Acquire image information of a group pen of a pig farm to be weighed by intelligent group breeding.
  • Step 102 Acquire image information of each pig in the group pen according to the image information of the group pen.
  • Step 103 Obtain the weight information of each pig in the group pen according to the image information of each pig in the group pen using a preset weight measurement model;
  • the preset weight measurement model is obtained after training based on the image sample input data of the pigs in the group pen and the weight sample result data.
  • yet another embodiment of the present application provides a computer-readable storage medium that stores a computer program on the computer-readable storage medium, and when the computer program is executed by a processor, implements all the steps of the first embodiment For example, when the processor executes the computer program, the following steps are realized:
  • Step 101 Acquire image information of a group pen of a pig farm to be weighed by intelligent group breeding.
  • Step 102 Acquire image information of each pig in the group pen according to the image information of the group pen.
  • Step 103 Obtain the weight information of each pig in the group pen according to the image information of each pig in the group pen using a preset weight measurement model;
  • the preset weight measurement model is obtained after training based on the image sample input data of the pigs in the group pen and the weight sample result data.

Landscapes

  • Life Sciences & Earth Sciences (AREA)
  • Environmental Sciences (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Biophysics (AREA)
  • Animal Husbandry (AREA)
  • Biodiversity & Conservation Biology (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Image Analysis (AREA)

Abstract

一种智能养猪群养测重方法、装置、电子设备及存储介质,测重方法包括:S1、获取待进行智能群养测重的养猪场群养栏的图像信息(101);S2、根据群养栏的图像信息,获取群养栏内每个猪只的图像信息(102);S3、根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;其中,预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的(103)。该智能养猪群养测重方法,可以有效解决现有养猪过程中因驱赶猪只至共有地秤上进行称重而带来的应激反应问题。

Description

一种智能养猪群养测重方法、装置、电子设备及存储介质
交叉引用
本申请引用于2018年12月13日提交的专利名称为“一种智能养猪群养测重方法、装置、电子设备及存储介质”的第2018115255707号中国专利申请,其通过引用被全部并入本申请。
技术领域
本申请涉及计算机技术领域,具体涉及一种智能养猪群养测重方法、装置、电子设备及存储介质。
背景技术
目前,养猪场在猪只的养殖过程中,需要在猪只的不同生长阶段获取猪只的体重信息,以便及时了解猪只的生长情况。
对于群养栏内猪只的体重测重,为了尽可能地降低共用地秤测重时因驱赶猪只而造成的应激反应,目前一般是在每只猪的群养栏上放置一台地秤用于称重,但是这样的方式缺点是成本较高,且由于猪只在地秤上来回走动,导致地秤晃动,测量数据不稳定,有时候会产生10-20公斤的误差。此外,由于猪场环境恶劣,地秤在猪场使用一段时间,往往由于粪尿侵蚀等原因造成地秤损坏,从而进一步增加了更换成本。
发明内容
针对现有技术中的缺陷,本申请提供一种智能养猪群养测重方法、装置、电子设备及存储介质。
具体地,本申请提供以下技术方案:
第一方面,本申请提供了一种智能养猪群养测重方法,包括:
S1、获取待进行智能群养测重的养猪场群养栏的图像信息;
S2、根据群养栏的图像信息,获取群养栏内每个猪只的图像信息;
S3、根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;
其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入 数据和体重样本结果数据进行训练后得到的。
进一步地,在所述S1之前,所述方法还包括:
S0、建立所述预设的体重测量模型。
进一步地,所述S0包括:
S01、准备训练样本数据;所述训练样本数据包括待进行智能群养测重的养猪场中指定群养栏内猪只的图像数据和体重数据;
S02、基于准备的训练样本数据,基于机器学习的方法对初始体重测量模型进行训练,得到所述预设的体重测量模型。
进一步地,所述S01包括:
定期采集所述待进行智能群养测重的养猪场内指定的群养栏内每个猪只的体重信息和图像信息,并将每次采集的体重信息和图像信息组成数据对后进行成对存储;
相应地,所述S02包括:
将采集得到的猪只体重信息和图像信息的数据对中的图像信息作为输入,将相应数据对中的体重信息作为输出,基于机器学习的方法对初始体重测量模型进行训练,得到所述预设的体重测量模型。
进一步地,所述定期采集所述待进行智能群养测重的养猪场内指定的群养栏内每个猪只的体重信息和图像信息,并将每次采集的体重信息和图像信息组成数据对后进行成对存储,包括:
进行群养栏内单猪多姿态数据采集;以及,进行群养栏内猪只综合位置数据采集;
其中,进行群养栏内单猪多姿态数据采集,包括:
采集群养栏内单个猪只在饮水采食过程中各姿态下的图像信息以及对应的体重信息;其中,群养栏内放置有预设数量的正方形地秤,每个地秤的上方设置有图像采集设备,每个地秤四周设有围栏,每个地秤内放一只猪,每个地秤的一侧安装食槽,另一侧安装水槽,猪只饮水采食时在地秤上来回走动;
其中,进行群养栏内猪只综合位置数据采集,包括:
采集群养栏内猪只在综合位置下的图像信息以及对应的体重信息;其中,群养栏周边预设位置处设置有图像采集设备和耳标识别设备;群养栏 中间采用栅栏分开为两个区域,食槽放置在一个区域,水槽放置在另外一个区域,在中间栅栏处放置两个电子地秤,两个电子地秤均设置单向门,其中一个地秤只能从饮水区进入采食区,另一个地秤只能从采食区进入饮水区,猪只饮水采食时在两个地秤上来回走动,群养栏内的每个猪只均佩戴有用于表示身份信息的耳标。
进一步地,所述S1包括:
利用安装在食槽或水槽上方的固定式图像采集设备或利用滑轨式图像采集设备获取待进行智能群养测重的养猪场群养栏的图像信息;其中,当需要采集单个群养栏内猪只的图像信息时,采用固定式图像采集设备;当需要采集多个群养栏内猪只的图像信息时,采用滑轨式图像采集设备,其中滑轨安装在养猪场多个群养栏的上方,图像采集设备在所述滑轨上滑动,每滑动到相应群养栏食槽或水槽上方位置时采集对应群养栏内猪只的图像信息;所述滑轨的一端或两端安装有充电仓,用于为所述图像采集设备充电;当所述图像采集设备在所述滑轨上滑动结束后,回至安装在滑轨一端或两端的充电仓内充电。
进一步地,所述方法还包括:
S4、根据群养栏内每个猪只的体重信息以及群养栏内的猪只数量信息获取栏均重信息;其中,在获取栏均重信息时,根据猪只的身份信息去除重复的猪只体重信息。
第二方面,本申请还提供了一种智能养猪群养测重装置,包括:
第一获取模块,被配置为获取待进行智能群养测重的养猪场群养栏的图像信息;
第二获取模块,被配置为根据群养栏的图像信息,获取群养栏内每个猪只的图像信息;
测重模块,被配置为根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;
其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的。
第三方面,本申请还提供了一种电子设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,所述处理器执行所述程 序时实现如第一方面所述智能养猪群养测重方法的步骤。
第四方面,本申请还提供了一种计算机可读存储介质,其上存储有计算机程序,该计算机程序被处理器执行时实现如第一方面所述智能养猪群养测重方法的步骤。
由上面技术方案可知,本申请提供的智能养猪群养测重方法,首先获取待进行智能群养测重的养猪场群养栏的图像信息,然后根据群养栏的图像信息,获取群养栏内每个猪只的图像信息,最后根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的。可见,本申请提供的智能养猪群养测重方法,克服了现有方式中采用地秤进行测重的各种弊端,尤其是因驱赶测重导致的应激反应,本申请解决了传统养猪过程中的痛点,实现了无应激测重。
附图说明
为了更清楚地说明本申请实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1是本申请一实施例提供的智能养猪群养测重方法的流程图;
图2是本申请一实施例提供的模型训练阶段单猪多姿态数据采集方式示意图;
图3是本申请一实施例提供的模型训练阶段猪只综合位置数据采集方式示意图;
图4是本申请一实施例提供的模型应用阶段位于食槽上方的固定式图像采集设备的设置示意图;
图5是本申请一实施例提供的模型应用阶段位于水槽上方的固定式图像采集设备的设置示意图;
图6是本申请一实施例提供的模型应用阶段位于群养栏上方的滑轨式图像采集设备的设置示意图;
图7是本申请另一实施例提供的智能养猪群养测重装置的结构示意图;
图8是本申请又一实施例提供的电子设备的结构示意图。
具体实施方式
为使本申请实施例的目的、技术方案和优点更加清楚,下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整的描述,显然,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
本申请一实施例提供了一种智能养猪群养测重方法,参见图1,该方法包括如下步骤:
步骤101:获取待进行智能群养测重的养猪场群养栏的图像信息。
在本步骤中,由于采用的是根据群养栏猪只图像预测群养栏内猪只重量的方法,因此,需要先获取群养栏的图像信息,这里群养栏的图像信息中包含有群养栏内猪只的图像信息。
步骤102:根据群养栏的图像信息,获取群养栏内每个猪只的图像信息。
在本步骤中,对群养栏的图像信息进行处理,如采用图像分割算法等,获取群养栏内每个猪只的图像信息。
步骤103:根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;
其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的。
在本步骤中,将群养栏内每个猪只的图像信息,输入至预先训练好的体重测量模型,获取群养栏内每个猪只的体重信息。
由上面技术方案可知,本实施例提供的智能养猪群养测重方法,首先获取待进行智能群养测重的养猪场群养栏的图像信息,然后根据群养栏的图像信息,获取群养栏内每个猪只的图像信息,最后根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的。可见,本实施例提供的智能养猪群养测重方法,根据猪只图像信息即可获知猪只重量信息,从而克 服了现有方式中采用地秤进行测重的各种弊端(费时费力,效率低下,测不准,产生应激反应等),尤其是因驱赶测重导致的应激反应,可见,本实施例解决了传统养猪过程中的痛点,实现了无应激测重,具有较高的市场意义和现实意义。
基于上述实施例的内容,在一种可选实施方式中,在所述步骤101之前,所述方法还包括:
步骤100:建立所述预设的体重测量模型。
在本实施方式中,所述智能养猪群养测重方法还包括了建立体重测量模型的过程。
基于上述实施例的内容,在一种可选实施方式中,所述步骤100包括:
步骤100a:准备训练样本数据;所述训练样本数据包括待进行智能群养测重的养猪场中指定群养栏内猪只的图像数据和体重数据;
步骤100b:基于准备的训练样本数据,基于机器学习的方法对初始体重测量模型进行训练,得到所述预设的体重测量模型。
在本实施方式中,在建立体重测量模型时,先准备训练样本数据,所述训练样本数据包括来自于同一养猪场中指定群养栏内猪只的图像数据和体重数据,这两种数据需要成对出现,也即一幅图像要对应采集该幅图像时猪只的重量信息,然后基于准备的训练样本数据,采用机器学习的方法对初始体重测量模型进行训练,直到精确度满足预设要求后得到所述预设的体重测量模型。
需要说明的是,这里在基于机器学习的方法对初始模型进行训练时,可以采用CNN或RNN进行模型学习和训练。
基于上述实施例的内容,在一种可选实施方式中,所述步骤100a具体可通过下面方式实现:
定期采集所述待进行智能群养测重的养猪场内指定的群养栏内每个猪只的体重信息和图像信息,并将每次采集的体重信息和图像信息组成数据对后进行成对存储;
相应地,所述步骤100b具体可通过下面方式实现:
将采集得到的猪只体重信息和图像信息的数据对中的图像信息作为输入,将相应数据对中的体重信息作为输出,基于机器学习的方法对初始 体重测量模型进行训练,得到所述预设的体重测量模型。
基于上述实施例的内容,在一种可选实施方式中,所述定期采集所述待进行智能群养测重的养猪场内指定的群养栏内每个猪只的体重信息和图像信息,并将每次采集的体重信息和图像信息组成数据对后进行成对存储,包括:
进行群养栏内单猪多姿态数据采集;以及,进行群养栏内猪只综合位置数据采集;
其中,进行群养栏内单猪多姿态数据采集,包括:
采集群养栏内单个猪只在饮水采食过程中各姿态下的图像信息以及对应的体重信息;其中,群养栏内放置有预设数量的正方形地秤,每个地秤的上方设置有图像采集设备,每个地秤四周设有围栏,每个地秤内放一只猪,每个地秤的一侧安装食槽,另一侧安装水槽,猪只饮水采食时在地秤上来回走动;
其中,进行群养栏内猪只综合位置数据采集,包括:
采集群养栏内猪只在综合位置下的图像信息以及对应的体重信息;其中,群养栏周边预设位置处设置有图像采集设备和耳标识别设备;群养栏中间采用栅栏分开为两个区域,食槽放置在一个区域,水槽放置在另外一个区域,在中间栅栏处放置两个电子地秤,两个电子地秤均设置单向门,其中一个地秤只能从饮水区进入采食区,另一个地秤只能从采食区进入饮水区,猪只饮水采食时在两个地秤上来回走动,群养栏内的每个猪只均佩戴有用于表示身份信息的耳标。
在本实施方式中,上述步骤100实际上属于模型训练阶段,在该阶段,主要目的是采集数据,用于模型训练。在本实施方式中,数据采集包括两方面的内容,一方面是单猪多姿态数据采集,另一方面是综合位置数据采集。需要说明的是,对于一方面提到的单猪多姿态数据采集,目的是采集群养环境中单只猪在自由状态下尽量多的图像;而对于另一方面提到的综合位置数据采集,目的是采集猪在实际群养应用场景下的图像;两种数据采集综合起来,能够采集到更丰富,更接近实际场景的数据,因而可以帮助建立更精确的模型。
具体来说,对于单猪多姿态数据采集,主要目的是采集猪只在尽量自 由活动情况下的各姿态视频、图像数据,包括尽量自由活动情况下行走的姿态图像数据、饮水时的姿态图像数据和采食时的姿态图像数据。具体实施时,在群养栏中隔离出一块区域,在区域内放置一定数目如图2所示正方形地秤,每个地秤四周设有围栏,在地秤一侧安装食槽,另一侧安装水槽,猪只饮水采食时来回走动,便于采集更丰富的图像数据。每个地秤内放一只猪,24小时采集该猪只视频、图像信息。配备数据采集软件系统,使整体采集过程完全自动化,无需人为干预。位于地秤中心上方的数据采集设备为固定式图像采集设备方案、滑轨式图像采集设备方案提供一部分建模数据。需要说明的是,图2中位于地秤食槽侧上方的纵向伸缩数据采集设备阵列能够以伸缩方式获取食槽侧上方的图像。
具体来说,对于综合位置数据采集,主要目的是采集实际群养场景下各姿态视频、图像数据,需要合适数目的群养栏,对群养栏做一定改造,在群养栏中间用栅栏分开为两个区域,将食槽放置在一个区域,水槽放置在另外一个区域,在中间栅栏处放置两个电子地秤(如图3中的上下两个电子地秤)。对采集猪只佩戴耳标并在猪背部做一定标记。配备数据采集软件系统,使整体采集过程完全自动化,无需人为干预。改造后采集场景如图3所示,其中两个电子地秤均设置单向门,即其中一个地秤只能从饮水区进入采食区,另一个地秤只能从采食区进入饮水区。在标有数字上方合适位置安装图像数据采集设备。在标有大写字母上方合适位置安装追踪设备,辅助数据采集。在标有小写字母附近合适位置安装个体识别数据采集设备。在食槽及电子地秤合适位置安装耳标读取设备辅助数据采集,在食槽耳标读取设备周围安装隔离设施,保证各读取设备互不干扰。
需要说明的是,通过上面两种数据采集方式,可以保证能够充分采集中猪只在行走、饮水、采食等各种姿态下的图像信息以及对应的重量信息,从而为模型训练提供了足够的样本数据,进而通过这些样本数据对模型训练后,可以使得模型能够适应各种场景下的猪只姿态图像。
基于上述实施例的内容,在一种可选实施方式中,上述步骤101可通过如下方式实现:
利用安装在食槽或水槽上方的固定式图像采集设备(参见图4和图5中的采集器),或,利用滑轨式图像采集设备获取待进行智能群养测重的 养猪场群养栏的图像信息(参见图6中的采集器);
其中,当需要采集单个群养栏内猪只的图像信息时,采用固定式图像采集设备;当需要采集多个群养栏内猪只的图像信息时,采用滑轨式图像采集设备;其中,滑轨安装在养猪场多个群养栏的上方,图像采集设备在所述滑轨上滑动,每滑动到相应群养栏食槽或水槽上方位置时采集对应群养栏内猪只的图像信息;所述滑轨的一端或两端安装有充电仓,用于为所述图像采集设备充电;当所述图像采集设备在所述滑轨上滑动结束后,回至安装在滑轨一端或两端的充电仓内充电。
在本实施方式中,上述步骤100的模型训练阶段结束后,进入了模型应用阶段,也即步骤101实际上属于模型应用阶段的内容。在模型应用阶段,将不再需要为群养栏安装地秤,而只需获取每个群养栏内猪只的图像信息,进而利用获取的图像信息以及训练好的模型即可获取相应群养栏内猪只的体重信息。
在本实施方式中,在模型应用阶段,由于需要获取养猪场群养栏内猪只的图像信息,因此可以在每个或每组水槽或食槽上方安装固定式图像采集设备(如图4和图5中的采集器),采集猪只饮水或采食过程中的图像信息。需要说明的是,当需要采集多个群养栏内猪只的图像信息时,可以如图6所示的方式,在滑轨上安装采集器,采集器在滑轨上移动,用于采集不同群养栏内猪只图像。其中,滑轨的一端或两端安装有充电仓(图6以两端均安装有充电仓为例进行举例),用于为所述图像采集设备充电,当所述图像采集设备在所述滑轨上滑动结束后,回至安装在滑轨一端或两端的充电仓内充电,以便进行下一次的滑动和图像采集。
需要说明的是,在水槽处还是食槽处进行数据采集,主要是从用户场景出发,一般来说,猪采食时姿态好于饮水时姿态。因此对于长方形食槽,可采用食槽位置处测量,但有些猪场采用圆形食槽,此种情况,就需要在水槽位置处进行数据采集。此外,如果对于大面积群养栏的猪只图像获取场景或多个群养栏的猪只图像获取场景,建议采用滑轨式图像采集设备。
需要说明的是,训练模型的阶段可以理解为定制化阶段,在这个阶段,需要采集体重信息和图像信息形成数据对,用于对模型进行训练,而模型训练阶段结束也即定制化阶段结束后,进入了模型应用阶段,也即进入了 产品推广阶段,在这个阶段,为特定用户实施可推广的智能群养测重系统,该系统中重要的测重算法模型来自于定制化系统获取的数据以及训练并优化后得到的模型。在模型应用阶段,不需要再为群养栏安装地秤,只需在群养栏的水槽或食槽上方安装固定采集设备(即图4和图5中的“采集器”),或在群养栏上方安装轨道,然后将轨道图像采集设备(即图6中的“采集器”)安装到滑轨上,在滑轨一侧或两侧安装无线充电仓,并将硬件设备接入远端云平台。工作时,轨道图像采集设备按照客户设置在云平台的工作策略,每天进行巡视,即沿滑轨进行滑动采集图像数据,并将符合要求的数据传送至云端,轨道图像采集设备沿滑轨滑动完毕后,回至充电仓内进行充电,沿途采集到的所有合格的数据均已传至云端,云端对图像进行猪只图像分割,然后将分割后的猪只图像输入至模型中,得到每个猪只的体重信息,返回至笔记本电脑、台式机、手机等客户端上,客户可实施得到每只猪的体重信息,还可获取栏均重信息,进而根据这些信息可为用户进行数据分析,评价猪只生长、饲喂水平,并对后续饲喂给出建议。
需要说明的是,图4中的追踪器,用于对栏内的每只猪进行实时跟踪,其作用是记录每只猪的行动轨迹,也可辅助个体识别进行身份鉴定。追踪器采用人工智能算法,通过数据训练,得到追踪模型,实现对特定范围内的所有猪只进行实时追踪。此外,图4中的个体识别设备,用于识别猪只个体,防止猪多次采食导致系统重复计算体重,从而导致整体栏均重的偏差。
基于上述实施例的内容,在一种可选实施方式中,所述方法还包括:
步骤104:根据群养栏内每个猪只的体重信息以及群养栏内的猪只数量信息获取栏均重信息;其中,在获取栏均重信息时,根据猪只的身份信息去除重复的猪只体重信息。
需要说明的是,由于群养栏内每个猪只都带有表示身份信息的耳标,因此,在获取栏均重信息时,根据猪只的身份信息去除重复的猪只体重信息,因此得到的均重信息更准确。
基于相同的发明构思,本申请另一实施例提供了一种智能养猪群养测重装置,参见图7,该装置包括:第一获取模块21、第二获取模块22和 测重模块23,其中:
第一获取模块21,被配置为获取待进行智能群养测重的养猪场群养栏的图像信息;
第二获取模块22,被配置为根据群养栏的图像信息,获取群养栏内每个猪只的图像信息;
测重模块23,被配置为根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;
其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的。
需要说明的是,本实施例提供的智能养猪群养测重装置可以用于执行上述实施例所述的智能养猪群养测重方法,其工作原理和技术效果类似,具体可参见上述实施例的描述,此处不再详述。
基于相同的发明构思,本申请又一实施例提供了一种电子设备,参见图8,所述电子设备具体包括如下内容:处理器801、存储器802、通信接口803和总线804;
其中,所述处理器801、存储器802和通信接口803通过所述总线804完成相互间的通信;所述通信接口803被配置为实现各建模软件及智能制造装备模块库等相关设备之间的信息传输;
所述处理器801被配置为调用所述存储器802中的计算机程序,所述处理器执行所述计算机程序时实现上述实施例一中的全部步骤,例如,所述处理器执行所述计算机程序时实现下述步骤:
步骤101:获取待进行智能群养测重的养猪场群养栏的图像信息。
步骤102:根据群养栏的图像信息,获取群养栏内每个猪只的图像信息。
步骤103:根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;
其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的。
基于相同的发明构思,本申请又一实施例提供了一种计算机可读存储介质,该计算机可读存储介质上存储有计算机程序,该计算机程序被处理 器执行时实现上述实施例一的全部步骤,例如,所述处理器执行所述计算机程序时实现下述步骤:
步骤101:获取待进行智能群养测重的养猪场群养栏的图像信息。
步骤102:根据群养栏的图像信息,获取群养栏内每个猪只的图像信息。
步骤103:根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;
其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的。
需要说明的是,在本文中,诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来,而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
以上实施例仅用于说明本申请的技术方案,而非对其限制;尽管参照前述实施例对本申请进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或替换,并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围。

Claims (10)

  1. 一种智能养猪群养测重方法,其特征在于,包括:
    S1、获取待进行智能群养测重的养猪场群养栏的图像信息;
    S2、根据群养栏的图像信息,获取群养栏内每个猪只的图像信息;
    S3、根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;
    其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的。
  2. 根据权利要求1所述的方法,其特征在于,在所述S1之前,所述方法还包括:
    S0、建立所述预设的体重测量模型。
  3. 根据权利要求2所述的方法,其特征在于,所述S0包括:
    S01、准备训练样本数据;所述训练样本数据包括待进行智能群养测重的养猪场中指定群养栏内猪只的图像数据和体重数据;
    S02、基于准备的训练样本数据,基于机器学习的方法对初始体重测量模型进行训练,得到所述预设的体重测量模型。
  4. 根据权利要求3所述的方法,其特征在于,所述S01包括:
    定期采集所述待进行智能群养测重的养猪场内指定的群养栏内每个猪只的体重信息和图像信息,并将每次采集的体重信息和图像信息组成数据对后进行成对存储;
    相应地,所述S02包括:
    将采集得到的猪只体重信息和图像信息的数据对中的图像信息作为输入,将相应数据对中的体重信息作为输出,基于机器学习的方法对初始体重测量模型进行训练,得到所述预设的体重测量模型。
  5. 根据权利要求4所述的方法,其特征在于,所述定期采集所述待进行智能群养测重的养猪场内指定的群养栏内每个猪只的体重信息和图像信息,并将每次采集的体重信息和图像信息组成数据对后进行成对存储,包括:
    进行群养栏内单猪多姿态数据采集;以及,进行群养栏内猪只综合位置数据采集;
    其中,进行群养栏内单猪多姿态数据采集,包括:
    采集群养栏内单个猪只在饮水采食过程中各姿态下的图像信息以及对应的体重信息;其中,群养栏内放置有预设数量的正方形地秤,每个地秤的上方设置有图像采集设备,每个地秤四周设有围栏,每个地秤内放一只猪,每个地秤的一侧安装食槽,另一侧安装水槽,猪只饮水采食时在地秤上来回走动;
    其中,进行群养栏内猪只综合位置数据采集,包括:
    采集群养栏内猪只在综合位置下的图像信息以及对应的体重信息;其中,群养栏周边预设位置处设置有图像采集设备和耳标识别设备;群养栏中间采用栅栏分开为两个区域,食槽放置在一个区域,水槽放置在另外一个区域,在中间栅栏处放置两个电子地秤,两个电子地秤均设置单向门,其中一个地秤只能从饮水区进入采食区,另一个地秤只能从采食区进入饮水区,猪只饮水采食时在两个地秤上来回走动,群养栏内的每个猪只均佩戴有用于表示身份信息的耳标。
  6. 根据权利要求5所述的方法,其特征在于,所述S1包括:
    利用安装在食槽或水槽上方的固定式图像采集设备或利用滑轨式图像采集设备获取待进行智能群养测重的养猪场群养栏的图像信息;其中,当需要采集单个群养栏内猪只的图像信息时,采用固定式图像采集设备;当需要采集多个群养栏内猪只的图像信息时,采用滑轨式图像采集设备,其中滑轨安装在养猪场多个群养栏的上方,图像采集设备在所述滑轨上滑动,每滑动到相应群养栏食槽或水槽上方位置时采集对应群养栏内猪只的图像信息;所述滑轨的一端或两端安装有充电仓,用于为所述图像采集设备充电;当所述图像采集设备在所述滑轨上滑动结束后,回至安装在滑轨一端或两端的充电仓内充电。
  7. 根据权利要求5所述的方法,其特征在于,所述方法还包括:
    S4、根据群养栏内每个猪只的体重信息以及群养栏内的猪只数量信息获取栏均重信息;其中,在获取栏均重信息时,根据猪只的身份信息去除重复的猪只体重信息。
  8. 一种智能养猪群养测重装置,其特征在于,包括:
    第一获取模块,被配置为获取待进行智能群养测重的养猪场群养栏的 图像信息;
    第二获取模块,被配置为根据群养栏的图像信息,获取群养栏内每个猪只的图像信息;
    测重模块,被配置为根据群养栏内每个猪只的图像信息,采用预设的体重测量模型,获取群养栏内每个猪只的体重信息;
    其中,所述预设的体重测量模型是基于群养栏内猪只的图像样本输入数据和体重样本结果数据进行训练后得到的。
  9. 一种电子设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,其特征在于,所述处理器执行所述程序时实现如权利要求1至7任一项所述智能养猪群养测重方法的步骤。
  10. 一种计算机可读存储介质,其上存储有计算机程序,其特征在于,该计算机程序被处理器执行时实现如权利要求1至7任一项所述智能养猪群养测重方法的步骤。
PCT/CN2019/124174 2018-12-13 2019-12-10 一种智能养猪群养测重方法、装置、电子设备及存储介质 WO2020119659A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU2019399813A AU2019399813A1 (en) 2018-12-13 2019-12-10 Intelligent pig group rearing weighing method and apparatus, electronic device and storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811525570.7 2018-12-13
CN201811525570.7A CN109632058B (zh) 2018-12-13 2018-12-13 一种智能养猪群养测重方法、装置、电子设备及存储介质

Publications (1)

Publication Number Publication Date
WO2020119659A1 true WO2020119659A1 (zh) 2020-06-18

Family

ID=66073671

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/124174 WO2020119659A1 (zh) 2018-12-13 2019-12-10 一种智能养猪群养测重方法、装置、电子设备及存储介质

Country Status (3)

Country Link
CN (1) CN109632058B (zh)
AU (2) AU2019399813A1 (zh)
WO (1) WO2020119659A1 (zh)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112106689A (zh) * 2020-10-16 2020-12-22 中国农业大学 对虾料台观测系统
CN112841074A (zh) * 2020-12-31 2021-05-28 京东数科海益信息科技有限公司 检测方法、装置、设备及存储介质
CN112990073A (zh) * 2021-03-31 2021-06-18 南京农业大学 基于边缘计算的哺乳期仔猪活动规律统计系统
CN113124971A (zh) * 2021-05-14 2021-07-16 深圳市路安仪器设备有限公司 一种线性称重方法
CN113780073A (zh) * 2021-08-03 2021-12-10 华南农业大学 一种辅助估测鸡群均匀度的装置与方法
CN114067364A (zh) * 2021-11-23 2022-02-18 江苏省家禽科学研究所 一种基于图像采集的鸡只自动称重装置
CN114543674A (zh) * 2022-02-22 2022-05-27 成都睿畜电子科技有限公司 一种基于图像识别的检测方法及系统
CN115226651A (zh) * 2022-08-19 2022-10-25 深圳进化动力数码科技有限公司 一种智慧养猪的生猪估重方法、装置、电子设备及介质
CN115226644A (zh) * 2022-08-05 2022-10-25 中国农业大学 一种猪只分群饲喂管理装置和系统
CN116058302A (zh) * 2023-03-15 2023-05-05 江苏省家禽科学研究所 一种自动称重的肉鸡养殖笼具
CN116091786A (zh) * 2023-04-11 2023-05-09 厦门农芯数字科技有限公司 猪只估重的全息体尺自编码方法、系统、设备及存储介质
CN117528034A (zh) * 2023-12-29 2024-02-06 常州辉途智能科技有限公司 一种牧场智能养殖系统及工作方法
CN117718244A (zh) * 2024-02-08 2024-03-19 中国农业科学院北京畜牧兽医研究所 一种针对牛群的分群称重系统

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109632058B (zh) * 2018-12-13 2020-06-02 北京小龙潜行科技有限公司 一种智能养猪群养测重方法、装置、电子设备及存储介质
CN110426112B (zh) * 2019-07-04 2022-05-13 平安科技(深圳)有限公司 一种生猪体重测量方法及装置
CN110580916B (zh) * 2019-08-13 2022-01-14 北京农业信息技术研究中心 体重声学测量模型的创建方法与体重测量方法、装置
CN110612921A (zh) * 2019-09-25 2019-12-27 农芯科技(广州)有限责任公司 一种定位栏母猪的监测系统及方法
CN110672189A (zh) * 2019-09-27 2020-01-10 北京海益同展信息科技有限公司 体重估计方法、装置、系统和存储介质
CN110558248B (zh) * 2019-09-27 2022-06-28 重庆大直科技有限公司 一种猪只数量监控装置
CN111183917B (zh) * 2019-11-14 2022-02-01 京东科技信息技术有限公司 一种动物异常监测、图像处理方法及装置
CN111418509A (zh) * 2020-01-16 2020-07-17 秒针信息技术有限公司 家畜的投喂方法和装置、存储介质及电子装置
CN111264405A (zh) * 2020-02-19 2020-06-12 北京海益同展信息科技有限公司 一种饲喂方法、系统、装置、设备和计算机可读存储介质
CN111369378A (zh) * 2020-02-25 2020-07-03 成都睿畜电子科技有限公司 一种基于计算机视觉识别的生猪监管方法及系统
CN111507432A (zh) * 2020-07-01 2020-08-07 四川智迅车联科技有限公司 农险理赔智能称重方法、系统、电子设备和存储介质
CN112903078B (zh) * 2021-01-21 2023-01-10 核工业四一六医院 一种实验鼠数据采集装置及采集方法
CN113063479A (zh) * 2021-03-16 2021-07-02 深圳市归位科技有限公司 一种结合跟踪系统和磅秤的畜牧估重装置和方法
CN115331266B (zh) * 2022-10-17 2023-02-10 天津大学四川创新研究院 一种猪只唯一化识别去重警报方法
CN116206342B (zh) * 2023-04-27 2023-07-25 广东省农业科学院动物科学研究所 一种猪只体重检测方法、装置、设备及存储介质

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011143711A1 (en) * 2010-05-19 2011-11-24 Australasian Pork Research Institute Ltd Image analysis for making animal measurements
CN103884280A (zh) * 2014-03-14 2014-06-25 中国农业大学 一种移动式多猪圈猪体尺体重监测系统
CN105784083A (zh) * 2016-04-05 2016-07-20 北京农业信息技术研究中心 基于立体视觉技术的奶牛体型测量方法和系统
CN206177429U (zh) * 2016-11-04 2017-05-17 北京农业信息技术研究中心 基于深度图像的肉鸡体重估测装置
CN107667903A (zh) * 2017-07-27 2018-02-09 北京奥云牧数据科技有限公司 基于物联网的畜牧养殖活体体重监测方法
CN108871520A (zh) * 2018-07-06 2018-11-23 平安科技(深圳)有限公司 牲畜体重测量方法及装置
CN109579956A (zh) * 2018-12-13 2019-04-05 北京小龙潜行科技有限公司 一种智能养猪限位测重方法、装置、电子设备及存储介质
CN109632058A (zh) * 2018-12-13 2019-04-16 北京小龙潜行科技有限公司 一种智能养猪群养测重方法、装置、电子设备及存储介质

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011143711A1 (en) * 2010-05-19 2011-11-24 Australasian Pork Research Institute Ltd Image analysis for making animal measurements
CN103884280A (zh) * 2014-03-14 2014-06-25 中国农业大学 一种移动式多猪圈猪体尺体重监测系统
CN105784083A (zh) * 2016-04-05 2016-07-20 北京农业信息技术研究中心 基于立体视觉技术的奶牛体型测量方法和系统
CN206177429U (zh) * 2016-11-04 2017-05-17 北京农业信息技术研究中心 基于深度图像的肉鸡体重估测装置
CN107667903A (zh) * 2017-07-27 2018-02-09 北京奥云牧数据科技有限公司 基于物联网的畜牧养殖活体体重监测方法
CN108871520A (zh) * 2018-07-06 2018-11-23 平安科技(深圳)有限公司 牲畜体重测量方法及装置
CN109579956A (zh) * 2018-12-13 2019-04-05 北京小龙潜行科技有限公司 一种智能养猪限位测重方法、装置、电子设备及存储介质
CN109632058A (zh) * 2018-12-13 2019-04-16 北京小龙潜行科技有限公司 一种智能养猪群养测重方法、装置、电子设备及存储介质

Cited By (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112106689A (zh) * 2020-10-16 2020-12-22 中国农业大学 对虾料台观测系统
CN112841074A (zh) * 2020-12-31 2021-05-28 京东数科海益信息科技有限公司 检测方法、装置、设备及存储介质
CN112990073A (zh) * 2021-03-31 2021-06-18 南京农业大学 基于边缘计算的哺乳期仔猪活动规律统计系统
CN113124971A (zh) * 2021-05-14 2021-07-16 深圳市路安仪器设备有限公司 一种线性称重方法
CN113780073A (zh) * 2021-08-03 2021-12-10 华南农业大学 一种辅助估测鸡群均匀度的装置与方法
CN113780073B (zh) * 2021-08-03 2023-12-05 华南农业大学 一种辅助估测鸡群均匀度的装置与方法
CN114067364B (zh) * 2021-11-23 2023-10-03 江苏省家禽科学研究所 一种基于图像采集的鸡只自动称重装置
CN114067364A (zh) * 2021-11-23 2022-02-18 江苏省家禽科学研究所 一种基于图像采集的鸡只自动称重装置
CN114543674A (zh) * 2022-02-22 2022-05-27 成都睿畜电子科技有限公司 一种基于图像识别的检测方法及系统
CN115226644A (zh) * 2022-08-05 2022-10-25 中国农业大学 一种猪只分群饲喂管理装置和系统
CN115226651B (zh) * 2022-08-19 2023-09-19 深圳进化动力数码科技有限公司 一种智慧养猪的生猪估重方法、装置、电子设备及介质
CN115226651A (zh) * 2022-08-19 2022-10-25 深圳进化动力数码科技有限公司 一种智慧养猪的生猪估重方法、装置、电子设备及介质
CN116058302B (zh) * 2023-03-15 2023-08-29 江苏省家禽科学研究所 一种自动称重的肉鸡养殖笼具
CN116058302A (zh) * 2023-03-15 2023-05-05 江苏省家禽科学研究所 一种自动称重的肉鸡养殖笼具
CN116091786A (zh) * 2023-04-11 2023-05-09 厦门农芯数字科技有限公司 猪只估重的全息体尺自编码方法、系统、设备及存储介质
CN116091786B (zh) * 2023-04-11 2023-06-20 厦门农芯数字科技有限公司 猪只估重的全息体尺自编码方法、系统、设备及存储介质
CN117528034A (zh) * 2023-12-29 2024-02-06 常州辉途智能科技有限公司 一种牧场智能养殖系统及工作方法
CN117528034B (zh) * 2023-12-29 2024-03-12 常州辉途智能科技有限公司 一种牧场智能养殖系统及工作方法
CN117718244A (zh) * 2024-02-08 2024-03-19 中国农业科学院北京畜牧兽医研究所 一种针对牛群的分群称重系统

Also Published As

Publication number Publication date
AU2019399813A1 (en) 2021-02-25
CN109632058B (zh) 2020-06-02
AU2019101786A4 (en) 2021-04-01
CN109632058A (zh) 2019-04-16

Similar Documents

Publication Publication Date Title
WO2020119659A1 (zh) 一种智能养猪群养测重方法、装置、电子设备及存储介质
CN108875647B (zh) 一种基于牲畜身份的移动轨迹监测方法和系统
Bao et al. Artificial intelligence in animal farming: A systematic literature review
CN109632059B (zh) 一种智能养猪方法、系统、电子设备及存储介质
Dohmen et al. Computer vision-based weight estimation of livestock: a systematic literature review
CN102333117B (zh) 基于物联网的家禽育种选种系统
CN105512949A (zh) 一种基于社交平台的养殖场信息化管理系统及方法
CN203535703U (zh) 牲畜体征信息自动采集通道门电控系统
CN201088576Y (zh) 一种pc遥控养殖业畜牧业流动监护终端/卫星定位匹配装置
Dongre et al. Comparative efficiency of artificial neural networks and multiple linear regression analysis for prediction of first lactation 305-day milk yield in Sahiwal cattle
CN109579956B (zh) 一种智能养猪限位测重方法、装置、电子设备及存储介质
Lu et al. An automatic splitting method for the adhesive piglets’ gray scale image based on the ellipse shape feature
CN108961330A (zh) 基于图像的猪体长测算方法及系统
CN102578031A (zh) 群养种鹅个体监控记录装置
CN113711944B (zh) 一种母猪发情鉴定方法、装置及系统
CN115294185B (zh) 猪只体重的估测方法以及相关设备
CN104599059A (zh) 一种奶牛移动数据管理系统及其方法
CN114898405A (zh) 基于边缘计算的便携式肉鸡异常监测系统
CN205593626U (zh) 畜牧养殖自动称重对比系统
CN204860479U (zh) 一种用于商品猪体重智能分群的系统
Watson-Haigh et al. Proximity loggers: data handling and classification for quality control
CN104573188B (zh) 一种基于大数据猪行为方式分析的育肥猪舍设计方法
CN116091962A (zh) 一种基于多层次行为分析的大熊猫产期预测方法
WO2021085806A1 (ko) 가축사양표준에 기초한 가축사양관리 가이드 제공 방법
Ositanwosu et al. Automatic measurement and prediction of Chinese Grown Pigs weight using multilayer perceptron neural networks

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: 19895502

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2019399813

Country of ref document: AU

Date of ref document: 20191210

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19895502

Country of ref document: EP

Kind code of ref document: A1