CN208188853U - A kind of milk cow face recognition device - Google Patents
A kind of milk cow face recognition device Download PDFInfo
- Publication number
- CN208188853U CN208188853U CN201820789669.7U CN201820789669U CN208188853U CN 208188853 U CN208188853 U CN 208188853U CN 201820789669 U CN201820789669 U CN 201820789669U CN 208188853 U CN208188853 U CN 208188853U
- Authority
- CN
- China
- Prior art keywords
- milk cow
- miniature webcam
- face
- image
- cloud platform
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
Landscapes
- Image Analysis (AREA)
Abstract
The utility model provides a kind of milk cow face recognition device, including miniature webcam, camera bracket, communication device and Internet of Things cloud platform；The miniature webcam is set on the camera bracket, and the camera bracket is fixed on milking area or feeds region or inspection area；The miniature webcam includes at least three, for shooting milk cow nappa, left side and the milk cow of right side face-image；The communication device connects the miniature webcam, for the milk cow face-image to be transmitted to Internet of Things cloud platform；The Internet of Things cloud platform is for identifying milk cow according to the milk cow face-image.The demands such as the individual goal identification, the action trail tracking that are precisely cultivated towards milk cow, realize that milk cow facial characteristics is accurate, quickly and automation, Intelligent information acquisition and monitoring, integrated micro camera, machine vision, small-sample learning and intelligent information processing technology, building systematization, quantification, standardization milk cow facial image acquisition device.
Description
Technical field
The utility model relates to automatic breeding technical fields, more particularly, to a kind of milk cow face recognition device.
Background technique
In recent years, milk cattle cultivating pasture is constantly developed to scale, automation and information-based direction, the ox in single pasture
Quantity can achieve 5000 or more, not adapt to by traditional manual type the need of development for the management of cows
It asks, carries out Group management especially in accordance with the breeding stage of milk cow, the high low index of the output of milk, constantly circulation becomes between each group
Change, thus to group carry out accurate counting and when individual identification fine-grained management important link.Computer vision and intelligence prison
Control has gradually been deep into milk cow fining cultivation, wherein individual archives are established, information collection and dairy products are traced to the source, is required to
Fast and accurately individual identification is carried out to milk cow.
Milk cow is numbered and writes or be printed upon on ear tag by traditional milk cow personal identification method, then by puncturing ear
Mark is installed on milk cow ear.Traditional ear tag needs artificial eye to read, and is simply limited to illumination and the limitation of distance, easily because of damage
It is bad, contamination spot and fail.With the development of intelligent perception technology, the cooperation of RFID electron ear tage is held or fixed reader quilt
For scale dairy cow farm, milk cow individual identity information can automatically, be quickly read using this technology, improves milk
The efficiency of ox information collection.But single ear tag price is higher, and large-scale plant that raises cost of investment is excessively high, and penetration type ear tag needs end
Body is installed on milk cow ear, generates damage physically to milk cow body, Rig up error results even in the tearing of milk ivory-white chirita herb；Ear
It indicates certain probability to fall off and lose, the loss of electron ear tage not only will increase the cost reinstalled, but also need a large amount of
Surcharge, which is spent, buys new electron ear tage.
Currently used ox face identification technology replicates face recognition technology mostly, and during milk cattle cultivating, milk cow will not
It is automatically aligned to camera as people, will lead to that collected ox face image is unintelligible, causes to extract from data set
Ox face local feature differs greatly, and must not little amplitude discarding sample, be unable to ensure large sample training dataset, therefore letter
The duplication of single face recognition technology does not ensure that the ox face under small sample mode identifies accuracy.
Utility model content
The utility model provides a kind of a kind of milk cow face for overcoming the above problem or at least being partially solved the above problem
Portion's identification device, solves in the prior art that electronic recognition milk cow is at high cost, the low problem of image recognition accuracy.
One aspect according to the present utility model provides a kind of milk cow face recognition device, including miniature webcam, camera shooting
Head bracket, communication device and Internet of Things cloud platform；The miniature webcam is set on the camera bracket, the camera branch
Frame is fixed on milking area or feeds region or inspection area；The miniature webcam includes at least three, for shooting milk cow
The milk cow face-image in front, left side and right side；The communication device connects the miniature webcam, is used for the milk
Ox face-image is transmitted to Internet of Things cloud platform；The Internet of Things cloud platform be used for according to the milk cow face-image to milk cow into
Row identification.
The utility model proposes a kind of milk cow face recognition devices, obtain target object front, a left side using miniature webcam
Acquired image is real-time transmitted to Internet of Things cloud platform by wireless network and carries out milk by the milk cow face-image of side, right side
Ox facial characteristics label is classified with milk cow face-image to be identified, realizes that milk cow facial characteristics is accurate, quickly and automation, intelligence
Information collection can be changed and monitored, the technologies such as integrated micro camera, machine vision, small-sample learning and Intelligent Information Processing,
Construct systematization, quantification, standardization milk cow facial image acquisition device, and the software and hardware ring of integrated internet of things data platform
Border realizes that remote transmission, management and the online recognition of data monitor, to improve the intelligent water of milk cow individual goal identification
It is flat.
Detailed description of the invention
Fig. 1 is a kind of milk cow face recognition device schematic diagram according to the utility model embodiment.
Specific embodiment
With reference to the accompanying drawings and examples, specific embodiment of the present utility model is described in further detail.Below
Embodiment is not intended to limit the scope of the present invention for illustrating the utility model.
Currently used ox face identification technology replicates face recognition technology mostly, first to collect ox face data set carry out
Characteristic point mark, then with the method for machine learning is trained the feature after mark and deep learning, obtains data set
Then required model or anticipation function carry out forecast analysis to unknown (unlabelled) sample, this method must
Wanting condition is number of samples abundance, and only sample number approach infinity is big, the accuracy rate of ability ensuring method.And milk cattle cultivating process
In, the acquisition of ox face data can be unfavorable for, and by video because ox companion be blocked, lick, the movements such as feed shelter from face
It is unintelligible to also result in collected ox face image for the limitation of quality, distance and shooting angle；Furthermore be illuminated by the light, state of ground,
The interference of the extraneous factors such as weather condition, causes the ox face local feature extracted from data set to differ greatly, and has to
Sample is significantly abandoned, is unable to ensure large sample training dataset, therefore the duplication of simple face recognition technology can not
Guarantee that the ox face under small sample mode identifies accuracy.
Therefore, a kind of milk cow face recognition device is provided in the present embodiment, obtains target object using miniature webcam
The milk cow face-image in front, left side, right side, to improve the intelligent level of milk cow individual goal identification.
Specifically, as shown in Figure 1, in the present embodiment, a kind of milk cow face recognition device, including miniature webcam 1, camera shooting
Head bracket 2, communication device and Internet of Things cloud platform；The miniature webcam 1 is set on the camera bracket 2, the camera shooting
Head bracket 2 is fixed on milking area or feeds region or inspection area；The miniature webcam 1 includes at least three, for clapping
Take the photograph milk cow nappa, left side and the milk cow of right side face-image；The communication device connects the miniature webcam 1, is used for
The milk cow face-image is transmitted to Internet of Things cloud platform；The Internet of Things cloud platform is used for according to the milk cow face-image
Milk cow is identified.
In the present embodiment, the demands such as individual goal identification, action trail tracking for precisely being cultivated towards milk cow, it is therefore intended that
Realize that milk cow facial characteristics is accurate, quickly and automation, Intelligent information acquisition and monitoring, integrated micro camera 1, machine
The technologies such as vision, small-sample learning and Intelligent Information Processing, building systematization, quantification, standardization milk cow face-image are adopted
Acquisition means, and the hardware environment of integrated internet of things data platform realize that remote transmission, management and the online recognition of data are supervised
It surveys, to improve the intelligent level of milk cow individual goal identification.
As a preferred embodiment, the camera bracket 2 includes base fixture, the fixed dress in the bottom
It sets and is fixed on milking area or feeds on the ground of region or inspection area, or be fixed on milking area or feed region or inspection
It looks on the railing device of region.
In the present embodiment, it in order to which the identification device of the present embodiment is more flexible, is filled by the way that transferable bottom is fixed
It sets and camera is directly anchored to specified region, specifically, the base fixture can be the fixed device of base type, directly move
It moves to milking area or feeds region or inspection area, it is mobile for convenience, it can be set on the fixed device of the base type
Universal wheel fixes universal wheel after identification device is moved to specified region by universal wheel；It certainly can also be using more
Add easy mode, will be inserted directly into milking area using inserted link as the indefinite device in bottom or feed region or inspection area
Ground, specifically, the base fixture be equipped with telescopic connecting rod, aspect for different height milk cow carry out
It takes pictures identification.
As a preferred embodiment, the camera bracket 2 includes 1 mounting rack of miniature webcam, it is described miniature
1 mounting rack of camera is U-shaped, V-arrangement or arc, in order to make miniature webcam 1 that can take different sides after mounting arrangements
Milk cow face-image, needing to guarantee the miniature webcams 1 of two sides, relatively intermediate miniature webcam 1 is symmetrical respectively, and two sides are miniature
Camera 1 with respectively with 1 connecting line of miniature webcam at middle part is not conllinear or miniature webcam 1 is all on same camber line,
On this basis, the other structures other than U-shaped, V-arrangement or arc can also.
As a preferred embodiment, the miniature webcam 1 includes three, it is respectively arranged in the miniature video camera
The middle part and two sides of first 1 mounting rack, and the 1 face milk cow frontal face of miniature webcam at middle part, the miniature webcam 1 at middle part with
Milk cow center of face is connected as a reference line, the miniature webcam 1 and milk cow center of face line and the reference line of two sides
Angle it is identical, be 30 °~60 °.
During milk cow milking, because four limbs are fixed motionless, the most common movement is that head rocks from side to side, therefore is acquiring
When milk cow face-image, on milk cow nappa, 30 ° of the left side, 30 ° of the right 3 miniature webcams 1 is installed respectively, acquires different angle
Milk cow face-image.Miniature webcam 1 and milk cow head distance are controlled at 0.5 meter, every 10 seconds, milk cow head front,
The left side, the right shoot photo 1 respectively and open, and device is communicated in real time to Internet of Things by wireless communication by the image data taken pictures
In server of cloud platform.
As a preferred embodiment, the Internet of Things cloud platform includes image data base, arest neighbors tagsort
Device；Described image database is for choosing milk cow face picture and carrying out feature mark；The arest neighbors feature classifiers are used for
Classify to milk cow facial image features to be identified.
Background server pre-processes the milk cow face-image of shooting, artificial to select the higher picture of clarity, into
Row feature mark, constitutes milk cow face image data library.The deficiency that sample size is made up by priori knowledge, according to milk cow face
Feature has symmetry, and only head swings ox frequent occurrence during the motion, facial in existing tape label milk cow
On the basis of sample essential characteristic remains unchanged, virtual sample is created by rotatable head angle, is realized to existing small sample face
The expansion of portion's sets of image data.The milk cow facial characteristics in exptended sample data is extracted, arest neighbors feature classifiers are constructed, it is right
Milk cow facial image features to be identified are classified, and are continued to choose after classification after the high milk cow face-image of confidence level is marked
It is stored in training sample set, expands marker samples set re -training and obtain new arest neighbors feature classifiers, to remaining
The milk cow face-image not being labeled is classified.Identification process is the process constantly recycled, will identify that and carrys out confidence level height
Image tape label training sample set is added, re -training obtains new nearest neighbor classifier, to remaining without labeled
Milk cow face-image classify.
As a preferred embodiment, further include handheld terminal, the handheld terminal is wirelessly connected described miniature take the photograph
As head 1 and the Internet of Things cloud platform, for controlling 1 shooting angle of miniature webcam, frequency, and obtains and record the object
The recognition result for cloud platform of networking.
In the present embodiment, handheld terminal connection miniature webcam 1 and Internet of Things cloud platform, can control miniature webcam
1 progress image acquisition simultaneously obtains recognition result from Internet of Things cloud platform, and then realizes milk cow information record and verifying；It can be with
The filming frequency of microcam is set by handheld terminal, in addition, motor can also be set in the connecting rod in above-described embodiment,
Controls connecting rod is flexible, which may be alternatively provided as directly controlling by handheld terminal, improve the intelligence of identification device.
As a preferred embodiment, further including identification trigger device, the identification trigger device is gravity triggering
Device or infrared trigger device, for the triggering unlatching when detecting that milk cow enters milking area or feeds region or inspection area
The miniature webcam 1 identifies milk cow.
In the present embodiment, in order to guarantee to economize on resources, which can be entered in identification range with milk cow for mark
Standard carries out the unlatching of identification device, and in the present embodiment, opening ways can be triggered by infrared triggering and gravity, i.e., by setting
Infrared inductor is set, senses that milk cow enters identification region, then controls identification device unlatching, miniature webcam 1 carries out milk cow
Face image capture；It can also be triggered by gravity sensing, when milk cow enters identification region, gravity trigger device senses weight
Power, then control identification device unlatching, and miniature webcam 1 carries out milk cow face image capture.
In conclusion a kind of milk cow face recognition device is provided in the present embodiment, according to milk cow head double swerve
Habit carries out offset to marker samples feature and obtains new textural characteristics and Feature Points, constitutes new virtual sample feature
Data.Milk cow face-image after expansion is indicated with vector, the characteristic vector for calculating normalized images collection covariance matrix obtains
Milk cow facial image features are got, and construct arest neighbors character separation device, calculate sample characteristics point to be identified to milk cow face figure
As the shortest distance of all characteristic points in library, the Accurate classification to milk cow to be identified is realized；The individual mesh precisely cultivated towards milk cow
Mark not, the demands such as action trail tracking, it is therefore intended that realize that milk cow facial characteristics is accurate, quickly and automation, intelligent
Information collection and monitoring, the technologies such as integrated micro camera 1, machine vision, small-sample learning and Intelligent Information Processing, building
Systematization, quantification, standardization milk cow facial image acquisition device, and the hardware environment of integrated internet of things data platform, it is real
Remote transmission, management and the online recognition of existing data monitor, to improve the intelligent level of milk cow individual goal identification.
Finally, the method for the utility model is only preferable embodiment, it is not intended to limit the protection of the utility model
Range.Within the spirit and principle of the utility model, any modification, equivalent replacement, improvement and so on should be included in
Within the protection scope of the utility model.
Claims (7)
1. a kind of milk cow face recognition device, which is characterized in that including miniature webcam, camera bracket, communication device and object
Networking cloud platform；The miniature webcam be set to the camera bracket on, the camera bracket be fixed on milking area or
Feed region or inspection area；The miniature webcam includes at least three, for shooting milk cow nappa, left side and right side
Milk cow face-image；The communication device connects the miniature webcam, for the milk cow face-image to be transmitted to object
Networking cloud platform；The Internet of Things cloud platform is for identifying milk cow according to the milk cow face-image.
2. milk cow face recognition device according to claim 1, which is characterized in that the camera bracket includes that bottom is solid
Determine device, the base fixture is fixed on milking area or feeds on the ground of region or inspection area, or is fixed on
Milking area is fed on region or inspection area railing device.
3. milk cow face recognition device according to claim 1, which is characterized in that the camera bracket includes miniature takes the photograph
As head mounting rack, the miniature webcam mounting rack is U-shaped, V-arrangement or arc.
4. milk cow face recognition device according to claim 3, which is characterized in that the miniature webcam includes three,
It is respectively arranged in the middle part and two sides of the miniature webcam mounting rack, and the miniature webcam face milk cow face at middle part is just
Face, the miniature webcam and milk cow center of face at middle part are connected as a reference line, the miniature webcam of two sides and milk cow face
The line of centres is identical as the angle of the reference line, is 30 °~60 °.
5. milk cow face recognition device according to claim 1, which is characterized in that the Internet of Things cloud platform includes image
Database, arest neighbors feature classifiers；Described image database is for choosing milk cow face picture and carrying out feature mark；It is described
Arest neighbors feature classifiers are for classifying to milk cow facial image features to be identified.
6. milk cow face recognition device according to claim 1, which is characterized in that it further include handheld terminal, it is described hand-held
Terminal wireless connects the miniature webcam and the Internet of Things cloud platform, for controlling miniature webcam shooting angle, frequency,
And obtain and record the recognition result of the Internet of Things cloud platform.
7. milk cow face recognition device according to claim 1, which is characterized in that it further include identification trigger device, it is described
Identify trigger device be gravity trigger device or infrared trigger device, for detect milk cow enter milking area or feed area
Triggering is opened the miniature webcam and is identified to milk cow when domain or inspection area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201820789669.7U CN208188853U (en) | 2018-05-24 | 2018-05-24 | A kind of milk cow face recognition device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201820789669.7U CN208188853U (en) | 2018-05-24 | 2018-05-24 | A kind of milk cow face recognition device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN208188853U true CN208188853U (en) | 2018-12-04 |
Family
ID=64427747
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201820789669.7U Active CN208188853U (en) | 2018-05-24 | 2018-05-24 | A kind of milk cow face recognition device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN208188853U (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109618961A (en) * | 2018-12-12 | 2019-04-16 | 北京京东金融科技控股有限公司 | A kind of intelligence of domestic animal feeds system and method |
| CN109886145A (en) * | 2019-01-29 | 2019-06-14 | 浙江泽曦科技有限公司 | Pet recognition algorithms and system |
| CN110169374A (en) * | 2019-06-19 | 2019-08-27 | 北京农业信息技术研究中心 | A kind of wearable monitoring device and monitoring system of cow feeding behavior and feed intake |
| CN110221588A (en) * | 2019-07-09 | 2019-09-10 | 哈尔滨商业大学 | A kind of wisdom cattle-breeding system based on Internet of Things |
| CN110347134A (en) * | 2019-07-29 | 2019-10-18 | 南京图玩智能科技有限公司 | A kind of AI intelligence aquaculture specimen discerning method and cultivating system |
| CN111291683A (en) * | 2020-02-08 | 2020-06-16 | 内蒙古大学 | Dairy cow individual identification system based on deep learning and identification method thereof |
| US10757914B1 (en) * | 2019-04-17 | 2020-09-01 | National Taiwan University | Feeding analysis system |
| CN113140047A (en) * | 2021-03-26 | 2021-07-20 | 深圳市博西尼电子有限公司 | Method and system for rapidly paying fee when parking lot is out |
| CN114569091A (en) * | 2022-03-03 | 2022-06-03 | 宁夏芯锐智联科技有限公司 | Milk cow respiratory heart rate monitoring system |
| CN117558034A (en) * | 2024-01-05 | 2024-02-13 | 四川智迅车联科技有限公司 | Multi-dimensional cow face recognition method and system based on image |
-
2018
- 2018-05-24 CN CN201820789669.7U patent/CN208188853U/en active Active
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020119184A1 (en) * | 2018-12-12 | 2020-06-18 | 京东数字科技控股有限公司 | Intelligent feeding system and method for livestock |
| CN109618961A (en) * | 2018-12-12 | 2019-04-16 | 北京京东金融科技控股有限公司 | A kind of intelligence of domestic animal feeds system and method |
| CN109886145A (en) * | 2019-01-29 | 2019-06-14 | 浙江泽曦科技有限公司 | Pet recognition algorithms and system |
| US10757914B1 (en) * | 2019-04-17 | 2020-09-01 | National Taiwan University | Feeding analysis system |
| CN110169374A (en) * | 2019-06-19 | 2019-08-27 | 北京农业信息技术研究中心 | A kind of wearable monitoring device and monitoring system of cow feeding behavior and feed intake |
| CN110169374B (en) * | 2019-06-19 | 2024-05-10 | 北京农业信息技术研究中心 | Wearable monitoring device and monitoring system for dairy cow feeding behavior and feeding capacity |
| CN110221588A (en) * | 2019-07-09 | 2019-09-10 | 哈尔滨商业大学 | A kind of wisdom cattle-breeding system based on Internet of Things |
| CN110347134A (en) * | 2019-07-29 | 2019-10-18 | 南京图玩智能科技有限公司 | A kind of AI intelligence aquaculture specimen discerning method and cultivating system |
| CN111291683A (en) * | 2020-02-08 | 2020-06-16 | 内蒙古大学 | Dairy cow individual identification system based on deep learning and identification method thereof |
| CN111291683B (en) * | 2020-02-08 | 2023-04-18 | 内蒙古大学 | Dairy cow individual identification system based on deep learning and identification method thereof |
| CN113140047A (en) * | 2021-03-26 | 2021-07-20 | 深圳市博西尼电子有限公司 | Method and system for rapidly paying fee when parking lot is out |
| CN114569091A (en) * | 2022-03-03 | 2022-06-03 | 宁夏芯锐智联科技有限公司 | Milk cow respiratory heart rate monitoring system |
| CN117558034A (en) * | 2024-01-05 | 2024-02-13 | 四川智迅车联科技有限公司 | Multi-dimensional cow face recognition method and system based on image |
| CN117558034B (en) * | 2024-01-05 | 2024-03-26 | 四川智迅车联科技有限公司 | Multi-dimensional cow face recognition method and system based on image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN208188853U (en) | A kind of milk cow face recognition device | |
| Chen et al. | An AIoT based smart agricultural system for pests detection | |
| CN108875647B (en) | Moving track monitoring method and system based on livestock identity | |
| Achour et al. | Image analysis for individual identification and feeding behaviour monitoring of dairy cows based on Convolutional Neural Networks (CNN) | |
| KR101984983B1 (en) | System and method for monitoring wild animals | |
| CN108990831B (en) | Livestock health monitoring method and system | |
| CN110046631A (en) | System and method for inferring the variation of time-space image automatically | |
| CN112131927B (en) | Sow delivery time prediction system based on gesture conversion characteristics in late gestation period | |
| KR102584357B1 (en) | Apparatus for identifying a livestock using a pattern, and system for classifying livestock behavior pattern based on images using the apparatus and method thereof | |
| CN109376584A (en) | A kind of poultry quantity statistics system and method for animal husbandry | |
| CN113223035A (en) | Intelligent inspection system for cage-rearing chickens | |
| Noe et al. | Automatic detection and tracking of mounting behavior in cattle using a deep learning-based instance segmentation model | |
| CN113569971B (en) | Image recognition-based catch target classification detection method and system | |
| CN114581948A (en) | Animal face identification method | |
| CN111354011A (en) | Multi-moving-target information capturing and tracking system and method | |
| CN108830222A (en) | A kind of micro- expression recognition method based on informedness and representative Active Learning | |
| CN111797831A (en) | BIM and artificial intelligence based parallel abnormality detection method for poultry feeding | |
| Rodriguez et al. | Automated video monitoring of unmarked and marked honey bees at the hive entrance | |
| Chowdhury et al. | Deep learning based computer vision technique for automatic heat detection in cows | |
| CN113724250A (en) | Animal target counting method based on double-optical camera | |
| WO2020258977A1 (en) | Object counting method and device | |
| CN116563758A (en) | Lion head goose monitoring method, device, equipment and storage medium | |
| Gao et al. | Adaptive cow movement detection using evolving spiking neural network models | |
| CN113221704A (en) | Animal posture recognition method and system based on deep learning and storage medium | |
| Zhang et al. | Automatic location and recognition of cow's collar ID based on machine learning. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant |