CN115760523A - Animal management method and system based on cloud platform - Google Patents
Animal management method and system based on cloud platform Download PDFInfo
- Publication number
- CN115760523A CN115760523A CN202211448386.3A CN202211448386A CN115760523A CN 115760523 A CN115760523 A CN 115760523A CN 202211448386 A CN202211448386 A CN 202211448386A CN 115760523 A CN115760523 A CN 115760523A
- Authority
- CN
- China
- Prior art keywords
- animal
- information
- detection
- server
- module
- 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.)
- Pending
Links
Images
Abstract
The invention discloses an animal management method and system based on a cloud platform, belonging to the technical field of animal management, wherein the management method comprises the following specific steps: (1) Inputting information of each animal and classifying and marking the input information; (2) Carrying out cascade detection on animals in each area and carrying out anomaly analysis; (3) Feeding back abnormal animal information and moving the animal to a control area; (4) The running information of the carousel whiteboard display server is periodically subjected to performance optimization; the invention can greatly improve the tracking precision of each animal, can ensure that full cross-monitoring matching is carried out, avoids missing matching caused by too few matching opportunities, ensures that workers can better manage, can carry out large-granularity compression on the memory of the server, improves the compression efficiency, effectively improves the response speed of a server port, and saves the time required by compressing the memory.
Description
Technical Field
The invention relates to the technical field of animal management, in particular to an animal management method and system based on a cloud platform.
Background
Animal management is a subject of applied technology. It takes the ecology of wild animals as its theoretical basis and focuses on the study of management practice. It is a comprehensive technical subject which guides the practice of wild animal management by applying the theory and technology of each subject, and the traditional 'feeding' mode is always used for the past animal management. The acquisition and processing of information such as vital sign information of animals, health information of the animals, movement and health of the animals, diet and health of the animals, breeding and health of the animals and the like are always problems to be solved urgently, along with the continuous development of science and technology, the protection cause of wild animals in China has obvious effect, and five management systems of field protection, rescue and breeding, scientific support, law enforcement supervision and control and epidemic disease monitoring and prevention and control of the terrestrial wild animals mainly in a natural protection area are formed;
through retrieval, the Chinese patent number CN113671892A discloses an animal management method and an animal management system based on a cloud platform, and although the invention provides data for health management, pathological explanation and medical treatment schemes of animals and plays a role in helping and guiding feeding of the animals, the tracking precision of the animals is low, and sufficient cross-monitoring matching cannot be carried out; in addition, the existing animal management method and system based on the cloud platform have the disadvantages that the response speed of a port is slowed down due to too long service time of a server, and meanwhile, the compression efficiency of a memory of the server is low; therefore, a cloud platform-based animal management method and system are provided.
Disclosure of Invention
The invention aims to solve the defects in the prior art and provides an animal management method and system based on a cloud platform.
In order to achieve the purpose, the invention adopts the following technical scheme:
an animal management method based on a cloud platform comprises the following specific steps:
(1) Inputting information of each animal and classifying and marking the input information;
(2) Carrying out cascade detection on animals in each area and carrying out anomaly analysis;
(3) Feeding back information of abnormal animals and moving the animals to a control area;
(4) And the performance of the carousel whiteboard display server is optimized periodically by the running information.
As a further scheme of the invention, the specific steps of the classification mark in the step (1) are as follows:
the method comprises the following steps: the server receives animal information uploaded by workers, classifies the animal information according to animal types to generate a corresponding animal type set, and numbers and marks each group of animal information in the classified animal type set, wherein the number information of each group of animals is unique;
step two: acquiring increase and decrease information of regional animals in real time, correspondingly adjusting the information stored in a concentrated manner according to the types of related animals, reserving a number corresponding to the animal information if the information of the animals needing to be decreased exists, and endowing the animal information with the number when waiting for the next increase of the corresponding animal information.
As a further scheme of the invention, the cascade detection in the step (2) comprises the following specific steps:
step I: after receiving each group of image information collected by each monitor, the server processes a single monitoring video or image sequence frame with a fixed frame rate, calculates the interval time of the actual video frame and records the calculated interval time of the actual video frame;
and step II: establishing a motion model through a Kalman filtering theory, simultaneously acquiring the motion state of a tracking target in real time through the established motion model, allocating one ID to all tracking targets, recording all tracking targets allocated with IDs as a target set after the allocation is finished, acquiring appearance characteristic vectors of all tracking targets, and integrating and concluding all the acquired characteristic vectors into a target characteristic set;
step III: defining the motion state of the tracking target in a video frame by a motion model according to the linear motion hypothesis of the tracking target, collecting the motion state of the tracking target in the current video frame, constructing a prediction equation to estimate the motion state of each tracking target in the next video frame, and recording the motion state of each group of tracking targets in the ith monitored current video frame as a characteristic position set and a characteristic covariance matrix set;
step IV: then collecting the detection results of all the targets in the ith monitoring current video frame calculated by the multi-target real-time detection algorithm to generate a detection set, recording the positions of the detection sets in the current video frame as a detection position set, and extracting appearance characteristic vectors one by one to obtain an appearance characteristic vector set of the detection results;
step V: calculating cosine distance between a detection result and a tracking target, filtering by using a threshold value to obtain a cost matrix represented by the cosine distance, filtering the cost matrix represented by the cosine distance again according to the mahalanobis distance matrix and a related constraint condition, and matching by adopting a Hungary algorithm to perform binary matching;
step VI: after matching is completed, the server sequentially processes video frame data of each path of video stream in parallel, and sequentially executes target marking, estimation of a motion state of a tracking target, matching association and cross-monitoring multi-target real-time tracking on video frames obtained after down-sampling in each path of video stream.
As a further scheme of the invention, the abnormality analysis in the step (2) comprises the following specific steps:
the first step is as follows: constructing and training a group of detection neural networks, simultaneously introducing the living states of various animals into the detection neural networks in real time, and then receiving various animal detection standards uploaded by workers by the detection neural networks;
the second step: the method comprises the steps that a detection neural network respectively processes animal detection standards and animal living states to generate a training set and a test set, then, the data of the training set are subjected to standardization processing to obtain training samples, specific parameters of a model are set, the detection neural network is trained by adopting a long-term iteration method, the test set is input into the trained model, an animal state prediction curve is drawn, and the animal state prediction curve is analyzed.
As a further scheme of the invention, the performance optimization in the step (4) comprises the following specific steps:
s1: the performance optimization module generates a start linked list for each port of the server, further links the head of each set of start linked lists according to the LRU sequence of the ports, collects the port information with the minimum interaction frequency, arranges the start linked list of the port at the head of the LRU linked list and sequences the port information in sequence;
s2: before the port is started, clearing access bits of all updated page table entries, rechecking the access bits of all the pages by a performance optimization module before the port starting time is finished, updating data of all groups of page information in a starting linked list after the check is finished, sequentially selecting the least active port from the head of the LRU linked list, and selecting a victim page from the corresponding starting linked list until enough pages are obtained;
s3: merging the selected victim pages into a block, marking the block, waking up a compression driver to analyze the marked block and obtain a physical page belonging to the block, copying the physical page into a buffer area, calling a compression algorithm to compress the physical page in the buffer area into a compression block, and storing the compression block into a performance optimization module.
An animal management system based on a cloud platform comprises a server, a monitoring module, an activity management module, a state prediction module, a working platform, an alarm module, a control module, a performance optimization module and a carousel white board;
the server is used for receiving and storing the animal data of each group;
the monitoring module is used for collecting the activity condition of each animal in the area in real time;
the activity management module is used for receiving the collected activity conditions of each animal and performing cascade detection;
the state prediction module is used for collecting data of each animal and performing state prediction;
the working platform is used for receiving the detection information and the prediction information for the staff to check;
the alarm module is used for receiving abnormal animal information and giving an alarm to prompt staff;
the management and control module is used for monitoring animals in a management and control area;
the performance optimization module is used for optimizing the running performance of the server;
the carousel whiteboard is used for circularly playing the animal information and the server operation information in the area.
Compared with the prior art, the invention has the beneficial effects that:
1. according to the animal management method based on the cloud platform, after a server receives each group of image information collected by each monitor, a single monitoring video or image sequence frame with a fixed frame rate is processed, the interval time of an actual video frame is calculated, the interval time of the calculated actual video frame is recorded, a motion model is established through a Kalman filtering theory, the motion state of a tracking target is obtained, meanwhile, the cosine distance between a detection result and the tracking target is detected and filtered, two-component matching is carried out through the Hungary algorithm for matching, after the matching is completed, the server sequentially processes video frame data of each path of video stream in parallel, and sequentially carries out target marking, estimation of the motion state of the tracking target, matching association and cross-monitoring multi-target real-time tracking on the video frames obtained after down sampling in each path of video stream, so that the tracking precision of each animal can be greatly improved, full cross-monitoring matching can be ensured, missing matching caused by too few matching opportunities can be avoided, and better management of workers can be ensured;
2. the animal management method based on the cloud platform comprises the steps of generating a starting linked list for each port of a server through a performance optimization module, further performing link sequencing on the head of each group of starting linked lists according to the LRU sequence of the ports, clearing access bits of all updated page table entries before the ports are started, rechecking the access bits of all pages before the port starting time is finished, updating data of each group of page information in the starting linked list after the check is finished, sequentially selecting the most inactive port from the head of the LRU linked list, selecting a victim page from the corresponding starting linked list until enough pages are obtained, combining the selected victim pages into one block, marking the block, awakening a compression driving program to analyze the marked block and obtain physical pages belonging to the block, copying the physical pages into a buffer area, then calling a compression algorithm to compress the physical pages in the buffer area into a compression block, and storing the compression block into the performance optimization module, so that large-size compression can be performed on a server memory, the compression efficiency is improved, the response speed of the server is effectively improved, and the time required by the compression of the ports is saved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.
Fig. 1 is a flow chart of an animal management method based on a cloud platform according to the present invention;
fig. 2 is a system block diagram of an animal management system based on a cloud platform according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
Example 1
Referring to fig. 1, the embodiment discloses an animal management method based on a cloud platform, which specifically includes the following steps:
and inputting information of each animal and classifying and marking the input information.
Specifically, the server receives animal information uploaded by staff, classifies the animal information according to animal types to generate a corresponding animal type set, numbers and marks animal information of each group in the classified animal type set, the animal number information of each group is unique, increase and decrease information of regional animals is collected in real time, corresponding adjustment is carried out on information stored in the relevant animal type set, if the animal information needing to be decreased exists, the number corresponding to the animal information is reserved, and when the corresponding animal information is added next time, the number is given to the animal information.
And carrying out cascade detection on animals in each area and carrying out anomaly analysis.
Specifically, after receiving each group of image information collected by each monitor, the server processes a single monitoring video or image sequence frame with a fixed frame rate, calculates the interval time of an actual video frame, records the interval time of the calculated actual video frame, establishes a motion model through a Kalman filtering theory, simultaneously acquires the motion state of a tracking target in real time through the established motion model, allocates one ID to all tracking targets, after the allocation is completed, marks all the tracking targets to which the IDs are allocated as a target set, collects the appearance characteristic vectors of all the tracking targets, integrates and summarizes all the collected groups of characteristic vectors into a target characteristic set, defines the motion state of the tracking targets in the video frame according to a linear motion hypothesis of the tracking targets, collects the motion state of the tracking targets in the current video frame, establishes a prediction equation to estimate the motion state of each tracking target in the next video frame, then marks each group of tracking targets as a cosine motion state characteristic position set and a characteristic covariance characteristic matrix set according to a linear motion hypothesis of the tracking targets, calculates all appearance characteristic distance detection results of the i th monitored current video frame by using a cosine detection matrix, and filters the appearance characteristic distance of the cosine detection result of the I of the tracking targets calculated by a cosine detection algorithm, and calculates again by using a two-half cosine detection cost of a cosine detection matrix, and a cosine detection cost of the appearance characteristic distance in the current video frame, and a cosine detection result, and a variance detection cost set, and a two-weighted cosine detection cost of the detection result of the sampling distance detection result obtained by a sampling method, and a sampling method, after matching is completed, the server sequentially processes the video frame data of each path of video stream in parallel, and sequentially executes target marking, estimation of the motion state of a tracking target, matching association and cross-monitoring multi-target real-time tracking on the video frames obtained after down-sampling in each path of video stream.
Specifically, a group of detection neural networks are constructed and trained, living states of various animals are imported into the detection neural networks in real time, then the detection neural networks receive various animal detection standards uploaded by workers, the detection neural networks process the animal detection standards and the living states of the animals respectively to generate a training set and a test set, then the data of the training set are subjected to standardized processing to obtain training samples, specific parameters of the models are set, the detection neural networks are trained by adopting a long-term iteration method, the test set is input into the trained models, animal state prediction curves are drawn, and the animal state prediction curves are analyzed.
And feeding back abnormal animal information and moving the animal to a control area.
And the running information of the carousel whiteboard display server is periodically subjected to performance optimization.
Specifically, the performance optimization module generates a start linked list for each port of the server, further links the head of each set of start linked lists according to the LRU sequence of the ports, collects port information with the minimum interaction frequency, arranges the start linked list of the port at the head of the LRU linked list, sequences the start linked list, clears the access bits of all updated page list items before the port starts, rechecks the access bits of all pages before the port start time is over, updates data of each set of page information in the start linked list after the check is over, sequentially selects the least active port from the head of the LRU linked list, selects a victim page from the corresponding start linked list until enough pages are obtained, combines the selected victim pages into a block, marks the block, wakes up a compression driver to analyze the marked block, obtains a physical page belonging to the block, copies the page into a buffer, calls a compression algorithm to compress the physical page in the buffer into a compression block, and stores the compression block into the performance optimization module.
Example 2
Referring to fig. 2, the embodiment discloses an animal management system based on a cloud platform, which includes a server, a monitoring module, an activity management module, a state prediction module, a working platform, an alarm module, a management and control module, a performance optimization module, and a carousel whiteboard.
The server is used for receiving and storing the animal data of each group; the monitoring module is used for collecting the activity condition of each animal in the area in real time; the activity management module is used for receiving the collected activity conditions of each animal and performing cascade detection; and the state prediction module is used for collecting data of each animal and performing state prediction.
The working platform is used for receiving the detection information and the prediction information for the staff to check; the alarm module is used for receiving the abnormal animal information and giving an alarm to prompt staff; the management and control module is used for monitoring animals in the management and control area.
And the performance optimization module is used for optimizing the running performance of the server.
The carousel whiteboard is used for circularly playing the animal information and the server operation information in the area.
Claims (6)
1. An animal management method based on a cloud platform is characterized by comprising the following specific steps:
(1) Inputting information of each animal and classifying and marking the input information;
(2) Carrying out cascade detection on animals in each area and carrying out anomaly analysis;
(3) Feeding back abnormal animal information and moving the animal to a control area;
(4) And the running information of the carousel whiteboard display server is periodically subjected to performance optimization.
2. The animal management method based on the cloud platform as claimed in claim 1, wherein the specific steps of the classification and marking in step (1) are as follows:
the method comprises the following steps: the server receives animal information uploaded by workers, classifies the animal information according to animal types to generate a corresponding animal type set, and numbers and marks each group of animal information in the classified animal type set, wherein the number information of each group of animals is unique;
step two: acquiring increase and decrease information of regional animals in real time, correspondingly adjusting the information stored in a concentrated manner according to the types of related animals, reserving a number corresponding to the animal information if the animal information needing to be decreased exists, and endowing the number with the animal information when waiting for the next increase of the corresponding animal information.
3. The animal management method based on the cloud platform as claimed in claim 1, wherein the cascade detection in step (2) specifically comprises the following steps:
step I: after receiving each group of image information collected by each monitor, the server processes a single monitoring video or image sequence frame with a fixed frame rate, calculates the interval time of the actual video frame and records the calculated interval time of the actual video frame;
step II: establishing a motion model through a Kalman filtering theory, simultaneously acquiring the motion state of a tracking target in real time through the established motion model, allocating one ID to all tracking targets, recording all tracking targets allocated with IDs as a target set after the allocation is finished, acquiring appearance characteristic vectors of all tracking targets, and integrating and concluding all the acquired characteristic vectors into a target characteristic set;
step III: defining the motion state of the tracking target in a video frame by a motion model according to the linear motion hypothesis of the tracking target, collecting the motion state of the tracking target in the current video frame, constructing a prediction equation to estimate the motion state of each tracking target in the next video frame, and recording the motion state of each group of tracking targets in the ith monitored current video frame as a characteristic position set and a characteristic covariance matrix set;
step IV: then collecting the detection results of all the targets in the ith monitoring current video frame calculated by the multi-target real-time detection algorithm to generate a detection set, recording the positions of the detection sets in the current video frame as a detection position set, and extracting appearance characteristic vectors one by one to obtain an appearance characteristic vector set of the detection results;
step V: calculating cosine distance between a detection result and a tracking target, filtering by using a threshold value to obtain a cost matrix represented by the cosine distance, filtering the cost matrix represented by the cosine distance again according to the mahalanobis distance matrix and a related constraint condition, and matching by adopting a Hungary algorithm to perform binary matching;
step VI: after matching is completed, the server sequentially processes the video frame data of each path of video stream in parallel, and sequentially executes target marking, estimation of the motion state of a tracking target, matching association and cross-monitoring multi-target real-time tracking on the video frames obtained after down-sampling in each path of video stream.
4. The animal management method based on the cloud platform as claimed in claim 3, wherein the abnormality analysis in step (2) comprises the following specific steps:
the first step is as follows: constructing and training a group of detection neural networks, simultaneously introducing the living states of various animals into the detection neural networks in real time, and then receiving various animal detection standards uploaded by workers by the detection neural networks;
the second step is that: the method comprises the steps that a detection neural network respectively processes an animal detection standard and an animal living state to generate a training set and a test set, then, the data of the training set are subjected to standardization processing to obtain a training sample, specific parameters of a model are set, the detection neural network is trained by adopting a long-term iteration method, the test set is input into the trained model, an animal state prediction curve is drawn, and the animal state prediction curve is analyzed.
5. The animal management method based on the cloud platform as claimed in claim 1, wherein the performance optimization in step (4) comprises the following specific steps:
s1: the performance optimization module generates a start linked list for each port of the server, further links the head of each set of start linked lists according to the LRU sequence of the ports, collects the port information with the minimum interaction frequency, arranges the start linked list of the port at the head of the LRU linked list and sequences the port information in sequence;
s2: before the port is started, clearing access bits of all updated page table entries, rechecking the access bits of all pages by a performance optimization module before the port starting time is finished, updating data of each group of page information in a starting linked list after the check is finished, sequentially selecting the least active port from the head of an LRU linked list, and selecting a victim page from the corresponding starting linked list until enough pages are obtained;
s3: merging the selected victim pages into a block, marking the block, waking up a compression driver to analyze the marked block and obtain a physical page belonging to the block, copying the physical page into a buffer area, calling a compression algorithm to compress the physical page in the buffer area into a compression block, and storing the compression block into a performance optimization module.
6. A cloud translation-based management system is characterized by comprising a server, a monitoring module, an activity management module, a state prediction module, a working platform, an alarm module, a management and control module, a performance optimization module and a carousel white board;
the server is used for receiving and storing the animal data of each group;
the monitoring module is used for collecting the activity condition of each animal in the area in real time;
the activity management module is used for receiving the collected activity conditions of each animal and performing cascade detection;
the state prediction module is used for collecting data of each animal and performing state prediction;
the working platform is used for receiving the detection information and the prediction information for the staff to check;
the alarm module is used for receiving abnormal animal information and giving an alarm to prompt staff;
the management and control module is used for monitoring animals in the management and control area;
the performance optimization module is used for optimizing the running performance of the server;
the carousel whiteboard is used for circularly playing the animal information and the server operation information in the area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211448386.3A CN115760523A (en) | 2022-11-18 | 2022-11-18 | Animal management method and system based on cloud platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211448386.3A CN115760523A (en) | 2022-11-18 | 2022-11-18 | Animal management method and system based on cloud platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115760523A true CN115760523A (en) | 2023-03-07 |
Family
ID=85373513
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211448386.3A Pending CN115760523A (en) | 2022-11-18 | 2022-11-18 | Animal management method and system based on cloud platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115760523A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116863368A (en) * | 2023-06-06 | 2023-10-10 | 深圳启示智能科技有限公司 | Artificial intelligent identification terminal |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160012748A1 (en) * | 2014-07-08 | 2016-01-14 | Nestec Sa | Systems and methods for providing animal health, nutrition, and/or wellness recommendations |
| CN105512640A (en) * | 2015-12-30 | 2016-04-20 | 重庆邮电大学 | Method for acquiring people flow on the basis of video sequence |
| CN113223059A (en) * | 2021-05-17 | 2021-08-06 | 浙江大学 | Weak and small airspace target detection method based on super-resolution feature enhancement |
| CN113671892A (en) * | 2021-08-18 | 2021-11-19 | 迟明海 | Animal management method and animal management system based on cloud platform |
| CN114283355A (en) * | 2021-12-06 | 2022-04-05 | 重庆邮电大学 | Multi-target endangered animal tracking method based on small sample learning |
| CN115277243A (en) * | 2022-08-04 | 2022-11-01 | 陈锦汇 | Network communication scheme based on LRU cache |
| CN115329676A (en) * | 2022-08-23 | 2022-11-11 | 山东工艺美术学院 | Digital twin virtual-real synchronous terminal equipment and system |
-
2022
- 2022-11-18 CN CN202211448386.3A patent/CN115760523A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160012748A1 (en) * | 2014-07-08 | 2016-01-14 | Nestec Sa | Systems and methods for providing animal health, nutrition, and/or wellness recommendations |
| CN105512640A (en) * | 2015-12-30 | 2016-04-20 | 重庆邮电大学 | Method for acquiring people flow on the basis of video sequence |
| CN113223059A (en) * | 2021-05-17 | 2021-08-06 | 浙江大学 | Weak and small airspace target detection method based on super-resolution feature enhancement |
| CN113671892A (en) * | 2021-08-18 | 2021-11-19 | 迟明海 | Animal management method and animal management system based on cloud platform |
| CN114283355A (en) * | 2021-12-06 | 2022-04-05 | 重庆邮电大学 | Multi-target endangered animal tracking method based on small sample learning |
| CN115277243A (en) * | 2022-08-04 | 2022-11-01 | 陈锦汇 | Network communication scheme based on LRU cache |
| CN115329676A (en) * | 2022-08-23 | 2022-11-11 | 山东工艺美术学院 | Digital twin virtual-real synchronous terminal equipment and system |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116863368A (en) * | 2023-06-06 | 2023-10-10 | 深圳启示智能科技有限公司 | Artificial intelligent identification terminal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112580552B (en) | Murine behavior analysis method and device | |
| CN108175389B (en) | Multi-source information fusion cow behavior monitoring system and method | |
| CN110889448A (en) | Electrocardiogram classification method based on convolutional neural network | |
| CN115760523A (en) | Animal management method and system based on cloud platform | |
| CN111161090B (en) | Method, device and system for determining containment column information and storage medium | |
| CN110610125A (en) | Ox face identification method, device, equipment and storage medium based on neural network | |
| CN114943917A (en) | Algorithm for visually identifying aeration rate of aerobic tank of sewage plant | |
| CN114627411A (en) | Crop growth period identification method based on parallel detection under computer vision | |
| CN114980011B (en) | Livestock and poultry body temperature monitoring method with cooperation of wearable sensor and infrared camera | |
| CN115937251A (en) | Multi-target tracking method for shrimps | |
| CN114581948A (en) | Animal face identification method | |
| CN116016869A (en) | Campus safety monitoring system based on artificial intelligence and Internet of things | |
| CN113569656B (en) | Examination room monitoring method based on deep learning | |
| CN111027385A (en) | Clustering visitor counting method, system, equipment and computer readable storage medium | |
| CN112528823A (en) | Striped shark movement behavior analysis method and system based on key frame detection and semantic component segmentation | |
| CN115547500B (en) | Hemodialysis patient's health monitoring equipment and system | |
| WO2023108782A1 (en) | Method and apparatus for training behavior recognition model, behavior recognition method, apparatus and system, and medium | |
| CN115857473A (en) | Intelligent driving and ADAS simulation test method and system based on satellite positioning | |
| CN113111847A (en) | Automatic monitoring method, device and system for process circulation | |
| CN113191196A (en) | Novel track analysis method and system in intelligent security system | |
| CN116740598B (en) | Method and system for identifying ability of old people based on video AI identification | |
| CN109002791A (en) | A kind of system and method automatically tracking milk cow Ruminant behavior based on video | |
| CN117612693B (en) | Patient real-time monitoring and early warning method, device, computer and storage medium | |
| CN114711790B (en) | Newborn electroconvulsive type determination method, newborn electroconvulsive type determination device, newborn electroconvulsive type determination equipment and storage medium | |
| JP7368916B1 (en) | Pollen identification device using learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |