CN110826406A - Child high-altitude protection method based on deep learning model - Google Patents
Child high-altitude protection method based on deep learning model Download PDFInfo
- Publication number
- CN110826406A CN110826406A CN201910948946.3A CN201910948946A CN110826406A CN 110826406 A CN110826406 A CN 110826406A CN 201910948946 A CN201910948946 A CN 201910948946A CN 110826406 A CN110826406 A CN 110826406A
- Authority
- CN
- China
- Prior art keywords
- human body
- face
- protection method
- deep learning
- child
- 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
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000013136 deep learning model Methods 0.000 title claims abstract description 17
- 238000001514 detection method Methods 0.000 claims abstract description 63
- 239000000284 extract Substances 0.000 claims abstract description 7
- 210000000746 body region Anatomy 0.000 claims abstract description 4
- 238000012544 monitoring process Methods 0.000 claims description 16
- 238000004458 analytical method Methods 0.000 claims description 10
- 238000005516 engineering process Methods 0.000 claims description 10
- 230000036544 posture Effects 0.000 claims description 10
- 230000005540 biological transmission Effects 0.000 claims description 4
- 240000007651 Rubus glaucus Species 0.000 claims description 3
- 235000011034 Rubus glaucus Nutrition 0.000 claims description 3
- 235000009122 Rubus idaeus Nutrition 0.000 claims description 3
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 230000035515 penetration Effects 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 230000000007 visual effect Effects 0.000 claims description 3
- 230000001815 facial effect Effects 0.000 claims description 2
- 230000009194 climbing Effects 0.000 abstract description 3
- 230000002265 prevention Effects 0.000 abstract description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 2
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000013100 final test Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000013095 identification testing Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000001681 protective effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/103—Static body considered as a whole, e.g. static pedestrian or occupant recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Signal Processing (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
- Alarm Systems (AREA)
Abstract
The invention discloses a child high-altitude protection method based on a deep learning model, and relates to a safety protection method. At present, how to prevent children from climbing out of a window and how to remind parents under the condition that the parents are not in place for supervision is a difficult point to solve. The invention comprises the following steps: when a moving object is detected in the picture, the human body information detection module detects human body information in each frame of image in real time; the human body information detection module extracts the height and position information of a human body; the target tracking module tracks the detected human body region to ensure that the detected human body does not lose frames, and the human body information detection module and the target tracking module are mutually fused by adopting a Hungarian algorithm; according to the technical scheme, the identity recognition model based on the human face, the age recognition model based on the human face and the height recognition model based on the human body key point recognition are fused, accurate recognition of children is carried out, and safety prevention and control of the children are achieved effectively.
Description
Technical Field
The invention relates to a safety protection method, in particular to a child high-altitude protection method based on a deep learning model.
Background
The children fall from high altitude, and the children die from high altitude, so that tragedies are all the year round. Survey statistics show that the tragedies are not close to unprotected balconies or windows when children approach the balcony or the European window, so that how to timely stop the children from climbing out of the balcony or the European window becomes the serious factor for preventing safety accidents.
The research shows that: the younger the child is, the more difficult it is to detect the high potential risk. This is because children are not aware of the outside world and experience little dangerous experience; meanwhile, the child has poor migration capability and lacks of specific perception and action balance coordination capability. Even if the children feel danger under the conditions of curiosity to drive, opening windows for playing and the like, the children are too late. By carefully analyzing the reasons for tragedies, in addition to subjective reasons such as carelessness of adults, it is obvious that safety facilities such as windows of high-rise residents and lack of protective guards on balconies are also one of the most direct factors causing children to fall from high altitude. How to prevent children from climbing out of a window and how to remind parents under the condition that the parents are not monitored in place is a difficult point to solve.
Disclosure of Invention
The invention aims to solve the technical problems and provide a technical task for improving and improving the prior technical scheme and providing a child high-altitude protection method based on a deep learning model so as to achieve the aim of improving safety. Therefore, the invention adopts the following technical scheme.
A child high-altitude protection method based on a deep learning model comprises the following steps:
1) the camera collects videos and pushes the collected video data to a public network and an intranet part in a video streaming mode;
2) the video analysis terminal acquires video stream, reads code stream data and detects moving objects;
3) when a moving object is detected in the picture, the human body information detection module detects human body information in each frame of image in real time; the human body information detection module extracts the height and position information of a human body; the target tracking module tracks the detected human body region to ensure that the detected human body does not lose frames, and the human body information detection module and the target tracking module are mutually fused by adopting a Hungarian algorithm, so that the accuracy of detection and target tracking is improved;
4) when a moving object is detected in the picture, the face information detection module extracts a face area, an age and a face identity matched with the face feature database for searching;
5) identifying the identity of one person and one face according to the face information and the human body information obtained in the step 3) and the step 4);
6) when the identified identity information is the child, monitoring whether the child crosses the warning line, and issuing an alarm instruction when the video analysis identifies that the child target crosses the warning line and no adult attends;
7) when a dangerous condition exists, issuing commands of closing doors and windows and/or opening an alarm buzzer and sending early warning short messages to a mobile terminal;
8) when the door and window closing device receives the issued closed door and window, closing the door and window;
9) the mobile terminal checks the field condition in real time through a network and can remotely realize the opening and closing of doors and windows and the alarm relief.
As a preferable technical means: in the step 2), the video analysis terminal analyzes the video in real time, and performs background adaptive learning and moving object detection on the monitoring picture by using a background subtractor GMG model.
As a preferable technical means: in the step 3), the human body information detection module adopts a Yolov3 target detection model to identify the human body, and gives a human body characteristic vector and gives a unique id to the human body.
As a preferable technical means: the target tracking module adopts a Kalman filtering target tracking algorithm to make up for missing detection of a part of frames of a human body detection algorithm.
As a preferable technical means: in step 3), human body information and human bodies under different postures are identified through the human body key point detection module, and the human body key point detection module adopts a human body key point detection model simple _ position _ resnet.
As a preferable technical means: in the step 4), the key point detection of the retinafee model is used for high-precision positioning of the facial features of the human face, a human face feature database is established, the human face of the target child is matched, and high-precision identification of the human face of the family child is achieved by inputting the picture of the child in the family; and the age of the face is identified to realize the face discrimination of the strange children.
As a preferable technical means: in step 5), a deep _ sort model architecture is adopted to match the detected human face with the detected human body, and matching is started from the moment the human body enters the camera.
As a preferable technical means: and in the step 6), monitoring the position of the target in real time, continuously updating target information, and quickly making a judgment and responding in time when the attention target exceeds an early warning line.
As a preferable technical means: the human body information detection module is provided with a height recognition model, when the height recognition model is trained, various features including the position of a human body, the maximum pixel length and the posture are extracted under the condition of a fixed monitoring visual angle, and when the final detection error is within 15cm, the training is finished.
As a preferable technical means: when internal and external network data transmission is carried out, the command of the internal network and the public network to the raspberry group is issued through an internal network penetration technology; and the current state of the door and the window is obtained through the data acquisition module.
Has the advantages that:
1. the technical scheme performs reliable monitoring through the recognition of human body information and face information in the video. The safety prevention and control of children are effectively realized.
2. Under the alarm state, the operations of alarming sound, door and window closing, mobile phone reminding and the like are simultaneously realized. The safety protection of children is ensured.
3. Can accomplish the intranet and read when video streaming transmission, with public network live broadcast two kinds of modes, effectively improve the security, can be outside long-range look over the real-time condition to safety prevention and control.
4. By adopting a backstreaming subtrectorGMG model and utilizing the characteristics that the model is insensitive to slow changes of light rays and the like and sensitive to foreground actions, the security is improved for monitoring the family video.
5. When the child singly crosses the set warning line, triggering an alarm; when the adult and the child cross the warning line at the same time, the alarm cannot be given; this reduces the impact of unnecessary alarms on life.
Drawings
FIG. 1 is a flow chart of the present invention.
Fig. 2 is a schematic diagram of the present invention.
Fig. 3 is a flowchart of a face recognition procedure of the present invention.
FIG. 4 is a flow chart of the height identification procedure of the present invention.
FIG. 5 is a three-layer neural network of the present invention.
Detailed Description
The technical scheme of the invention is further explained in detail by combining the drawings in the specification.
As shown in fig. 1, the present invention comprises the steps of:
1) the camera collects videos and pushes the collected video data to a public network and an intranet part in a video streaming mode;
2) the video analysis terminal acquires video stream, reads code stream data and detects moving objects; the video analysis terminal analyzes the video in real time, and performs background adaptive learning and moving object detection on the monitoring picture by using a background subtractor GMG model;
3) when a moving object is detected in the picture, the human body information detection module detects human body information in each frame of image in real time; the human body information detection module extracts the height and position information of a human body; the target tracking module tracks the detected human body region to ensure that the detected human body does not lose frames, and the human body information detection module and the target tracking module are mutually fused by adopting a Hungarian algorithm, so that the accuracy of detection and target tracking is improved; identifying human body information and human bodies under different postures through a human body key point detection module, wherein the human body key point detection module adopts a human body key point detection model simple _ position _ resnet; the human body information detection module adopts a YOLOV3 target detection model to identify a human body, and gives a human body characteristic vector to give a unique id to the human body; the target tracking module adopts a Kalman filtering target tracking algorithm to make up for missing detection of a part of frames of a human body detection algorithm;
4) when a moving object is detected in the picture, the face information detection module extracts a face area, an age and a face identity matched with the face feature database for searching; the method comprises the steps of detecting 128 key points of a retinafee model to carry out high-precision positioning on face features, establishing a face feature database, matching the face of a target child, and recording photos of the child in a family to realize high-precision identification of the face of the family child; identifying the age of the face to realize the face discrimination of the strange children;
5) identifying the identity of one person and one face according to the face information and the human body information obtained in the step 3) and the step 4); matching the detected human face with the detected human body by adopting a deep _ sort model architecture, and starting matching from the moment the human body enters the camera;
6) when the identified identity information is the child, monitoring whether the child crosses the warning line, and issuing an alarm instruction when the video analysis identifies that the child target crosses the warning line and no adult attends; monitoring the position of a target in real time, continuously updating target information, quickly making a judgment when a focus target exceeds an early warning line, and responding in time;
7) when a dangerous condition exists, issuing commands of closing doors and windows and/or opening an alarm buzzer and sending early warning short messages to a mobile terminal;
8) when the door and window closing device receives the issued closed door and window, closing the door and window;
9) the mobile terminal checks the field condition in real time through a network and can remotely realize the opening and closing of doors and windows and the alarm relief.
According to the technical scheme, three detection strategies are fused, namely an identity recognition model based on the human face, an age recognition model based on the human face, a height recognition model based on human body key point recognition and the accurate recognition of the current children through the fusion of the three models.
The following problems are overcome:
the method for realizing face recognition only by using a monitoring camera has the following problems that when a child enters a monitoring range, the method for immediately recognizing the face recognition is as follows: such as: when the child faces back or the side faces face the camera, the face recognition cannot be performed, and the identity of the strange child cannot be performed.
Second, a technology for identifying an age only by a human face includes: for the case that the face recognition cannot be performed in a special posture.
In order to improve the accuracy of recognition, the human body information detection module is provided with a height recognition model, when the height recognition model is trained, various features including the position of a human body, the maximum pixel length and the posture are extracted under the condition of fixing a monitoring visual angle, and when the final detection error is within 15cm, the training is finished.
When internal and external network data transmission is carried out, the command of the internal network and the public network to the raspberry group is issued through an internal network penetration technology; and the current state of the door and the window is obtained through the data acquisition module.
As shown in figure 2, the system that this method relied on includes the camera, the video analysis terminal, parts such as removal end APP, actuating device includes electric control terminal, parts such as fence structure, with traditional rail, the window is upgraded and is reformed transform, through the camera control, utilize image recognition technology discernment children to be close to the windowsill, and in time report an emergency and ask for help or increased vigilance the recognition result, close safety grid or window through APP control windowsill, to children and its action being close to the balcony, the window carry out intelligent discernment and management, realize thing and thing, thing and people's ubiquitous connection and science and technology, safety, comfortable organic combination, solve the problem of children safety protection, provide a means that effectively prevents children from falling from the building for the head of a family.
The following specific examples are given for some embodiments:
firstly, face identity recognition:
by utilizing the face recognition technology, the face (generally 1-3 persons) of the child in the family is read and recorded, meanwhile, the face is collected by the camera for comparison, and when the similarity is higher, the child is considered to be protected, and an alarm signal is sent.
In a face detection experiment, the two models are excellent in performance, the face can be completely detected under the condition that the face faces a lens, the corresponding detection time is short, and the detection can be completed within 30 ms.
The flow of the face recognition procedure is shown in fig. 3.
In this embodiment, the face information detection module adopts the euclidean distance in the face similarity calculation, and selects the empirical threshold value of 0.33 as the threshold value for face matching. When the threshold value is exceeded, the face matching is successful.
The experimental results are as follows: the results of the actual tests are shown in the table below.
TABLE 2 face recognition test results
| Identification method | Recognition rate |
| Insightface | 98% |
| Dilb | 93% |
Through the face identification technology, the identity of the child in the image can be accurately and efficiently identified under the condition of facing the lens.
Second, human body recognition
In the model for pedestrian recognition, a target tracking technology is added to make up for the defects of the detection model, and experiments show that the recognition effect of the human body target is greatly improved after target tracking is added.
The experimental results are as follows: as shown in the table 3 below, the following examples,
TABLE 3 posture identification test results
| Posture recognition method | Unadditized target tracking technology | Joining target tracking techniques |
| ssd_mobilenet1.0 | 90% | 99% |
| yolo3_darknet | 91% | 99% |
Third, height recognition
The detection method comprises the following steps: one is a method based on camera calibration, which obtains a high-precision target size under a limited condition through projection matrix calculation, and the other is a method which obtains maximum likelihood estimation of a three-dimensional attitude coordinate of a measured target through learning of a known sample through deep learning. The second method is adopted in the technical scheme, because:
a first height detection method comprises the following steps:
the camera monitors the picture, is in a fixed position, is easy to calibrate the relative position of the camera and the ground, and determines the position of the human body on the ground according to the position of the foot of the human body, so that the number of pixel points occupied by the longest edge of the human body can be obtained, the height of the human body can be calculated according to the obtained number of the pixel points, and finally, the height result obtained by the human body on a multi-frame picture is utilized to select the maximum value to be judged as the height identification result. However, the method has a long execution flow and a complex calculation process, and once the field environment is slightly changed, the camera needs to be calibrated again, and the robustness is poor due to the fact that parameters such as the focal length of the camera lens are fixed and unchanged.
A second height detection method:
the human body position is obtained through selection and detection by adopting an approximate detection method experiment of a neural network regression model based on the human body posture position, and the coordinate data of key points, and a three-layer full-connection neural network is constructed, as shown in figure 5, so that the height identification and detection are carried out. The height identification detection procedure is shown in FIG. 4. Through final test, the human body appearing in the picture can be stably identified, and the maximum error of the height identification part is below 15cm, so that the use requirement can be met.
The method for protecting children from high altitude based on deep learning model shown in fig. 1-4 is a specific embodiment of the present invention, which already embodies the essential features and advances of the present invention, and can be modified equivalently in shape, structure and the like according to the practical needs and with the teaching of the present invention, and is within the scope of protection of the present invention.
Claims (10)
1. A child high-altitude protection method based on a deep learning model is characterized by comprising the following steps: the method comprises the following steps:
1) the camera collects videos and pushes the collected video data to a public network and an intranet part in a video streaming mode;
2) the video analysis terminal acquires video stream, reads code stream data and detects moving objects;
3) when a moving object is detected in the picture, the human body information detection module detects human body information in each frame of image in real time; the human body information detection module extracts the height and position information of a human body; the target tracking module tracks the detected human body region to ensure that the detected human body does not lose frames, and the human body information detection module and the target tracking module are mutually fused by adopting a Hungarian algorithm, so that the accuracy of detection and target tracking is improved;
4) when a moving object is detected in the picture, the face information detection module extracts a face area, an age and a face identity matched with the face feature database for searching;
5) identifying the identity of one person and one face according to the face information and the human body information obtained in the step 3) and the step 4);
6) when the identified identity information is the child, monitoring whether the child crosses the warning line, and issuing an alarm instruction when the video analysis identifies that the child target crosses the warning line and no adult attends;
7) when a dangerous condition exists, issuing commands of closing doors and windows and/or opening an alarm buzzer and sending early warning short messages to a mobile terminal;
8) when the door and window closing device receives the issued closed door and window, closing the door and window;
9) the mobile terminal checks the field condition in real time through a network and can remotely realize the opening and closing of doors and windows and the alarm relief.
2. The high-altitude protection method for children based on the deep learning model as claimed in claim 1, characterized in that: in the step 2), the video analysis terminal analyzes the video in real time, and performs background adaptive learning and moving object detection on the monitoring picture by using a background subtractor GMG model.
3. The high-altitude protection method for children based on the deep learning model as claimed in claim 1, characterized in that: in the step 3), the human body information detection module adopts a Yolov3 target detection model to identify the human body, and gives a human body characteristic vector and gives a unique id to the human body.
4. The high-altitude protection method for children based on the deep learning model as claimed in claim 3, characterized in that: the target tracking module adopts a Kalman filtering target tracking algorithm to make up for missing detection of a part of frames of a human body detection algorithm.
5. The high-altitude protection method for children based on the deep learning model as claimed in claim 4, wherein the high-altitude protection method comprises the following steps: in step 3), human body information and human bodies under different postures are identified through the human body key point detection module, and the human body key point detection module adopts a human body key point detection model simple _ position _ resnet.
6. The high-altitude protection method for children based on the deep learning model as claimed in claim 1, characterized in that: in the step 4), the key point detection of the retinafee model is used for high-precision positioning of the facial features of the human face, a human face feature database is established, the human face of the target child is matched, and high-precision identification of the human face of the family child is achieved by inputting the picture of the child in the family; and the age of the face is identified to realize the face discrimination of the strange children.
7. The high-altitude protection method for children based on the deep learning model as claimed in claim 1, characterized in that: in step 5), a deep _ sort model architecture is adopted to match the detected human face with the detected human body, and matching is started from the moment the human body enters the camera.
8. The high-altitude protection method for children based on the deep learning model as claimed in claim 1, characterized in that: and in the step 6), monitoring the position of the target in real time, continuously updating target information, and quickly making a judgment and responding in time when the attention target exceeds an early warning line.
9. The high-altitude protection method for children based on the deep learning model as claimed in claim 1, characterized in that: the human body information detection module is provided with a height recognition model, when the height recognition model is trained, various features including the position of a human body, the maximum pixel length and the posture are extracted under the condition of a fixed monitoring visual angle, and when the final detection error is within 15cm, the training is finished.
10. The high-altitude protection method for children based on the deep learning model as claimed in claim 1, characterized in that: when internal and external network data transmission is carried out, the command of the internal network and the public network to the raspberry group is issued through an internal network penetration technology; and the current state of the door and the window is obtained through the data acquisition module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910948946.3A CN110826406A (en) | 2019-10-08 | 2019-10-08 | Child high-altitude protection method based on deep learning model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910948946.3A CN110826406A (en) | 2019-10-08 | 2019-10-08 | Child high-altitude protection method based on deep learning model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110826406A true CN110826406A (en) | 2020-02-21 |
Family
ID=69548700
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910948946.3A Pending CN110826406A (en) | 2019-10-08 | 2019-10-08 | Child high-altitude protection method based on deep learning model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110826406A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111325279A (en) * | 2020-02-26 | 2020-06-23 | 福州大学 | Pedestrian and personal sensitive article tracking method fusing visual relationship |
| CN111898511A (en) * | 2020-07-23 | 2020-11-06 | 北京以萨技术股份有限公司 | High-altitude parabolic detection method, device and medium based on deep learning |
| CN112130463A (en) * | 2020-09-30 | 2020-12-25 | 宁波久婵物联科技有限公司 | Intelligent home linkage system based on face lock |
| CN112257549A (en) * | 2020-10-19 | 2021-01-22 | 中国电子科技集团公司第五十八研究所 | Floor danger detection early warning method and system based on computer vision |
| CN114187738A (en) * | 2020-08-24 | 2022-03-15 | 浙江宇视科技有限公司 | Safety early warning method, device, equipment and medium |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN206236185U (en) * | 2016-10-09 | 2017-06-09 | 歌尔科技有限公司 | A kind of intelligent door system |
| CN107045623A (en) * | 2016-12-30 | 2017-08-15 | 厦门瑞为信息技术有限公司 | A kind of method of the indoor dangerous situation alarm based on human body attitude trace analysis |
| CN107273822A (en) * | 2017-05-26 | 2017-10-20 | 西安电子科技大学 | A kind of method for secret protection based on monitor video multiple target tracking and recognition of face |
| CN107358149A (en) * | 2017-05-27 | 2017-11-17 | 深圳市深网视界科技有限公司 | A kind of human body attitude detection method and device |
| US20180039837A1 (en) * | 2014-06-05 | 2018-02-08 | Ainemo Inc | Device and method for automatic monitoring and autonomic response |
| CN108289202A (en) * | 2018-01-29 | 2018-07-17 | 北京大椽科技有限公司 | A kind of intelligent network monitoring method and system |
| WO2018133666A1 (en) * | 2017-01-17 | 2018-07-26 | 腾讯科技（深圳）有限公司 | Method and apparatus for tracking video target |
| CN109035685A (en) * | 2018-06-26 | 2018-12-18 | 浙江树人学院 | A kind of child's anti-falling overhead system and method |
| CN109166272A (en) * | 2018-07-20 | 2019-01-08 | 北京联合大学 | A kind of home monitoring system for children |
| CN109186584A (en) * | 2018-07-18 | 2019-01-11 | 浙江臻万科技有限公司 | A kind of indoor orientation method and positioning system based on recognition of face |
| CN109376679A (en) * | 2018-11-05 | 2019-02-22 | 绍兴文理学院 | A kind of face identification system and method based on deep learning |
| CN109785564A (en) * | 2019-03-21 | 2019-05-21 | 安徽威尔信通信科技有限责任公司 | A kind of home intelligent safety defense monitoring system |
-
2019
- 2019-10-08 CN CN201910948946.3A patent/CN110826406A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180039837A1 (en) * | 2014-06-05 | 2018-02-08 | Ainemo Inc | Device and method for automatic monitoring and autonomic response |
| CN206236185U (en) * | 2016-10-09 | 2017-06-09 | 歌尔科技有限公司 | A kind of intelligent door system |
| CN107045623A (en) * | 2016-12-30 | 2017-08-15 | 厦门瑞为信息技术有限公司 | A kind of method of the indoor dangerous situation alarm based on human body attitude trace analysis |
| WO2018133666A1 (en) * | 2017-01-17 | 2018-07-26 | 腾讯科技（深圳）有限公司 | Method and apparatus for tracking video target |
| CN107273822A (en) * | 2017-05-26 | 2017-10-20 | 西安电子科技大学 | A kind of method for secret protection based on monitor video multiple target tracking and recognition of face |
| CN107358149A (en) * | 2017-05-27 | 2017-11-17 | 深圳市深网视界科技有限公司 | A kind of human body attitude detection method and device |
| CN108289202A (en) * | 2018-01-29 | 2018-07-17 | 北京大椽科技有限公司 | A kind of intelligent network monitoring method and system |
| CN109035685A (en) * | 2018-06-26 | 2018-12-18 | 浙江树人学院 | A kind of child's anti-falling overhead system and method |
| CN109186584A (en) * | 2018-07-18 | 2019-01-11 | 浙江臻万科技有限公司 | A kind of indoor orientation method and positioning system based on recognition of face |
| CN109166272A (en) * | 2018-07-20 | 2019-01-08 | 北京联合大学 | A kind of home monitoring system for children |
| CN109376679A (en) * | 2018-11-05 | 2019-02-22 | 绍兴文理学院 | A kind of face identification system and method based on deep learning |
| CN109785564A (en) * | 2019-03-21 | 2019-05-21 | 安徽威尔信通信科技有限责任公司 | A kind of home intelligent safety defense monitoring system |
Non-Patent Citations (4)
| Title |
|---|
| BIN XIAO, ET AL.: "Simple Baselines for Human Pose Estimation and Tracking" * |
| JIANKANG DENG, ET AL.: "RetinaFace: Single-stage Dense Face Localisation in the Wild" * |
| 张梦: "安防巡逻机器人行人感知***研究" * |
| 李敏等: "基于智能视频分析的人流量态势感知方法研究" * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111325279A (en) * | 2020-02-26 | 2020-06-23 | 福州大学 | Pedestrian and personal sensitive article tracking method fusing visual relationship |
| CN111325279B (en) * | 2020-02-26 | 2022-06-10 | 福州大学 | Pedestrian and personal sensitive article tracking method fusing visual relationship |
| CN111898511A (en) * | 2020-07-23 | 2020-11-06 | 北京以萨技术股份有限公司 | High-altitude parabolic detection method, device and medium based on deep learning |
| CN114187738A (en) * | 2020-08-24 | 2022-03-15 | 浙江宇视科技有限公司 | Safety early warning method, device, equipment and medium |
| CN112130463A (en) * | 2020-09-30 | 2020-12-25 | 宁波久婵物联科技有限公司 | Intelligent home linkage system based on face lock |
| CN112257549A (en) * | 2020-10-19 | 2021-01-22 | 中国电子科技集团公司第五十八研究所 | Floor danger detection early warning method and system based on computer vision |
| CN112257549B (en) * | 2020-10-19 | 2022-08-02 | 中国电子科技集团公司第五十八研究所 | Floor danger detection early warning method and system based on computer vision |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110826406A (en) | Child high-altitude protection method based on deep learning model | |
| CN103839373B (en) | A kind of unexpected abnormality event Intelligent Recognition alarm device and warning system | |
| US10657749B2 (en) | Automatic system access using facial recognition | |
| US10235822B2 (en) | Automatic system access using facial recognition | |
| US20190236342A1 (en) | Face concealment detection | |
| CN103246869B (en) | Method is monitored in crime based on recognition of face and behavior speech recognition | |
| KR101164228B1 (en) | A security system and a method using multiplex biometrics of face and body | |
| CN103198595A (en) | Intelligent door and window anti-invasion system | |
| US20120188370A1 (en) | Surveillance systems and methods to monitor, recognize, track objects and unusual activities in real time within user defined boundaries in an area | |
| CN110795963A (en) | Monitoring method, device and equipment based on face recognition | |
| CN111540152A (en) | Intelligent security early warning system | |
| CN111739248B (en) | Artificial intelligent Internet of things security system and control method | |
| EP3918586B1 (en) | Smart surveillance system for swimming pools | |
| Shoaib et al. | View-invariant fall detection for elderly in real home environment | |
| CN109035685B (en) | High-altitude falling prevention system and method for infants | |
| CN108765921A (en) | View-based access control model lexical analysis is applied to the intelligent patrol method of patrol robot | |
| CN116563797B (en) | Monitoring management system for intelligent campus | |
| US11509831B2 (en) | Synchronous head movement (SHMOV) detection systems and methods | |
| CN107416630A (en) | The detection method and system of the improper closing of elevator | |
| CN116994389A (en) | Monitoring alarm driving system and method based on artificial intelligence and image recognition | |
| CN111681355A (en) | Security and protection system applicable to smart campus | |
| Handalage et al. | Computer vision enabled drowning detection system | |
| CN203885510U (en) | Driver fatigue detection system based on infrared detection technology | |
| Nagamani et al. | Anti-Theft Monitoring for a Smart Home | |
| WO2022204153A1 (en) | Image based tracking system |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200221 |
|
| WD01 | Invention patent application deemed withdrawn after publication |