CN113392706A - Device and method for detecting smoking and using mobile phone behaviors - Google Patents
Device and method for detecting smoking and using mobile phone behaviors Download PDFInfo
- Publication number
- CN113392706A CN113392706A CN202110525138.3A CN202110525138A CN113392706A CN 113392706 A CN113392706 A CN 113392706A CN 202110525138 A CN202110525138 A CN 202110525138A CN 113392706 A CN113392706 A CN 113392706A
- Authority
- CN
- China
- Prior art keywords
- image
- smoking
- mobile phone
- data
- classifier
- 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
- 230000000391 smoking effect Effects 0.000 title claims abstract description 46
- 230000006399 behavior Effects 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 21
- 238000001514 detection method Methods 0.000 claims abstract description 48
- 238000012545 processing Methods 0.000 claims abstract description 3
- 235000019504 cigarettes Nutrition 0.000 claims description 7
- 238000005520 cutting process Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 239000000779 smoke Substances 0.000 description 11
- 238000010276 construction Methods 0.000 description 7
- 238000004519 manufacturing process Methods 0.000 description 5
- 241000282412 Homo Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a device for detecting smoking and using mobile phone behaviors, which comprises a camera, a display and a controller, wherein the camera is used for obtaining a human body image of a working site; the network is used for transmitting the image to the remote server; the human body posture detector identifies a human body image based on a pre-trained yolov3-SPP algorithm, and sends the human body image into a trained key point detection network for processing; the classifier classifies the images according to the distance between the hands and the head according to the key point detection network, and identifies the images for smoking or using the mobile phone. The invention can construct a high-precision detection device and a high-precision identification method for identifying the smoking or mobile phone using behaviors of an operator in real time for the illegal behaviors of staff.
Description
Technical Field
The invention relates to the field of computer vision, in particular to a device and a method for detecting smoking and using mobile phone behaviors.
Background
Factory or worksite safety is of paramount importance to factory-type companies, particularly in the environment within a chemical plant. Such as an operator smoking a cigarette in a production environment, is highly likely to cause an unrecoverable accident. Serious accidents caused by mishandling of fire sources can be found from a lot of historical news. On the other hand, with the development of smart phones, the stickiness of people to the phones is increasing day by day, and people can see that sales staff carelessly sell products in many places but play the phones with low heads. The same situation may also occur in a factory production environment, but the situation is more severe because the production environment may be greatly lost to humans or to the economy due to a single loss. Without upper leaders in the field, operators are easy to have a lacked behavior, such as talking on a phone, playing the phone with a lower head, and the like. Therefore, the method provides a full-intelligent supervision for the plant in an artificial intelligence mode based on the improvement of the safety and the efficiency of the production environment of the plant, so that the potential dangerous behaviors of the plant can be prevented, and the behaviors of operators on the working lackluster and the like can be improved.
In the prior art (application number: 201921343086.2 a smoke behavior detection device for construction site), in a construction site scene, whether an operator smokes smoke or not is judged based on a sensor or a simple camera, and the smoke behavior detection device mainly comprises a guide rail, a support, a fixing part, a smoke sensor, a camera and the like. According to the scene, the camera is installed on the fixing piece, the plurality of smoke sensors are installed on the support, finally, the smoke sensors and the camera are connected with the server through the Ethernet, and the server is installed in the construction shed. The scheme needs great customization in a use scene, needs to consider field construction environments in many ways, and has low extensibility. Under the open environment of a construction site, smoke is diluted by the surrounding environment, so that the smoke sensor has poor performance (the smoke needs to reach a certain concentration and then alarms). Many devices are difficult to maintain, for example smoke sensors require regular cleaning to ensure accuracy of detection. Even the camera can be in time clap down the workman and smoke, also there is not timely feedback of the way and reports. Precaution against dangerous behavior is the one most needed by the plant. And the inspection is performed at the later stage of manual work, so that the efficiency is extremely low.
In other prior arts (application number: 201911073341.0 construction safety detection method and system based on tiny-YOLOv 3), the computer vision technology is also utilized, and the data labeling and sequential training are directly carried out on the acquired images based on the target detection mode. This scheme can not take actions such as safety helmet to the operator smoking to see through cell-phone APP with the result of detection and send for the operator and warn. In the scheme, the detection precision of yolov3-tiny belongs to an extremely light detection model in a deep learning detection model, the detection precision of the yolov3-tiny is lower than that of other models, and the algorithm is not enough to provide high accuracy for a factory environment with special attention to safety. The yolov3-tiny detector is only used for judging, and the detection of the whole image is easy to generate false alarm. To achieve a higher and wider field of view, cameras at a worksite are generally installed at higher positions, which means that the captured image scene can be very complex, containing many objects in addition to important targets (people), which results in the possibility that objects other than the targets (people) may increase the probability of being recognized wrongly. For example, the captured image is a 1920 × 1080 pixel image, the smoking to be detected may only account for 20 × 20 pixels, which is a very small proportion of the whole image, and after a plurality of times of feature extraction and downsampling, the amount of information useful for detection is very small.
The detection process in the prior art is not deep enough, and low false alarm and high accuracy cannot be achieved. Therefore, a reliable scheme is needed to solve the problems of low universality of applicable scenes and low detection precision and easy misjudgment in the prior art.
Disclosure of Invention
The invention aims to solve the problems of low universality of applicable scenes and low detection precision and easy misjudgment of illegal behaviors of staff, such as smoking or playing mobile phones, and builds a high-precision identification device capable of identifying the behaviors of smoking or using mobile phones of operators in real time on the illegal behaviors.
The invention provides a device for detecting smoking and using mobile phone behaviors, which comprises a camera, a network and a human posture detector, wherein the camera is used for obtaining human body images of a working site, the network is used for transmitting the images to a remote server, the human posture detector identifies the human body images based on a pre-trained yolov3-SPP algorithm, and the human body images are sent to a trained key point detection network for processing; the classifier classifies the images according to the distance between the hands and the head according to the key point detection network, and identifies the images for smoking or using the mobile phone.
Wherein the human gesture detector changes the detected image to 256 x 192 pixels in size.
The key point detection network is a well-trained network based on the MSCOCO2017 key point data. The key point detection network is the backbone network ResNet 50. The classifier is based on the medium lightweight model reseest 50 or EfficientNetB 4. The classifier partitions the data into three categories: 1. normal class 2, smoking 3 and using the mobile phone, the data occupation ratios of the three scenes are respectively about 0.6, 0.2 and 0.2.
The data of the normal class of the classifier, the smoking class and the data of the three classes of the mobile phone are head or hand data, and the resolution of the original image is not lower than 60 x 60 pixels.
Further, the classifier can be a detector, and not only can classify that the target in the image is a cigarette or a mobile phone, but also can find the position.
Another object of the present invention is to provide a method for detecting smoking and using a mobile phone by constructing a high-precision recognition device capable of recognizing smoking and using a mobile phone behavior of an operator in real time with respect to illegal behaviors such as smoking and playing a mobile phone in a workplace, comprising the steps of: an image A acquired by a camera firstly enters a human body posture detector in a digital image format, the human body posture detector detects the image A to obtain coordinate information of a person on the image, and the image A is cut out according to the coordinate information and is represented by an image B; the image B enters a key point detection model for detection, and the key point detection model selects key points of the head, the right hand and the left hand based on the MSCOCO data; the distances between the key points of the three parts are calculated to judge whether violation is taking place, the logic that the hand is close to the head represents that smoking or using a mobile phone is possible, and the two-hand distance is close represents that the mobile phone is also possible to play, so that three distances can be calculated: RH: right hand-to-head distance, LH: left-hand-to-head distance, RL: distance between the hands; screening out a target with a short distance as a potential violation through the RH, LH and RL by using a threshold, and then cutting out a regional image of the head or the hand and sending the regional image into a classifier for judgment; the cut images are sent to a pre-trained classifier for recognition, and then the images for smoking or using the mobile phone can be distinguished.
Wherein the human pose detector changes the detected human image size to 256 x 192 pixels based on the pre-trained yolov3-SPP algorithm. The key point detection model is a well-trained network based on the MSCOCO2017 key point data. The key point detection model is backbone network ResNet 50.
Wherein, the classifier is based on a medium-light weight model ResNest50 or EfficientNetB 4. The classifier divides the data into three categories 1, normal category 2, smoking 3, and the data ratio of the three scenes is about 0.6, 0.2 and 0.2 respectively by using a mobile phone. The data of the normal class of the classifier, the smoking class and the data of the three classes of the mobile phone are head or hand data, and the resolution of the original image is not lower than 60 x 60 pixels.
Further, the classifier can be a detector, and not only can classify that the target in the image is a cigarette or a mobile phone, but also can find the position.
The invention has the beneficial effects that in order to solve the problems of low universality of applicable scenes and low detection precision and easy misjudgment in the prior art, a high-precision identification device and a method capable of identifying the actions of smoking or using a mobile phone of an operator in real time are constructed for illegal actions.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flow chart of the smoking and behavior detection method using a mobile phone according to the present invention.
Figure 2 is a flow chart of another embodiment of a method for smoking a cigarette and detecting behavior using a mobile phone according to the present invention.
Detailed Description
The technical solution 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. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
Factory or construction site safety is the most important of factory type companies, for example, the operator smoking in the production environment is very likely to cause an unrecoverable accident, and under the condition of no top leader supervision on site, the operator is easy to have a lacked behavior such as talking a phone by using a mobile phone, playing the mobile phone with a head down, and the like. The prior art has the defects of low universality of applicable scenes, low detection precision and easy misjudgment. The invention constructs a high-precision identification device and method capable of identifying the smoking or mobile phone using behaviors of an operator in real time according to the illegal behaviors.
The present invention is described in further detail below to enable those skilled in the art to practice the invention with reference to the description.
Referring to fig. 1, fig. 1 is a flow chart of a smoking and behavior detection method using a mobile phone according to the present invention. In fig. 1, the detection flow of the smoking and the method for detecting behavior of using a mobile phone in this embodiment is as follows:
the method comprises the steps that an lmage A collected by a camera firstly enters a human body posture detector in a digital image format, the human body posture detector detects a human body, coordinate information of the human body on an image is obtained after the human body is detected, and the image is cut out according to the coordinate information and is represented by ImageB. ImageB enters a human body key point model for detection, the human body key point detection model is trained on the basis of MSCOCO data to generate 17 key points, and only 3 key point coordinates such as 1, a head key point (H)2, a right hand key point (R)3 and a left hand key point (L) need to be obtained according to requirements. The distances between the key points of the three parts are calculated to judge whether the violation is carried out, the logic that the hand is close to the head is that smoking or using a mobile phone is possible, and the logic that the hand is close to the head is that the mobile phone is possibly played. And screening out the targets with shorter distances, which are considered as potential illegal behaviors, through the threshold values of the RH, LH and RL, and then cutting out the area image of the head or the hand and sending the image into a classifier for discrimination. The image (about 60 × 60 pixels in size) of the head area or the hand area is cut out and sent to a pre-trained classifier for recognition, so that the current head area can be judged to be one of the three types of normal, smoking and mobile phone. And finally, the detection result is fed back in a form required by the product and the client.
Referring to fig. 2, fig. 2 is a flowchart illustrating a method for detecting smoking and behavior of a mobile phone according to another embodiment of the present invention.
In some embodiments, the classifier may be a detector that not only classifies the target in the image as a cigarette or a cell phone, but also finds the location.
While embodiments of the invention have been disclosed above, it is not intended to be limited to the uses set forth in the specification and examples. It can be applied to all kinds of fields suitable for the present invention. Additional modifications will readily occur to those skilled in the art. Therefore, the invention is not to be limited to the specific details and illustrations shown and described herein, without departing from the general concept defined by the claims and their equivalents.
Claims (18)
1. Smoking and use cell-phone action detection device includes:
the camera is used for obtaining a human body image of a working site;
the network is used for transmitting the image to the remote server;
the human body posture detector identifies a human body image based on a pre-trained yolov3-SPP algorithm, and sends the human body image into a trained key point detection network for processing;
the classifier classifies the images according to the distance between the hands and the head according to the key point detection network, and identifies the images for smoking or using the mobile phone.
2. The apparatus according to claim 1, wherein the human gesture detector changes the size of the detected image to 256 x 192 pixels.
3. The apparatus according to claim 1, wherein the keypoint detection model is a trained network based on MSCOCO2017 keypoint data.
4. The smoking and use handset behavior detection device of claim 1, wherein the keypoint detection network is backbone network ResNet 50.
5. The smoking and use handset behavior detection device of claim 1, wherein the classifier is based on a medium-to-light weight model reseest 50.
6. The smoking and use handset behavior detection device of claim 1, wherein the classifier is based on the medium-to-lightweight model EfficientNetB 4.
7. The apparatus for testing smoking and using handset behavior according to claim 1, wherein the classifier segments the data into three categories: 1. smoking 2, normal class, smoking 3, using the mobile phone; the data occupation ratios of the three scenes are about 0.6, 0.2 and 0.2 respectively.
8. The apparatus according to claim 1, wherein the data of the normal class of the classifier, the data of the smoking class and the data of the mobile phone class are head or hand data, and the resolution of the original image is not less than 60 x 60 pixels.
9. The apparatus according to claim 1, wherein the classifier is a detector, and not only can classify the object in the image as a cigarette or a mobile phone, but also can find the position.
10. The method for detecting smoking and using mobile phone behaviors is characterized by comprising the following steps of:
an image A acquired by a camera firstly enters a human body posture detector in a digital image format, the human body posture detector detects the image A to obtain coordinate information of a person on the image, and the image A is cut out according to the coordinate information and is represented by an image B;
the image B enters a key point detection model for detection, and the key point detection model selects key points of the head, the right hand and the left hand based on the MSCOCO data;
the distances between the key points of the three parts are calculated to judge whether violation is taking place, the logic that the hand is close to the head represents that smoking or using a mobile phone is possible, and the two-hand distance is close represents that the mobile phone is also possible to play, so that three distances can be calculated:
RH: right hand to head distance
LH: left hand to head distance
RL: distance between both hands
Screening out a target with a short distance as a potential violation through the RH, LH and RL by using a threshold, and then cutting out a regional image of the head or the hand and sending the regional image into a classifier for judgment;
the cut images are sent to a pre-trained classifier for recognition, and then the images of smoking or using the mobile phone can be distinguished.
11. The method of claim 10, wherein the human gesture detector changes the size of the detected human image to 256 x 192 pixels based on a pre-trained yolov3-SPP algorithm.
12. The method of claim 10, wherein the keypoint detection model is a network trained based on MSCOCO2017 keypoint data.
13. The apparatus according to claim 10, wherein the key point detection model is a backbone network ResNet 50.
14. The method of claim 10, wherein the classifier is based on the lightweight medium model reseest 50.
15. The method of claim 10, wherein the classifier is based on the medium-light model EfficientNetB 4.
16. The method according to claim 10, wherein the classifier classifies the data into three categories 1, normal category 2, smoking 3, using mobile phone; the data occupation ratios of the three scenes are respectively about 0.6, 0.2 and 0.2.
17. The method according to claim 10, wherein the data of the normal class of the classifier, the data of the smoking class and the data of the mobile phone class are head or hand data, and the resolution of the original image is not less than 60 x 60 pixels.
18. The method according to claim 10, wherein the classifier is a detector, and not only can classify the object in the image as a cigarette or a mobile phone, but also can find the position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110525138.3A CN113392706A (en) | 2021-05-13 | 2021-05-13 | Device and method for detecting smoking and using mobile phone behaviors |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110525138.3A CN113392706A (en) | 2021-05-13 | 2021-05-13 | Device and method for detecting smoking and using mobile phone behaviors |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113392706A true CN113392706A (en) | 2021-09-14 |
Family
ID=77617970
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110525138.3A Pending CN113392706A (en) | 2021-05-13 | 2021-05-13 | Device and method for detecting smoking and using mobile phone behaviors |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113392706A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114170677A (en) * | 2021-11-12 | 2022-03-11 | 深圳先进技术研究院 | Network model training method and equipment for detecting smoking behavior |
| CN115457518A (en) * | 2022-08-30 | 2022-12-09 | 淮阴工学院 | Driver behavior recognition method and system based on attitude perception and geometric constraint |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109858403A (en) * | 2019-01-16 | 2019-06-07 | 深圳壹账通智能科技有限公司 | A kind of method and photographic device that monitoring driver drives using mobile phone |
| CN110705383A (en) * | 2019-09-09 | 2020-01-17 | 深圳市中电数通智慧安全科技股份有限公司 | Smoking behavior detection method and device, terminal and readable storage medium |
| CN110738186A (en) * | 2019-10-23 | 2020-01-31 | 德瑞姆创新科技(深圳)有限公司 | driver smoking detection method and system based on computer vision technology |
| CN112115775A (en) * | 2020-08-07 | 2020-12-22 | 北京工业大学 | Smoking behavior detection method based on computer vision in monitoring scene |
| CN112163469A (en) * | 2020-09-11 | 2021-01-01 | 燊赛(上海)智能科技有限公司 | Smoking behavior recognition method, system, equipment and readable storage medium |
| CN112200092A (en) * | 2020-10-13 | 2021-01-08 | 深圳龙岗智能视听研究院 | Intelligent smoking detection method based on variable-focus movement of dome camera |
| CN112528960A (en) * | 2020-12-29 | 2021-03-19 | 之江实验室 | Smoking behavior detection method based on human body posture estimation and image classification |
-
2021
- 2021-05-13 CN CN202110525138.3A patent/CN113392706A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109858403A (en) * | 2019-01-16 | 2019-06-07 | 深圳壹账通智能科技有限公司 | A kind of method and photographic device that monitoring driver drives using mobile phone |
| CN110705383A (en) * | 2019-09-09 | 2020-01-17 | 深圳市中电数通智慧安全科技股份有限公司 | Smoking behavior detection method and device, terminal and readable storage medium |
| CN110738186A (en) * | 2019-10-23 | 2020-01-31 | 德瑞姆创新科技(深圳)有限公司 | driver smoking detection method and system based on computer vision technology |
| CN112115775A (en) * | 2020-08-07 | 2020-12-22 | 北京工业大学 | Smoking behavior detection method based on computer vision in monitoring scene |
| CN112163469A (en) * | 2020-09-11 | 2021-01-01 | 燊赛(上海)智能科技有限公司 | Smoking behavior recognition method, system, equipment and readable storage medium |
| CN112200092A (en) * | 2020-10-13 | 2021-01-08 | 深圳龙岗智能视听研究院 | Intelligent smoking detection method based on variable-focus movement of dome camera |
| CN112528960A (en) * | 2020-12-29 | 2021-03-19 | 之江实验室 | Smoking behavior detection method based on human body posture estimation and image classification |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114170677A (en) * | 2021-11-12 | 2022-03-11 | 深圳先进技术研究院 | Network model training method and equipment for detecting smoking behavior |
| WO2023082407A1 (en) * | 2021-11-12 | 2023-05-19 | 深圳先进技术研究院 | Network model training method for detecting smoking behavior and device thereof |
| CN115457518A (en) * | 2022-08-30 | 2022-12-09 | 淮阴工学院 | Driver behavior recognition method and system based on attitude perception and geometric constraint |
| CN115457518B (en) * | 2022-08-30 | 2024-01-26 | 淮阴工学院 | Driver behavior recognition method and system based on gesture perception and geometric constraint |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109040693B (en) | Intelligent alarm system and method | |
| CN107437318B (en) | Visible light intelligent recognition algorithm | |
| KR102149832B1 (en) | Automated Violence Detecting System based on Deep Learning | |
| CN109544838B (en) | Artificial intelligence cognitive recognition system for special area | |
| CN113392706A (en) | Device and method for detecting smoking and using mobile phone behaviors | |
| CN109657626B (en) | Analysis method for recognizing human body behaviors | |
| CN111079694A (en) | Counter assistant job function monitoring device and method | |
| CN111047824B (en) | Indoor child nursing linkage control early warning method and system | |
| JP7282186B2 (en) | situational awareness surveillance | |
| CN109544870A (en) | Alarm decision method and intelligent monitor system for intelligent monitor system | |
| CN115660297A (en) | Automatic AI early warning system and method for construction site safety | |
| KR20230039468A (en) | Interaction behavior detection apparatus between objects in the image and, method thereof | |
| CN116129490A (en) | Monitoring device and monitoring method for complex environment behavior recognition | |
| CN104574729A (en) | Alarming method, device and system | |
| CN113095160B (en) | Power system personnel safety behavior identification method and system based on artificial intelligence and 5G | |
| CN113128414A (en) | Personnel tracking method and device, computer readable storage medium and electronic equipment | |
| CN113554364A (en) | Disaster emergency management method, device, equipment and computer storage medium | |
| CN110533889B (en) | Sensitive area electronic equipment monitoring and positioning device and method | |
| KR101552564B1 (en) | Fusion security system based on gas sensor and IP network camera | |
| CN106803937B (en) | Double-camera video monitoring method, system and monitoring device with text log | |
| CN115860979A (en) | Artificial intelligence management system for field operation of power grid | |
| CN116597501A (en) | Video analysis algorithm and edge device | |
| CN210895500U (en) | Dangerous behavior action recognition system | |
| CN114419774A (en) | Intelligent epidemic prevention method and system | |
| CN116863399B (en) | Network security monitoring system and method based on artificial intelligence |
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 |