CN116030391A - Intelligent monitoring method for personnel risk of coal discharge port - Google Patents
Intelligent monitoring method for personnel risk of coal discharge port Download PDFInfo
- Publication number
- CN116030391A CN116030391A CN202310030524.4A CN202310030524A CN116030391A CN 116030391 A CN116030391 A CN 116030391A CN 202310030524 A CN202310030524 A CN 202310030524A CN 116030391 A CN116030391 A CN 116030391A
- Authority
- CN
- China
- Prior art keywords
- personnel
- discharge opening
- coal
- behavior
- preset
- 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.)
- Granted
Links
Images
Landscapes
- Emergency Alarm Devices (AREA)
Abstract
The invention discloses an intelligent monitoring method for personnel risk of a coal discharge opening, which is mainly characterized in that an auxiliary lighting warning lamp and a plurality of monitoring cameras are arranged at the coal discharge opening, and mechanisms such as object contour recognition, human body detection, personnel behavior judgment and the like are applied to analyze and process dangerous behaviors of personnel in the area near the discharge opening and close to the discharge opening, and shooting scene change is realized by combining the flickering of the warning lamp, so that the complexity of personnel detection and behavior recognition at the discharge opening in a monitoring video is reduced, early warning and shutdown signals are sent in time, and safety accidents are avoided. The invention avoids the defect that the conventional laser and infrared sensing modes are easily affected by dust and other environments, and by adopting an artificial intelligent image analysis strategy, the monitoring precision and the response speed can be greatly improved, and the situation of false detection and false report is effectively avoided.
Description
Technical Field
The invention relates to the field of energy chemical production, in particular to an intelligent monitoring method for personnel risk of a coal discharge port.
Background
Coal is a main raw material for coking and coal chemical production, and after the coal is transported to a factory production workshop through an automobile, the coal is unloaded at a discharge opening, and a coal crusher is started to crush massive coal. Because the discharge opening of the coal breaker is an open place, personnel can not be shielded from peripheral operation, and personal injury can be directly caused if personnel fall carelessly during operation.
In addition, the coal breaker generally does not have automatic interlocking protection measures, and is usually started and stopped by manual switch control equipment, so that casualties are easily caused when personnel carelessly fall into the coal breaker and cannot automatically stop running, the necessary safety equipment is added to realize equipment control interlocking, but more dust can appear on site during unloading or crushing operation, adverse environmental effects are caused, and conventional sensors such as laser and infrared are easy to misreport, so that equipment is frequently in misoperation, and the production efficiency and the service life of the equipment are seriously influenced. Therefore, the technical scheme of finding a method capable of guaranteeing the risk identification effect under dust interference and easy to upgrade and maintain is a direction for solving the problem.
Disclosure of Invention
In view of the above, the present invention aims to provide an intelligent monitoring method for risk of coal discharge personnel, so as to solve the above-mentioned technical problems.
The technical scheme adopted by the invention is as follows:
the invention provides an intelligent monitoring method for personnel risk of a coal discharge port, which comprises the following steps:
controlling a warning lamp arranged at the discharge opening to flash according to a preset period;
the monitoring camera is controlled to periodically collect video information of a preset area at the discharge port;
extracting object contour features from the video information;
if the coal transportation vehicle is judged according to the object contour features, continuing to execute the next period of video acquisition;
if the object is judged not to belong to the coal-carrying vehicle according to the object contour characteristics, extracting image gray scale characteristics from the current video information and the video information collected later;
and starting human body identification based on the image gray features, and triggering an alarm and/or interlocking with sudden stop and controlling the coal breaker to stop after determining that a human body appears in a preset area of the discharge opening.
In at least one possible implementation manner, the starting human body recognition based on the image gray scale features includes:
storing gray features in a preset area in the current video to a preset queue;
storing gray features in video information acquired in a subsequent video acquisition time period into the preset queue;
calculating the similarity of adjacent gray features in the preset queue to form a similarity array;
when at least two peaks appear in the similarity array, triggering the human body identification strategy to start.
In at least one possible implementation manner, the human body recognition adopts a pre-constructed human body detection model, and the training manner of the human body detection model comprises:
setting a training target as to whether personnel appear in a preset area of the discharge opening;
collecting data samples for training and testing a model by combining the set video shooting scene angle and a preset true and false scene; the method comprises the steps of,
and classifying and labeling all photographed scenes, and dividing the detection model according to different scenes.
In at least one possible implementation manner, the intelligent monitoring method further comprises determining that the human behavior appears in the video image of the human body through a human body behavior recognition model.
In at least one possible implementation manner, the construction manner of the human behavior recognition model comprises:
selecting a feature extraction network for identifying personnel behaviors;
constructing a behavior data set offline, and training a personnel behavior recognition model by using deep learning;
inputting the image to be identified into the human body behavior identification model, and outputting and obtaining the behavior label and the confidence information of the target person;
and carrying out logic reasoning based on the behavior label, the confidence information and preset personnel behavior priori knowledge, judging and finally determining whether the behavior of the personnel entering the preset area at the discharge opening occurs.
In at least one possible implementation manner, the logic reasoning process includes:
when the identification result of the behavior label in each shooting device is that a person appears, confidence judgment is carried out;
based on video data continuously collected by a plurality of shooting devices, respectively generating corresponding confidence sequences;
when all confidence coefficient sequences have confidence coefficient values exceeding a preset threshold value, judging that the behavior of entering a preset area at the discharge opening by personnel exists;
when confidence coefficient values exceeding a set threshold value do not appear in at least one confidence coefficient sequence, consistency judgment of analysis results of different shooting scenes in the video acquired by the shooting equipment is carried out;
if the identification results of different scenes are inconsistent, carrying out human behavior identification again; if the identification results of different scenes are consistent, judging that no behavior of entering a preset area at the discharge opening exists.
In at least one possible implementation manner, the determining whether the vehicle is a coal-carrying vehicle according to the object profile features includes:
and comparing whether the distance between the object contour feature vector and the preset coal-carrying vehicle feature vector reaches a preset distance threshold value, and if so, determining that the object appearing in the video is the coal-carrying vehicle.
In at least one possible implementation manner, the frequency corresponding to the flashing period of the warning light is less than half of the image acquisition frequency of the monitoring camera.
The main design concept of the invention is that an auxiliary illumination warning lamp and a plurality of monitoring cameras are arranged at the coal discharge opening, and the mechanisms of object contour recognition, human body detection, personnel behavior judgment and the like are applied to analyze and process dangerous behaviors of personnel in the area near the discharge opening and close to the discharge opening, and shooting scene change is realized by combining the flickering of the warning lamp, so that the complexity of personnel detection and behavior recognition at the discharge opening in the monitoring video is reduced, early warning and shutdown signals are timely sent, and safety accidents are avoided. The invention avoids the defect that the conventional laser and infrared sensing modes are easily affected by dust and other environments, adopts an artificial intelligent image analysis strategy, expands the acquisition and detection of laser, infrared points and line dimensions to the acquisition and detection of surface dimensions, has higher tolerance to dust interference, can further improve the image recognition effect by utilizing technologies such as video image enhancement and recognition, can ensure the monitoring precision and response speed, and effectively eliminates the false detection and false alarm.
Drawings
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described with reference to the accompanying drawings, in which:
fig. 1 is a schematic flow chart of a method for intelligently monitoring risk of coal discharge personnel according to an embodiment of the invention.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the invention.
The invention provides an embodiment of a coal discharge opening personnel risk intelligent monitoring method, which is specifically shown in fig. 1 and comprises the following steps:
s1, controlling a warning lamp arranged at a discharge port to flash according to a preset period;
specifically, the warning lamp is triggered to be turned on and blinks according to a fixed period T, and the frequency f=1/T corresponding to the blinking period T is less than half of the frequency of the image collected by the monitoring camera.
Step S2, controlling a monitoring camera to periodically acquire video information of a preset area (surrounding dangerous area) at the discharge port;
s3, extracting object contour features from the video information;
it will be appreciated that the captured video information may also be pre-processed and image enhanced, including but not limited to removing video noise, prior to object contour feature extraction.
Step S4, if the coal-carrying vehicle is judged according to the object contour features, continuing to execute the next period of video acquisition; the next cycle here may refer to the aforementioned blinking cycle T of the warning light.
In actual operation, whether the distance between the object contour feature vector and the preset coal-carrying vehicle feature vector reaches a preset threshold value or not may be compared, if so, it is determined that the object appears in the video as the coal-carrying vehicle, and at this time, the process returns to the step S2 to continue to collect the image.
Step S5, if the object is judged not to belong to the coal-carrying vehicle according to the object contour characteristics, extracting image gray features from the current video information and the subsequently acquired video information;
and S6, starting human body recognition based on the image gray features, and triggering an alarm and/or interlocking with sudden stop and controlling the coal breaker to stop after recognizing and determining that personnel appear in a preset area of the discharge opening.
Enabling human body recognition based on the image gray scale features includes:
step S61, storing gray features in a preset area to a preset queue Qg;
step S62, storing gray features in video information acquired in a subsequent video acquisition time period (the gray features can be stored in the preset queue Qg according to the flashing period of the two continuous warning lamps, namely 2T);
step S63, calculating the similarity of adjacent gray features in the preset queue to form a similarity array As;
step S64, triggering the human body identification strategy to start when at least two wave peaks appear in the similarity array; otherwise, the step S2 is executed again to collect and analyze the video continuously.
It should be noted that, there is a difference in video characteristics in lighting scenes where at least two peak-indicating warning lamps are turned on or off, and at least one stably appearing object to be recognized exists in different lighting scenes, so that the possibility of personnel appearing after the vehicle is excluded is maximum.
Preferably, the human body recognition is implemented by using a pre-constructed human body detection model, and a training architecture of the human body detection model may be referred to as follows, including:
setting a training target as to whether personnel appear in a preset area of the discharge opening;
collecting data samples for training and testing a model by combining the set video shooting scene angle and a preset true and false scene; and classifying and labeling all photographed scenes, and dividing the detection model according to different scenes.
Based on this concept, in order to reliably identify the behavior of a person entering a dangerous area near a coal discharge opening, based on the above-mentioned person detection strategy, the present invention proposes that, more preferably, the above-mentioned model may be improved to enable further identification of the person behavior (of course, a model specially used for predicting the person behavior may be additionally built), which may specifically include:
step S01, selecting a characteristic extraction network for identifying personnel behaviors;
in particular, a depth separable convolutional neural network can be used as a feature extraction network of the algorithm model, and the model has fewer parameters and higher performance than the traditional network.
Step S02, an action data set is built offline, and a human action recognition model is trained by using a depth separable convolutional neural network (and as described above, the ability of action prediction can be added on the basis of human body detection, or a special human action recognition model is built);
in practice, the process of constructing the behavioral data set may include: data acquisition, data screening, data marking, data enhancement and the like.
Regarding data collection, the personnel behavior can be recorded under different environments and illumination conditions, namely, different personnel enter a preset area (defined risk area) at the discharge opening, and of course, shooting angle changes and human body posture changes with different degrees are preferably considered during recording.
With respect to data screening, it will be appreciated that personnel entering the predetermined area at the discharge opening is a continuous process and that there may be redundancy between successive video frames, so that employing data screening can re-acquire 1 frame of images every few frames and process them using algorithms, such as processing image files using a human physical examination algorithm first, leaving only images that can be detected as having a human being present as an effective training sample.
With respect to data labeling, labeling the aforementioned valid training samples may be primarily framing target elements in image regions, where the elements in the images may comprise several different levels: the presence of personnel, the presence of personnel access, and the characteristics of the predetermined area at the discharge opening.
For data enhancement, in order to improve the robustness and generalization capability of the model, a data enhancement method is used to further increase the number and diversity of samples involved in training, for example, the data enhancement can be achieved by performing operations such as image scaling, horizontal mirror image flipping, and random brightness and hue adjustment on the noted effective samples in the training stage.
The offline structure and the data processing method can refer to the mature related technology, and the description of the present invention is omitted.
Step S03, inputting the image to be identified into the human body behavior identification model, and outputting and obtaining the label and the confidence information (the position information can be included) of the target personnel behavior; the behavior recognition model also adopts a deep separable convolutional neural network, and in the specific implementation, a MobileNet model is adopted, and the model can still achieve the accuracy similar to the traditional network and greatly reduce the operation times under the condition that the parameter quantity is only one tenth or even less than that of the traditional network, and meanwhile, the model has the characteristics of light weight, has low requirements on hardware equipment configuration, and has lower maintenance and upgrading costs.
And step S04, carrying out logic reasoning based on the behavior label, the confidence information and preset personnel behavior priori knowledge, judging and finally determining whether the behavior of the personnel entering the preset area at the discharge port occurs.
In some preferred embodiments, for the logical reasoning, reference may be made specifically to the following manner:
step S040, when the identification result of the behavior label in each shooting device is that a person appears, confidence judgment is carried out;
step S041, respectively generating corresponding confidence sequences based on video data continuously collected by a plurality of shooting devices;
taking two monitoring cameras for personnel entering behavior recognition as an example, two confidence sequences can be correspondingly obtained:
P I {P I1 ，P I2 ，P I3 …P IN and P r {P r1 ，P r2 ，P r3 …P rN }。
Step S042, judging that the behavior of a person entering a preset area at the discharge opening exists when all confidence coefficient sequences have confidence coefficient values exceeding a preset threshold value;
for example, when P I P r When confidence values larger than the confidence threshold value appear in the area, the behavior that people enter the area is judged, and the confidence threshold value can be set according to experience.
Step S043, when confidence coefficient values exceeding a set threshold value do not appear in at least one confidence coefficient sequence, consistency judgment of analysis results of different shooting scenes in the video acquired by the shooting equipment is carried out;
step S044, if the identification results of different scenes are inconsistent, human body behavior identification is carried out again; if the identification results of different scenes are consistent, judging that no behavior of entering a preset area at the discharge opening exists.
For example, when P I P r When no confidence value larger than the confidence threshold value appears in the two, the two are compared to judge the scene consistency, specifically, P can be respectively calculated I P r And solving the variance, if the difference value between the variance and the variance is larger than a preset difference value threshold value, indicating that the scene is inconsistent, and re-acquiring videos of the flicker periods of two continuous warning lamps for subsequent judgment. Otherwise, the scene is consistent, and the situation that the confidence threshold exceeds the limit does not occur, the behavior that no person enters the dangerous area is characterized, and the behavior identification of the round is ended.
The foregoing embodiments may be implemented in the practical application stage with reference to the following schemes: the warning lamp can be arranged at the ceiling of the discharge opening workshop close to the gate, irradiates outwards, is used for giving out danger warning to external personnel and provides auxiliary illumination near the gate of the workshop. The monitoring camera is arranged on the inner wall opposite to the gate of the discharge opening factory building, collects videos of the inside of the discharge opening factory building and the gate, and transmits video information to a server which is pre-configured with an AI identification algorithm. The server acquires the video stream of the discharge port camera, performs real-time AI analysis processing, judges whether personnel are moving in the monitoring video picture, sends out an alarm signal according to the judgment result, and starts an alarm and/or makes the crusher suddenly stop. Namely, an auxiliary lighting warning lamp and a plurality of monitoring cameras are arranged at the discharge opening, and an artificial intelligent recognition technology (human body detection and personnel behavior recognition algorithm) is applied to analyze and judge dangerous behaviors of personnel in the area near the discharge opening and close to the discharge opening. The flashing of the warning lamp is utilized, the scene of the discharge opening can be changed, the complexity of the recognition of the discharge opening personnel in the monitoring video is reduced through scene change, and therefore rapid and real-time judgment of the discharge opening personnel is carried out, and early warning and/or shutdown signals are timely sent out.
Specifically, two warning lamps are installed at the ceiling of the discharge opening factory building near the exit, the warning lamps emit red light, flash at fixed frequency, the irradiation direction of the warning lamps is from the inside of the discharge opening factory building to the outside, and meanwhile, the irradiation direction of the warning lamps forms 45 degrees to the ground, so that the light rays of the warning lamps can be observed by people near the outside of the discharge opening factory building, meanwhile, the effect of providing illumination near the gate of the discharge opening factory building can be achieved, and the brightness of the warning lamps can be used for enabling a camera to shoot objects within the range of 3 meters outside the gate of the discharge opening when larger dust is produced by working at night and in a crusher.
Two cameras are installed on the inner wall of the discharge opening factory building, the lens faces to the outside from the inside of the discharge opening factory building, no dead angle is guaranteed in discharge opening monitoring, and the gate of the discharge opening factory building is in the video image range of the cameras. The video signals of the cameras are converged to a field weak current box switch and are transmitted back to an inner core switch of a relevant machine room through a gigabit network.
Furthermore, an audible and visual alarm is arranged at the striking position of the discharge opening, and can be connected into a field weak current control box through a 485 bus and then converted into a network signal through a TCP serial server module to be connected into an information network. The switching value output terminal of the TCP serial port server module can be connected with the scram switch input terminal of the coal breaker control box in a wiring way, so that the switching value output signal can control the scram of the coal breaker.
The server can be deployed in a factory floor machine room, acquires a camera RTSP video stream to perform AI video analysis, and can send out an scram control signal and/or an on-site alarm prompt signal to the coal pulverizer according to a video analysis result.
According to the scheme in the different aspects, through practical application verification, compared with a pure end-to-end personnel behavior detection method based on deep learning, the method has the advantages of high identification precision and high detection speed, and can effectively solve the problem of false detection, so that the detection and identification precision is improved. Because the method is convenient for realizing the on-site data acquisition and timely training of the deep network model, the method can be popularized to detect other specific personnel behaviors, is convenient for deployment to different occasions and meets specific application requirements.
In summary, the main design concept of the invention is that the warning lamp and the monitoring cameras for auxiliary illumination are arranged at the coal discharge opening, the mechanisms of object contour recognition, human body detection, personnel behavior judgment and the like are applied to analyze and process dangerous behaviors of personnel in the area near the discharge opening and close to the discharge opening, and shooting scene change is realized by combining with the flickering of the warning lamp, so that the complexity of personnel detection and behavior recognition at the discharge opening in the monitoring video is reduced, early warning and shutdown signals are timely sent, and safety accidents are avoided. The invention avoids the defect that the conventional laser and infrared sensing modes are easily affected by dust and other environments, and by adopting an artificial intelligent image analysis strategy, the monitoring precision and the response speed can be greatly improved, and the situation of false detection and false report is effectively avoided.
In the embodiments of the present invention, "at least one" means one or more, and "a plurality" means two or more. "and/or", describes an association relation of association objects, and indicates that there may be three kinds of relations, for example, a and/or B, and may indicate that a alone exists, a and B together, and B alone exists. Wherein A, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of the following" and the like means any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c may represent: a, b, c, a and b, a and c, b and c or a and b and c, wherein a, b and c can be single or multiple.
The construction, features and effects of the present invention are described in detail according to the embodiments shown in the drawings, but the above is only a preferred embodiment of the present invention, and it should be understood that the technical features of the above embodiment and the preferred mode thereof can be reasonably combined and matched into various equivalent schemes by those skilled in the art without departing from or changing the design concept and technical effects of the present invention; therefore, the invention is not limited to the embodiments shown in the drawings, but is intended to be within the scope of the invention as long as changes made in the concept of the invention or modifications to the equivalent embodiments do not depart from the spirit of the invention as covered by the specification and drawings.
Claims (8)
1. The intelligent monitoring method for the personnel risk of the coal discharge opening is characterized by comprising the following steps of:
controlling a warning lamp arranged at the discharge opening to flash according to a preset period;
the monitoring camera is controlled to periodically collect video information of a preset area at the discharge port;
extracting object contour features from the video information;
if the coal transportation vehicle is judged according to the object contour features, continuing to execute the next period of video acquisition;
if the object is judged not to belong to the coal-carrying vehicle according to the object contour characteristics, extracting image gray scale characteristics from the current video information and the video information collected later;
and starting human body identification based on the image gray features, and triggering an alarm and/or interlocking with sudden stop and controlling the coal breaker to stop after determining that a human body appears in a preset area of the discharge opening.
2. The intelligent monitoring method for risk of coal discharge personnel according to claim 1, wherein the starting human body identification based on the image gray scale features comprises:
storing gray features in a preset area in the current video to a preset queue;
storing gray features in video information acquired in a subsequent video acquisition time period into the preset queue;
calculating the similarity of adjacent gray features in the preset queue to form a similarity array;
when at least two peaks appear in the similarity array, triggering the human body identification strategy to start.
3. The intelligent monitoring method for risk of coal discharge opening personnel according to claim 1, wherein the human body identification adopts a pre-constructed human body detection model, and the training mode of the human body detection model comprises the following steps:
setting a training target as to whether personnel appear in a preset area of the discharge opening;
collecting data samples for training and testing a model by combining the set video shooting scene angle and a preset true and false scene; the method comprises the steps of,
and classifying and labeling all photographed scenes, and dividing the detection model according to different scenes.
4. The intelligent monitoring method for coal discharge opening personnel risk according to claim 1, further comprising determining the personnel behavior in the video image of the human body through a human body behavior recognition model.
5. The intelligent monitoring method for risk of coal discharge opening personnel according to claim 4, wherein the construction mode of the human behavior recognition model comprises the following steps:
selecting a feature extraction network for identifying personnel behaviors;
constructing a behavior data set offline, and training a personnel behavior recognition model by using deep learning;
inputting the image to be identified into the human body behavior identification model, and outputting and obtaining the behavior label and the confidence information of the target person;
and carrying out logic reasoning based on the behavior label, the confidence information and preset personnel behavior priori knowledge, judging and finally determining whether the behavior of the personnel entering the preset area at the discharge opening occurs.
6. The intelligent monitoring method for risk of coal discharge personnel according to claim 5, wherein the process of logical reasoning comprises:
when the identification result of the behavior label in each shooting device is that a person appears, confidence judgment is carried out;
based on video data continuously collected by a plurality of shooting devices, respectively generating corresponding confidence sequences;
when all confidence coefficient sequences have confidence coefficient values exceeding a preset threshold value, judging that the behavior of entering a preset area at the discharge opening by personnel exists;
when confidence coefficient values exceeding a set threshold value do not appear in at least one confidence coefficient sequence, consistency judgment of analysis results of different shooting scenes in the video acquired by the shooting equipment is carried out;
if the identification results of different scenes are inconsistent, carrying out human behavior identification again; if the identification results of different scenes are consistent, judging that no behavior of entering a preset area at the discharge opening exists.
7. The intelligent monitoring method for risk of coal discharge opening personnel according to any one of claims 1 to 6, wherein the judging of whether the coal vehicle is a coal-carrying vehicle or not according to the object profile features comprises:
and comparing whether the distance between the object contour feature vector and the preset coal-carrying vehicle feature vector reaches a preset distance threshold value, and if so, determining that the object appearing in the video is the coal-carrying vehicle.
8. The intelligent monitoring method for personnel risk of coal discharge openings according to any one of claims 1 to 6, wherein the frequency corresponding to the flashing period of the warning lamp is less than half of the image acquisition frequency of the monitoring camera.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310030524.4A CN116030391B (en) | 2023-01-06 | 2023-01-06 | Intelligent monitoring method for personnel risk of coal discharge port |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310030524.4A CN116030391B (en) | 2023-01-06 | 2023-01-06 | Intelligent monitoring method for personnel risk of coal discharge port |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116030391A true CN116030391A (en) | 2023-04-28 |
| CN116030391B CN116030391B (en) | 2023-07-21 |
Family
ID=86091085
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310030524.4A Active CN116030391B (en) | 2023-01-06 | 2023-01-06 | Intelligent monitoring method for personnel risk of coal discharge port |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116030391B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109376673A (en) * | 2018-10-31 | 2019-02-22 | 南京工业大学 | A kind of coal mine down-hole personnel unsafe acts recognition methods based on human body attitude estimation |
| CN111126193A (en) * | 2019-12-10 | 2020-05-08 | 枣庄矿业（集团）有限责任公司蒋庄煤矿 | Artificial intelligence recognition system based on deep learning coal mine underground unsafe behavior |
| CN111814601A (en) * | 2020-06-23 | 2020-10-23 | 国网上海市电力公司 | Video analysis method combining target detection and human body posture estimation |
| CN214175107U (en) * | 2020-09-17 | 2021-09-10 | 华洋通信科技股份有限公司 | AI intelligent video identification control system for coal mine |
| CN113850144A (en) * | 2021-08-30 | 2021-12-28 | 湖南黑鲸数据科技有限公司 | Video reliability automatic inspection system based on image recognition |
| CN114202886A (en) * | 2021-12-09 | 2022-03-18 | 安徽海螺信息技术工程有限责任公司 | Mine blasting safety monitoring and early warning system |
| CN114565886A (en) * | 2021-12-16 | 2022-05-31 | 高新兴创联科技有限公司 | Method for detecting dangerous behavior of personnel crossing station tracks in railway industry |
-
2023
- 2023-01-06 CN CN202310030524.4A patent/CN116030391B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109376673A (en) * | 2018-10-31 | 2019-02-22 | 南京工业大学 | A kind of coal mine down-hole personnel unsafe acts recognition methods based on human body attitude estimation |
| CN111126193A (en) * | 2019-12-10 | 2020-05-08 | 枣庄矿业（集团）有限责任公司蒋庄煤矿 | Artificial intelligence recognition system based on deep learning coal mine underground unsafe behavior |
| CN111814601A (en) * | 2020-06-23 | 2020-10-23 | 国网上海市电力公司 | Video analysis method combining target detection and human body posture estimation |
| CN214175107U (en) * | 2020-09-17 | 2021-09-10 | 华洋通信科技股份有限公司 | AI intelligent video identification control system for coal mine |
| CN113850144A (en) * | 2021-08-30 | 2021-12-28 | 湖南黑鲸数据科技有限公司 | Video reliability automatic inspection system based on image recognition |
| CN114202886A (en) * | 2021-12-09 | 2022-03-18 | 安徽海螺信息技术工程有限责任公司 | Mine blasting safety monitoring and early warning system |
| CN114565886A (en) * | 2021-12-16 | 2022-05-31 | 高新兴创联科技有限公司 | Method for detecting dangerous behavior of personnel crossing station tracks in railway industry |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116030391B (en) | 2023-07-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103839373B (en) | A kind of unexpected abnormality event Intelligent Recognition alarm device and warning system | |
| JP6867153B2 (en) | Abnormality monitoring system | |
| CN105809679B (en) | Mountain railway side slope rockfall detection method based on visual analysis | |
| CN109637068A (en) | Intelligent pyrotechnics identifying system | |
| CN104599427B (en) | A kind of intelligent image type fire alarm system for vcehicular tunnel | |
| KR102195706B1 (en) | Method and Apparatus for Detecting Intruder | |
| CN109544859A (en) | A kind of active safety household camera intelligent realization method | |
| CN111523397B (en) | Intelligent lamp post visual identification device, method and system and electronic equipment thereof | |
| CN103839346A (en) | Intelligent door and window anti-intrusion device and system and intelligent door control system | |
| CN102013147A (en) | Intelligent burglary prevention monitoring method and device for high-voltage power transmission tower | |
| CN105976365A (en) | Nocturnal fire disaster video detection method | |
| CN108259846B (en) | Industrial patrol robot system | |
| Toreyin et al. | Wildfire detection using LMS based active learning | |
| CN110867046A (en) | Intelligent car washer video monitoring and early warning system based on cloud computing | |
| EP2000998B1 (en) | Flame detecting method and device | |
| CN107688793A (en) | A kind of outside transformer substation fire automatic monitoring method for early warning | |
| CN104102921A (en) | Video detection method and video detection device based on face recognition | |
| JP2003219399A (en) | Supervisory apparatus for identifying supervisory object | |
| CN112149576A (en) | Elevator safety real-time monitoring management system based on image analysis | |
| WO2013093771A1 (en) | Monitoring a scene | |
| CN116030391B (en) | Intelligent monitoring method for personnel risk of coal discharge port | |
| CN101106727A (en) | Method for fire detection based on flame color template | |
| CN111908288A (en) | TensorFlow-based elevator safety system and method | |
| KR102204015B1 (en) | Lifting gate based on image and vehicle control method through the same | |
| KR101300130B1 (en) | System and method for detecting smoke using surveillance camera |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |