CN110647819A - Method and device for detecting abnormal behavior of underground personnel crossing belt - Google Patents
Method and device for detecting abnormal behavior of underground personnel crossing belt Download PDFInfo
- Publication number
- CN110647819A CN110647819A CN201910802003.XA CN201910802003A CN110647819A CN 110647819 A CN110647819 A CN 110647819A CN 201910802003 A CN201910802003 A CN 201910802003A CN 110647819 A CN110647819 A CN 110647819A
- Authority
- CN
- China
- Prior art keywords
- belt
- lambda
- theta
- image
- human body
- 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
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
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
-
- 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)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Emergency Management (AREA)
- Image Analysis (AREA)
- Alarm Systems (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method and a device for detecting abnormal behavior of underground personnel crossing a belt, wherein a high-definition camera head is arranged in an area near the belt to monitor the behavior of the personnel near the belt in real time; the method comprises the following steps: preprocessing a monitoring video image at the belt, including low-light enhancement and noise reduction processing of the image; the preprocessed image is subjected to a bottom-up posture estimation method to obtain human body joint point information, and posture estimation can be more accurately carried out under the condition that a small part of the human body is shielded; calculating the relative distance lambda between the ankle joint point and the knee joint point between different legsaAnd the same leg-leg bending angle thetaaWhen lambda isaNot less than threshold value lambda and thetaaAnd when the value is less than or equal to the threshold value theta, determining that the abnormal behavior of the person crossing the belt exists in the image, and sending out an alarm signal. The method can detect the action of personnel crossing the belt under the condition of a plurality of people in the underground with weak light, and has the advantages of high real-time performance, small error and high running speed.
Description
Technical Field
The invention relates to an abnormal behavior detection method and device, in particular to an abnormal behavior detection method and device for underground personnel crossing a belt, and belongs to the technical field of underground safety.
Background
The belt transmission is used as the most main transportation mode of a mine and is widely applied to coal mines. In the process of coal transportation, the running speed of the belt is high, and the randomness of the coal block occurrence time and the size of the coal block on the belt is high. The underground personnel cross the belt, clothes are coiled into the belt or fall down on the conveying surface of the belt, so that serious casualties or equipment damage can be caused, and safety accidents can occur.
The current abnormal behavior detection aiming at underground personnel is influenced by the underground complex environment, the anti-interference capability of sensing equipment is poor, and the sensitivity and the accuracy are not ideal enough. The abnormal behavior detection based on the video is more stable and efficient compared with a sensor method, and the investment cost is lower. However, under the constraint of lighting, shading and other conditions, the method for monitoring abnormal behaviors by video needs to be improved. In the prior human body posture estimation work, a top-down method is adopted, the limb joints of the whole human body are concerned, each detected person in an image is subjected to independent posture estimation, and when problems such as shielding and the like are encountered in some scenes, the posture of some persons cannot be estimated due to the fact that some persons cannot be detected frequently. The most important thing for the alarm device is real-time, and the traditional image detection method is difficult to estimate the action information of a plurality of people in a short time under the environment of a plurality of people, so that the application in the actual scene is not strong.
Summarizing the above process, the main drawbacks of the current similar processes are:
1. when the illumination is weak, the human body action cannot be accurately identified, and further the human body action cannot be detected;
2. the posture of the human body cannot be accurately estimated under the condition that a small part of the human body is shielded;
3. the motion detection accuracy rate is low and the real-time performance is poor under the condition that a plurality of people exist in the image;
4. the detection method has low accuracy and is easy to generate false alarm.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides the method and the device for detecting the abnormal behavior of the underground personnel crossing the belt, which can detect the personnel crossing the belt near the belt in real time and send out an alarm in time, are not influenced by light, and have the advantages of high accuracy, high real-time property and low maintenance cost.
The invention discloses a method for detecting abnormal behavior of underground personnel crossing a belt, which comprises the following steps:
1) carrying out real-time video monitoring near the belt, and extracting video frame images;
2) preprocessing the image extracted in the step (1), wherein the preprocessing comprises low light enhancement and noise reduction processing on the image;
3) estimating the posture of the personnel in the image of the preprocessed image to acquire the joint point information of the personnel, thereby acquiring the posture information of the human body in real time;
4) extracting information of ankle joint points and knee joint points of the persons in the image according to the human body posture information acquired in real time in the step (3), and calculating the relative distance lambda between the ankle joint points and the knee joint points of different legs of the same personaAnd the same asBending included angle theta of leg part of one legaI.e. the included angle between the thigh and the shank of the same leg;
5) setting the relative distance threshold between ankle joint points and knee joint points of different legs of a person as lambda and the leg bending included angle threshold of the same leg as theta, and calculating the distance lambda calculated in the step (4)aAnd angle thetaaComparing the distance with a distance threshold lambda and a threshold theta in real time, and determining when lambda isaIs not less than lambda and thetaaAnd when the angle is less than or equal to theta, judging that the person in the image has abnormal behavior crossing the belt, and sending out an alarm signal.
An abnormal behavior detection device for underground personnel crossing over a belt comprises
The acquisition module is used for acquiring image information near the underground belt;
the preprocessing calculation module is used for performing low-light enhancement and noise reduction processing on the image transmitted by the acquisition module;
the human body posture estimation module is used for receiving the preprocessed images transmitted by the preprocessing calculation module, generating heat maps and confidence maps for the preprocessed images to obtain joint point information, performing joint point matching by using an affinity field method, and combining limbs into a human body posture by using a minimum dichotomy method;
a calculation module for extracting ankle joint points and knee joint points according to the human body posture information transmitted by the human body posture estimation module, and calculating the distance lambda between the ankle joint points and the knee joint points between different legs of the same personaAnd the same leg-leg bending angle thetaa；
The determining module is internally stored with a relative distance threshold lambda of ankle joint points and knee joint points between different legs based on the same human body and a bending included angle threshold theta of the same leg and leg, and is used for judging the distance lambda transmitted by the calculating moduleaWhether the angle is larger than the threshold lambda and the included angle thetaaIf it is less than threshold value theta, if lambdaaIs not less than lambda and thetaaIf the speed is less than or equal to theta, judging that abnormal behaviors of the personnel crossing the belt exist in the collected underground image, and transmitting a signal to an alarm module;
and the alarm module is used for sending out an alarm signal according to the signal transmitted by the determination module.
Compared with the prior art, the invention has the following advantages:
1) the invention performs low light enhancement and noise reduction preprocessing on the image with weak illumination, can still perform action recognition on the preprocessed image under the dim condition, and has high recognition accuracy.
2) The method of the affinity field is used, real-time multi-person posture estimation is realized under the condition that multiple persons exist in the image, real-time detection whether the multiple persons have abnormal behaviors crossing the belt or not is realized, and the real-time performance is high.
3) The invention uses a bottom-up posture estimation method, namely, the human body joint point information is detected firstly and then the whole human body posture is combined through the joint point information, and the posture of the human body can be accurately estimated under the condition that the human body is shielded by a small part.
4) The detection method is simple, high in accuracy and not easy to generate false alarm.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a schematic view of the apparatus of the present invention;
wherein, 1, high definition digtal camera, 2 audible-visual annunciator, 3 the belt.
Detailed Description
The invention will be further explained with reference to the drawings.
As shown in fig. 1, the invention relates to a method for detecting abnormal behavior of downhole personnel crossing a belt,
1) carrying out real-time video monitoring near the belt 3, and extracting video frame images;
2) preprocessing the image extracted in the step (1), wherein the preprocessing comprises low light enhancement and noise reduction processing on the image;
3) carrying out posture estimation on the preprocessed image, and acquiring the joint point information of the person so as to acquire the posture information of the human body;
4) extracting the information of ankle joint points and knee joint points of the person in the image according to the human body posture information acquired in real time in the step (3), and doing so when the information of the joint points is extractedCoordinate normalization processing, namely adapting to joint point information of different persons at different positions far from and near the camera during calculation, and calculating the relative distance lambda between the joint point of the left (right) ankle and the joint point of the right (left) knee of the same personaAnd the angle theta between the left (right) thigh and the left (right) calfa；
In the step, coordinate normalization processing is performed when the joint point information is extracted, so that the technical problem that people with different heights use the same threshold value for comparison is effectively solved, the accuracy of detecting whether people cross the belt or not by the system is further improved, and the normalization processing is a ubiquitous method in image identification, so that the method is not repeated.
5) Setting a relative distance threshold value of a left (right) ankle joint point and a right (left) knee joint point of a person as lambda and an included angle threshold value between a left (right) thigh and a left (right) shank as theta, setting the threshold values theta and lambda by the coordinate normalization processing in the step (4) without being influenced by the distance from the camera to the camera, and setting the distance lambda calculated in the step (4)aAnd angle thetaaComparing with a threshold lambda and a threshold theta when lambda is detectedaIs not less than lambda and thetaaAnd when the angle is less than or equal to theta, judging that the person in the image has abnormal behavior crossing the belt 3, and sending out an alarm signal.
As shown in FIG. 2, an apparatus for applying the above method comprises
The acquisition module is erected near the belt 3 by adopting a high-definition camera 1 and is used for acquiring image information near the underground belt 3;
the preprocessing calculation module is used for performing low-light enhancement and noise reduction processing on the image transmitted by the acquisition module;
the human body posture estimation module is used for receiving the preprocessed images transmitted by the preprocessing calculation module, generating heat maps and confidence maps for the preprocessed images to obtain joint point information, performing joint point matching by using an affinity field method, and combining limbs into a human body posture by using a minimum dichotomy method;
a calculation module for calculating the distance between the joint point of the left (right) ankle and the joint point of the right (left) knee of the same person according to the human body posture information transmitted by the human body posture estimation moduleλaAnd the angle theta between the left (right) thigh and the left (right) calfa；
The determining module is internally stored with a relative distance threshold lambda of ankle joint points and knee joint points between different legs based on the same human body and a bending included angle threshold theta of the same leg and leg, and is used for judging the distance lambda transmitted by the calculating moduleaWhether the angle is larger than the threshold lambda and the included angle thetaaIf it is less than threshold value theta, if lambdaaIs not less than lambda and thetaaIf the speed is less than or equal to theta, judging that the abnormal behavior of the personnel crossing the belt 3 exists in the collected underground image, and transmitting a signal to an alarm module;
the alarm module adopts an audible and visual alarm device, consists of a buzzer and a warning lamp, wherein the buzzer is electrically connected with the warning lamp and is used for sending alarm sound according to the signal transmitted by the determination module and lighting the warning lamp at the same time.
The specific detection method comprises the following steps:
1) the high-definition camera 1 carries out real-time video monitoring on the position near the belt 3, extracts video frame images and transmits the video frame images to the preprocessing calculation module;
2) the preprocessing calculation module performs low light enhancement and processing for removing noise data in the image extracted in the step (1), and the processing method is as follows:
2.1) estimating an illumination component T in each frame image of a monitoring video by using a stacked automatic encoder, calculating a reflectivity component beta through the obtained illumination component T, wherein the reflectivity component beta is obtained through the following formula:
wherein g is the image of the original monitoring video with dark illumination acquired in the step (1);
meanwhile, the stacking automatic encoder converts the input image into one-dimensional information and then carries out denoising treatment, but the original two-dimensional information of the image is lost;
2.2) using a convolution automatic encoder to make up information of converting a low-light image into a one-dimensional structure loss when a stacking automatic encoder removes noise, the specific method is as follows: the convolution automatic encoder trains two-dimensional information of an image, reduces noise amplification while keeping edge information, reconstructs a reflection component beta after convolution, keeps the two-dimensional information of an original image while realizing enhancement of image contrast through the reconstructed reflection component beta, and eliminates noise data amplified in brightness enhancement.
3) The human body posture estimation module carries out posture estimation on the personnel A in the image on the preprocessed image, and the method comprises the following steps:
3.1) generating a confidence map and a heat map of the preprocessed image, wherein pixel values in the confidence map represent the confidence degrees of the human body joint points, calculating the pixel values in each frame of image to obtain the human body joint point information in the confidence map, the actual human body joint point position is the peak value of the heat map joint points, and the pixel peak value S of the joint points is calculated by a non-maximum inhibition method:
where P denotes the position in the figure, δ controls the peak spread, and X denotes the actual position of the joint point.
3.2) after finding all the peaks of each heat map of the joint points, selecting the maximum peak S of all the peaksjMaximum peak value S as actual joint point informationjObtained by the following formula:
Sj＝maxSj，k
where j denotes a limb j and k denotes a person k in the image.
3.3) matching the joint points to form a limb by using an Affinity field (Affinity Fields), wherein the Affinity field belongs to a 2D vector code between joint points of a certain limb and points to the other joint point from one joint point of the limb, and the specific method is as follows:
is provided with a joint point j1And a joint point j2Form a limb C, with a point p between two joint points, then j1And j2Between the affinity field LcComprises the following steps:
calculating the degree of association between two joint points by calculating the line integral of the affinity field, the joint point with the highest degree of association matching the component limb:
wherein d isj1Denotes j1Position of (d)j2Denotes j2The position of (a).
When analyzing the posture of a human body in implementation, the limbs of the human body are taken as a skeleton spanning tree, all joint points are taken as the nodes of the tree, and all the limbs are taken as the edges of the tree; selecting a minimum number of edges to form a spanning tree of the human body posture, solving the limb matching problem by using a minimum dichotomy, and finally assembling limbs sharing the same joint point into a whole body skeleton of the human body so as to obtain the human body posture information in real time;
4) the calculation module calculates the distance lambda between the left ankle joint point and the right knee joint point of the person A in the image according to the human posture information transmitted by the human posture estimation moduleAAnd the angle theta between the left thigh and the left calfA；
5) The determining module is used for determining the distance lambda calculated in the step (4)AAnd angle thetaAComparing with threshold lambda and threshold theta stored in the interior in real time, and when lambda is detectedAIs not less than lambda and thetaAWhen theta is less than or equal to theta, the fact that the person A has abnormal behaviors crossing the belt 3 in the image is judged, an alarm signal is sent to the audible and visual alarm 2, the audible and visual alarm 2 sends out alarm sound after receiving the signal, and meanwhile, the warning lamp is lightened.
If the image has multi-person information, respectively calculating the relative distance lambda between the ankle joint point and the knee joint point between different legs of each person in the imageaAnd the angle theta between the thigh and the calf of the same legaAnd comparing the image with a threshold lambda and a threshold theta respectively, and comparing the image with the threshold lambda and the threshold theta respectivelyaIs not less than lambda and thetaaTheta or less, judging that the abnormal behavior of the personnel crossing the belt 3 exists in the collected underground imageAnd transmits the signal to an alarm module to give an alarm, and notifies monitoring personnel in time, so that the monitoring personnel can stop the personnel crossing the belt 3.
Claims (5)
1. A method for detecting abnormal behavior of underground personnel crossing a belt is characterized by comprising the following steps:
1) carrying out real-time video monitoring near the belt, and extracting video frame images;
2) preprocessing the image extracted in the step (1), wherein the preprocessing comprises low light enhancement and noise reduction processing on the image;
3) estimating the posture of the personnel in the image of the preprocessed image to acquire the joint point information of the personnel, thereby acquiring the posture information of the human body in real time;
4) extracting information of ankle joint points and knee joint points of the persons in the image according to the human body posture information acquired in real time in the step (3), and calculating the relative distance lambda between the ankle joint points and the knee joint points of different legs of the same personaAnd the same leg-leg bending angle thetaa；
5) Setting the relative distance threshold between ankle joint points and knee joint points of different legs of a person as lambda and the leg bending included angle threshold of the same leg as theta, and calculating the distance lambda calculated in the step (4)aAnd angle thetaaComparing with threshold lambda and threshold theta in real time when lambda is detectedaIs not less than lambda and thetaaAnd when the angle is less than or equal to theta, judging that the person in the image has abnormal behavior crossing the belt, and sending out an alarm signal.
2. The method for detecting abnormal behavior of underground personnel across a belt according to claim 1, wherein in the step (4), coordinate normalization processing is performed when joint point information is extracted.
3. A detection device for abnormal behavior of underground personnel crossing a belt is characterized by comprising
The acquisition module is used for acquiring image information near the underground belt;
the preprocessing calculation module is used for performing low-light enhancement and noise reduction processing on the image transmitted by the acquisition module;
the human body posture estimation module is used for receiving the preprocessed images transmitted by the preprocessing calculation module, generating heat maps and confidence maps for the preprocessed images to obtain joint point information, performing joint point matching by using an affinity field method, and combining limbs into a human body posture by using a minimum dichotomy method;
a calculation module for extracting ankle joint points and knee joint points according to the human body posture information transmitted by the human body posture estimation module, and calculating the relative distance lambda between the ankle joint points and the knee joint points between different legs of the same personaAnd the same leg-leg bending angle thetaa(ii) a You are
The determining module is internally stored with a relative distance threshold lambda of ankle joint points and knee joint points between different legs based on the same human body and a bending included angle threshold theta of the same leg and leg, and is used for judging the distance lambda transmitted by the calculating moduleaWhether the angle is larger than the threshold lambda and the included angle thetaaIf it is less than threshold value theta, if lambdaaIs not less than lambda and thetaaIf the speed is less than or equal to theta, judging that abnormal behaviors of the personnel crossing the belt exist in the collected underground image, and transmitting a signal to an alarm module;
and the alarm module is used for sending out an alarm signal according to the signal transmitted by the determination module.
4. The abnormal behavior detection device for the crossing of the belt by the underground personnel as claimed in claim 3, characterized in that the acquisition module is a high-definition camera (1).
5. The device for detecting the abnormal behavior of the underground personnel crossing the belt according to the claim 3 or 4, characterized in that the alarm module is an audible and visual alarm (2) which comprises a buzzer and a warning lamp, and the buzzer is electrically connected with the warning lamp.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910802003.XA CN110647819B (en) | 2019-08-28 | 2019-08-28 | Method and device for detecting abnormal behavior of underground personnel crossing belt |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910802003.XA CN110647819B (en) | 2019-08-28 | 2019-08-28 | Method and device for detecting abnormal behavior of underground personnel crossing belt |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110647819A true CN110647819A (en) | 2020-01-03 |
| CN110647819B CN110647819B (en) | 2022-02-01 |
Family
ID=69009938
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910802003.XA Active CN110647819B (en) | 2019-08-28 | 2019-08-28 | Method and device for detecting abnormal behavior of underground personnel crossing belt |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110647819B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111291728A (en) * | 2020-03-18 | 2020-06-16 | 北京华能新锐控制技术有限公司 | Detection system, detection equipment and detection method for illegal crossing of transmission belt behavior |
| CN111476609A (en) * | 2020-04-10 | 2020-07-31 | 广西中烟工业有限责任公司 | Retail data acquisition method, system, device and storage medium |
| CN113155070A (en) * | 2021-03-25 | 2021-07-23 | 深圳英飞拓科技股份有限公司 | Scaffold climbing safety detection method and system |
| WO2021164644A1 (en) * | 2020-02-18 | 2021-08-26 | 深圳市商汤科技有限公司 | Violation event detection method and apparatus, electronic device, and storage medium |
| CN113537166A (en) * | 2021-09-15 | 2021-10-22 | 北京科技大学 | Alarm method, alarm device and storage medium |
| CN113610072A (en) * | 2021-10-11 | 2021-11-05 | 精英数智科技股份有限公司 | Method and system for identifying person crossing belt based on computer vision |
| CN113705274A (en) * | 2020-05-20 | 2021-11-26 | 杭州海康威视数字技术股份有限公司 | Climbing behavior detection method and device, electronic equipment and storage medium |
| CN114120368A (en) * | 2021-11-30 | 2022-03-01 | 青岛海信智慧生活科技股份有限公司 | Target detection method and detection equipment |
| WO2022099824A1 (en) * | 2020-11-16 | 2022-05-19 | 深圳技术大学 | Human risk pose recognition method and system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105033997A (en) * | 2015-09-15 | 2015-11-11 | 北京理工大学 | Visual-sense-based rapid working whole-body planning and control method of humanoid robot |
| WO2016027342A1 (en) * | 2014-08-21 | 2016-02-25 | 三菱電機株式会社 | Handrail abnormality detection system for passenger conveyor |
| CN106491097A (en) * | 2016-12-19 | 2017-03-15 | 尹志军 | Human body is across detection means and system |
| CN107563320A (en) * | 2017-08-24 | 2018-01-09 | 中南大学 | Human body sitting posture bearing method of testing and its system based on spatial positional information |
| CN107886069A (en) * | 2017-11-10 | 2018-04-06 | 东北大学 | A kind of multiple target human body 2D gesture real-time detection systems and detection method |
| CN108805032A (en) * | 2018-05-17 | 2018-11-13 | 郑州大学 | Fall detection method based on depth convolutional network |
| CN108958478A (en) * | 2018-06-14 | 2018-12-07 | 吉林大学 | Action recognition and appraisal procedure are ridden in a kind of operation of Virtual assemble |
| CN109376673A (en) * | 2018-10-31 | 2019-02-22 | 南京工业大学 | Method for identifying unsafe behaviors of underground coal mine personnel based on human body posture estimation |
-
2019
- 2019-08-28 CN CN201910802003.XA patent/CN110647819B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016027342A1 (en) * | 2014-08-21 | 2016-02-25 | 三菱電機株式会社 | Handrail abnormality detection system for passenger conveyor |
| CN105033997A (en) * | 2015-09-15 | 2015-11-11 | 北京理工大学 | Visual-sense-based rapid working whole-body planning and control method of humanoid robot |
| CN106491097A (en) * | 2016-12-19 | 2017-03-15 | 尹志军 | Human body is across detection means and system |
| CN107563320A (en) * | 2017-08-24 | 2018-01-09 | 中南大学 | Human body sitting posture bearing method of testing and its system based on spatial positional information |
| CN107886069A (en) * | 2017-11-10 | 2018-04-06 | 东北大学 | A kind of multiple target human body 2D gesture real-time detection systems and detection method |
| CN108805032A (en) * | 2018-05-17 | 2018-11-13 | 郑州大学 | Fall detection method based on depth convolutional network |
| CN108958478A (en) * | 2018-06-14 | 2018-12-07 | 吉林大学 | Action recognition and appraisal procedure are ridden in a kind of operation of Virtual assemble |
| CN109376673A (en) * | 2018-10-31 | 2019-02-22 | 南京工业大学 | Method for identifying unsafe behaviors of underground coal mine personnel based on human body posture estimation |
Non-Patent Citations (5)
| Title |
|---|
| LI CHENG 等: "EADetection: An efficient and accurate sequential behavior anomaly detection approach over data streams", 《INTERNATIONAL JOURNAL OF DISTRIBUTED SENSOR NETWORKS》 * |
| MING LI 等: "Intelligent Detection System for Mine Belt Tearing Based on Machine Vision", 《2011 CHINESE CONTROL AND DECISION CONFERENCE (CCDC)》 * |
| YONG ZHOU 等: "A User Behavior Anomaly Detection Approach based on Sequence Mining over Data Streams", 《2016 17TH INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED COMPUTING, APPLICATIONS AND TECHNOLOGIES》 * |
| 陈娜 等: "脑卒中患者跨越不同高度障碍物的运动学分析", 《中国康复医学集志》 * |
| 陈岩 等: "有跌倒史老年人跨越障碍步态特征的研究", 《山东体育学院学报》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021164644A1 (en) * | 2020-02-18 | 2021-08-26 | 深圳市商汤科技有限公司 | Violation event detection method and apparatus, electronic device, and storage medium |
| CN111291728A (en) * | 2020-03-18 | 2020-06-16 | 北京华能新锐控制技术有限公司 | Detection system, detection equipment and detection method for illegal crossing of transmission belt behavior |
| CN111476609A (en) * | 2020-04-10 | 2020-07-31 | 广西中烟工业有限责任公司 | Retail data acquisition method, system, device and storage medium |
| CN113705274A (en) * | 2020-05-20 | 2021-11-26 | 杭州海康威视数字技术股份有限公司 | Climbing behavior detection method and device, electronic equipment and storage medium |
| CN113705274B (en) * | 2020-05-20 | 2023-09-05 | 杭州海康威视数字技术股份有限公司 | Climbing behavior detection method and device, electronic equipment and storage medium |
| WO2022099824A1 (en) * | 2020-11-16 | 2022-05-19 | 深圳技术大学 | Human risk pose recognition method and system |
| CN113155070A (en) * | 2021-03-25 | 2021-07-23 | 深圳英飞拓科技股份有限公司 | Scaffold climbing safety detection method and system |
| CN113537166A (en) * | 2021-09-15 | 2021-10-22 | 北京科技大学 | Alarm method, alarm device and storage medium |
| CN113610072A (en) * | 2021-10-11 | 2021-11-05 | 精英数智科技股份有限公司 | Method and system for identifying person crossing belt based on computer vision |
| CN114120368A (en) * | 2021-11-30 | 2022-03-01 | 青岛海信智慧生活科技股份有限公司 | Target detection method and detection equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110647819B (en) | 2022-02-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110647819B (en) | Method and device for detecting abnormal behavior of underground personnel crossing belt | |
| CN104038738B (en) | Intelligent monitoring system and intelligent monitoring method for extracting coordinates of human body joint | |
| US20220114839A1 (en) | Behavior prediction method and apparatus, gait recognition method and apparatus, electronic device, and computer-readable storage medium | |
| CN114067358A (en) | Human body posture recognition method and system based on key point detection technology | |
| CN102389361A (en) | Blindman outdoor support system based on computer vision | |
| Shoaib et al. | View-invariant fall detection for elderly in real home environment | |
| CN104574351B (en) | A kind of method for detecting parking stalls based on Video processing | |
| WO2018161849A1 (en) | Alarm system for falling in water based on image water texture and method therefor | |
| CN109805936B (en) | Human body tumbling detection system based on ground vibration signal | |
| KR101310464B1 (en) | Biometric surveillance system and biometric surveillance method using the same system | |
| CN108491762A (en) | A kind of detection method and device that human body is fallen | |
| CN114155492A (en) | High-altitude operation safety belt hanging rope high-hanging low-hanging use identification method and device and electronic equipment | |
| CN112270807A (en) | Old man early warning system that tumbles | |
| KR20220106580A (en) | Method and System for detecting fall situation by using deep learning model | |
| KR20100085337A (en) | Intelligent security system | |
| CN116612374A (en) | Structured light-based human face living body detection model safety test method and system | |
| CN101930540A (en) | Video-based multi-feature fusion flame detecting device and method | |
| CN113378702B (en) | Multi-feature fusion fatigue monitoring and identifying method for pole climbing operation | |
| CN114758414A (en) | Pedestrian behavior detection method, device, equipment and computer storage medium | |
| KR20080026326A (en) | Apparatus and method for image based-monitoring elderly people with principal component analysis | |
| CN104392201A (en) | Human fall identification method based on omnidirectional visual sense | |
| CN111291728A (en) | Detection system, detection equipment and detection method for illegal crossing of transmission belt behavior | |
| CN107273801B (en) | Method for detecting abnormal points by video multi-target tracking | |
| KR102572782B1 (en) | System and method for identifying the type of user behavior | |
| CN114170685A (en) | RGBD image-based detection method for falling behavior of pedestrian riding escalator |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230517 Address after: No. 30 Kunlun Road, Xuzhou High tech Industrial Development Zone, Xuzhou City, Jiangsu Province, 221000 Patentee after: GLORIA TECHNOLOGY LLC Address before: 221000 No. 1 University Road, Xuzhou City, Jiangsu Province Patentee before: CHINA University OF MINING AND TECHNOLOGY |