CN116524540A - Coal mine underground people counting method based on bipartite graph - Google Patents
Coal mine underground people counting method based on bipartite graph Download PDFInfo
- Publication number
- CN116524540A CN116524540A CN202310489568.3A CN202310489568A CN116524540A CN 116524540 A CN116524540 A CN 116524540A CN 202310489568 A CN202310489568 A CN 202310489568A CN 116524540 A CN116524540 A CN 116524540A
- Authority
- CN
- China
- Prior art keywords
- personnel
- image
- coal mine
- matching
- taking
- 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
- 239000003245 coal Substances 0.000 title claims abstract description 24
- 238000000034 method Methods 0.000 title claims abstract description 14
- 238000001514 detection method Methods 0.000 claims abstract description 16
- 230000003190 augmentative effect Effects 0.000 claims description 6
- 238000010586 diagram Methods 0.000 claims description 4
- 230000003321 amplification Effects 0.000 claims description 3
- 230000007423 decrease Effects 0.000 claims description 3
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000005286 illumination Methods 0.000 description 4
- 239000000428 dust Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000000779 smoke Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention relates to a coal mine underground people counting method based on a bipartite graph, and belongs to the field of artificial intelligence. Taking a coal mine well entrance area or underground key place image as input data, taking yolov5 as a main network for target detection, taking an interval time T to acquire images, forming a coal mine well entrance area personnel bipartite graph, and carrying out T by using a Huo Puke Luofet-Kapp algorithm 1 And T 2 Matching personnel at the moment, and counting the number of people entering and exiting the well; the invention is based on the yolov5 light personnel detection model, and can realize rapid iteration and implementation deployment in field application under the condition of meeting the prediction precision.
Description
Technical Field
The invention belongs to the field of artificial intelligence, and relates to a coal mine underground people counting method based on a bipartite graph.
Background
In order to ensure the safe production of coal mines, the number of people entering the well is one of important indexes. An explosion-proof camera is generally configured at a coal mine auxiliary wellhead, a main wellhead, a wind wellhead or a downhole key place, personnel entering the pit are identified at edge equipment by using a related algorithm, and the number of people entering the pit is counted.
The underground environment of the coal mine is complex, such as underground roadways, mining working surfaces and the like, and the conditions of insufficient illumination, intensive personnel, dust or smoke shielding and the like are more frequently caused, so that a two-part graph matching method is adopted, and the accurate matching of targets in images under a time sequence is an important method for realizing underground crowd statistics.
At present, the method for counting the number of people under the condition of better illumination is described in the literature, but the method is not proposed or clearly indicated to be based on the underground coal mine with complex background environment.
Disclosure of Invention
Therefore, the invention aims to provide the coal mine underground people counting method based on the bipartite graph, which can accurately track coal mine personnel entering under the conditions of insufficient illumination and certain dust fog shielding, and realize the statistics of the coal mine underground people.
In order to achieve the above purpose, the present invention provides the following technical solutions:
a coal mine underground people counting method based on two graphs comprises the following steps:
taking a coal mine well entrance area or underground key place image as input data, taking yolov5 as a main network for target detection, taking an interval time T to acquire images, forming a coal mine well entrance area personnel bipartite graph, and carrying out T by using a Huo Puke Luofet-Kapp algorithm 1 And T 2 Matching personnel at the moment, and counting the number of people entering and exiting the well;
the input image is a picture with personnel at an observation site, the image is sliced, yolov5 is taken as a target detection network, the sliced image is input into a backbone network, and the position information P of the personnel target is generated through convolution, residual error and up-sampling i (x 1 ,x 2 ,y 1 ,y 2 )；x 1 Is the left upper corner abscissa of the mark frame, x 2 Is the horizontal coordinate of the right lower corner of the mark frame, y 1 Is the vertical coordinate of the upper left corner of the mark frame, y 2 Is the ordinate of the lower right corner of the mark frame;
to determine the number of people entering and exiting from the well, whether the person passes through a straight line L (x) =kx+b in the picture is identified, the line L is taken as an area line for entering and exiting detection, k and b are constants, and x is an independent variable; acquisition time T 1 And T 2 The target detection image is used for judging whether personnel in the image move in two moments and cross the regional line L;
graph G 1 (v 1 ,v 2 ,…,v i ) And G 2 (u 1 ,u 2 ,…,u i ) Respectively time T 1 And T 2 Lower staff identification schematic diagram, G 1 Midpoint v i Represents the center point of the position before the movement of the ith person, G 2 Midpoint u i Representing the center point of the position after the movement of the ith personnel; for matching sets (v 1 ,v 2 ,…,v i ) Sum set (u) 1 ,u 2 ,…,u i ) The maximum matching is realized by adopting a Hopklover-Kapu algorithm;
in each iteration, the Hopklovir-Kapp algorithm searches for an augmented path based on a current solution, which is a current match; if an augmented path exists, the current solution is not the optimal solution; after a certain iteration is finished, if any amplification path cannot be found, the current optimal solution, namely the maximum matching and the matching T, is found 1 And T 2 Personnel in the moment image;
the matching mode is assumed person P i For uniform movement or stationary, v i And u i The distance between the position coordinates is taken as a judgment basis, and the minimum distance s is defaulted i For person P i From v i And u i Is a moving distance of (2); graph G 1 (v 1 ,v 2 ,…,v i ) And G 2 (u 1 ,u 2 ,…,u i ) Each store in the database is subjected to iterative matching to find each pair v i And u i Best match between E i (v i ,u i )；
Discovery of T 1 And T 2 After the best matching combination of the moments, the computing personnel P i Is provided with a moving direction and a moving distance; if when P i Not crossing the regional line L, P i Is not put into or put out of the well; when P is detected i The cross-domain area line L increases or decreases the population statistics according to the cross-domain direction.
Optionally, the slicing processing of the image is: the 608×608×3 image is processed as a 304×304×12 feature map.
The invention has the beneficial effects that:
(1) The two-part diagram principle is introduced to construct a personnel position information model at different moments, so that the personnel position information at different moments can be effectively analyzed, and the personnel position relation is excavated.
(2) The Hopklover-Kapu algorithm is adopted to search paths and match the maximum paths in the two-part graph of the person, so that the association speed can be improved, and the real-time requirement of people counting is ensured.
(3) Based on the yolov5 light-weight personnel detection model, quick iteration and implementation deployment can be realized in field application under the condition of meeting prediction precision.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objects and other advantages of the invention may be realized and obtained by means of the instrumentalities and combinations particularly pointed out in the specification.
Drawings
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in the following preferred detail with reference to the accompanying drawings, in which:
FIG. 1 is a flow chart of a coal mine underground population statistics method based on two graphs;
FIG. 2 is a graph of two matching people at different times.
Detailed Description
Other advantages and effects of the present invention will become apparent to those skilled in the art from the following disclosure, which describes the embodiments of the present invention with reference to specific examples. The invention may be practiced or carried out in other embodiments that depart from the specific details, and the details of the present description may be modified or varied from the spirit and scope of the present invention. It should be noted that the illustrations provided in the following embodiments merely illustrate the basic idea of the present invention by way of illustration, and the following embodiments and features in the embodiments may be combined with each other without conflict.
Wherein the drawings are for illustrative purposes only and are shown in schematic, non-physical, and not intended to limit the invention; for the purpose of better illustrating embodiments of the invention, certain elements of the drawings may be omitted, enlarged or reduced and do not represent the size of the actual product; it will be appreciated by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The same or similar reference numbers in the drawings of embodiments of the invention correspond to the same or similar components; in the description of the present invention, it should be understood that, if there are terms such as "upper", "lower", "left", "right", "front", "rear", etc., that indicate an azimuth or a positional relationship based on the azimuth or the positional relationship shown in the drawings, it is only for convenience of describing the present invention and simplifying the description, but not for indicating or suggesting that the referred device or element must have a specific azimuth, be constructed and operated in a specific azimuth, so that the terms describing the positional relationship in the drawings are merely for exemplary illustration and should not be construed as limiting the present invention, and that the specific meaning of the above terms may be understood by those of ordinary skill in the art according to the specific circumstances.
As shown in fig. 1, a main network with a coal mine entrance well head or underground key place image as input data and yolov5 as target detection is adopted, an interval time T is taken to acquire images, a coal mine entrance area personnel bipartite graph is formed, and a Huo Puke Luofet-Kapp algorithm is used for T 1 And T 2 Personnel at the moment are matched, so that the number of people entering and exiting the well is counted.
The input image is a picture with a person at the observation point, and the image is subjected to slicing processing, for example, 608×608×3 image is processed as a feature map 304×304×12. Taking yolov5 as a target detection network, inputting the sliced image into a backbone network, and processing the sliced image by continuous combination of modules such as convolution, residual error, up-sampling and the like to generate the position information P of a personnel target i (x 1 ,x 2 ,y 1 ,y 2 )。
As shown in fig. 2, in order to determine the number of people entering and exiting from the well, it is necessary to identify whether the people pass through a key area L (x, y) in the screen, and the line L is used as an area line for entering and exiting detection. Acquisition time T 1 And T 2 The target detection image is used for judging whether the personnel in the image move in two moments and cross the regional line L. WellThe lower environment is worse than the upper environment, the judgment of personnel is more fuzzy, and under the poor illumination and shielding, the identified target personnel are matched, and then the theory of maximum matching of the two images is introduced to realize accurate matching.
Graph G 1 (v 1 ,v 2 ,…,v i ) And G 2 (u 1 ,u 2 ,…,u i ) Respectively time T 1 And T 2 Lower staff identification schematic diagram, G 1 The midpoint represents the information before the person moves, G 2 The midpoint represents information after the person moves. For an exact match set (v 1 ,v 2 ,…,v i ) Sum set (u) 1 ,u 2 ,…,u i ) The maximum matching is realized by adopting a Hopklover-Kapu algorithm.
The hopcalite-kapu algorithm finds an augmented path based on the current solution (i.e., the current match) in each iteration, and if an augmented path exists, it indicates that the current solution is not the optimal solution yet. After a certain iteration is finished, no amplification path is found, and then the current optimal solution (namely maximum matching) is found, thereby matching T 1 And T 2 Personnel in the time of day image.
The matching pattern here is assumed person P i For uniform movement or stationary, v i And u i The distance between the position coordinates is taken as a judgment basis, and the minimum distance s is defaulted i For person P i From v i And u i Is a moving distance of the moving object). Graph G 1 (v 1 ,v 2 ,…,v i ) And G 2 (u 1 ,u 2 ,…,u i ) Each store in the database is subjected to iterative matching to find each pair v i And u i Best match between E i (v i ,u i )。
Discovery of T 1 And T 2 After the best matching combination of the moments, the computing personnel P i Is provided, and a moving direction and a moving distance of the same. If when P i Not crossing the regional line L, P i Is not put into or put out of the well; when P is detected i The cross-domain area line L increases or decreases the population statistics according to the cross-domain direction.
By combining the target personnel detection method based on yolov5 and the principle of introducing two graphs, personnel matching is realized, thus realizing real-time judgment of the entry and exit conditions of underground personnel in a coal mine, and counting the number of people in a key underground area in real time.
Finally, it is noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made thereto without departing from the spirit and scope of the present invention, which is intended to be covered by the claims of the present invention.
Claims (2)
1. A coal mine underground people counting method based on two graphs is characterized by comprising the following steps of: the method comprises the following steps:
taking a coal mine well entrance area or underground key place image as input data, taking yolov5 as a main network for target detection, taking an interval time T to acquire images, forming a coal mine well entrance area personnel bipartite graph, and carrying out T by using a Huo Puke Luofet-Kapp algorithm 1 And T 2 Matching personnel at the moment, and counting the number of people entering and exiting the well;
the input image is a picture with personnel at an observation site, the image is sliced, yolov5 is taken as a target detection network, the sliced image is input into a backbone network, and the position information P of the personnel target is generated through convolution, residual error and up-sampling i (x 1 ,x 2 ,y 1 ,y 2 )；x 1 Is the left upper corner abscissa of the mark frame, x 2 Is the horizontal coordinate of the right lower corner of the mark frame, y 1 Is the vertical coordinate of the upper left corner of the mark frame, y 2 Is the ordinate of the lower right corner of the mark frame;
to determine the number of people entering and exiting from the well, whether the person passes through a straight line L (x) =kx+b in the picture is identified, the line L is taken as an area line for entering and exiting detection, k and b are constants, and x is an independent variable; acquisition time T 1 And T 2 The target detection image is used for judging whether personnel in the image move in two moments and cross the regional line L;
graph G 1 (v 1 ,v 2 ,…,v i ) And G 2 (u 1 ,u 2 ,…,u i ) Respectively time T 1 And T 2 Lower staff identification schematic diagram, G 1 Midpoint v i Represents the center point of the position before the movement of the ith person, G 2 Midpoint u i Representing the center point of the position after the movement of the ith personnel; for matching sets (v 1 ,v 2 ,…,v i ) Sum set (u) 1 ,u 2 ,…,u i ) The maximum matching is realized by adopting a Hopklover-Kapu algorithm;
in each iteration, the Hopklovir-Kapp algorithm searches for an augmented path based on a current solution, which is a current match; if an augmented path exists, the current solution is not the optimal solution; after a certain iteration is finished, if any amplification path cannot be found, the current optimal solution, namely the maximum matching and the matching T, is found 1 And T 2 Personnel in the moment image;
the matching mode is assumed person P i For uniform movement or stationary, v i And u i The distance between the position coordinates is taken as a judgment basis, and the minimum distance s is defaulted i For person P i From v i And u i Is a moving distance of (2); graph G 1 (v 1 ,v 2 ,…,v i ) And G 2 (u 1 ,u 2 ,…,u i ) Each store in the database is subjected to iterative matching to find each pair v i And u i Best match between E i (v i ,u i )；
Discovery of T 1 And T 2 After the best matching combination of the moments, the computing personnel P i Is provided with a moving direction and a moving distance; if when P i Not crossing the regional line L, P i Is not put into or put out of the well; when P is detected i The cross-domain area line L increases or decreases the population statistics according to the cross-domain direction.
2. The bipartite graph-based underground coal mine people counting method as claimed in claim 1, wherein: the slicing processing of the image is as follows: the 608×608×3 image is processed as a 304×304×12 feature map.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310489568.3A CN116524540A (en) | 2023-05-04 | 2023-05-04 | Coal mine underground people counting method based on bipartite graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310489568.3A CN116524540A (en) | 2023-05-04 | 2023-05-04 | Coal mine underground people counting method based on bipartite graph |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116524540A true CN116524540A (en) | 2023-08-01 |
Family
ID=87397222
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310489568.3A Pending CN116524540A (en) | 2023-05-04 | 2023-05-04 | Coal mine underground people counting method based on bipartite graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116524540A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116680545A (en) * | 2023-08-02 | 2023-09-01 | 西安核音智言科技有限公司 | Coal mine well exit personnel prediction method based on Markov random field |
-
2023
- 2023-05-04 CN CN202310489568.3A patent/CN116524540A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116680545A (en) * | 2023-08-02 | 2023-09-01 | 西安核音智言科技有限公司 | Coal mine well exit personnel prediction method based on Markov random field |
| CN116680545B (en) * | 2023-08-02 | 2023-10-20 | 西安核音智言科技有限公司 | Coal mine well exit personnel prediction method based on Markov random field |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111027525B (en) | Method, device and system for tracking potential infected persons in public places during epidemic situation | |
| CN108710868B (en) | Human body key point detection system and method based on complex scene | |
| CN106412501B (en) | A kind of the construction safety behavior intelligent monitor system and its monitoring method of video | |
| CN106127204B (en) | A kind of multi-direction meter reading Region detection algorithms of full convolutional neural networks | |
| CA3094424C (en) | Safety monitoring and early-warning method for man-machine interaction behavior of underground conveyor belt operator | |
| CN110866887A (en) | Target situation fusion sensing method and system based on multiple sensors | |
| CN109544563B (en) | Passive millimeter wave image human body target segmentation method for security inspection of prohibited objects | |
| US20110115909A1 (en) | Method for tracking an object through an environment across multiple cameras | |
| CN110070010A (en) | A kind of face character correlating method identified again based on pedestrian | |
| CN109657592A (en) | A kind of face identification system and method for intelligent excavator | |
| Szwoch | Extraction of stable foreground image regions for unattended luggage detection | |
| CN109255286A (en) | A kind of quick detection recognition method of unmanned plane optics based on YOLO deep learning network frame | |
| CN112669350A (en) | Adaptive feature fusion intelligent substation human body target tracking method | |
| CN102254394A (en) | Antitheft monitoring method for poles and towers in power transmission line based on video difference analysis | |
| Guo et al. | Evaluation-oriented façade defects detection using rule-based deep learning method | |
| CN116524540A (en) | Coal mine underground people counting method based on bipartite graph | |
| CN106709938A (en) | Multi-target tracking method based on improved TLD (tracking-learning-detected) | |
| CN112861673A (en) | False alarm removal early warning method and system for multi-target detection of surveillance video | |
| Wang et al. | Multiple-human tracking by iterative data association and detection update | |
| Papaioannou et al. | Accurate positioning via cross-modality training | |
| Chen et al. | Motion propagation detection association for multi-target tracking in wide area aerial surveillance | |
| CN109389177B (en) | Tunnel vehicle re-identification method based on cooperative cascade forest | |
| Zhou et al. | Street-view imagery guided street furniture inventory from mobile laser scanning point clouds | |
| CN112613668A (en) | Scenic spot dangerous area management and control method based on artificial intelligence | |
| Wieneke et al. | Combined person tracking and classification in a network of chemical sensors |
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 |