CN112836643A - Specific scene smoking and calling identification method - Google Patents
Specific scene smoking and calling identification method Download PDFInfo
- Publication number
- CN112836643A CN112836643A CN202110154950.XA CN202110154950A CN112836643A CN 112836643 A CN112836643 A CN 112836643A CN 202110154950 A CN202110154950 A CN 202110154950A CN 112836643 A CN112836643 A CN 112836643A
- Authority
- CN
- China
- Prior art keywords
- smoking
- human body
- calling
- video image
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000000391 smoking effect Effects 0.000 title claims abstract description 34
- 238000000034 method Methods 0.000 title claims abstract description 19
- 238000001514 detection method Methods 0.000 claims abstract description 16
- 230000006399 behavior Effects 0.000 claims abstract description 14
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 11
- 230000009191 jumping Effects 0.000 claims abstract description 3
- 238000007635 classification algorithm Methods 0.000 claims description 9
- 239000000779 smoke Substances 0.000 abstract description 7
- 230000008859 change Effects 0.000 abstract description 4
- 239000003245 coal Substances 0.000 description 2
- 230000003137 locomotive effect Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000012774 insulation material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/107—Static hand or arm
-
- 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
- G06V40/28—Recognition of hand or arm movements, e.g. recognition of deaf sign language
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a specific scene smoking and calling identification method, which comprises the following steps: s1: collecting a video image shot by a gunlock; s2: carrying out coarse positioning and identification processing on the video image by using a human body detection algorithm; s3: if a plurality of human body targets are detected, selecting one with the largest target area to enter S4, and if a single human body target is detected, directly jumping to S4; s4: calculating the position coordinates of the central point, and sending the position coordinates of the central point to a ball machine; s5: the ball machine enlarges a human body target area to the whole picture in an automatic zooming mode and tracks the area in real time; s6: sending the video image collected by the ball machine into a target classifier; s7: and (6) ending. The invention combines engineering means and technical means, realizes intelligent smoke calling behavior recognition without dead angle, can be used in scenes with complex background and large change, improves the effective pixel number of small target smoke calling, improves recognition rate, and reduces false alarm rate and false alarm rate.
Description
Technical Field
The invention relates to the technical field of behavior recognition, in particular to a specific scene smoking and calling recognition method.
Background
In places such as locomotive cabs, public security inquiries, coal mines, gas stations and the like, because of the storage of heat insulation materials and combustible objects, fire prevention work is very important, and great life and property loss can be caused by smoking or calling, so that smoking or calling is prohibited in key places such as the locomotive cabs, the public security inquiries, the coal mines, the gas stations and the like, workers with good luck psychology can possibly avoid safety patrol, and the work needs to be carried out by dead-corner-free monitoring.
In the field of computer vision, smoking or calling detection is a challenging subject, and because the target is small, the smoke or calling detection is easily interfered by light and background and is difficult to detect, the existing smoking or calling documents and patents are more, most of schemes are realized based on algorithms, and are not applied in a project, so that the false alarm rate and the missing report rate of smoking or calling are higher, the application range is limited, and the detection is only applied in a closed scene with a single background; the method is not suitable for places with complex background and large scene change. Although the existing document adopts a cutout mode to detect whether smoking or calling exists, effective pixels are not increased essentially, and the false detection is higher particularly under the condition of low image resolution.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, combine engineering realization, adopt the existing gun-ball linkage system, combine an algorithm and equipment, improve the effective pixel number of a smoking or calling area, and provide a smoking and calling identification method in a specific scene.
The purpose of the invention is realized by the following technical scheme:
a specific scene smoking and calling identification method comprises the following steps:
s1: acquiring a video image of a fixed area through a gunlock;
s2: carrying out rough positioning and identification processing on the video image by using a human body detection algorithm to obtain a candidate suspected smoking or calling human body target area;
s3: if a plurality of human body targets are detected, selecting one with the largest target area to enter S4; if a single human body target is detected, directly jumping to S4;
s4: calculating the position coordinates of the central point according to the obtained human body target area, and sending the position coordinates of the central point to a ball machine;
s5: the ball machine enlarges a human body target area to the whole picture in an automatic zooming mode, and positions and tracks the area in real time;
s6: sending the video image collected by the ball machine into a target classifier, and finishing the judgment of smoking or calling behaviors by the target classifier
S7: and (6) ending.
Further, the object classifier includes a smoking classification algorithm and a calling classification algorithm.
Further, when the smoking behavior is detected in the rough positioning mode, the method specifically comprises the following steps: a fixed part is formed by placing the hands of the human body at the mouth side and the head in the video image, and the fixed part is identified as a rigid body.
Further, when the coarse positioning detection is performed to detect a call-making behavior, the method specifically comprises the following steps: a fixed part is formed by the head and the hand of the human body placed beside the ear in the video image, and the fixed part is identified as a rigid body.
Further, the human body target area is a fixed part area obtained after rough positioning.
The invention has the beneficial effects that: the invention combines engineering means and technical means, realizes intelligent smoke calling behavior recognition without dead angle, can be used in scenes with complex background and large change, improves the effective pixel number of small target smoke calling, improves the recognition rate, and reduces the false alarm rate and the missing alarm rate.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
In order to more clearly understand the technical features, objects, and effects of the present invention, embodiments of the present invention will now be described with reference to the accompanying drawings.
In this embodiment, as shown in fig. 1, a method for identifying smoking and making a call in a specific scene includes the following steps:
s1: acquiring a shooting video image of a gunlock;
s2: processing by using a human body detection algorithm to obtain a candidate suspected target area;
s3: if a single suspected target is detected, directly entering S4; if a plurality of suspected calling targets are detected, selecting one of the candidate targets with the largest target area to enter S4;
s4: calculating the position coordinates of the central point according to the obtained suspected target area, and sending the position coordinates of the central point to a ball machine;
s5: the ball machine enlarges the suspected target area to the whole picture in an automatic zooming mode and positions and tracks the area in real time;
s6: sending the video image collected by the ball machine into a target classifier to finish the judgment of smoking or calling behaviors
S7: and (6) ending.
The human body detection algorithm comprises a suspected smoking detection algorithm and a suspected calling detection algorithm, and can be realized by adopting a general target detection algorithm such as YOLO, SSD and fasterRcnn.
The target classifier comprises a smoking classification algorithm and a calling classification algorithm, the smoking classification algorithm is used for identifying the extraction behavior, the calling classification algorithm is used for identifying the calling behavior, and common classification algorithms such as SVM, SqueezeNet, MobileNet and ShuffleNet are adopted to realize the classification.
In this embodiment, when the human body detection algorithm processes the video image, the behavior of smoking or making a call is preliminarily identified; meanwhile, when a human body in a video image acquired by the gunlock performs smoking action or calls, a hand is placed at the mouth or the ear to form a fixed part with the head, and the fixed part is used as a rigid body for identification; and obtaining a fixed part area through rough positioning, and further calculating the position coordinate of the central point.
The invention combines engineering means and technical means, realizes intelligent smoke calling behavior recognition without dead angle, can be used in scenes with complex background and large change, improves the effective pixel number of small target smoke calling, improves the recognition rate, and reduces the false alarm rate and the missing alarm rate.
The foregoing shows and describes the general principles and broad features of the present invention and advantages thereof. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (5)
1. A specific scene smoking and calling identification method is characterized by comprising the following steps:
s1: acquiring a video image of a fixed area through a gunlock;
s2: carrying out rough positioning and identification processing on the video image by using a human body detection algorithm to obtain a candidate suspected smoking or calling human body target area;
s3: if a plurality of human body targets are detected, selecting one with the largest target area to enter S4; if a single human body target is detected, directly jumping to S4;
s4: calculating the position coordinates of the central point according to the obtained human body target area, and sending the position coordinates of the central point to a ball machine;
s5: the ball machine enlarges a human body target area to the whole picture in an automatic zooming mode, and positions and tracks the area in real time;
s6: sending the video image acquired by the ball machine into a target classifier, and finishing the judgment of smoking or calling behaviors by the target classifier;
s7: and (6) ending.
2. The scene-specific smoking and call recognition method of claim 1, wherein the object classifier comprises a smoking classification algorithm and a call classification algorithm.
3. The specific-scene smoking and call recognition method according to claim 1, wherein when the rough positioning detection smoking behavior is performed, the method specifically comprises the following steps: a fixed part is formed by placing the hands of the human body at the mouth side and the head in the video image, and the fixed part is identified as a rigid body.
4. The specific-scene smoking and call recognition method according to claim 1, wherein when the rough positioning detection smoking behavior is performed, the method specifically comprises the following steps: a fixed part is formed by placing the hands of the human body at the mouth side and the head in the video image, and the fixed part is identified as a rigid body.
5. The specific scene smoking and call recognition method of claim 1 or 3, wherein the human target area is a fixed part area obtained after rough positioning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110154950.XA CN112836643A (en) | 2021-02-04 | 2021-02-04 | Specific scene smoking and calling identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110154950.XA CN112836643A (en) | 2021-02-04 | 2021-02-04 | Specific scene smoking and calling identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112836643A true CN112836643A (en) | 2021-05-25 |
Family
ID=75932037
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110154950.XA Pending CN112836643A (en) | 2021-02-04 | 2021-02-04 | Specific scene smoking and calling identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112836643A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102447835A (en) * | 2011-10-29 | 2012-05-09 | 合肥博微安全电子科技有限公司 | Non-blind-area multi-target cooperative tracking method and system |
| CN110287906A (en) * | 2019-06-26 | 2019-09-27 | 四川长虹电器股份有限公司 | Method and system based on image/video detection people " playing mobile phone " |
| CN111062319A (en) * | 2019-12-16 | 2020-04-24 | 武汉极目智能技术有限公司 | Driver call detection method based on active infrared image |
| CN111898514A (en) * | 2020-07-24 | 2020-11-06 | 燕山大学 | Multi-target visual supervision method based on target detection and action recognition |
| CN112052815A (en) * | 2020-09-14 | 2020-12-08 | 北京易华录信息技术股份有限公司 | Behavior detection method and device and electronic equipment |
| CN112115775A (en) * | 2020-08-07 | 2020-12-22 | 北京工业大学 | Smoking behavior detection method based on computer vision in monitoring scene |
| CN112257643A (en) * | 2020-10-30 | 2021-01-22 | 天津天地伟业智能安全防范科技有限公司 | Smoking behavior and calling behavior identification method based on video streaming |
-
2021
- 2021-02-04 CN CN202110154950.XA patent/CN112836643A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102447835A (en) * | 2011-10-29 | 2012-05-09 | 合肥博微安全电子科技有限公司 | Non-blind-area multi-target cooperative tracking method and system |
| CN110287906A (en) * | 2019-06-26 | 2019-09-27 | 四川长虹电器股份有限公司 | Method and system based on image/video detection people " playing mobile phone " |
| CN111062319A (en) * | 2019-12-16 | 2020-04-24 | 武汉极目智能技术有限公司 | Driver call detection method based on active infrared image |
| CN111898514A (en) * | 2020-07-24 | 2020-11-06 | 燕山大学 | Multi-target visual supervision method based on target detection and action recognition |
| CN112115775A (en) * | 2020-08-07 | 2020-12-22 | 北京工业大学 | Smoking behavior detection method based on computer vision in monitoring scene |
| CN112052815A (en) * | 2020-09-14 | 2020-12-08 | 北京易华录信息技术股份有限公司 | Behavior detection method and device and electronic equipment |
| CN112257643A (en) * | 2020-10-30 | 2021-01-22 | 天津天地伟业智能安全防范科技有限公司 | Smoking behavior and calling behavior identification method based on video streaming |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101465033B (en) | Automatic tracking recognition system and method | |
| CN201278180Y (en) | Automatic tracking recognition system | |
| CN103726879B (en) | Utilize camera automatic capturing mine ore deposit to shake and cave in and the method for record warning in time | |
| CN102833478B (en) | Fault-tolerant background model | |
| CN109872341A (en) | A kind of throwing object in high sky detection method based on computer vision and system | |
| KR101191844B1 (en) | Image processing apparatus and image processing method | |
| KR102195706B1 (en) | Method and Apparatus for Detecting Intruder | |
| CN106341661B (en) | Patrol robot | |
| CN110633612B (en) | Monitoring method and system for inspection robot | |
| CN102117484B (en) | Processing system, processing method and image classification method using image color information | |
| CN107036687A (en) | The grain storage Monitoring of Quantity method and device of view-based access control model | |
| CN111428617A (en) | Video image-based distribution network violation maintenance behavior identification method and system | |
| WO2018161849A1 (en) | Alarm system for falling in water based on image water texture and method therefor | |
| US20110123067A1 (en) | Method And System for Tracking a Target | |
| CN112396658A (en) | Indoor personnel positioning method and positioning system based on video | |
| KR101442669B1 (en) | Method and apparatus for criminal acts distinction using intelligent object sensing | |
| CN111461078A (en) | Anti-fishing monitoring method based on computer vision technology | |
| CN201726494U (en) | Device and system which utilize image color information to conduct image comparison | |
| CN110544271A (en) | Parabolic motion detection method and related device | |
| CN109299700A (en) | Subway group abnormality behavioral value method based on crowd density analysis | |
| CN114913323B (en) | Charging pile area night open fire detection method | |
| CN111223131B (en) | Monitoring and tracking method for continuous movement of specific target and suspected target | |
| CN112836643A (en) | Specific scene smoking and calling identification method | |
| CN109799844B (en) | Dynamic target tracking system and method for pan-tilt camera | |
| CN113554682B (en) | Target tracking-based safety helmet detection method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210525 |
|
| RJ01 | Rejection of invention patent application after publication |