CN110096942A - A kind of Smoke Detection algorithm based on video analysis - Google Patents
A kind of Smoke Detection algorithm based on video analysis Download PDFInfo
- Publication number
- CN110096942A CN110096942A CN201811591402.8A CN201811591402A CN110096942A CN 110096942 A CN110096942 A CN 110096942A CN 201811591402 A CN201811591402 A CN 201811591402A CN 110096942 A CN110096942 A CN 110096942A
- Authority
- CN
- China
- Prior art keywords
- cigarette district
- area
- smog
- doubtful
- formula
- 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
Classifications
-
- 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
- G06V20/49—Segmenting video sequences, i.e. computational techniques such as parsing or cutting the sequence, low-level clustering or determining units such as shots or scenes
-
- 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
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/10—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in agriculture
- Y02A40/28—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in agriculture specially adapted for farming
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computing Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Fire-Detection Mechanisms (AREA)
Abstract
The present invention relates to a kind of Smoke Detection algorithm based on video analysis, the production of training set, the picture comprising smog is obtained first from existing forest smog video, mark includes smoke region manually, makes label image, the training of full volume machine network model, using trained network model, the doubtful cigarette district in single-frame images, doubtful cigarette district area dynamic Analysis in Growth are obtained, the formula that continuous videos sequence corresponds to doubtful cigarette district area sum is calculated, the formula of the relative growth of doubtful cigarette district area is calculated.The present invention is realized based on rotating lens forest Smoke Detection, is alarmed for forest fire, achievees the effect that carry out early prediction to forest fire, and economic loss caused by reducing because of fire brings better prospect.
Description
Technical field
The invention belongs to deep learnings and artificial intelligence field, particularly relate to a kind of image recognition algorithm, are used for Forest Fire
Calamity early warning.
Background technique
The cause of fire and generation place have diversity, cause obstruction to fire alarm and fighting work.For many years,
It is dedicated to fire study on prevention both at home and abroad, has carried out various trials.All kinds of fire detectors at this stage mainly have thermoinduction
Formula detector, photoinduction formula detector and cigarette induction type detector.These traditional detectors are cheap, accuracy is high,
But generally existing some insoluble defects.Such as since phenomena such as smoke propagation and temperature rising, is required to relatively
Longer time, so that traditional sensors inevitably result from operating lag, in addition, traditional sensors needs have been mounted on
The factors such as neighbouring, the Long Term Contact a large amount of dust of fire point, so that traditional sensors easily trouble or failure, and be particularly unsuitable for
The fire detection in these high-hazard contents places of large space or outdoor scene.Smoke Detection technology based on video analysis is one
Kind is based on machine vision, non-contacting fire detection technology, and the fire detection for being especially suitable for solving the places such as large space, outdoor is difficult
Topic.Such methods have the advantages such as response is fast, is not easily affected by environmental factors, is widely applicable, is at low cost.
Summary of the invention
In view of the studies above background, it is an object of the invention to: a kind of extraction of smog suspicious region based on FCN is provided
Algorithm avoids traditional algorithm because monitoring probe is shaken or rotates, and can completely does not extract smog suspicious region, improves prison
Control the utilization rate of probe.Specific step is as follows:
Step 1: the production of training set obtains the picture comprising smog first from existing forest smog video, marks manually
Note includes smoke region, makes label image.
Step 2: the training of full volume machine network model
Step 3: using trained network model, obtain the doubtful cigarette district in single-frame images.
Step 4: doubtful cigarette district area dynamic Analysis in Growth
Increased using the dynamic of cigarette district, removes interference region.Early stage fire, due to fire generate heat and itself
Diffusion motion so that cigarette district area becomes larger over time.And common chaff interferent (pedestrian that such as wears gray suit,
The automobile etc. of grey) profile be fixed, so the area of doubtful cigarette district to be increased to the standard as differentiation smog herein.By
It is slower in smog movement, so that the area that adjacent two frame corresponds to smoke region is very nearly the same, by weather or systematic error
It influences, cigarette district area is it is possible that negative growth.Herein for problem mentioned above, doubtful cigarette district area dynamic is proposed
The method of Analysis in Growth.
It calculates continuous videos sequence and corresponds to the formula of doubtful cigarette district area sum and be
In formula: At is that n frame corresponds to the sum of smoke region area before t moment;ArkIt is the area of kth frame smoke region.
The formula for calculating the relative growth of doubtful cigarette district area is
|At+1-At|>Th6 (2)
The diffusance of smog can be measured by formula (2).
Compared with prior art, the invention has the following beneficial effects:
In the present invention, the Smoke Detection technology based on video analysis is a kind of based on machine vision, the spy of non-contacting fire
Survey technology, is especially suitable for solving the fire detection problem of large space, the places such as outdoor, such methods have response it is fast, not vulnerable to
The advantages such as such environmental effects, widely applicable, at low cost, bring better prospect of the application.
Detailed description of the invention
Fig. 1 is cigarette district area dynamic Analysis in Growth figure of the invention.
Fig. 2 is scene application drawing of the invention.
Specific embodiment
The invention will be further described below in conjunction with the accompanying drawings, and following embodiment is only used for clearly illustrating the present invention
Technical solution, and not intended to limit the protection scope of the present invention.
Embodiment 1
Embodiment 1
The following steps are included:
Cigarette district in training sample is carried out manual markings, generates its corresponding label by step S1, the production of marker samples
Figure.
Step S2, the full convolutional network of training, obtains doubtful cigarette district.
Step S3 carries out area dynamic Analysis in Growth to doubtful cigarette district.
This model is detected in different test sets, as a result (Figure of description), show this model can adapt in
Multiple and different scenes.
The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention.The technology of the industry
Personnel are it should be appreciated that the present invention is not limited to the above embodiments, and the above embodiments and description only describe this
The principle of invention, without departing from the spirit and scope of the present invention, various changes and improvements may be made to the invention, these changes
Change and improvement all fall within the protetion scope of the claimed invention.The claimed scope of the invention by appended claims and its
Equivalent thereof.
Claims (1)
1. a kind of Smoke Detection algorithm based on video analysis, it is characterised in that this method comprises the following steps:
Step 1: the production of training set obtains the picture comprising smog, manually mark packet first from existing forest smog video
Containing smoke region, label image is made;
Step 2: the training of full volume machine network model;
Step 3: using trained network model, obtain the doubtful cigarette district in single-frame images;
Step 4: doubtful cigarette district area dynamic Analysis in Growth,
Early stage fire, due to heat and the diffusion motion of itself that fire generates, so that cigarette district area pushing away with the time
Shifting becomes larger.And the profile of common chaff interferent (such as wearing the pedestrian of gray suit, the automobile of grey) is fixed, so herein
The area of doubtful cigarette district is increased as the standard for differentiating smog.Since smog movement is slower, so that adjacent two frame is corresponding
The area of smoke region is very nearly the same, is influenced by weather or systematic error, and cigarette district area is it is possible that negative growth, this paper needle
To problem mentioned above, the method for doubtful cigarette district area dynamic Analysis in Growth is proposed,
It calculates continuous videos sequence and corresponds to the formula of doubtful cigarette district area sum and be
In formula: At is that n frame corresponds to the sum of smoke region area before t moment;ArkIt is the area of kth frame smoke region,
The formula for calculating the relative growth of doubtful cigarette district area is
|At+1-At| > Th6 (2)
The diffusance of smog can be measured by formula (2).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811591402.8A CN110096942A (en) | 2018-12-20 | 2018-12-20 | A kind of Smoke Detection algorithm based on video analysis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811591402.8A CN110096942A (en) | 2018-12-20 | 2018-12-20 | A kind of Smoke Detection algorithm based on video analysis |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110096942A true CN110096942A (en) | 2019-08-06 |
Family
ID=67443718
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811591402.8A Pending CN110096942A (en) | 2018-12-20 | 2018-12-20 | A kind of Smoke Detection algorithm based on video analysis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110096942A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111126478A (en) * | 2019-12-19 | 2020-05-08 | 北京迈格威科技有限公司 | Convolutional neural network training method, device and electronic system |
CN112257523A (en) * | 2020-10-09 | 2021-01-22 | 营口新山鹰报警设备有限公司 | Smoke identification method and system of image type fire detector |
CN113378629A (en) * | 2021-04-27 | 2021-09-10 | 阿里云计算有限公司 | Method and device for detecting abnormal vehicle in smoke discharge |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2011032117A1 (en) * | 2009-09-13 | 2011-03-17 | Delacom Detection Systems, Llc | Method and system for wildfire detection using a visible range camera |
CN104794486A (en) * | 2015-04-10 | 2015-07-22 | 电子科技大学 | Video smoke detecting method based on multi-feature fusion |
CN107025443A (en) * | 2017-04-06 | 2017-08-08 | 江南大学 | Stockyard smoke monitoring and on-time model update method based on depth convolutional neural networks |
CN107609470A (en) * | 2017-07-31 | 2018-01-19 | 成都信息工程大学 | The method of outdoor fire disaster early-stage smog video detection |
CN107749067A (en) * | 2017-09-13 | 2018-03-02 | 华侨大学 | Fire hazard smoke detecting method based on kinetic characteristic and convolutional neural networks |
-
2018
- 2018-12-20 CN CN201811591402.8A patent/CN110096942A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2011032117A1 (en) * | 2009-09-13 | 2011-03-17 | Delacom Detection Systems, Llc | Method and system for wildfire detection using a visible range camera |
CN104794486A (en) * | 2015-04-10 | 2015-07-22 | 电子科技大学 | Video smoke detecting method based on multi-feature fusion |
CN107025443A (en) * | 2017-04-06 | 2017-08-08 | 江南大学 | Stockyard smoke monitoring and on-time model update method based on depth convolutional neural networks |
CN107609470A (en) * | 2017-07-31 | 2018-01-19 | 成都信息工程大学 | The method of outdoor fire disaster early-stage smog video detection |
CN107749067A (en) * | 2017-09-13 | 2018-03-02 | 华侨大学 | Fire hazard smoke detecting method based on kinetic characteristic and convolutional neural networks |
Non-Patent Citations (6)
Title |
---|
FEINIU YUAN 等: "Deep Smoke Segmentation", 《ARXIV》 * |
THOU-HO (CHAO-HO) CHEN 等: "An Early Fire-Detection Method Based on Image Processing", 《ICIP 2004》 * |
史劲亭 等: "视频烟雾检测研究进展", 《中国图象图形学报》 * |
吴冬梅 等: "基于多特征融合的烟雾检测", 《图学学报》 * |
杨冰 等: "基于监控平台的火灾烟雾监测***设计", 《中国优秀硕士学位论文全文数据库(电子期刊) 信息科技辑》 * |
王娜娜: "基于视频的火灾烟雾检测算法研究", 《中国优秀硕士学位论文全文数据库(电子期刊) 信息科技辑》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111126478A (en) * | 2019-12-19 | 2020-05-08 | 北京迈格威科技有限公司 | Convolutional neural network training method, device and electronic system |
CN111126478B (en) * | 2019-12-19 | 2023-07-07 | 北京迈格威科技有限公司 | Convolutional neural network training method, device and electronic system |
CN112257523A (en) * | 2020-10-09 | 2021-01-22 | 营口新山鹰报警设备有限公司 | Smoke identification method and system of image type fire detector |
CN113378629A (en) * | 2021-04-27 | 2021-09-10 | 阿里云计算有限公司 | Method and device for detecting abnormal vehicle in smoke discharge |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Gong et al. | A Real‐Time Fire Detection Method from Video with Multifeature Fusion | |
CN108734125B (en) | Smoking behavior identification method for open space | |
CN110096942A (en) | A kind of Smoke Detection algorithm based on video analysis | |
CN101738394B (en) | Method and system for detecting indoor smog | |
CN106097346B (en) | A kind of video fire hazard detection method of self study | |
CN102867386B (en) | Intelligent video analysis-based forest smoke and fire detection method and special system thereof | |
CN111626188B (en) | Indoor uncontrollable open fire monitoring method and system | |
CN102673979B (en) | Method and device for judging deviation of conveying belt | |
CN101373553B (en) | Early-stage smog video detecting method capable of immunizing false alarm in dynamic scene | |
CN104598895A (en) | Method and device for flame detection based on video image analysis | |
CN106463043A (en) | Intrusion detection with motion sensing | |
CN103150549A (en) | Highway tunnel fire detecting method based on smog early-stage motion features | |
CN105574468B (en) | Video flame detection method, device and system | |
WO2006115676A3 (en) | Video ghost detection by outline | |
CN104715559B (en) | A kind of Smoke Detection and fire alarm method based on track identification | |
CN104811586A (en) | Scene change video intelligent analyzing method, device, network camera and monitoring system | |
CN108280953A (en) | Video detecting alarm method and device, storage medium, camera | |
CN106709905A (en) | Vibration-proof hammer fault online detection and identification method based on binocular vision image | |
CN105528581B (en) | Video smoke event intelligent detecting method based on bionical color reaction model | |
CN107093179A (en) | Localization method is searched for for the two waveband fire in wide area space | |
CN108363992B (en) | Fire early warning method for monitoring video image smoke based on machine learning | |
CN112906674A (en) | Mine fire identification and fire source positioning method based on binocular vision | |
CN114120171A (en) | Fire smoke detection method, device and equipment based on video frame and storage medium | |
CN108830161A (en) | Smog recognition methods based on video stream data | |
CN102609685B (en) | Shadowing judging method of image type fire detector |
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: 20190806 |
|
RJ01 | Rejection of invention patent application after publication |