CN109147259A - A kind of remote fire detection system and method based on video image - Google Patents
A kind of remote fire detection system and method based on video image Download PDFInfo
- Publication number
- CN109147259A CN109147259A CN201811386160.9A CN201811386160A CN109147259A CN 109147259 A CN109147259 A CN 109147259A CN 201811386160 A CN201811386160 A CN 201811386160A CN 109147259 A CN109147259 A CN 109147259A
- Authority
- CN
- China
- Prior art keywords
- flame
- region
- image
- video
- remote
- 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
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000001514 detection method Methods 0.000 title claims abstract description 31
- 238000004891 communication Methods 0.000 claims abstract description 19
- 238000000605 extraction Methods 0.000 claims abstract description 11
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 5
- 230000004069 differentiation Effects 0.000 claims description 4
- 238000000151 deposition Methods 0.000 claims 1
- 230000035945 sensitivity Effects 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 2
- 239000000779 smoke Substances 0.000 description 2
- 231100000768 Toxicity label Toxicity 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000036632 reaction speed Effects 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B17/00—Fire alarms; Alarms responsive to explosion
- G08B17/12—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions
- G08B17/125—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions by using a video camera to detect fire or smoke
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Fire-Detection Mechanisms (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
The invention discloses a kind of remote fire detection system and method based on video image, the system includes: image capture module, the characteristics of acquiring infrared image by two-waveband video acquisition technique, and being imaged according to infrared camera, is sent into control module by Video Decoder for infrared image；Control module realizes the feature extraction to infrared image by DSP, carries out flame identification to the characteristics of image extracted by recognizer, recognition result is uploaded to remote control center；Network communication module, control module are connected by physical layer transceiver inside network communication module and network interface with remote control center, if fire generation, remote control center assign control command, pass through network communication module, realization ethernet communication function；Wireless remote alarming module realizes wireless network remote alarms.The present invention improves accuracy, sensitivity and the reliability of flame identification algorithm, provides effective ways for the detection of large space remote fire.
Description
Technical field
The present invention relates to fire and flame detection field more particularly to a kind of remote fire detection systems based on video image
And method.
Background technique
Traditional fire detector mainly includes smoke alarm, temperature detector, combustible gas probe and red at present
Outer correlative detector etc., however these detectors competence exertion generally in City Building acts on.Smoke alarm is general
It is just to issue alarm after the smog of detection reaches a certain concentration, accurate fire prediction can not be carried out, while being also easy to appear
Situations such as reporting by mistake, failing to report.Temperature detector, it determines warm type temperature-sensing element using analog switch amount formula, generally at 68 DEG C or less
Lose benefit, while can not show and predict current temperature value and rate that temperature rises, other fire detectors all or
Mostly or less there is external interference factor.Simultaneously when the space of detection becomes larger, and fire occurs, the product of burning rises to sky
In, a certain height that product rises to space will be cooled down by air, be stagnated in the sky, these visit traditional fire
The Detection Techniques for surveying device are hindered, and also it are regularly repaired and be maintained, so traditional fire detection side
Their detection efficient is not also high while method needs to spend a large amount of man power and material, loses more than gain.
Summary of the invention
The technical problem to be solved in the present invention is that for traditional fire identification technology in the prior art in sensitivity and
It cannot balance, and the defect that traditional flame identification algorithm discrimination is not high, provide a kind of based on video image in reliability
Remote fire detection system and method.
The technical solution adopted by the present invention to solve the technical problems is:
The present invention provides a kind of remote fire detection system based on video image, comprising: image capture module, control mould
Block, network communication module, wireless remote alarming module and remote control center；Wherein:
Image capture module, including infrared camera and Video Decoder, infrared camera are acquired by two-waveband video
Technology acquires infrared image, and the characteristics of be imaged according to infrared camera, and infrared image is sent into control by Video Decoder
Module；
Control module realizes the feature extraction to infrared image by DSP, special to the image extracted by recognizer
Sign carries out flame identification, and recognition result is uploaded to remote control center；
Network communication module, including physical layer transceiver and network interface, control module pass through inside network communication module
Physical layer transceiver and network interface be connected with remote control center, if fire occur, remote control center assign control life
It enables, by network communication module, realizes ethernet communication function；
Wireless remote alarming module, is connected by way of wireless connection with control module, realizes that wireless network is remotely reported
It is alert.
Further, the method for image is obtained in image capture module of the invention specifically:
Infrared image is acquired by two-waveband video acquisition technique, by the way that threshold value, on the image position of high brightness is arranged
Pixel is formed, the region high to brightness height, that is, temperature carries out feature extraction；Wherein, the color of each pixel can use RnIt is red
Color, GnGreen, BnBlue three components indicate, the average brightness T of the pixelnAre as follows:
Tn=0.22 × Rn+0.587×Gn+0.114×Bn
Pixel value is subjected to hysteresis binaryzation to avoid the pixel value in video from generating runout error, is set higher than the upper limit
It is non-targeted point lower than lower limit for target point, is the pixel of previous frame among bound.
The present invention provides a kind of remote fire detection method based on video image, method includes the following steps:
S1, video infrared image is obtained；
S2, binary conversion treatment is carried out to infrared image, obtains multiple high temperature pixels, high temperature pixel is on infrared image
Form high temperature suspicious region；
S3, by doubtful flame region recognizer, judge whether there is doubtful flame region；If it exists, step is executed
S4；If it does not exist, video infrared image is reacquired；
S4, pass through profile scan algorithm, doubtful flame region is scanned；
S5, flame characteristic extraction is carried out, obtains doubtful flame region information；
S6, according to doubtful flame region information, doubtful flame region is judged again, if being confirmed as doubtful flame zone
Domain executes step S7；If it is not, then reacquiring video infrared image；
S7, be confirmed after doubtful flame region；
S8, pass through visible images distinguished number, compare doubtful flame region and extract visible images；
S9, binary conversion treatment is carried out to visible images；
S10, profile scan algorithm is executed to the visible images after binary conversion treatment；
S11, the flame characteristic information extracted；
S12, flame is judged whether it is by differentiating flame algorithm according to flame characteristic information；If flame executes step
Rapid S13；If not flame, reacquires video infrared image；
S13, flame pattern judgement is carried out to flame, judges whether it is stable flame；
S14, if flame is stablized, execute rudimentary alarm；If unstable flame, advanced alarm is executed.
Further, profile scan algorithm in this method of the invention method particularly includes:
After image carries out binaryzation, the panel region that high temperature pixel is linked to be is obtained；Profile scan algorithm is from picture
It scans in order left to bottom right, obtains the position that high temperature pixel is linked to be the boundary point in region, and save it in boundary bit
It sets in caching, the calculating for characteristic value below is prepared；Boundary length and the inside of the high-temperature area scanned are calculated simultaneously
The upper left corner of all pixels point and the coordinate value of bottom right angle point are stored in boundary position caching respectively and neutralize scanning area caching
In；After being scanned every time, each high-temperature area can carry out zone number since 1, which is scanning number.
Further, doubtful flame region recognizer in this method of the invention method particularly includes:
The high temperature suspicious region of present frame is matched with the high temperature suspicious region of former frame, if two frames are mutually matched,
Then the identiflication number of former frame is transferred in present frame；If not being mutually matched, just freeze the scanning number in the region, from number
Scanning number the smallest and without frozen and used number, as the region is selected in pond；It will be regarded by the algorithm
Two field pictures in frequency are matched, and determine whether identified region is the same object in two field pictures.
Further, the differentiation flame algorithm in this method of the invention method particularly includes:
Step 1: saving data in real time: the infrared image that infrared camera extracts is calculated by the identification of doubtful flame region
Method and profile scan algorithm obtain the doubtful flame region on image after calculating, and each numbering area is acquired by two algorithms
To real time data be put into caching, data include doubtful flame region average brightness, bottom position, height and its variation,
Perimeter and its variation, area and its variation, wedge angle number and its variation and circularity；These data all will be later as differentiation
Flame discriminant information, wherein the average height μ of doubtful flame region are as follows:
Wherein, HiFor the height value in same doubtful pyrotechnics region in each frame image in continuous several frames；
The standard deviation σ of height are as follows:
The standard rate δ of height are as follows:
δ=σ/μ
Doubtful flame region perimeter, area standard rate calculation method it is identical as the calculating of Height Standard rate, point
It is not replaced with perimeter value Li, the height value Si of same doubtful flame region in each frame image；
The round value C of doubtful flame region nnCalculation method are as follows:
Wherein Sn、LnFor the area and perimeter in the doubtful fireworks region；
The data of each number acquired above are divided into buffer zone small one by one in buffer area by number；
Step 2: handling in real time data cached: the content in will be data cached in real time, which be calculated, to be used to differentiate
Data content；When operation, for the cache contents of number N, while take out correspond in each criterion it is data cached into
Row operation；In calculating process, data mean value, the mark of buffer area are corresponded to according to each criterion that the demand of criterion calculates number N
Quasi- difference or standard rate, are differentiated for step 3；
Step 3: flame differentiates: using the data of step 2 acquisition for numbering the region for being N, gone out according to threshold decision
The type of flame, including nonflame, stable three kinds of flame conditions of small fire and unstable high fire；When there is unstable high fire, hair
Flame is alarmed out, while marking the different types of fire conditions occurred in different regions.
Further, the method for flame pattern judgment threshold is set in this method of the invention specifically:
The selection of threshold value is according to extraneous every interference source, including candle, lighter, newspaper, as stable and unstable fire
The characteristic parameter in source carries out experimental data statistics, and summary data forms threshold decision flame pattern.
Further, visible images distinguished number in this method of the invention method particularly includes:
Doubtful flame region is obtained under infrared image, and by under the area maps to visible images, is then passed through
YGrCb color criterion carries out binary conversion treatment to the visible images in the region using hysteresis binarization method；
Y (x, y) > Cb (x, y)
Cr (x, y) > Cb (x, y)
Cr(x,y)-Cb(x,y)|≥τ
Wherein Y (x, y), Cb (x, y) and Cr (x, y) are the brightness of the point on position (x, y), blue color difference and red respectively
Color value of chromatism；τ is constant.
The beneficial effect comprise that: the remote fire detection system and method for the invention based on video image,
Using two-waveband video acquisition technique and image processing techniques, infrared and visible light visual field information is comprehensively utilized, fire is improved
Calamity discrimination and anti-interference ability.Whole system performance under the action of several modules is more efficient, can be more acurrate, rapidly
Fire is detected and is alarmed.Efficiently solving traditional fire identification technology cannot put down in sensitivity and reliability
Weighing apparatus problem proposes a kind of new flame identification algorithm, improves the accuracy of flame identification algorithm, realizes under most environment
The generation that fire is detected by incipient fire both image change characteristics provides certain effective to the detection of large space remote fire
Method.
Detailed description of the invention
Present invention will be further explained below with reference to the attached drawings and examples, in attached drawing:
Fig. 1 is the remote fire detection system structural schematic diagram the present invention is based on video image.
Fig. 2 is the remote fire detection system recognition methods flow chart the present invention is based on video image.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that described herein, specific examples are only used to explain the present invention, not
For limiting the present invention.
As shown in Figure 1, the remote fire detection system based on video image of the embodiment of the present invention, system includes: figure
As acquisition module, control module, network communication module and wireless remote alarming module.The image capture module passes through double wave
Section video capture technology, the characteristics of being imaged according to infrared camera, are sent into control module by Video Decoder, and control module is logical
The feature extraction that DSP realizes acquisition image is crossed, the characteristics of image extracted is subjected to flame identification, recognition result is uploaded to far
Process control center, network communication module are connect by network interface with physical layer transceiver, if fire occurs, remote control center
Control command is assigned, by network communication module, realizes that ethernet communication function, wireless remote alarming module realize voice transfer
And messaging, serial port communication thread is run immediately using the variation of alarm switch, realizes wireless network remote alarming function.
Two-waveband video fire detector is put into the region that fire will occur, acquires two-waveband video.
Referring to fig. 2, two-waveband video acquisition technique is used in step 1, since the area of more infrared rays can be reflected at the scene
Brightness of the domain on infrared camera image is high, according to this advantage, passes through certain threshold value position of high brightness on the image
Pixel is formed, the region high to brightness height, that is, temperature carries out feature extraction.Wherein, the color of each pixel can use Rn
(red), Gn(green), Bn(blue) three components indicate, the average brightness T of that pixelnAre as follows:
Tn=0.22 × Rn+0.587×Gn+0.114×Bn
Then pixel value is subjected to hysteresis binaryzation to avoid the pixel value in video from generating runout error, be set higher than
Limit is target point, is non-targeted point lower than lower limit, is the pixel of previous frame among bound.
In step 2, after image carries out binaryzation, many high temperature pixels are obtained, high temperature pixel is formed on the image
High temperature pixel region.Image is scanned by profile scan algorithm from left to bottom right, if encountering the high temperature of doubtful flame
Pixel then starts to carry out profile scan, successively scans according to up, down, left and right four directions four field clockwise, scanning process
In if it find that new high temperature pixel, then new high temperature pixel is considered as current pixel, and by its transverse and longitudinal coordinate
In information deposit caching, and the moving direction of the pixel at this time is recorded, using its opposite direction as the beginning starting point of scanning.If no
It was found that new pixel, then scan to opposite direction, restart four field scanning modes of current pixel point, until side is arrived in scanning
Boundary's starting point, the number that the surface sector scanning terminates to record the high-temperature area is 1, then carries out next sector scanning in order
Number.
In step 3, judge whether there is doubtful flame region, if there is no restart obtain infrared image, if
In the presence of progress step 4.
In step 4, the high temperature suspicious region of present frame is matched with the high temperature suspicious region of former frame, if two frames
It is mutually matched, then the identiflication number of former frame corresponding region is transformed into current region.If not being mutually matched, just freeze the area
The scanning in domain is numbered, and the smallest and no frozen and used number, the scanning as the region are selected from number pond
Number.
In step 5, which mainly obtains image by flame pixels extraction algorithm and profile scan algorithm doubtful
Flame region information, including boundary information and region confidence etc. carry out calculate extract characteristic value, because of doubtful flame region
In include identification pixel, it may be possible to true flame, it is also possible to meet the object of binaryzation alternative condition, simultaneously as
The performance of no flame, small fire and high fire in characteristic value also has the difference of significant difference, so can identify fire by characteristic value
The state of flame.In addition, there is provision of the numbering area prestored a caching.
Calculated flame characteristic value is handled, treatment process is as follows:
1, save data in real time: the image that camera extracts is calculated by doubtful flame pixels extraction algorithm and profile scan
Method obtains the doubtful flame region on image after calculating, and each numbering area is put by the collected real time data of two algorithms
Enter in caching, data include average brightness, bottom position, height and its variation, perimeter and its variation, face of doubtful flame region
Product and its variation, wedge angle number and its variation and circularity etc..These data all will differentiate letter as differentiation flame later
Breath.The wherein average height μ of doubtful flame region are as follows:
Wherein, HiFor the height value in same doubtful pyrotechnics region in each frame image in continuous several frames.
The standard deviation σ of height are as follows:
The standard rate δ of height are as follows:
δ=σ/μ
Doubtful flame region perimeter, area standard rate calculation method with Height Standard rate, respectively with each frame figure
Perimeter value Li, the height value Si of same doubtful flame region are replaced as in.
The round value C of doubtful flame region nnCalculation method are as follows:
Wherein Sn、LnFor the area and perimeter in the doubtful fireworks region.
The data of each number acquired above will be divided into buffer zone small one by one in buffer area by number.
2, handle in real time data cached: the content in will be data cached in real time carries out that the data for being used for differentiating are calculated
Content.When operation, for the cache contents of number N, while the data cached carry out operation corresponded in each criterion is taken out.
In calculating process, according to each criterion that the demand of criterion calculates number N correspond to the data mean value of buffer area, standard deviation or
Standard rate, is differentiated for S3.
3, flame differentiates: using the data of S2 acquisition for numbering the region for being N, goes out flame according to certain threshold decision
Type, that is, nonflame, stablize three kinds of flame conditions of small fire and unstable high fire.When there is unstable high fire, flame report is issued
It is alert, while console marks the different types of fire conditions occurred in different regions.It, can basis about the constituency of threshold value
The characteristic parameter of extraneous items interference source such as candle, lighter, newspaper etc. stable and unstable fire source carries out experimental data
Statistics, summary data form threshold decision flame kenel.
In step 8, the visible images of doubtful flame region are extracted, doubtful flame region is obtained under infrared image, and
By under the area maps to visible images, then pass through YGrCb color criterion, using hysteresis binarization method in the region
Visible images carry out binary conversion treatment.
Y (x, y) > Cb (x, y)
Cr (x, y) > Cb (x, y)
Cr(x,y)-Cb(x,y)|≥τ
Wherein Y (x, y), Cb (x, y) and Cr (x, y) are the brightness of the point on position (x, y), blue color difference and red respectively
Color value of chromatism；τ is a constant.
In step 10, by profile scan algorithm, the position size information of suspicious region under visible images is obtained, by it
Area information is compared with the suspicious region under corresponding infrared image, judges whether it is real flame.Finally, by flame zone
Field mark comes out and shows, records t at the time of can becoming totally visible flame in the video image of video display devices1。
In step 14, flame indicator light is flashed when alarming with the frequency of 2Hz, the police instruction of warning device
Lamp is also flashed with the frequency of 2Hz, and issues high frequency alarm.Circularity and fire angle number of variations criterion are used simultaneously
To flame is stablized and unstable flame is distinguished, rudimentary alarm is carried out to fire source is stablized, advanced report is carried out to unstable fire source
It is alert, realize the classifying alarm to fire.The red-label frame moment t of record video image Flame at this time2.Record t3=t2-
t1.Repeat experiment three times.It calculates and records longest time of fire alarming and average σ-K width.
It can be found that the time of fire alarming of fire detecting system designed by the present invention is significantly shorter than traditional Detection Techniques
Time of fire alarming.
To sum up, the present invention establishes a kind of novel remote fire detection system.Using two-waveband video fire detector
With two-waveband video acquisition and image processing techniques, infrared and visible light field data is comprehensively utilized, algorithmically flame
Recognizer calculating speed is obviously faster than recognizer, present system guarantees that the accuracy and reliability of fire detection, to steady
Determine flame and unstable flame carries out classifying alarm, reaction speed is rapid, has strong anti-interference ability.
It should be understood that for those of ordinary skills, it can be modified or changed according to the above description,
And all these modifications and variations should all belong to the protection domain of appended claims of the present invention.
Claims (8)
1. a kind of remote fire detection system based on video image characterized by comprising image capture module, control mould
Block, network communication module, wireless remote alarming module and remote control center；Wherein:
Image capture module, including infrared camera and Video Decoder, infrared camera pass through two-waveband video acquisition technique
The characteristics of acquiring infrared image, and being imaged according to infrared camera, is sent into control module by Video Decoder for infrared image；
Control module realizes feature extraction to infrared image by DSP, by recognizer to the characteristics of image extracted into
Recognition result is uploaded to remote control center by row flame identification；
Network communication module, including physical layer transceiver and network interface, control module pass through the object inside network communication module
Reason layer transceiver and network interface are connected with remote control center, if fire occurs, remote control center assigns control command, lead to
Network communication module is crossed, realizes ethernet communication function；
Wireless remote alarming module, is connected by way of wireless connection with control module, realizes wireless network remote alarms.
2. the remote fire detection system according to claim 1 based on video image, which is characterized in that Image Acquisition mould
The method of image is obtained in block specifically:
Infrared image is acquired by two-waveband video acquisition technique, by the way that threshold value is arranged, position of high brightness is formed on the image
Pixel, the region high to brightness height, that is, temperature carry out feature extraction；Wherein, the color of each pixel can use RnRed, Gn
Green, BnBlue three components indicate, the average brightness T of the pixelnAre as follows:
Tn=0.22 × Rn+0.587×Gn+0.114×Bn
Pixel value is subjected to hysteresis binaryzation to avoid the pixel value in video from generating runout error, being set higher than the upper limit is mesh
Punctuate, is non-targeted point lower than lower limit, is the pixel of previous frame among bound.
3. a kind of remote fire detection side of the remote fire detection system using described in claim 1 based on video image
Method, which is characterized in that method includes the following steps:
S1, video infrared image is obtained；
S2, binary conversion treatment is carried out to infrared image, obtains multiple high temperature pixels, high temperature pixel is formed on infrared image
High temperature suspicious region；
S3, by doubtful flame region recognizer, judge whether there is doubtful flame region；If it exists, step S4 is executed；If
It is not present, reacquires video infrared image；
S4, pass through profile scan algorithm, doubtful flame region is scanned；
S5, flame characteristic extraction is carried out, obtains doubtful flame region information；
S6, according to doubtful flame region information, doubtful flame region is judged again, if being confirmed as doubtful flame region,
Execute step S7；If it is not, then reacquiring video infrared image；
S7, be confirmed after doubtful flame region；
S8, pass through visible images distinguished number, compare doubtful flame region and extract visible images；
S9, binary conversion treatment is carried out to visible images；
S10, profile scan algorithm is executed to the visible images after binary conversion treatment；
S11, the flame characteristic information extracted；
S12, flame is judged whether it is by differentiating flame algorithm according to flame characteristic information；If flame, step is executed
S13；If not flame, reacquires video infrared image；
S13, flame pattern judgement is carried out to flame, judges whether it is stable flame；
S14, if flame is stablized, execute rudimentary alarm；If unstable flame, advanced alarm is executed.
4. the remote fire detection method according to claim 3 based on video image, which is characterized in that in this method
Profile scan algorithm method particularly includes:
After image carries out binaryzation, the panel region that high temperature pixel is linked to be is obtained；Profile scan algorithm is from the upper left of picture
It is scanned in order to bottom right, obtains the position that high temperature pixel is linked to be the boundary point in region, and it is slow to save it in boundary position
In depositing, the calculating for characteristic value below is prepared；Boundary length and the inside for calculating the high-temperature area scanned simultaneously are all
The upper left corner of pixel and the coordinate value of bottom right angle point are stored in boundary position caching respectively and neutralize in scanning area caching；Often
It is secondary be scanned after, each high-temperature area can carry out zone number since 1, which is scanning number.
5. the remote fire detection method according to claim 4 based on video image, which is characterized in that in this method
Doubtful flame region recognizer method particularly includes:
The high temperature suspicious region of present frame is matched with the high temperature suspicious region of former frame, it, will if two frames are mutually matched
The identiflication number of former frame is transferred in present frame；If not being mutually matched, just freeze the scanning number in the region, from number pond
Select scanning number the smallest and without frozen and used number, as the region；It will be in video by the algorithm
Two field pictures matched, determine whether identified region is the same object in two field pictures.
6. the remote fire detection method according to claim 3 based on video image, which is characterized in that in this method
Differentiate flame algorithm method particularly includes:
Step 1: saving data in real time: the infrared image that infrared camera extracts by doubtful flame region recognizer and
Profile scan algorithm obtains the doubtful flame region on image after calculating, and each numbering area is collected by two algorithms
Real time data is put into caching, and data include average brightness, bottom position, height and its variation of doubtful flame region, perimeter
And its variation, area and its variation, wedge angle number and its variation and circularity；These data all will be later as differentiation flame
Discriminant information, wherein the average height μ of doubtful flame region are as follows:
Wherein, HiFor the height value in same doubtful pyrotechnics region in each frame image in continuous several frames；
The standard deviation σ of height are as follows:
The standard rate δ of height are as follows:
δ=σ/μ
Doubtful flame region perimeter, area standard rate calculation method it is identical as the calculating of Height Standard rate, respectively with
The perimeter value Li of same doubtful flame region, height value Si are replaced in each frame image；
The round value C of doubtful flame region nnCalculation method are as follows:
Wherein Sn、LnFor the area and perimeter in the doubtful fireworks region；
The data of each number acquired above are divided into buffer zone small one by one in buffer area by number；
Step 2: handling in real time data cached: the content in will be data cached in real time carries out that the number for being used for differentiating is calculated
According to content；When operation, for the cache contents of number N, while taking out correspond in each criterion data cached and being transported
It calculates；In calculating process, the data mean value of buffer area, standard deviation are corresponded to according to each criterion that the demand of criterion calculates number N
Or standard rate, differentiated for step 3；
Step 3: flame differentiates: using the data of step 2 acquisition for numbering the region for being N, go out flame according to threshold decision
Type, including nonflame, stablize three kinds of flame conditions of small fire and unstable high fire；When there is unstable high fire, fire is issued
Flame alarm, while marking the different types of fire conditions occurred in different regions.
7. the remote fire detection method according to claim 6 based on video image, which is characterized in that set in this method
The method for setting flame pattern judgment threshold specifically:
The selection of threshold value is according to extraneous every interference source, including candle, lighter, newspaper, as stable and unstable fire source
Characteristic parameter carries out experimental data statistics, and summary data forms threshold decision flame pattern.
8. the remote fire detection method according to claim 3 based on video image, which is characterized in that can in this method
Light-exposed image discriminating algorithm method particularly includes:
Doubtful flame region is obtained under infrared image, and by under the area maps to visible images, then passes through YGrCb face
Color criterion carries out binary conversion treatment to the visible images in the region using hysteresis binarization method；
Y (x, y) > Cb (x, y)
Cr (x, y) > Cb (x, y)
Cr(x,y)-Cb(x,y)|≥τ
Wherein Y (x, y), Cb (x, y) and Cr (x, y) are brightness, blue color difference and the red color of the point on position (x, y) respectively
Difference；τ is constant.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811386160.9A CN109147259B (en) | 2018-11-20 | 2018-11-20 | Remote fire detection system and method based on video image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811386160.9A CN109147259B (en) | 2018-11-20 | 2018-11-20 | Remote fire detection system and method based on video image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109147259A true CN109147259A (en) | 2019-01-04 |
| CN109147259B CN109147259B (en) | 2021-10-08 |
Family
ID=64806103
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811386160.9A Active CN109147259B (en) | 2018-11-20 | 2018-11-20 | Remote fire detection system and method based on video image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109147259B (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109801470A (en) * | 2019-03-22 | 2019-05-24 | 京东方科技集团股份有限公司 | Alert device and monitoring and alarming system |
| CN109872493A (en) * | 2019-01-28 | 2019-06-11 | 上海得舟信息科技有限公司 | A kind of flame location detection device based on video image processing |
| CN109919120A (en) * | 2019-03-15 | 2019-06-21 | 江苏鼎集智能科技股份有限公司 | A flame detection method based on near-infrared spectral imaging |
| CN110244011A (en) * | 2019-06-26 | 2019-09-17 | 熊颖郡 | The river blowdown of unmanned plane monitors analyzing and alarming system automatically |
| CN110320843A (en) * | 2019-07-19 | 2019-10-11 | 云南北方驰宏光电有限公司 | Intelligent building security protection monitoring system and method based on double spectrum |
| CN110363952A (en) * | 2019-08-20 | 2019-10-22 | 宝钢湛江钢铁有限公司 | A kind of contactless early warning system of fire-fighting early stage |
| CN110652685A (en) * | 2019-10-17 | 2020-01-07 | 广东德臻消防机电工程有限公司 | Indoor fire extinguishing method and system |
| CN111739248A (en) * | 2020-06-11 | 2020-10-02 | 湖北美和易思教育科技有限公司 | Artificial intelligent Internet of things security system and control method |
| CN111899458A (en) * | 2020-07-27 | 2020-11-06 | 山东工商学院 | Artificial intelligence-based fire smoke image identification method |
| CN111899460A (en) * | 2020-07-27 | 2020-11-06 | 山东工商学院 | Remote fire detection system and method based on video image |
| CN111914689A (en) * | 2020-07-15 | 2020-11-10 | 营口新山鹰报警设备有限公司 | Flame identification method of image type fire detector |
| CN112287849A (en) * | 2020-10-30 | 2021-01-29 | 武汉理工光科股份有限公司 | Fire early warning method and device for high-rise building |
| CN113076797A (en) * | 2021-02-24 | 2021-07-06 | 江苏濠汉信息技术有限公司 | Charging station electric vehicle fire alarm method and system based on intelligent video identification |
| CN113158719A (en) * | 2020-11-30 | 2021-07-23 | 齐鲁工业大学 | Image identification method for fire disaster of photovoltaic power station |
| CN113516091A (en) * | 2021-07-27 | 2021-10-19 | 福建工程学院 | A method for identifying electrical spark images in substations |
| CN114821289A (en) * | 2022-01-17 | 2022-07-29 | 电子科技大学 | Forest fire picture real-time segmentation and fire edge point monitoring algorithm |
| CN116091959A (en) * | 2022-11-21 | 2023-05-09 | 武汉坤达安信息安全技术有限公司 | Double-light linkage identification method and device based on all-weather smoke and fire |
| US11651670B2 (en) | 2019-07-18 | 2023-05-16 | Carrier Corporation | Flame detection device and method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6057549A (en) * | 1996-07-31 | 2000-05-02 | Fire Sentry Corporation | Fire detector with multi-level response |
| CN101908142A (en) * | 2010-08-04 | 2010-12-08 | 丁天 | Feature analysis-based video flame detecting method |
| CN202306757U (en) * | 2011-06-21 | 2012-07-04 | 沈阳天目科技有限公司 | Embedded flame detection device based on double-waveband |
| CN204155388U (en) * | 2014-11-07 | 2015-02-11 | 莆田学院 | A kind of two-waveband video double-core fire detector and fore device thereof |
| CN105512667A (en) * | 2014-09-22 | 2016-04-20 | 中国石油化工股份有限公司 | Method for fire identification through infrared and visible-light video image fusion |
| CN105788142A (en) * | 2016-05-11 | 2016-07-20 | 中国计量大学 | Video image processing-based fire detection system and detection method |
| CN107123227A (en) * | 2017-07-06 | 2017-09-01 | 合肥科大立安安全技术股份有限公司 | A kind of embedded image flame detector and its recognition methods based on two waveband |
-
2018
- 2018-11-20 CN CN201811386160.9A patent/CN109147259B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6057549A (en) * | 1996-07-31 | 2000-05-02 | Fire Sentry Corporation | Fire detector with multi-level response |
| CN101908142A (en) * | 2010-08-04 | 2010-12-08 | 丁天 | Feature analysis-based video flame detecting method |
| CN202306757U (en) * | 2011-06-21 | 2012-07-04 | 沈阳天目科技有限公司 | Embedded flame detection device based on double-waveband |
| CN105512667A (en) * | 2014-09-22 | 2016-04-20 | 中国石油化工股份有限公司 | Method for fire identification through infrared and visible-light video image fusion |
| CN204155388U (en) * | 2014-11-07 | 2015-02-11 | 莆田学院 | A kind of two-waveband video double-core fire detector and fore device thereof |
| CN105788142A (en) * | 2016-05-11 | 2016-07-20 | 中国计量大学 | Video image processing-based fire detection system and detection method |
| CN107123227A (en) * | 2017-07-06 | 2017-09-01 | 合肥科大立安安全技术股份有限公司 | A kind of embedded image flame detector and its recognition methods based on two waveband |
Non-Patent Citations (1)
| Title |
|---|
| 刘媛珺: "双波段野外火灾图像识别及目标定位方法研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109872493A (en) * | 2019-01-28 | 2019-06-11 | 上海得舟信息科技有限公司 | A kind of flame location detection device based on video image processing |
| CN109919120A (en) * | 2019-03-15 | 2019-06-21 | 江苏鼎集智能科技股份有限公司 | A flame detection method based on near-infrared spectral imaging |
| CN109801470A (en) * | 2019-03-22 | 2019-05-24 | 京东方科技集团股份有限公司 | Alert device and monitoring and alarming system |
| CN110244011A (en) * | 2019-06-26 | 2019-09-17 | 熊颖郡 | The river blowdown of unmanned plane monitors analyzing and alarming system automatically |
| US11651670B2 (en) | 2019-07-18 | 2023-05-16 | Carrier Corporation | Flame detection device and method |
| CN110320843A (en) * | 2019-07-19 | 2019-10-11 | 云南北方驰宏光电有限公司 | Intelligent building security protection monitoring system and method based on double spectrum |
| CN110363952A (en) * | 2019-08-20 | 2019-10-22 | 宝钢湛江钢铁有限公司 | A kind of contactless early warning system of fire-fighting early stage |
| CN110652685A (en) * | 2019-10-17 | 2020-01-07 | 广东德臻消防机电工程有限公司 | Indoor fire extinguishing method and system |
| CN111739248A (en) * | 2020-06-11 | 2020-10-02 | 湖北美和易思教育科技有限公司 | Artificial intelligent Internet of things security system and control method |
| CN111739248B (en) * | 2020-06-11 | 2022-04-01 | 湖北美和易思教育科技有限公司 | Artificial intelligent Internet of things security system and control method |
| CN111914689A (en) * | 2020-07-15 | 2020-11-10 | 营口新山鹰报警设备有限公司 | Flame identification method of image type fire detector |
| CN111914689B (en) * | 2020-07-15 | 2024-03-22 | 营口新山鹰报警设备有限公司 | Flame identification method of image type fire detector |
| CN111899458A (en) * | 2020-07-27 | 2020-11-06 | 山东工商学院 | Artificial intelligence-based fire smoke image identification method |
| CN111899460A (en) * | 2020-07-27 | 2020-11-06 | 山东工商学院 | Remote fire detection system and method based on video image |
| CN112287849A (en) * | 2020-10-30 | 2021-01-29 | 武汉理工光科股份有限公司 | Fire early warning method and device for high-rise building |
| CN113158719A (en) * | 2020-11-30 | 2021-07-23 | 齐鲁工业大学 | Image identification method for fire disaster of photovoltaic power station |
| CN113076797A (en) * | 2021-02-24 | 2021-07-06 | 江苏濠汉信息技术有限公司 | Charging station electric vehicle fire alarm method and system based on intelligent video identification |
| CN113076797B (en) * | 2021-02-24 | 2022-01-18 | 江苏濠汉信息技术有限公司 | Charging station electric vehicle fire alarm method and system based on intelligent video identification |
| CN113516091A (en) * | 2021-07-27 | 2021-10-19 | 福建工程学院 | A method for identifying electrical spark images in substations |
| CN113516091B (en) * | 2021-07-27 | 2024-03-29 | 福建工程学院 | Method for identifying electric spark image of transformer substation |
| CN114821289A (en) * | 2022-01-17 | 2022-07-29 | 电子科技大学 | Forest fire picture real-time segmentation and fire edge point monitoring algorithm |
| CN114821289B (en) * | 2022-01-17 | 2023-10-17 | 电子科技大学 | A real-time segmentation and fire edge point monitoring algorithm for forest fire pictures |
| CN116091959A (en) * | 2022-11-21 | 2023-05-09 | 武汉坤达安信息安全技术有限公司 | Double-light linkage identification method and device based on all-weather smoke and fire |
| CN116091959B (en) * | 2022-11-21 | 2024-03-22 | 武汉坤达安信息安全技术有限公司 | Double-light linkage identification method and device based on all-weather smoke and fire |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109147259B (en) | 2021-10-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109147259A (en) | A kind of remote fire detection system and method based on video image | |
| CN111091072A (en) | YOLOv 3-based flame and dense smoke detection method | |
| WO2020098195A1 (en) | Ship identity recognition method based on fusion of ais data and video data | |
| CN104599427B (en) | A kind of intelligent image type fire alarm system for vcehicular tunnel | |
| CN105788142B (en) | A kind of fire detection system and detection method based on Computer Vision | |
| KR101822924B1 (en) | Image based system, method, and program for detecting fire | |
| CN107085714B (en) | A Video-Based Forest Fire Detection Method | |
| CN113393486B (en) | Abnormal event monitoring method, intelligent monitoring terminal and system | |
| CN101393603A (en) | A method for identifying and detecting fire flames in tunnels | |
| CN101751744A (en) | Detection and early warning method of smoke | |
| CN105306892B (en) | A kind of generation of ship video of chain of evidence form and display methods | |
| CN109635750A (en) | A kind of compound convolutional neural networks images of gestures recognition methods under complex background | |
| CN105139429A (en) | Fire detecting method based on flame salient picture and spatial pyramid histogram | |
| CN108230607B (en) | An image fire detection method based on regional feature analysis | |
| CN101316371A (en) | Flame detection method and device | |
| CN112633292A (en) | Method for measuring temperature of oxide layer on metal surface | |
| CN109741565A (en) | Coal mine fire identification system and method | |
| Tao et al. | Smoky vehicle detection based on range filtering on three orthogonal planes and motion orientation histogram | |
| CN102646191B (en) | Method applied to recognition of flame image generated by gas combustion associated in oil drilling | |
| CN112613483A (en) | Outdoor fire early warning method based on semantic segmentation and recognition | |
| CN114283367B (en) | Artificial intelligent open fire detection method and system for garden fire early warning | |
| CN111539264A (en) | A kind of ship flame detection and positioning system and detection and positioning method | |
| CN117132949B (en) | An all-weather fall detection method based on deep learning | |
| CN101540891A (en) | Luggage delivery warehouse human body detecting system based on monitoring video | |
| CN109028234B (en) | Range hood capable of identifying smoke grade |
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 |