CN104598895B - A kind of flame detecting method and device based on video image analysis - Google Patents
A kind of flame detecting method and device based on video image analysis Download PDFInfo
- Publication number
- CN104598895B CN104598895B CN201510068450.9A CN201510068450A CN104598895B CN 104598895 B CN104598895 B CN 104598895B CN 201510068450 A CN201510068450 A CN 201510068450A CN 104598895 B CN104598895 B CN 104598895B
- Authority
- CN
- China
- Prior art keywords
- image
- suspicious region
- flame
- threshold value
- region
- 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.)
- Active
Links
Landscapes
- Closed-Circuit Television Systems (AREA)
- Fire-Detection Mechanisms (AREA)
Abstract
A kind of flame detecting method and device based on video image analysis, the video image of monitor area is obtained by video image acquiring unit first and current frame image type is judged by image information judging unit, subsequent treatment can use different processing modes according to image type difference；Then extract doubtful flame region on current frame image using suspicious region extraction unit and ineligible false areas is rejected by suspicious region screening unit；Afterwards using match updating block to the suspicious region preserved after screening carry out matching renewal operation；Alarm unit is finally judged according to whether the matching times of each history suspicious region reach threshold value to judge whether the suspicious region occurs the condition of a fire.The present invention classifies and judged according to the different characteristic that night on daytime flame is shown, and greatly improves alarm accuracy rate.
Description
Technical field
The present invention relates to computer vision field, more particularly, to the fire defector applied to field of video monitoring.
Background technology
Fire is always to endanger one of most important multiple disaster of people's security of the lives and property.With social productive forces
Development, the wealth of society increasingly increase, and economic loss caused by fire and casualties are in expansion trend.And people are on fire
Research in terms of detection also never stops.Up to the present most widely used mainly temperature sensitive type fire detector and sense cigarette type
Fire detector.
Temperature sensitive type fire detector and smoke-sensitive fire detector are the temperature and smokescope by perceiving flame periphery,
Compared according to the information content perceived with threshold value to judge whether alert.Two thus be present, when by
In can not accurately provide warning message by the single feature of flame, wrong report is more；Second, detection range is limited, it is high indoors
Large space and outdoor spacious place, which can not play, should effect.And with the fast development of social economy, there is large space
Build increasing, the needs of conventional fire detector is difficult in terms of meeting it for fire prevention detection.
In recent years, opened along with the fast development of computer vision and image processing techniques, increasing researcher
Beginning sets about studying the flame detecting based on video analysis.But it have been found that the image-type detection skill based on this technology
Art is not widely applied to our daily life, and its reason is that achievement in research at this stage is difficult to meet us
Requirement for the real-time of early stage condition of a fire detection, validity and accuracy.
The content of the invention
It is an object of the invention to propose a kind of flame detecting method and device based on video image analysis.The present invention is logical
Alarm can fast and effectively be provided when early stage occurs for the condition of a fire by crossing the means of video analysis.In order to achieve the above object, this hair
It is bright to use following technological means：
A kind of flame detecting device based on video image analysis, including such as lower unit：
Video image acquiring unit, for obtaining the video image information of monitor area；
Image information judging unit, for detecting current frame image type, judge current frame image be coloured image or
Gray level image, then recorded if gray level image and continuously gray level image frame number occur；
Suspicious region extraction unit, it is different according to image type for extracting the flame suspicious region of current frame image, adopt
Different feature extraction modes is taken, if present image is coloured image, it is doubtful that the color property according to flame extracts flame
Region, go out flame suspicious region according to the highlighted feature extraction of flame if present image is gray level image；
Suspicious region screening unit, for excluding ineligible flame suspicious region, preserve qualified doubtful
Region relevant parameter is as historical data；
Updating block is matched, for present frame flame suspicious region to be carried out into matching operation with history suspicious region and updated
The relevant parameter of history suspicious region；
Alarm unit is judged, for judging whether each history suspicious region meets final alert if.
A kind of flame detecting method based on video image analysis, comprises the following steps：
First step, the video image information in monitor area is obtained using image collecting device；
Second step, current frame image type is detected, judge that current frame image is coloured image or gray level image, if
Gray level image, which then records, continuously there is gray level image frame number；
Third step, the flame suspicious region of current frame image is extracted, it is different according to image type, take different features
Extracting mode, if present image is coloured image, the color property according to flame extracts flame suspicious region, if current figure
Go out flame suspicious region as then highlighting feature extraction according to flame for gray level image；
Four steps, ineligible flame suspicious region is excluded, preserve qualified suspicious region relevant parameter
As historical data；
5th step, present frame flame suspicious region is subjected to matching operation with history suspicious region and more new historical is doubtful
The relevant parameter in region；
6th step, judges whether each history suspicious region meets final alert if.
The beneficial effects of the invention are as follows：Day and night, the difference of the feature performance of flame, the present invention gather during the day
Image is coloured image, night collection for gray level image, different feature extracting methods is carried out according to different image types,
Greatly improve the accuracy rate of detection.
Brief description of the drawings
The shared accompanying drawing six of the present invention is opened：
Fig. 1 shows the schematic block diagram of the flame detecting device based on video image analysis according to the present invention；
Fig. 2 is shown according to the flame detecting device suspicious region extraction unit based on video image analysis of the invention
Schematic block diagram；
Fig. 3 shows the signal of the matching updating block of the flame detecting device based on video image analysis according to the present invention
Block diagram；
Fig. 4 shows the overall flow figure of the flame detecting method based on video image analysis according to the present invention；
Fig. 5 shows the third step flow chart of the flame detecting method based on video image analysis according to the present invention；
Fig. 6 shows the 5th flow chart of steps of the flame detecting method based on video image analysis according to the present invention.
Embodiment
With reference to the accompanying drawings and the present invention is described in detail with reference to instantiation.The 352 of embodiment use ×
The image of 288 pixel sizes.On daytime, collection for YUV coloured images, perceive Lighting information amount to dusk photo resistance and be less than
During threshold value, infrared light compensating lamp receives open signal, now gather for gray level image.It should be noted that described example is only
It is for the ease of the understanding of the present invention, is not intended to limit the scope of the present invention.
Fig. 1 shows the schematic block diagram of the flame detecting device based on video image analysis according to the present invention.Following institute
Stating device each unit can all realize separately through microprocessor, but can also have high-performance using one to be cost-effective
CPU processor is realized.As shown in figure 1, include according to the flame detecting device based on video image analysis of the present invention：
Part I, video image acquiring unit 101, utilize the video figure of video image acquisition device acquisition monitor area
As information, image collecting device shines according to ambient light used by this embodiment intensity gathers the type of image to change,
Collected during daytime for coloured image, night collect for gray level image.
Part II, image information judging unit 102, judge that the image that video image acquiring unit 101 gathers is colored
Image or gray level image.Then recorded if gray level image and continuously gray level image frame number occur.
Part III, suspicious region extraction unit 103 is different according to the judged result of image information judging unit 102, uses
Different feature extraction modes, extracts present frame flame suspicious region, and concrete composition is as shown in Figure 2：
Feature point extraction unit 1031, for extracting scattered flame characteristic point under present image；Region connection unit 1032,
The characteristic point mutually adjoined for feature point extraction unit 1031 to be extracted carries out region connection；Parameter calculation unit
1033, for calculating the relevant parameter of the flame suspicious region after connecting.
Part IV, suspicious region screening unit 104, it is not inconsistent for exclude that suspicious region extraction unit 103 extracted
The flame suspicious region of conjunction condition, qualified suspicious region relevant parameter is preserved as historical data.
Part V, updating block 105 is matched, present frame flame suspicious region and history suspicious region are subjected to matching behaviour
Make and more the relevant parameter of new historical suspicious region, concrete composition are as shown in Figure 3：
Suspicious region morphological feature matching unit 1051, the morphological feature according to flame to current flame suspicious region and are gone through
History suspicious region carries out matching operation；Suspicious region temporal signatures matching unit 1052, the temporal signatures according to flame are to current
Flame suspicious region and history suspicious region carry out matching operation；Suspicious region motion feature matching unit 1053, according to flame
Motion feature matching operation is carried out to current flame suspicious region and history suspicious region.
Part VI, alarm unit 106 is judged, according to matching updating block 105 to each history flame suspicious region
The comparison of frame number and threshold value is matched, to judge whether the suspicious region is flame region.
Fig. 4 shows the overall flow figure of the flame detecting method based on video image analysis according to the present invention, such as Fig. 4
Shown, overall flow of the present invention is divided into 6 steps：
Step 201, the video image information in monitor area is obtained by image collecting device.The present invention exists according to flame
The feature that daytime and night show is different, and obtains video image information by the way of corresponding.Daytime, flame colour was believed
Breath is abundant, and image collecting device obtains the color image information of monitor area；Night flame is relative to other object brightness information
More prominent, image collecting device obtains the gray level image information of monitor area.
Step 202, the acquired image type of judgment step 201, then recorded if gray level image and continuously gray-scale map occur
The frame number of picture, when the numerical value of record is more than threshold value T1, then show that present image harvester starts stable collection gray-scale map
Picture.Its given threshold T1 spans are [10,30], it is therefore an objective to prevent other factors interference image collecting device from causing collection
A certain two field picture is gray level image, and the setting of the threshold value ensures that image collecting device starts stable collection gray level image.
Step 203, the result judged according to step 202, the doubtful flame zone of present frame is extracted in different ways
Domain.
As shown in figure 5, step 203 comprises the steps of：
Step 2031, scattered doubtful flame characteristic point is extracted according to frame difference method.
If step 202 judges present image for coloured image, YUV image is first converted into rgb format image, then according to fire
The color feature extracted of flame goes out to meet the characteristic point of condition, and feature point extraction condition is C1：1、R>G>B；2、R>220.It will meet
The pixel of condition is labeled as 1, and the point for being unsatisfactory for condition is labeled as 0, the current frame image marked is preserved, with previous frame mark
The image progress frame remembered is poor, seeks out scattered doubtful flame characteristic point, frame difference condition is：In two frame correspondence positions, mark
It is different.The point for meeting frame difference condition is marked on current frame image correspondence position and is.The YUV conversions RGB methods can
Select any-mode.
If step 202 judges that present image reaches threshold value T1 for the frame number that gray level image and gray level image continuously occur,
Scattered doubtful flame characteristic point is extracted according to the gray feature of flame and using frame difference method, frame difference condition C 2 is：1st, two frame figure
As gray scale difference value T2>50；2nd, the background model pixel value of the correspondence position is less than 150.The point for meeting frame difference condition C 2 is existed
Mark is on current frame image correspondence position.The method that the background model uses image averaging, calculate video image sequence
Continuous N two field picture average values are as background model in row.
Step 2032, doubtful flame characteristic point step 2031 extracted carries out region connection.
If step 202 judges present image for coloured image, present image is labeled as True according in step 2031
Point and present image in meet that the point of condition C 1 is re-flagged, and the point mutually adjoined after re-flagging carries out area
Domain connects, as doubtful flame region.
If step 202 judges present image for gray level image, present image is labeled as True according in step 2031
Point and present image in meet that gray value is re-flagged more than 220 point, and mutually adjoining after re-flagging
Point carries out region connection, as doubtful flame region.
Step 2033, the doubtful flame region after being connected to step 2032 carries out parameter calculating.
If step 202 judges present image for coloured image, the following parameter of doubtful flame region is calculated：Area, week
Length, filling rate, the ratio of width to height, centroid position, suspicious region RGB average value.
If step 202 judges present image for gray level image, the following parameter of doubtful flame region is calculated：Area, week
Length, filling rate, the ratio of width to height, centroid position, the average gray of suspicious region.
The filling rate is the ratio of suspicious region area and its minimum enclosed rectangle.
Step 204, the doubtful flame region extracted to step 203 screens.Screening conditions include：Fill suspicious region
Rate is more than threshold value T3 and is less than threshold value T4；Suspicious region the ratio of width to height meets that the wide ratio with high ratio and Gao Yukuan is respectively less than
T5.Wherein T3 spans are [0.50 ~ 0.60], and T4 spans are [0.90 ~ 0.95], and T5 spans are [2 ~ 4].
Step 205, matching renewal is carried out to the doubtful flame region retained after screening.
As shown in fig. 6, step 205 comprises the steps of：
Step 2051, matching operation is carried out according to suspicious region color or gray feature.If image is coloured image, when
The RGB average values of previous frame suspicious region are less than threshold value T6 with history suspicious region RGB mean variations difference；If image is gray scale
Image, then present frame suspicious region average gray and history suspicious region average intensity change difference are less than threshold value T7.T6 values
Scope is [20 ~ 30], and T7 spans are [10 ~ 20].
Step 2052, matching operation is carried out according to suspicious region temporal signatures.In sequence of video images, flame region
Area change should be in gradual change, i.e. area change between two frames is not over certain limit, this embodiment given threshold
For T8=[0.5 ~ 1.5], i.e. present frame suspicious region area is more than 0.5 with history suspicious region area ratio and is less than 1.5.
Step 2053, matching operation is carried out according to suspicious region motion feature.In sequence of video images, flame is continuous
Its motion of move but condition of a fire early period of origination can limit within a certain range again, therefore limit its single barycenter displacement and do not surpass
Threshold value T9 is crossed, the displacement threshold value that adds up is T10.T9 spans are [10 ~ 15], and T10 spans are [20 ~ 50].
History suspicious region matching times corresponding to the present frame suspicious region that above-mentioned all matching steps will be met add 1
And its parameters is updated with present frame suspicious region.It is if there is wherein any one mismatch, the history is doubtful
The loss number in region adds 1, when number is lost in history suspicious region has reached threshold value T11, empties the history suspicious region
Parameters.Wherein, T11 spans are [5 ~ 10].
Step 206, according to step 205 matching result, the matching times and given threshold of each history suspicious region are counted
Compare, when reaching threshold value T12, then judge the region for flame region alert.Wherein, T12 spans are
[10~16]。
It is described above, it is only the embodiment in the present invention, but protection scope of the present invention is not limited thereto, and is appointed
What be familiar with the researcher of the technology disclosed herein technical scope in, a variety of changes for being made according to actual conditions,
It should all be included within the scope of the present invention.Therefore, protection scope of the present invention should be with the protection model of claims
Enclose and be defined.
Claims (6)
- A kind of 1. flame detecting device based on video image analysis, it is characterised in that including：Video image acquiring unit, for obtaining the video image information of monitor area；What video image acquiring unit obtained regards Frequency image information includes coloured image or gray level image, and the coloured image and gray level image are obtained single by the video image Member obtains respectively according to ambient light according to intensity, wherein, the gray level image is that Lighting information amount is less than certain threshold value, is being opened The image gathered under infrared light compensating lamp；Image information judging unit, for detecting current frame image type, it is coloured image or gray scale to judge current frame image Image, then recorded if gray level image and continuously gray level image frame number occur；Suspicious region extraction unit, for extracting the flame suspicious region of current frame image；Suspicious region screening unit, for excluding ineligible flame suspicious region, preserve qualified suspicious region Relevant parameter is as historical data；Wherein, screening conditions include：Suspicious region filling rate is more than the 3rd threshold value and is less than the 4th threshold Value；Suspicious region the ratio of width to height meets that the wide ratio with high ratio and Gao Yukuan is respectively less than the 5th threshold value, wherein the 3rd threshold value takes It is [0.50 ~ 0.60] to be worth scope, and the 4th threshold value span is [0.90 ~ 0.95], and the 5th threshold value span is [2 ~ 4]；Updating block is matched, for present frame flame suspicious region and history suspicious region to be carried out into matching operation and more new historical The relevant parameter of suspicious region；Alarm unit is judged, for judging whether each history suspicious region meets final alert if.
- 2. according to the device described in claim 1, it is characterised in that suspicious region extraction unit includes：A feature point extraction units, for extracting scattered flame characteristic point under present image；B regions connection unit, the characteristic point mutually adjoined for feature point extraction unit to be extracted carry out region connection；C parameter calculation units, for calculating the relevant parameter of the flame suspicious region after connecting.
- 3. according to the device described in claim 1, it is characterised in that matching updating block includes：Suspicious region morphological feature matching unit, the morphological feature according to flame is to current flame suspicious region and the doubtful area of history Domain carries out matching operation；Suspicious region temporal signatures matching unit, the temporal signatures according to flame are to current flame suspicious region and the doubtful area of history Domain carries out matching operation；Suspicious region motion feature matching unit, the motion feature according to flame is to current flame suspicious region and the doubtful area of history Domain carries out matching operation.
- A kind of 4. flame detecting method based on video image analysis, it is characterised in that including：First step, the video image information in monitor area is obtained using image collecting device；The video image information of acquisition Including coloured image or gray level image, the coloured image and gray level image are by the video image acquiring unit according to the external world Intensity of illumination obtains respectively, wherein, the gray level image is that Lighting information amount be less than certain threshold value, in unlatching infrared light compensating lamp The image of lower collection；Second step, current frame image type is detected, it is coloured image or gray level image to judge current frame image, if gray scale Image, which then records, continuously there is gray level image frame number；Third step, the flame suspicious region of current frame image is extracted, it is different according to image type, take different feature extractions Mode, if present image is coloured image, the color property according to flame extracts flame suspicious region, if present image is Gray level image then highlights feature extraction according to flame and goes out flame suspicious region；Four steps, ineligible flame suspicious region is excluded, preserve qualified suspicious region relevant parameter conduct Historical data；5th step, present frame flame suspicious region and history suspicious region are subjected to matching operation and more new historical suspicious region Relevant parameter；6th step, judges whether each history suspicious region meets final alert if；Screening conditions include in four steps：Suspicious region filling rate is more than the 3rd threshold value and is less than the 4th threshold value；Suspicious region The ratio of width to height meets that the wide ratio with high ratio and Gao Yukuan is respectively less than the 5th threshold value, wherein the 3rd threshold value span is [0.50 ~ 0.60], the 4th threshold value span are [0.90 ~ 0.95], and the 5th threshold value span is [2 ~ 4].
- 5. in accordance with the method for claim 4, it is characterised in that third step includes following sub-step：Step A, extract scattered doubtful flame characteristic point according to frame difference method, doubtful flame characteristic point extraction conditions under coloured image For：1、R>G>B；2、R>220, doubtful flame characteristic point extraction conditions are under gray level image：1st, two field pictures gray scale difference value is more than 50,2, the background model pixel value of the correspondence position is less than 150；Step B, the doubtful flame characteristic point of extraction is subjected to region Connection；Step C, parameter calculating is carried out to the doubtful flame region after connection.
- 6. in accordance with the method for claim 4, it is characterised in that the 5th step includes following sub-step：Step D, matching operation is carried out according to suspicious region color or gray feature, if image is coloured image, present frame is doubted It is less than the 6th threshold value with history suspicious region RGB mean variations difference like the RGB average values in region；If image is gray-scale map Picture, then present frame suspicious region average gray and history suspicious region average intensity change difference are less than the 7th threshold value, the 6th threshold It is [20 ~ 30] to be worth span, and the 7th threshold value span is [10 ~ 20]；Step E, carried out according to suspicious region temporal signatures Matching operation, in sequence of video images, the area change of flame region should be in gradual change, i.e. area change between two frames Not over certain limit, i.e. present frame suspicious region area is set to the 8th threshold value with history suspicious region area ratio, and it takes It is [0.5 ~ 1.5] to be worth scope；Step F, matching operation is carried out according to suspicious region motion feature, in sequence of video images, fire Flame be constantly move but condition of a fire early period of origination its motion and can limit within a certain range, therefore limit its single barycenter movement Distance is no more than the 9th threshold value, and the displacement that adds up is the tenth threshold value, and the 9th threshold value span is [10 ~ 15], the tenth threshold value Span is [20 ~ 50].
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510068450.9A CN104598895B (en) | 2015-02-10 | 2015-02-10 | A kind of flame detecting method and device based on video image analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510068450.9A CN104598895B (en) | 2015-02-10 | 2015-02-10 | A kind of flame detecting method and device based on video image analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104598895A CN104598895A (en) | 2015-05-06 |
| CN104598895B true CN104598895B (en) | 2018-03-16 |
Family
ID=53124668
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510068450.9A Active CN104598895B (en) | 2015-02-10 | 2015-02-10 | A kind of flame detecting method and device based on video image analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104598895B (en) |
Families Citing this family (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106169218B (en) * | 2015-09-10 | 2019-04-05 | 河南联纵消防科技有限公司 | Fire locating method based on infrared thermal imaging and visible images technology |
| CN105752352B (en) * | 2015-11-03 | 2018-04-06 | 天津艾思科尔科技有限公司 | Detector carries the unmanned plane of anti-static device |
| CN105761275A (en) * | 2015-11-03 | 2016-07-13 | 天津艾思科尔科技有限公司 | Fire-fighting early warning aircraft with binocular visual structure |
| CN105764304B (en) * | 2015-11-03 | 2018-10-16 | 天津艾思科尔科技有限公司 | Detector carries the aircraft of steady type radiator structure |
| CN105763841A (en) * | 2015-11-03 | 2016-07-13 | 天津艾思科尔科技有限公司 | Aircraft provided with detector with protector |
| CN105741477B (en) * | 2015-11-03 | 2018-06-12 | 天津艾思科尔科技有限公司 | Aircraft with intelligent fire voice assistant |
| CN105763842A (en) * | 2015-11-03 | 2016-07-13 | 天津艾思科尔科技有限公司 | Aircraft with fire-fighting detection function |
| CN106096603A (en) * | 2016-06-01 | 2016-11-09 | 中国科学院自动化研究所 | A kind of dynamic flame detection method merging multiple features and device |
| CN107330441B (en) * | 2017-05-26 | 2019-07-09 | 天津大学 | Flame image foreground extraction algorithm |
| CN108629342A (en) * | 2017-11-28 | 2018-10-09 | 广东雷洋智能科技股份有限公司 | Binocular camera flame distance measurement method and device |
| CN108038510A (en) * | 2017-12-22 | 2018-05-15 | 湖南源信光电科技股份有限公司 | A kind of detection method based on doubtful flame region feature |
| CN108334855A (en) * | 2018-02-24 | 2018-07-27 | 南瑞集团有限公司 | A kind of substation's flame identification algorithm using enhancing RGB component feature |
| CN108765461B (en) * | 2018-05-29 | 2022-07-12 | 青鸟消防股份有限公司 | Fire-fighting fire image block extraction and identification method and device |
| CN109915887B (en) * | 2018-06-04 | 2021-05-04 | 嵊州市悦鑫五金配件经营部 | Electric heater fire extinguishing strategy customization mechanism |
| CN109410222B (en) * | 2018-11-01 | 2023-02-03 | 浩云科技股份有限公司 | Flame detection method and device |
| CN110580449B (en) * | 2019-08-09 | 2023-05-02 | 北京准视科技有限公司 | Image type flame identification and detection method |
| CN111202933A (en) * | 2020-01-09 | 2020-05-29 | 海达建设集团有限公司 | Indoor automatic water spraying fire extinguishing system and fire extinguishing method |
| CN111723663B (en) * | 2020-05-18 | 2024-01-23 | 中国农业科学院农业环境与可持续发展研究所 | Mulch film identification method, device and storage medium |
| CN111814638B (en) * | 2020-06-30 | 2024-04-30 | 成都睿沿科技有限公司 | Security scene flame detection method based on deep learning |
| 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 |
| CN114724287A (en) * | 2021-01-06 | 2022-07-08 | 中国石油天然气股份有限公司 | On-duty system and on-duty method of oil and gas station |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5289275A (en) * | 1991-07-12 | 1994-02-22 | Hochiki Kabushiki Kaisha | Surveillance monitor system using image processing for monitoring fires and thefts |
| US5937077A (en) * | 1996-04-25 | 1999-08-10 | General Monitors, Incorporated | Imaging flame detection system |
| CN101587622A (en) * | 2009-06-18 | 2009-11-25 | 任芳 | Forest rocket detection and recognition methods and equipment based on video image intelligent analysis |
| CN103593938A (en) * | 2013-11-20 | 2014-02-19 | 无锡北洋清安物联科技有限公司 | Fire detection method based on video image lengthwise characters |
| CN104079872A (en) * | 2014-05-16 | 2014-10-01 | 大连理工大学 | Video image processing and human-computer interaction method based on content |
-
2015
- 2015-02-10 CN CN201510068450.9A patent/CN104598895B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5289275A (en) * | 1991-07-12 | 1994-02-22 | Hochiki Kabushiki Kaisha | Surveillance monitor system using image processing for monitoring fires and thefts |
| US5937077A (en) * | 1996-04-25 | 1999-08-10 | General Monitors, Incorporated | Imaging flame detection system |
| CN101587622A (en) * | 2009-06-18 | 2009-11-25 | 任芳 | Forest rocket detection and recognition methods and equipment based on video image intelligent analysis |
| CN103593938A (en) * | 2013-11-20 | 2014-02-19 | 无锡北洋清安物联科技有限公司 | Fire detection method based on video image lengthwise characters |
| CN104079872A (en) * | 2014-05-16 | 2014-10-01 | 大连理工大学 | Video image processing and human-computer interaction method based on content |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104598895A (en) | 2015-05-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104598895B (en) | A kind of flame detecting method and device based on video image analysis | |
| CN110135269B (en) | Fire image detection method based on mixed color model and neural network | |
| CN104580896B (en) | A kind of video camera diurnal pattern switching method and apparatus | |
| CN102306304B (en) | Face occluder identification method and device | |
| CN101090482B (en) | Driver fatigue monitoring system and method based on image process and information mixing technology | |
| CN107622258A (en) | A kind of rapid pedestrian detection method of combination static state low-level image feature and movable information | |
| CN106650584B (en) | Flame detecting method and system | |
| CN107133564B (en) | Tooling cap detection method | |
| US20160260306A1 (en) | Method and device for automated early detection of forest fires by means of optical detection of smoke clouds | |
| CN105160297B (en) | Masked man's event automatic detection method based on features of skin colors | |
| CN111626188B (en) | Indoor uncontrollable open fire monitoring method and system | |
| CN112069975A (en) | Comprehensive flame detection method based on ultraviolet, infrared and vision | |
| CN104778686B (en) | A kind of tissue culture method for early warning monitored based on scab | |
| CN108388838B (en) | Unmanned aerial vehicle ground crowd monitoring system and monitoring method | |
| CN108629316A (en) | A kind of video accident detection method of various visual angles | |
| CN111179279A (en) | Comprehensive flame detection method based on ultraviolet and binocular vision | |
| CN107067412A (en) | A kind of video flame smog detection method of Multi-information acquisition | |
| CN103258232A (en) | Method for estimating number of people in public place based on two cameras | |
| CN106127292B (en) | Flow method of counting and equipment | |
| CN209029110U (en) | Chinese medicine facial diagnosis is health management system arranged | |
| CN106686377A (en) | Algorithm for determining video key area based on deep neural network | |
| CN111242096B (en) | People number gradient-based people group distinguishing method | |
| CN109040711A (en) | A kind of image-pickup method, monitoring method and imaging sensor | |
| CN108737785B (en) | Indoor automatic detection system that tumbles based on TOF 3D camera | |
| CN109034038B (en) | Fire identification device based on multi-feature fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 300457 10 North Zone, TEDA 57, Second Avenue, Binhai New Area Development Zone, Tianjin, Applicant after: TIANJIN?ISECURE?TECHNOLOGY?CO., LTD. Address before: 300457 building 20, No. sixth, 1 Avenue, Tianjin Binhai New Area Development Zone Applicant before: TIANJIN?ISECURE?TECHNOLOGY?CO., LTD. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |