CN105488941B - Double spectrum fire monitoring method and devices based on Infrared-Visible image - Google Patents
Double spectrum fire monitoring method and devices based on Infrared-Visible image Download PDFInfo
- Publication number
- CN105488941B CN105488941B CN201610029028.7A CN201610029028A CN105488941B CN 105488941 B CN105488941 B CN 105488941B CN 201610029028 A CN201610029028 A CN 201610029028A CN 105488941 B CN105488941 B CN 105488941B
- Authority
- CN
- China
- Prior art keywords
- image
- infrared
- region
- interest
- visible images
- 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
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000012544 monitoring process Methods 0.000 title claims abstract description 35
- 238000001228 spectrum Methods 0.000 title claims abstract description 26
- 238000012795 verification Methods 0.000 claims abstract description 22
- 239000011159 matrix material Substances 0.000 claims description 18
- 230000033001 locomotion Effects 0.000 claims description 17
- 238000009826 distribution Methods 0.000 claims description 15
- 238000013507 mapping Methods 0.000 claims description 12
- 230000007246 mechanism Effects 0.000 claims description 8
- 238000003709 image segmentation Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 238000012806 monitoring device Methods 0.000 claims description 4
- 238000001514 detection method Methods 0.000 abstract description 28
- 230000008901 benefit Effects 0.000 abstract description 9
- 238000004364 calculation method Methods 0.000 abstract description 6
- 230000005855 radiation Effects 0.000 abstract description 5
- 238000010586 diagram Methods 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 8
- 238000012545 processing Methods 0.000 description 7
- 238000012360 testing method Methods 0.000 description 6
- 238000003384 imaging method Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000002485 combustion reaction Methods 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000008030 elimination Effects 0.000 description 2
- 238000003379 elimination reaction Methods 0.000 description 2
- 238000000691 measurement method Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000000877 morphologic effect Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 238000011982 device technology Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000011049 filling Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000011895 specific detection Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 230000003313 weakening effect Effects 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
- G08B17/005—Fire alarms; Alarms responsive to explosion for forest fires, e.g. detecting fires spread over a large or outdoors area
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- 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)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Theoretical Computer Science (AREA)
- Fire-Detection Mechanisms (AREA)
- Closed-Circuit Television Systems (AREA)
- Image Analysis (AREA)
Abstract
Double spectrum fire monitoring method and devices based on Infrared-Visible image that the embodiment of the invention discloses a kind of.This method includes:Obtain the infrared image and visible images of current scene;Determine the doubtful fire point marked region in infrared image;Doubtful fire is put into marked region and is mapped to the corresponding position in visible images, obtains the area-of-interest in visible images;Image in area-of-interest is verified, is determined in current scene with the presence or absence of true fiery point according to verification result.Infrared image detection identification advantage and visible images detection identification advantage are combined by the present invention by using above-mentioned technical proposal, can be eliminated interference of the appearance from the object that flame similarity is high but radiation energy is different with flame to judging result, be promoted accuracy.Also, only the image in area-of-interest is verified in visible images, calculation amount can be reduced, raising efficiency ensures the real-time of fire monitoring.
Description
Technical field
The present embodiments relate to mobile device technology field more particularly to a kind of pairs based on Infrared-Visible image
Spectrum fire monitoring method and device.
Background technology
Forest fire is a kind of sudden strong, destructive natural calamity big, disposition relief is more difficult.Therefore, in woods
Fire feelings monitor in link, it is found that the promptness of dangerous situation and accuracy are particularly important, relevant staff can be made to take as early as possible
Emergency measures reduces the loss of forest fire to greatest extent.
Currently, holder ocean weather station observation is the main forest fires detection means that China uses, in order to save manpower, generally use regards
Frequency image automatic measurement technique monitors forest fire.Common video image automatic measurement technique includes being based on visible images
Fire point detection technique and based on remote sensing images fire point detection technique.Wherein, the fire point detection algorithm based on visible images
In, the visible light feature (such as color characteristic, textural characteristics and motion feature) of flame is utilized, it is difficult to filter out appearance with
The high object of flame similarity (red flag and light that such as wave), alarm accuracy rate is low, and algorithm moderate heat point is carried out with background
Same processing, needs to calculate larger data volume, less efficient, is particularly unsuitable under remote and low light environment
Detection.In fire point detection algorithm based on remote sensing images, mainly utilize certain wave bands of particular satellite by different modes
Thresholding obtains fire point position after False color image, wave band algebraic operation, it usually needs man-machine interactively processing;In addition, remote sensing number
According to what is come out for ground station reception, parsing, processing and distribution, and remote sensing images resolution ratio is low, it is found that ground is possible when fire point
It has been caused disaster that, therefore more inferior in terms of real-time.
Invention content
The purpose of the embodiment of the present invention is to provide a kind of double spectrum fire monitoring sides based on Infrared-Visible image
Method and device, to solve the problems, such as that existing fire monitoring scheme accuracy is low and real-time is poor.
On the one hand, an embodiment of the present invention provides a kind of double spectrum fire monitorings based on Infrared-Visible image
Method, including:
Obtain the infrared image and visible images of current scene;
Determine the doubtful fire point marked region in the infrared image;
The doubtful fire point marked region is mapped to the corresponding position in the visible images, obtains the visible light
Area-of-interest in image;
Image in the area-of-interest is verified, determines in the current scene whether deposit according to verification result
In true fiery point.
On the other hand, an embodiment of the present invention provides a kind of, and double spectrum forest fires based on Infrared-Visible image are supervised
Device is surveyed, including:
Image collection module, infrared image and visible images for obtaining current scene;
Doubtful fire point marked region determining module, for determining the doubtful fire point marked region in the infrared image;
Mapping block, for the doubtful fire point marked region to be mapped to the corresponding position in the visible images,
Obtain the area-of-interest in the visible images;
True fire point determining module, it is true according to verification result for being verified to the image in the area-of-interest
With the presence or absence of true fiery point in the fixed current scene.
A kind of double spectrum fire monitoring schemes based on Infrared-Visible image provided in the embodiment of the present invention,
After the infrared image and visible images that obtain current scene, the doubtful fire point marked region in infrared image is first determined, then will
Doubtful fire point marked region is mapped to the corresponding position in the visible images, obtains the region of interest in visible images
Domain, and the image in area-of-interest is verified, finally determined in current scene with the presence or absence of true according to verification result
Fiery point.By using above-mentioned technical proposal, infrared image detection identification advantage and visible images detection identification advantage are carried out
In conjunction with, it is first to determine doubtful fire point marked region according to the imaging features of infrared image, then doubtful fire is put into marked region and is mapped to
In visible images, appearance can be eliminated, judging result is done from the object that flame similarity is high but radiation energy is different with flame
It disturbs, promotes accuracy.Also, only the image in area-of-interest is verified in visible images, is avoided at night
The low background of brightness accounts for the overwhelming majority in forest zone image under scene, all pixels participate in calculating lacking for waste vast resources
It falls into, calculation amount can be reduced, raising efficiency ensures the real-time of fire monitoring.In addition, above-mentioned technical proposal can either overcome list can
Light-exposed forest fire algorithm is also applied for extensive search fire point and closely supervises in deficiency remote, under low light environment
Fire detecting feelings, with the mode ratio that remote sensing images are combined with visible images, the simplicity and real-time of the program are equally also very
It is prominent.
Description of the drawings
Fig. 1 is a kind of double spectrum fire monitorings based on Infrared-Visible image that the embodiment of the present invention one provides
The flow diagram of method;
Fig. 2 is a kind of double spectrum fire monitorings based on Infrared-Visible image provided by Embodiment 2 of the present invention
The flow diagram of method;
Fig. 3 is a kind of double spectrum Forest Fires being preferably based on Infrared-Visible image that the embodiment of the present invention three provides
The flow diagram of feelings monitoring method;
Fig. 4 is a kind of infrared heat point detection algorithm flow diagram that the embodiment of the present invention three provides;
Fig. 5 is the doubtful fire point verification algorithm flow schematic diagram of a kind of visible light that the embodiment of the present invention three provides;
Fig. 6 is a kind of double spectrum fire monitorings based on Infrared-Visible image that the embodiment of the present invention four provides
The structure diagram of device.
Specific implementation mode
Technical solution to further illustrate the present invention below with reference to the accompanying drawings and specific embodiments.It is appreciated that
It is that specific embodiment described herein is used only for explaining the present invention rather than limitation of the invention.It further needs exist for illustrating
, only the parts related to the present invention are shown for ease of description, in attached drawing rather than entire infrastructure.
It should be mentioned that some exemplary embodiments are described as before exemplary embodiment is discussed in greater detail
The processing described as flow chart or method.Although each step is described as the processing of sequence, many of which by flow chart
Step can be implemented concurrently, concomitantly or simultaneously.In addition, the sequence of each step can be rearranged.When its operation
The processing can be terminated when completion, it is also possible to the additional step being not included in attached drawing.The processing can be with
Corresponding to method, function, regulation, subroutine, subprogram etc..
Embodiment one
Fig. 1 is a kind of double spectrum fire monitorings based on Infrared-Visible image that the embodiment of the present invention one provides
The flow diagram of method, this method can be executed by double spectrum fire monitoring devices based on Infrared-Visible image,
Wherein the device can be generally integrated in fire monitoring system by software and or hardware realization.As shown in Figure 1, the party
Method includes:
Step 110, the infrared image and visible images for obtaining current scene.
Illustratively, the fire monitoring system may include billiard table (also known as holder), infrared camera and visible light phase
Machine may also include ball machine control device.Before starting monitoring, cruising manner and presetting bit that billiard table is rationally set are needed, to cover
All regions to be monitored are covered, infrared camera and Visible Light Camera be installed on billiard table, and to infrared camera and Visible Light Camera
Position and parameters etc. are adjusted.After above-mentioned preparation, can billiard table be controlled by ball machine control device and turned to
Presetting bit position, obtains the data of infrared camera and Visible Light Camera in real time, i.e., acquisition current scene in this step it is infrared
Image and visible images.
It should be noted that the viewfinder range size of infrared camera can be identical with the viewfinder range size of Visible Light Camera
It can be different.If it is different, preferred, it is seen that the viewfinder range of light camera is more than the viewfinder range of infrared camera.That is, can
Comprising all scenes in infrared image in light-exposed image, it also may include other scenes being not present in infrared image.
Step 120 determines that the doubtful fire in infrared image puts marked region.
The imaging characteristics of infrared image are the difference that can reflect the outside radiation energy of different objects, also can reflection body
Heat signature.Illustratively, because the heat signature of object can be weighed with the brightness value in infrared image, heat is higher
Brightness value of the object in infrared image it is higher, and brightness value of the lower object of heat in infrared image is relatively low, so
The region for the heat signature for meeting flame can be tentatively picked out according to the brightness value characteristic distributions in infrared image, so that it is determined that
Doubtful fire point marked region in infrared image.The scenery of doubtful fire point is contained in doubtful fire point marked region.
Illustratively, the shapes and sizes of the doubtful fire point marked region are not especially limited, and can be configured in advance,
Also it can be automatically adjusted according to actual brightness value characteristic distributions.Preferably, for easy analysis, the doubtful fire point label
Concretely doubtful fire puts rectangle marked region in region.
Further, this step is preferably:According to the brightness value characteristic distributions in infrared image by infrared Image Segmentation
At multiple regions, region of the rectangle marked by brightness value higher than predetermined luminance threshold value is used to draw a circle to approve as doubtful Huo Dian rectangle markeds area
Domain.Illustratively, when carrying out region segmentation, the pixel that can be closer to brightness value is divided into the same region, in detail
Partitioning scheme is not construed as limiting herein.At fire point rectangle marked region doubtful using rectangle marked delineation, current region can be calculated
Whether the average value of middle brightness value is higher than predetermined luminance threshold value, if so, being that doubtful fire puts rectangle marked by current region delineation
Region.
Doubtful fire point marked region is mapped to the corresponding position in visible images by step 130, obtains visible images
In area-of-interest.
Illustratively, infrared image and visible images contain current scene simultaneously, so being determined from infrared image
It is doubtful fire point marked region in include scenery be similarly present in visible images.It should be noted that the present embodiment
It is middle doubtful fire is put into marked region to be mapped to the corresponding position in visible images and refer to, doubtful fire is put to the boundary of marked region
It is mapped to the corresponding position of visible images, rather than doubtful fire is put into the picture material for including in marked region and copies to visible light
The corresponding position of image.The doubtful fire point marked region in infrared image is mapped in visible images according to preset algorithm
Corresponding position obtains the area-of-interest (region of interest, ROI) in visible images, so area-of-interest
In equally include doubtful fire point scenery, resolution ratio higher in visible images of scenery of the doubtful fire point is more advantageous to
Whether do further is the true fiery verification put.It should be noted that the preset algorithm is not unique, for example, can join
According to Infrared-Visible camera combined calibrating technology, the corresponding mapping square of Pixel-level between infrared image and visible images is obtained
Battle array, using the mapping matrix can will in infrared image it is doubtful fire point marked region compound mapping to visible images on.This
Field technology personnel can choose suitable pre- imputation according to the configuring condition of practical situations and fire monitoring system
Method, the present embodiment are not especially limited.
It should be noted that the doubtful fire point marked region in infrared image can be multiple, so corresponding visible light
Area-of-interest in image may be multiple.
Step 140 verifies the image in area-of-interest, determines in current scene whether deposit according to verification result
In true fiery point.
Illustratively, the mode verified to the image in the area-of-interest in visible images can refer to existing
Verification mode.For example, can be verified to color characteristic, Luminance Distribution feature and motion feature etc..It is tested with existing
The main distinction of card mode is, in existing verification mode, needs to verify entire visible images, to fire point and
Background is all similarly handled, and needs to calculate larger data volume, less efficient;And in the present embodiment, only to comprising doubtful
Image in the area-of-interest of fire point is verified, and calculative data volume is greatly reduced, and raising efficiency ensures real-time
Property.
Double spectrum fire monitoring methods based on Infrared-Visible image that the embodiment of the present invention one provides obtain
After the infrared image and visible images of current scene, the doubtful fire point marked region in infrared image is first determined, then will be doubtful
Fire point marked region is mapped to the corresponding position in the visible images, obtains the area-of-interest in visible images, and
Image in area-of-interest is verified, is finally determined in current scene with the presence or absence of true fiery point according to verification result.
By using above-mentioned technical proposal, infrared image detection identification advantage and visible images detection identification advantage are combined,
First doubtful fire point marked region is determined according to the imaging features of infrared image, then doubtful fire is put into marked region and is mapped to visible light
In image, interference of the appearance from the object that flame similarity is high but radiation energy is different with flame to judging result can be eliminated, is carried
Rise accuracy.Also, only the image in area-of-interest is verified in visible images, is avoided under night scenes
Forest zone image in the low background of brightness accounts for the overwhelming majority, all pixels participate in calculating the defects of waste vast resources, can subtract
Few calculation amount, raising efficiency ensure the real-time of fire monitoring.In addition, above-mentioned technical proposal can either overcome single visible flare up
Point recognizer is also applied for extensive search fire point and closely monitoring fire behavior in deficiency remote, under low light environment,
With the mode ratio that remote sensing images are combined with visible images, the simplicity and real-time of the program are equally also very prominent.
Embodiment two
Fig. 2 is a kind of double spectrum fire monitorings based on Infrared-Visible image provided by Embodiment 2 of the present invention
The flow diagram of method, the present embodiment are optimized based on above-described embodiment, are drawn a circle to approve and are doubted in particular by rectangle marked
Region is put like fire, and multiframe infrared image is handled using voting mechanism, improves and determines doubtful fire point rectangle marked region
Accuracy, take into account fire point it is that may be present flicker by a small margin movement under the premise of, eliminate the movement objects such as car light and cause
Interference.
Specifically, the method for the present embodiment includes the following steps:
Step 210, the infrared image set and visible images for obtaining current scene include n frames in infrared image set
The infrared image of corresponding n different moments.
Step 220, for each frame infrared image in infrared image set, will be current red according to brightness value characteristic distributions
Outer image segmentation uses region of the rectangle marked by brightness value higher than predetermined luminance threshold value to draw a circle to approve as rectangle marked at multiple regions
Region;Pixel in rectangle marked region in current infrared image is labeled as 1, the pixel outside rectangle marked region is labeled as
0。
Specifically, infrared image set can be denoted as A1, A2, A3 ... An | n>0 }, for every frame infrared image Ai (0<i<
N) it is marked, the pixel in rectangle marked region is labeled as 1, the pixel outside rectangle marked region is labeled as 0.
The pixel at same position in the different infrared images of step 230, statistics in infrared image set is labeled
For 1 number, the number for being marked as 1 is denoted as 1 not less than the pixel of preset times using voting mechanism, other pixels are denoted as
0, form target bianry image.
Preferably, the preset times are 0.8n.
Step 240 carries out target bianry image contours extract, and using the doubtful fire point rectangle marked of rectangle marked delineation
Region.
Illustratively, executing step 210- steps 240 is advantageous in that, the doubtful fire point square obtained to n frame infrared images
Shape marked region, using the voting mechanism of Pixel-level as final differentiation as a result, caused by the objects such as movement car light can be eliminated
Interference, while taking into account fire and putting movement that may be present of flickering by a small margin, finally obtain the higher doubtful fire point rectangle of accuracy
Marked region.
Step 250, using image transform model by it is doubtful fire put rectangle marked area maps to visible images in phase
Position is answered, the area-of-interest in visible images is obtained.
Specifically, described image transformation model is specially:
Wherein, C1Indicate infrared camera, C2Indicate Visible Light Camera;R is C2Relative to C1Spin matrix, t C2Relatively
In C1Motion vector;(u1,v1) it is coordinate of the spatial point in infrared image, (u2,v2) be the spatial point in visible light figure
Coordinate as in;K1For C1Intrinsic Matrix, K2For C2Intrinsic Matrix;Zc1For the spatial point to C1Image plane away from
From Zc2For the spatial point to C2The distance of image plane.
Illustratively, which gives the geometry obtained with different view between the two images comprising Same Scene and becomes
Relationship is changed, the image transform model based on camera motion can be referred to as.According to the description of equipment that camera manufacturer provides, obtain
Intrinsic parameter information of Visible Light Camera and infrared camera, including focal length, picture centre coordinate, pixel dimension etc. are taken, and is made
For intrinsic parameter initial value, i.e. matrix K1For and K2。
Spin matrix R in above-mentioned model formation can indicate with 3 Eulerian angles, i.e., the angle [alpha] rotated around X-axis, around Y
The angle beta of axis rotation, the expression formula of the angle γ rotated about the z axis, R are
The expression formula of translation vector t in above-mentioned model formation is t=[t1 t2 t3]T。
Preferably, it before the corresponding position during doubtful fire to be put to rectangle marked area maps to visible images, can incite somebody to action
The pixel of the outside continuation preset quantity of profile in doubtful fire point rectangle marked region forms final doubtful fire point marked region.With
Afterwards, final doubtful fire point marked region is mapped to the corresponding position in visible images, the sense obtained in visible images is emerging
Interesting region.This have the advantage that can eliminate mapping process generates error and the influence that brings.It should be noted that
Initial region of interest can be obtained after the corresponding position during doubtful fire to be put to rectangle marked area maps to visible images
Domain, then by the pixel of the outside continuation preset quantity of initial area-of-interest profile, final area-of-interest is formed, it can equally arrive
Up to the effect for eliminating error influence.
Step 260 verifies the image in area-of-interest, determines in current scene whether deposit according to verification result
In true fiery point.
Specifically, this step may include:Current interest region is matched with default kidney-yang feelings color space model,
Obtain the first matching degree;The current interest region is matched with default pseudo- fire behavior color space model, obtains second
Matching degree;If the first matching degree is higher than the second matching degree, it is determined that current interest region is that doubtful fire puts region;To visible light
All doubtful fire point regions in image are verified, and are determined in current scene with the presence or absence of true fiery point according to verification result.
Illustratively, the default kidney-yang feelings color space model and default pseudo- fire behavior color space model are concretely sharp
With mixed Gauss model technology respectively to true ignition point (such as trees, grassland) and interference heat source (such as vehicle, light) point
It carry out not be obtained from color space modeling.Matching degree, which is particularly used in, to be weighed area-of-interest and presets kidney-yang feelings color space
The degree of closeness of model or default pseudo- fire behavior color space model, matching degree is higher, then illustrates closer with corresponding model.When
It is higher than corresponding to the first matching degree for presetting kidney-yang feelings color space model and corresponds to the of default pseudo- fire behavior color space model
When two matching degrees, illustrate that current interest region closer to kidney-yang feelings color space model is preset, then can determine current interest
Region is that doubtful fire puts region.
Then, all doubtful fire point regions in visible images are verified, front court is worked as according to verification result determination
With the presence or absence of true fiery point in scape.For example, Luminance Distribution detection and motion feature detection etc. can be carried out.Carrying out Luminance Distribution inspection
When survey, the regularity of distribution of flame pixels brightness can be analyzed, by combustion centre's point of internal flame to flame envelope edge, due to ignition temperature
Difference, colour brightness can successively successively decrease, and be screened using luminance histogram statistics grey scale change rule, and to doubtful fire point;
When carrying out motion feature detection, using gauss hybrid models (Gaussian Mixture Model, GMM;Also known as mixing is high
This model) the doubtful fire point screening of background modeling technology progress, interference caused by moving the objects such as car light with further elimination, simultaneously
It takes into account fire and puts movement that may be present of flickering by a small margin, motion feature also needs multiframe visible images when detecting, preferably
, it is same to obtain visible images set when obtaining the infrared image set of current scene.
The embodiment of the present invention two using rectangle marked delineation doubtful fire point region, and utilizes on the basis of embodiment one
Voting mechanism handles multiframe infrared image, improves the accuracy for determining doubtful fire point rectangle marked region, is taking into account
Under the premise of fire puts movement that may be present of flickering by a small margin, interferes, can further carry caused by eliminating the objects such as movement car light
The accuracy of high fire monitoring.Also, it eliminates after moving the interference of the objects such as car light, the gross area in doubtful fire point rectangle marked region
Also it can accordingly reduce, to make the gross area of area-of-interest reduce, calculation amount can be further reduced, raising efficiency ensures fire
The real-time of feelings monitoring.
Embodiment three
Fig. 3 is a kind of double spectrum Forest Fires being preferably based on Infrared-Visible image that the embodiment of the present invention three provides
The flow diagram of feelings monitoring method, as shown in figure 3, this method specifically comprises the following steps:
Step 310, billiard table turn to current presetting bit.
Step 320, the infrared image and visible images for obtaining the corresponding current scene of current presetting bit.
Step 330 runs infrared heat point detection algorithm, judges whether to can be derived that doubtful fire point rectangle marked region, if
It is to then follow the steps 350;Otherwise, step 340 is executed.
Further, Fig. 4 is a kind of infrared heat point detection algorithm flow diagram that the embodiment of the present invention three provides, and is such as schemed
Shown in 4, which includes the following steps:
Step 331 pre-processes infrared image.
Specifically, the operations such as enhancing can be filtered to infrared image, weakening or some for eliminating the weaker generation of light are dry
It disturbs.
Step 332, to the infrared image after pretreatment into row threshold division.
Using the imaging characteristics of infrared image, i.e. otherwise the high object of heat brightness value in infrared image is high then low, right
Per frame infrared image into row threshold division, hot spot target can be separated from background.
Step 333 carries out morphological operation to the infrared image after Threshold segmentation.
Illustratively, in step 332 to infrared image into after row threshold division, binary map can be formed, by binary map
It carries out the morphological operation such as expanding, to achieve the purpose that filling cavity, come out the entirety of hot spot target rather than extracting section.
Step 334 draws a circle to approve rectangle marked region using rectangle marked.
Illustratively, hot spot target profile is found, the outmost boundary of rectangle marked is used in combination, obtains the rectangle of doubtful fire point
Marked region.
Step 335 judges whether current rectangle marked region meets the condition of filtering out, if so, thening follow the steps 336, otherwise
Execute step 337.
Illustratively, the priori for filtering out condition and may include forest fires.
Step 336 filters out current rectangle marked region, executes step 338.
Step 337 retains current rectangle marked region, executes step 338.
Step 338 judges whether current rectangle marked region is the last one, if so, terminating flow;Otherwise, it executes
Step 339.
Step 339 sets next rectangle marked region to new current rectangle marked region, returns to step
335。
It is doubtful fire point rectangle marked region by filtering out the rectangle marked region remained.
Preferably, it can also carry out, as handled multiframe infrared image using voting mechanism in embodiment two, obtaining
Final doubtful fire point rectangle marked region, to further increase the accuracy for determining doubtful fire point rectangle marked region.Specifically
Details can refer to the associated description in the embodiment of the present invention two.
After infrared heat point detection algorithm flow, step 350 can perform.If can not be obtained according to infrared heat point detection algorithm
Go out doubtful fire point rectangle marked region, illustrates to can perform step 340 there is no doubtful fiery point in current scene.
Next presetting bit is set as current presetting bit, and returns to step 310 by step 340.
Doubtful fire point marked region is mapped to the corresponding position in visible images by step 350, obtains visible images
In area-of-interest.
Specifically, image transform model can be used by the phase in doubtful fire point rectangle marked area maps to visible images
Position is answered, the area-of-interest in visible images is obtained.
Described image transformation model is specially:
Wherein, C1Indicate infrared camera, C2Indicate Visible Light Camera;R is C2Relative to C1Spin matrix, t C2Relatively
In C1Motion vector;(u1,v1) it is coordinate of the spatial point in infrared image, (u2,v2) be the spatial point in visible light figure
Coordinate as in;K1For C1Intrinsic Matrix, K2For C2Intrinsic Matrix;Zc1For the spatial point to C1Image plane away from
From Zc2For the spatial point to C2The distance of image plane.
Step 360 verifies the image in area-of-interest according to the doubtful fire point verification algorithm of visible light, according to testing
Card result is determined with the presence or absence of true fiery point in current scene, if so, thening follow the steps 370;Otherwise, step 340 is executed.
Further, Fig. 5 is the doubtful fire point verification algorithm flow signal of a kind of visible light that the embodiment of the present invention three provides
Figure, as shown in figure 5, the algorithm specifically may include following steps:
Step 361 pre-processes current interest region.
Specifically, the pretreatment may include, by the pixel of the outside continuation preset quantity of area-of-interest profile, being formed most
Whole area-of-interest can eliminate the influence that produced error band comes in follow-up mapping process.
Step 362 carries out color space model detection to the image in current interest region, whether judges testing result
Pass through, if so, thening follow the steps 365;Otherwise, step 363 is executed.
Specifically, can by current interest region respectively with default kidney-yang feelings color space model and default pseudo- fire behavior color
Spatial model is matched, and the first matching degree and the second matching degree are obtained;If the first matching degree is higher than the second matching degree, it is determined that
The testing result in current interest region is to pass through.Specific detection mode can refer to the phase in two step 260 of the embodiment of the present invention
Close description.
Step 363 judges whether current interest region is the last one area-of-interest, if so, terminating flow;It is no
Then, step 364 is executed.
Step 364 sets next area-of-interest to new current interest region, returns to step 361.
Step 365 carries out Luminance Distribution detection to the image in current interest region, judges whether testing result leads to
It crosses, if so, thening follow the steps 366;Otherwise, step 363 is executed.
Illustratively, the regularity of distribution that flame pixels brightness can be analyzed, by combustion centre's point of internal flame to flame envelope edge, by
In the difference of ignition temperature, colour brightness can successively successively decrease, using luminance histogram statistics grey scale change rule, if currently feeling emerging
Interesting region meets the grey scale change rule, then it is believed that detection passes through.
Step 366 carries out motion feature detection to the image in current interest region, judges whether testing result leads to
It crosses, if so, thening follow the steps 367;Otherwise, step 363 is executed.
Illustratively, using gauss hybrid models (Gaussian Mixture Model, GMM;Also known as mixed Gaussian mould
Type) background modeling technology carries out doubtful fire point screening, and interference caused by moving the objects such as car light with further eliminations takes into account simultaneously
Movement that may be present of flickering by a small margin is put to fire.For example, current interest region and utilization gauss hybrid models background modeling
When the master pattern matching that technology is established, it is believed that detection passes through.
Step 367 determines in current interest region there is true fiery point, and executes step 363.
It should be noted that when being determined in this step in current region in the presence of true fire point, step 370 progress can be first carried out
Fire point alarm, warning relevant staff take fire extinguishing related measure or control relevant device to enter fire extinguishing flow, continue simultaneously
Step 363 is executed, verifies in next area-of-interest whether there is also true fiery points.
Step 370 carries out fire point alarm.
A kind of double spectrum fire monitorings being preferably based on Infrared-Visible image that the embodiment of the present invention three provides
Method filters out the doubtful fiery point that heat signature meets testing conditions, then by doubtful fiery point using infrared heat point detection algorithm
It is mapped in visible images, then carries out the ROI region after the continuation of several pixels as visible images, finally utilize visible
The doubtful fire point verification algorithm of light verifies ROI region, to filter out true fire point and carry out fire point alarm.This method
Accuracy it is high, and calculation amount is small, and efficiency can be improved, and ensures the real-time of fire monitoring.
Example IV
Fig. 6 is a kind of double spectrum fire monitorings based on Infrared-Visible image that the embodiment of the present invention four provides
The structure diagram of device, the device can be generally integrated in fire monitoring system by software and or hardware realization, can
Forest fire is monitored by executing double spectrum fire monitoring methods based on Infrared-Visible image.As shown in fig. 6,
The device includes that image collection module 601, doubtful fire point marked region determining module 602, mapping block 603 and true fire point are true
Cover half block 604.
Wherein, image collection module 601, infrared image and visible images for obtaining current scene;Doubtful fire point
Marked region determining module 602, for determining the doubtful fire point marked region in the infrared image;Mapping block 603, is used for
The doubtful fire point marked region is mapped to the corresponding position in the visible images, is obtained in the visible images
Area-of-interest;True fire point determining module 604, for being verified to the image in the area-of-interest, according to verification
As a result it determines in the current scene with the presence or absence of true fiery point.
Double spectrum fire monitoring devices based on Infrared-Visible image that the embodiment of the present invention four provides, will be red
Outer image detection identification advantage is combined with visible images detection identification advantage, first true according to the imaging features of infrared image
Fixed doubtful fire point marked region, then doubtful fire is put into marked region and is mapped in visible images, appearance and flame phase can be eliminated
Interference like Du Gao but the radiation energy object different from flame to judging result promotes accuracy.Also, in visible images
In only the image in area-of-interest is verified, calculation amount can be reduced, raising efficiency ensures the real-time of fire monitoring.
On the basis of the above embodiments, the doubtful fire puts marked region concretely doubtful Huo Dian rectangle markeds area
Domain;The doubtful fire point marked region determining module is particularly used in:According to the brightness value characteristic distributions in the infrared image
By the infrared Image Segmentation at multiple regions, use rectangle marked by brightness value higher than the region of predetermined luminance threshold value draw a circle to approve for
Doubtful fire point rectangle marked region.
On the basis of the above embodiments, described image acquisition module is particularly used in:Obtain the infrared figure of current scene
Image set closes and visible images, corresponds to the infrared image of n different moments in the infrared image set comprising n frames.It is described to doubt
It may include that marking unit, target bianry image form unit and doubtful Huo Dian rectangle markeds area like fire point marked region determining module
Domain determination unit.Wherein, marking unit is used for for each frame infrared image in the infrared image set, according to brightness
Brightness value at multiple regions, is higher than predetermined luminance threshold value by current infrared Image Segmentation by Distribution value feature using rectangle marked
Region delineation is rectangle marked region;Pixel in the rectangle marked region in the current infrared image is labeled as 1,
Pixel outside rectangle marked region is labeled as 0.Target bianry image forms unit, for counting in the infrared image set
Different infrared images in same position at pixel be marked as 1 number, using voting mechanism by be marked as 1 time
Number is denoted as 1 not less than the pixel of preset times, other pixels are denoted as 0, forms target bianry image.Doubtful Huo Dian rectangle markeds area
Domain determination unit, for carrying out contours extract to the target bianry image, and using the doubtful fire point rectangle of rectangle marked delineation
Marked region.
On the basis of the above embodiments, the mapping block is particularly used in:It is doubted described using image transform model
It is mapped to the corresponding position in the visible images like fire point marked region, obtains the region of interest in the visible images
Domain;
Described image transformation model is specially:
Wherein, C1Indicate infrared camera, C2Indicate Visible Light Camera;R is C2Relative to C1Spin matrix, t C2Relatively
In C1Motion vector;(u1,v1) it is coordinate of the spatial point in infrared image, (u2,v2) be the spatial point in visible light figure
Coordinate as in;K1For C1Intrinsic Matrix, K2For C2Intrinsic Matrix;Zc1For the spatial point to C1Image plane away from
From Zc2For the spatial point to C2The distance of image plane.
On the basis of the above embodiments, the mapping block is particularly used in:By the doubtful fire point marked region
Default pixel of the outside continuation of profile forms final doubtful fire point marked region;The final doubtful fire point marked region is reflected
It is mapped to the corresponding position in the visible images, obtains the area-of-interest in the visible images.
On the basis of the above embodiments, the true fire point determining module is particularly used in:By current interest region
It is matched with default kidney-yang feelings color space model, obtains the first matching degree;By the current interest region and default puppet
Fire behavior color space model is matched, and the second matching degree is obtained;If first matching degree is higher than second matching degree,
Determine that current interest region is that doubtful fire puts region;All doubtful fire point regions in the visible images are tested
Card determines in the current scene according to verification result with the presence or absence of true fiery point.
The double spectrum fire monitoring devices based on Infrared-Visible image provided in above-described embodiment are executable originally
Double spectrum fire monitoring methods based on Infrared-Visible image that invention any embodiment is provided, having execution should
The corresponding function module of method and advantageous effect.The not technical detail of detailed description in the above-described embodiments, reference can be made to of the invention
Double spectrum fire monitoring methods based on Infrared-Visible image that any embodiment is provided.
Note that above are only presently preferred embodiments of the present invention and institute's application technology principle.It will be appreciated by those skilled in the art that
The present invention is not limited to specific embodiments described here, can carry out for a person skilled in the art it is various it is apparent variation,
It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out to the present invention by above example
It is described in further detail, but the present invention is not limited only to above example, without departing from the inventive concept, also
May include other more equivalent embodiments, and the scope of the present invention is determined by scope of the appended claims.
Claims (6)
1. a kind of double spectrum fire monitoring methods based on Infrared-Visible image, which is characterized in that including:
It is a not comprising n frames to correspond to n in the infrared image set for the infrared image set and visible images for obtaining current scene
Infrared image in the same time;
For each frame infrared image in the infrared image set, current infrared image is divided according to brightness value characteristic distributions
Multiple regions are cut into, region of the rectangle marked by brightness value higher than predetermined luminance threshold value is used to draw a circle to approve as rectangle marked region;It will
The pixel in the rectangle marked region in the current infrared image is labeled as 1, the element marking outside rectangle marked region
It is 0;
Count time that the pixel at the same position in the different infrared images in the infrared image set is marked as 1
The number for being marked as 1 is denoted as 1 by number using voting mechanism not less than the pixel of preset times, other pixels are denoted as 0, are formed
Target bianry image;
Contours extract is carried out to the target bianry image, and using the doubtful fire point marked region of rectangle marked delineation, wherein institute
It is specially doubtful fire point rectangle marked region to state doubtful fire point marked region;
The doubtful fire point marked region is mapped to the corresponding position in the visible images, obtains the visible images
In area-of-interest;
Image in the area-of-interest is verified, is determined in the current scene with the presence or absence of true according to verification result
Excess fire point.
2. according to the method described in claim 1, it is characterized in that, the doubtful fire point marked region is mapped to described visible
Corresponding position in light image obtains the area-of-interest in the visible images, including:
The doubtful fire point marked region is mapped to by the corresponding position in the visible images using image transform model, is obtained
To the area-of-interest in the visible images;
Described image transformation model is specially:
Wherein, C1Indicate infrared camera, C2Indicate Visible Light Camera;R is C2Relative to C1Spin matrix, t C2Relative to C1
Motion vector;(u1,v1) it is coordinate of the spatial point in infrared image, (u2,v2) be the spatial point in visible images
Coordinate;K1For C1Intrinsic Matrix, K2For C2Intrinsic Matrix;Zc1For the spatial point to C1The distance of image plane, Zc2
For the spatial point to C2The distance of image plane.
3. according to the method described in claim 1, it is characterized in that, the doubtful fire point marked region is mapped to described visible
Corresponding position in light image obtains the area-of-interest in the visible images, including:
By the pixel of the outside continuation preset quantity of profile of the doubtful fire point marked region, final doubtful fire point mark zone is formed
Domain;
The final doubtful fire point marked region is mapped to the corresponding position in the visible images, obtains the visible light
Area-of-interest in image.
4. according to the method described in claim 1, it is characterized in that, being verified to the image in the area-of-interest, root
It is determined with the presence or absence of true fiery point in the current scene according to verification result, including:
Current interest region is matched with default kidney-yang feelings color space model, obtains the first matching degree;Work as by described in
Preceding area-of-interest is matched with default pseudo- fire behavior color space model, obtains the second matching degree;
If first matching degree is higher than second matching degree, it is determined that current interest region is that doubtful fire puts region;
All doubtful fire point regions in the visible images are verified, the current scene is determined according to verification result
In with the presence or absence of true fire point.
5. a kind of double spectrum fire monitoring devices based on Infrared-Visible image, which is characterized in that including:
Image collection module, the infrared image set for obtaining current scene and visible images, the infrared image set
In comprising n frames correspond to the infrared images of n different moments;
Doubtful fire point marked region determining module, including marking unit, target bianry image form unit and doubtful fire point square
Shape marked region determination unit;
The marking unit is used for for each frame infrared image in the infrared image set, special according to brightness Distribution value
Point by current infrared Image Segmentation at multiple regions, draw a circle to approve by the region using rectangle marked by brightness value higher than predetermined luminance threshold value
For rectangle marked region;Pixel in the rectangle marked region in the current infrared image is labeled as 1, rectangle marked
Pixel outside region is labeled as 0;
The target bianry image forms unit, for counting the phase in the different infrared images in the infrared image set
It is marked as 1 number with the pixel at position, the number for being marked as 1 is not less than to the picture of preset times using voting mechanism
Element is denoted as 1, other pixels are denoted as 0, forms target bianry image;
The doubtful fire point rectangle marked area determination unit, for carrying out contours extract to the target bianry image, and is adopted
With rectangle marked delineation doubtful fire point rectangle marked region, wherein the doubtful fire point marked region is specially doubtful fire point square
Shape marked region;
Mapping block is obtained for the doubtful fire point marked region to be mapped to the corresponding position in the visible images
Area-of-interest in the visible images;
True fire point determining module determines institute for being verified to the image in the area-of-interest according to verification result
It states in current scene with the presence or absence of true fiery point.
6. device according to claim 5, which is characterized in that the mapping block is specifically used for:
The doubtful fire point marked region is mapped to by the corresponding position in the visible images using image transform model, is obtained
To the area-of-interest in the visible images;
Described image transformation model is specially:
Wherein, C1Indicate infrared camera, C2Indicate Visible Light Camera;R is C2Relative to C1Spin matrix, t C2Relative to C1
Motion vector;(u1,v1) it is coordinate of the spatial point in infrared image, (u2,v2) be the spatial point in visible images
Coordinate;K1For C1Intrinsic Matrix, K2For C2Intrinsic Matrix;Zc1For the spatial point to C1The distance of image plane, Zc2
For the spatial point to C2The distance of image plane.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610029028.7A CN105488941B (en) | 2016-01-15 | 2016-01-15 | Double spectrum fire monitoring method and devices based on Infrared-Visible image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610029028.7A CN105488941B (en) | 2016-01-15 | 2016-01-15 | Double spectrum fire monitoring method and devices based on Infrared-Visible image |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105488941A CN105488941A (en) | 2016-04-13 |
CN105488941B true CN105488941B (en) | 2018-10-30 |
Family
ID=55675901
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610029028.7A Active CN105488941B (en) | 2016-01-15 | 2016-01-15 | Double spectrum fire monitoring method and devices based on Infrared-Visible image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105488941B (en) |
Families Citing this family (30)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106056594A (en) * | 2016-05-27 | 2016-10-26 | 四川桑莱特智能电气设备股份有限公司 | Double-spectrum-based visible light image extraction system and method |
CN105931409A (en) * | 2016-05-30 | 2016-09-07 | 重庆大学 | Infrared and visible light camera linkage-based forest fire monitoring method |
CN106228732A (en) * | 2016-08-22 | 2016-12-14 | 无锡信大气象传感网科技有限公司 | A kind of forest fire intelligent identifying system |
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 |
CN109670388B (en) * | 2017-10-17 | 2021-04-23 | 杭州萤石网络有限公司 | Target behavior detection method and device, electronic equipment and storage medium |
CN108090495A (en) * | 2017-12-22 | 2018-05-29 | 湖南源信光电科技股份有限公司 | A kind of doubtful flame region extracting method based on infrared light and visible images |
CN108460804A (en) * | 2018-03-20 | 2018-08-28 | 重庆大学 | A kind of Three Degree Of Freedom position and posture detection method of transhipment docking mechanism and transhipment docking mechanism based on machine vision |
CN109146904A (en) * | 2018-08-13 | 2019-01-04 | 合肥英睿系统技术有限公司 | The method and apparatus of infrared image object profile is shown in visible images |
CN109243135A (en) * | 2018-09-26 | 2019-01-18 | 北京环境特性研究所 | A kind of intelligence fire detection and localization method, apparatus and system |
CN109375068B (en) * | 2018-09-26 | 2021-02-05 | 北京环境特性研究所 | Target identification method and device based on ultraviolet imaging corona detection |
CN110060444A (en) * | 2019-03-11 | 2019-07-26 | 视联动力信息技术股份有限公司 | A kind of fire early-warning system and method based on view networking |
CN109854964B (en) * | 2019-03-29 | 2021-03-19 | 沈阳天眼智云信息科技有限公司 | Steam leakage positioning system and method based on binocular vision |
CN110009530A (en) * | 2019-04-16 | 2019-07-12 | 国网山西省电力公司电力科学研究院 | A kind of nerve network system and method suitable for portable power inspection |
CN110097030A (en) * | 2019-05-14 | 2019-08-06 | 武汉高德红外股份有限公司 | It is a kind of based on infrared and visible images protrusion identification methods and system |
CN110244011A (en) * | 2019-06-26 | 2019-09-17 | 熊颖郡 | The river blowdown of unmanned plane monitors analyzing and alarming system automatically |
CN110390788A (en) * | 2019-08-21 | 2019-10-29 | 深圳云感物联网科技有限公司 | A kind of forest fire protection firework identification method and its system |
CN110619293A (en) * | 2019-09-06 | 2019-12-27 | 沈阳天眼智云信息科技有限公司 | Flame detection method based on binocular vision |
CN110569797B (en) * | 2019-09-10 | 2023-05-26 | 云南电网有限责任公司带电作业分公司 | Method, system and storage medium for detecting mountain fire of geostationary orbit satellite image |
CN111199629B (en) * | 2020-02-18 | 2021-11-23 | 普宙科技(深圳)有限公司 | Heat source identification device, unmanned aerial vehicle and heat source identification method |
CN113537204A (en) * | 2020-04-20 | 2021-10-22 | 富华科精密工业(深圳)有限公司 | Small flame detection method based on infrared features and machine learning and computer device |
CN113626377A (en) * | 2020-05-06 | 2021-11-09 | 杭州海康微影传感科技有限公司 | Bare data storage control method, device and equipment and storage medium |
CN112257554B (en) * | 2020-10-20 | 2021-11-05 | 南京恩博科技有限公司 | Forest fire recognition method, system, program and storage medium based on multiple spectra |
CN112257667A (en) * | 2020-11-12 | 2021-01-22 | 珠海大横琴科技发展有限公司 | Small ship detection method and device, electronic equipment and storage medium |
CN112614302B (en) * | 2020-12-03 | 2022-10-04 | 杭州海康微影传感科技有限公司 | Fire detection method, device and system and electronic equipment |
CN112560657B (en) * | 2020-12-12 | 2023-05-30 | 南方电网调峰调频发电有限公司 | Method, device, computer device and storage medium for identifying smoke and fire |
CN112860946B (en) * | 2021-01-18 | 2023-04-07 | 四川弘和通讯集团有限公司 | Method and system for converting video image information into geographic information |
CN113283322A (en) * | 2021-05-14 | 2021-08-20 | 柳城牧原农牧有限公司 | Livestock trauma detection method, device, equipment and storage medium |
CN113237556A (en) * | 2021-05-18 | 2021-08-10 | 深圳市沃特沃德信息有限公司 | Temperature measurement method and device and computer equipment |
CN114627124B (en) * | 2022-05-16 | 2022-07-26 | 江西武大扬帆科技有限公司 | Deep learning-based bubble spring detection method and interactive feedback system |
CN115546727A (en) * | 2022-10-20 | 2022-12-30 | 浙江华感科技有限公司 | Method and system for judging fire condition and electronic equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN201629019U (en) * | 2009-09-30 | 2010-11-10 | 青岛科恩锐通信息技术有限公司 | Forest fire detecting system |
CN102693603A (en) * | 2012-06-26 | 2012-09-26 | 山东神戎电子股份有限公司 | Dual spectrum based intelligent monitoring system for forest fire prevention |
JP2014093002A (en) * | 2012-11-05 | 2014-05-19 | Hochiki Corp | Flame detection apparatus and flame detection method |
CN104867265A (en) * | 2015-04-22 | 2015-08-26 | 深圳市佳信捷技术股份有限公司 | Camera apparatus, and fire detection alarm system and method |
CN204667578U (en) * | 2015-06-24 | 2015-09-23 | 山东神戎电子股份有限公司 | The two spectrum observation instrument of a kind of hand-held forest fire protection |
-
2016
- 2016-01-15 CN CN201610029028.7A patent/CN105488941B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN201629019U (en) * | 2009-09-30 | 2010-11-10 | 青岛科恩锐通信息技术有限公司 | Forest fire detecting system |
CN102693603A (en) * | 2012-06-26 | 2012-09-26 | 山东神戎电子股份有限公司 | Dual spectrum based intelligent monitoring system for forest fire prevention |
JP2014093002A (en) * | 2012-11-05 | 2014-05-19 | Hochiki Corp | Flame detection apparatus and flame detection method |
CN104867265A (en) * | 2015-04-22 | 2015-08-26 | 深圳市佳信捷技术股份有限公司 | Camera apparatus, and fire detection alarm system and method |
CN204667578U (en) * | 2015-06-24 | 2015-09-23 | 山东神戎电子股份有限公司 | The two spectrum observation instrument of a kind of hand-held forest fire protection |
Also Published As
Publication number | Publication date |
---|---|
CN105488941A (en) | 2016-04-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105488941B (en) | Double spectrum fire monitoring method and devices based on Infrared-Visible image | |
CN108596101B (en) | Remote sensing image multi-target detection method based on convolutional neural network | |
CN106225787B (en) | Unmanned aerial vehicle visual positioning method | |
CN104778721B (en) | The distance measurement method of conspicuousness target in a kind of binocular image | |
CN103442209B (en) | Video monitoring method of electric transmission line | |
CN107202982A (en) | A kind of beacon arrangement calculated based on UAV position and orientation and image processing method | |
CN111462128B (en) | Pixel-level image segmentation system and method based on multi-mode spectrum image | |
CN111126325A (en) | Intelligent personnel security identification statistical method based on video | |
CN112686172B (en) | Airport runway foreign matter detection method, device and storage medium | |
CN109190444A (en) | A kind of implementation method of the lane in which the drivers should pay fees vehicle feature recognition system based on video | |
CN112308156B (en) | Two-stage image change detection method based on counterstudy | |
CN110197259A (en) | Wafer defect detection method based on small lot data set Yu deep learning algorithm | |
CN114202646A (en) | Infrared image smoking detection method and system based on deep learning | |
CN109740522A (en) | A kind of personnel's detection method, device, equipment and medium | |
CN116311078A (en) | Forest fire analysis and monitoring method and system | |
CN105701805A (en) | Pork intramuscular fat content nondestructive testing method based on computer vision | |
CN104573662B (en) | A kind of cloud sentences method and system | |
CN106251337A (en) | A kind of drogue space-location method and system | |
CN106526558B (en) | Gust front automatic identifying method based on Doppler weather radar data | |
CN114202695A (en) | Remote sensing image automatic identification system based on artificial intelligence technology | |
CN109064444A (en) | Track plates Defect inspection method based on significance analysis | |
CN114037910A (en) | Unmanned aerial vehicle forest fire detecting system | |
CN108520255A (en) | A kind of method for detecting infrared puniness target and device | |
CN102194249B (en) | Water current modeling data capturing device with combination of infrared rays and visible light | |
CN111898427A (en) | Multispectral pedestrian detection method based on feature fusion deep neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |