CN201117049Y - Double-waveband image identification fire hazard detecting alarming system - Google Patents
Double-waveband image identification fire hazard detecting alarming system Download PDFInfo
- Publication number
- CN201117049Y CN201117049Y CNU2007200261613U CN200720026161U CN201117049Y CN 201117049 Y CN201117049 Y CN 201117049Y CN U2007200261613 U CNU2007200261613 U CN U2007200261613U CN 200720026161 U CN200720026161 U CN 200720026161U CN 201117049 Y CN201117049 Y CN 201117049Y
- Authority
- CN
- China
- Prior art keywords
- detector
- image
- video
- infrared
- waveband
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Fire-Detection Mechanisms (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
The utility model discloses a dual range image recognition fire-detecting alarm system, which solves the problem of low accuracy of alarm work which is caused by that current image alarms are easy to alarm falsely and has the advantages of simple structure and convenient use, and can effectively improve the accuracy rate of alarms and the like. The structure of the dual range image recognition fire-detecting alarm system is that the system at least comprises a detector and a power supply, wherein the detector is a dual range image detector, the output end of the detector is connected with a video distributor, one output of the video distributor is connected with a conventional closed circuit monitor system in one way and the other output is connected with an image acquisition device in the other way, the image acquisition device is connected with a micro-computer, the micro-computer is connected with a movement controller and the movement controller is respectively connected with an alarm device and a rescue equipment.
Description
Technical field
The utility model relates to a kind of warning device, relates in particular to a kind of two waveband image recognition fire detection alarm system.
Background technology
Along with the fast development of China's economy, novel, large-scale, high-rise, inferior all kinds of volumed space building emerge in multitudes, the storage scale enlarges day by day.The fireproofing installation of building interior, traditional method are to use temperature-sensitive, sense cigarette, sensitive detector and infrared to taking the photograph detector, but for the building and the open-air occasion of large space, traditional detection method can not adapt to the fire-fighting requirement in these places.For the volumed space building that surveillance equipment is installed, if can be by the image information of camera acquisition being done carry out detection when certain analysis is discerned, then both effectively overcome the shortcoming that Conventional detectors descends in the sensitivity of large space occasion, simultaneously can realize monitoring function, therefore, the image recognition fire detection technology arises at the historic moment.The smog that produces when fire takes place, flame color, brightness, shape, the trend that spreads, the visual signatures such as infrared and ultraviolet radiation of self, become the main foundation of image-type detector, it is fast that image-type fire detector has response speed, survey in time, advantages such as abundant information, existing image-type fire detection technology, angle from color is arranged, utilize the color characteristic of flame to differentiate (as Chinese patent CN1112702A), perhaps utilize the characteristics that the flame region area increases in the fire spread process to differentiate, there is the infrared light of the Fire Radiation of utilizing to differentiate (as Chinese patent CN1334476A), has smog in the collection monitoring scene to discern to judge whether exist (as the Chinese patent CN1427253A) of fire by image.But list is from color or infrared radiation goes to analyze the flame vision signal, lack reliability in many occasions, a lot of scenes or object has the same color of flame, even has a feature that brightness changes, be not to have only flame just to produce infrared and ultraviolet light equally, when these disturbing factors occur, tend to cause the wrong report or the slow newspaper of detector, influenced the accuracy of warning work.
The utility model content
The purpose of this utility model is reported by mistake easily in order to solve present images category alarm exactly, the not high problem of work accuracy causes reporting to the police, provide a kind of have simple in structure, easy to use, can effectively improve the two waveband image recognition fire detecting and alarm device and the method for supervising thereof of advantages such as warning accuracy rate.
For achieving the above object, the utility model adopts following technical scheme:
A kind of two waveband image recognition fire detection alarm system, it comprises at least one detector and power supply, described detector is the two waveband image detector; The detector output terminal is connected with video distributor, video distributor one tunnel outputs to conventional closed monitor system, and another road output is connected with image capture device, and image capture device is connected with microcomputer, microcomputer then is connected with linkage controller, and linkage controller is connected with warning, rescue aid respectively.
Described two waveband image detector comprises a pair of CCD camera head of arranging side by side, and infrared fileter is installed before one of them CCD camera head, forms the vision signal of infrared image and two kinds of different-wavebands of coloured image.
Described infrared fileter be installed in before the CCD photo-sensitive cell or camera head lens before or before the detector window or between camera and the CCD photo-sensitive cell; Or be attached on the CCD photo-sensitive cell.
Be provided with video switcher between described video distributor and image capture device.
Detection method of the present utility model is,
(1) at least one detector is fixed in the monitored area, the two-path video signal of each detector is after video distributor is handled, and one the tunnel outputs to conventional closed monitor system; Another road outputs to the video inputs of switch;
(2) switch adopts the two-way way of output, is respectively color channel and infrared channel, directly delivers to the image pick-up card input end;
(3) image pick-up card is converted to digital picture with vision signal and deposits in the calculator memory, carries out the flame comparison and judges, in case determine breaking out of fire, then sends the interlock control signal by computing machine, carries out fire alarm.
Flame comparison deterministic process is in the described step (3), the data formation image that computing machine reads in the specified memory shows on screen, simultaneously each pixel in the image is scanned, each pixel value is decomposed into three kinds of color components of RGB, compare with the flame color threshold value that is provided with in the software, judge this color of pixel feature, read different view data constantly, analyzing same pixel changes in difference brightness constantly, with change frequency and given threshold ratio, judge whether this pixel meets the feature of flicker, result in conjunction with twice judgement, there is not suspicious part if think in this scene, then computing machine two output terminals sending the instruction control video switcher switch to the colour and the infrared channel of next detector, if think that certain piece zone exists flame unusual, then computer program begins the infrared channel image of this detector is analyzed, infrared channel need not to form image and shows in screen, the data that computing machine directly reads in the internal memory are made background process, the intensity of infrared radiation often shows as the power of brightness in image, the gradation of image value is carried out continuous sweep, by determining suspicious region in the image with given threshold ratio, the abnormal area that two passages are determined carries out correlation analysis at last, judge that with this whether really to have flame in this detector monitors scope unusual, in case determine that the abnormal area that two passages provide is the same area, promptly be identified as fire.
The fire alarm procedure is in the step (3), computing machine sends the interlock control signal, linkage controller is connected to a plurality of output terminals, control panalarm, fire xegregating unit, the equipment of putting out a fire to save life and property, video recording apparatus etc. respectively, simultaneous computer sends signal makes video switcher lock this detector, and realization warning picture is monitored in real time.
The utility model is made of at least one pair of colored CCD video element, filtering apparatus, control circuit, switching device, image pick-up card and mini-computer, filtering apparatus is installed in wherein a CCD front end or camera lens front end, make it can filter the visible light composition, only survey the near-infrared radiation of scenery.According to the quantity that detector is installed, if detector, the two-path video output signal is the incoming image capture card directly, many detectors can switch one tunnel vision signal by order successively by video switching apparatus gives image pick-up card, this moment, the two-path video signal adopted the also mode of line output time-division processing, have only one tunnel vision signal to be input to capture card, capture card is saved in view data in the memory field that computing machine distributes in advance by pci bus.Computing machine switches to the color channel of certain detector earlier according to the instruction control video switching apparatus, and read data in the memory field, on monitor, form chromatic image, while is according to the video features of fire, by color to chromatic image, texture, the judgement of flicker number of times, in time judge whether color channel has fire to take place, if it is unusual that color channel detects fire, with recognition result mainly is that positional information is saved in the computing machine, computing machine sends the infrared channel that the instruction control video switching apparatus switches to this detector immediately then, the infrared channel that has filtering apparatus can absorb imaging to the infrared radiation light in the monitoring scene, computing machine is after obtaining the video of this passage, according to radiation intensity and radiation variation feature view data is analyzed, exist fire unusually then mark off the fire suspicious region if be judged as, and color channel identified region and the infrared channel identified region preserved compared, after being defined as same target, send alarm output signal, this signal will be controlled a plurality of linkages, start warning device respectively, the equipment of putting out a fire to save life and property, video recording apparatus etc.
Video display process passage of the present utility model can be realized in the following way: adopt the video switching apparatus with the output of multichannel input multichannel, the parallel control of video output channel is not disturbed mutually, can control the vision signal that it switches to appointment respectively, such as controlling the vision signal that output channel 1 shows input channel 3, control output channel 2 shows the vision signal of input channel 4, output channel 1 and 2 can be exported simultaneously, need not to switch.Make switching device output two-path video signal, export the color video and the infrared video of same detector respectively.Corresponding image pick-up card adopts supports the multichannel input to show processing capacity simultaneously simultaneously, the output of switch two-path video is connected to the two-path video input end of capture card respectively, handle when computing machine can be to the two-path video information of same, only monitor at screen display color channel video for convenient, hide infrared channel, data to infrared channel are made background process, and twin-channel data analysis and identification determination methods are same as above.
Detector installation method of the present utility model can be in the following way: with the detector monomer solid around monitoring of environmental; quantity according to the visual field and the protection domain of every detector determine to install detector makes all detectors can protect whole monitored area effectively.Determine the input channel of video switching apparatus according to detector quantity, too much detector can adopt the switching device cascade system to carry out video switch.The advantage of this mode is each detector independent operating, and input signal is stable, has reduced interference, has improved system reliability, and shortcoming is need more relatively equipment.
Another kind of mode can be installed in detector on the rotatable The Cloud Terrace, The Cloud Terrace is fixed on monitoring of environmental central authorities or all around, detector can rotate with The Cloud Terrace, the rotation of The Cloud Terrace is realized by cradle head control device drive motor, the cradle head control device is connected with computing machine, and computing machine rotates or the locking current location according to recognition result operation cradle head control device.The advantage of this access way is that the monitoring range of every detector is big, needs equipment few, and shortcoming is to have increased the cradle head control part, and the rotation of The Cloud Terrace has reduced video quality and has been not easy to identification.
The infrared imaging principle that the utility model used: in theory, as long as its temperature is higher than absolute zero (273.15 ℃), just there is the random motion of molecule and atom in all objects of occurring in nature, and its surface is infrared radiation constantly.Any object that has temperature, except that sending wavelength the visible light of 380～770nm, can also launch is not the infrared ray of 770～1350nm scope for the being seen wavelength of human eye.
Infrared radiation during flame combustion mainly concentrates on 950～2000nm wave band, and when using CCD to carry out the infrared view collection as sensitive component, CCD self has possessed the photoperceptivity of 400～1200nm, has contained ultrared fundamental wavelength spectrum.Just under general situation since the light quantity of visible light much larger than infrared light, thereby the image that collects can not be found out ultra-red ray effect.So native system adopts and installs CCD camera front end additional infrared fileter, this optical filter is the filtering visible light wave range effectively, make the infrared radiation of monitoring in the scene form image by optical filter smoothly, and a large amount of infrared rays of radiation just become a foundation of infrared image identification during flame combustion.
The beneficial effects of the utility model are:
1. monitoring camera-shooting and detection double dominant are arranged, colour TV camera can provide the on-the-spot convenient monitoring of image/video, show scope and particular location that fire takes place, under the support of recognizer, can reach the purpose of detection of fires simultaneously again in conjunction with the two waveband image information.
2. the video image of visible light wave range and infrared band can be provided simultaneously,, make that carrying out detection from many aspects such as color, brightness, texture, flicker, infrared radiations simultaneously becomes possibility for the identification of fire provides abundant image information.
3. this apparatus structure compactness, simple, control are easily, switch speed is fast, security performance good, failure rate is low.
4. the picked-up of infrared image adopts common CCD to add the mode of infrared fileter combination, conveniently install, simple in structure, need not the volume additional supply; Colored and infrared video is exported simultaneously, can arbitrarily switch.
5. can work in multiple mounting means, monitor area is wide, and protection domain is big, is subjected to space constraint little, is fit to large space and the interior fire monitorings of buildings on a large scale such as shop, large-scale Room factories and miness.
Description of drawings
Fig. 1 is a structured flowchart of the present utility model;
Fig. 2 is the detector front view;
Fig. 3 is Fig. 2 side sectional view.
Wherein, 1. protective cover, 2. camera lens, 3.CCD, 4. optical filter, 5. leading-out terminal, 6. detector, 7. video distributor, 8. conventional closed monitor system, 9. image capture device, 10. microcomputer, 11. linkage controllers, 12. video switchers.
Embodiment
The utility model is described in further detail below in conjunction with accompanying drawing and embodiment.
The utility model total system is made of equipment such as two waveband image detector 6, mini-computer 10, image capture device 9 (as the DH-CG series of company of Daheng), video switcher 12, conventional closed monitor system 8, video distributor 7 and linkage controllers 11.Detector 6 output termination video distributors 7, video distributor 7 outputs divide two-way, one the tunnel send conventional closed monitor system 8, video switcher 12 is sent on another road, video switcher 12 is connected with image capture device 9, image capture device 9 connects microcomputer 10, and 10 of microcomputers are connected with linkage controller 11, and linkage controller 11 is controlled warning device respectively, the equipment of putting out a fire to save life and property, video recording apparatus etc.
Wherein, two waveband image detector 6 comprises the camera head of a pair of employing CCD3 photo-sensitive cell of arranging side by side, infrared fileter 4 is installed before the CCD3 photo-sensitive cell of one of them camera head, is formed the vision signal of infrared image and two kinds of different-wavebands of coloured image.Infrared fileter 4 can be installed in before the CCD3 photo-sensitive cell or camera head lens before or before the detector window or between camera and the CCD photo-sensitive cell 3; Or be attached on the CCD3 photo-sensitive cell.Being provided with protective cover 1 before camera lens 2, is leading-out terminal 5 in the rear end of detector 6.Before optical filter 4 is placed on the CCD3 photo-sensitive cell, its area size is as the criterion to cover the CCD3 photo-sensitive cell fully, because Infrared wavelength is more much longer than wavelength of visible light, its focal length can not be consistent with visible light, therefore need carry out infrared region to camera lens specially and proofread and correct, be that object of reference is adjusted camera lens 2 positions and made color video and Infrared video image clear with the naked light.
Detection method of the present utility model is:
Several detectors 6 are fixed in around the monitored area, detector 6 is by the unified power supply of 12V direct supply, with the two-path video signal of every detector 6 after video distributor 7 is handled, one the tunnel sends into conventional closed monitor system 8, video switcher 12 is sent on another road, if 6 of separate unit detectors are directly sent into image capture device 9.
Video switcher 12 adopts the two-way way of output, be respectively color channel and infrared channel, the input end of image capture device 9 is directly delivered in video switcher 12 outputs, image capture device 9 is connected with microcomputer 10 by the PCI expansion slot of microcomputer 10, image capture device 9 is converted to digital picture with vision signal and deposits in microcomputer 10 internal memories, microcomputer 10 shows on screen by the data formation image that reads in the specified memory, simultaneously each pixel in the image is scanned, each pixel value is decomposed into three kinds of color components of RGB, compare with the flame color threshold value that is provided with in the software, judge this color of pixel feature, read different view data constantly, analyzing same pixel changes in difference brightness constantly, with change frequency and given threshold ratio, judge whether this pixel meets the feature of flicker, result in conjunction with twice judgement, there is not suspicious part if think in this scene, then microcomputer 10 two output terminals sending the instruction control video switcher switch to the colour and the infrared channel of next detector, if think that certain piece zone exists flame unusual, then computer program begins the infrared channel image of this detector is analyzed, infrared channel need not to form image and shows in screen, the data that computing machine directly reads in the internal memory are made background process, the intensity of infrared radiation often shows as the power of brightness in image, the gradation of image value is carried out continuous sweep, by determining suspicious region in the image with given threshold ratio, the abnormal area that two passages are determined carries out correlation analysis at last, judge that with this whether really to have flame in this detector monitors scope unusual, in case determine that the abnormal area that two passages provide is the same area, promptly be identified as fire, computing machine sends the interlock control signal, linkage controller is connected to a plurality of output terminals, control panalarm respectively, the fire xegregating unit, the equipment of putting out a fire to save life and property, video recording apparatus etc., simultaneous computer sends signal makes video switcher lock this detector, and realization warning picture is monitored in real time.
Claims (4)
1, a kind of two waveband image recognition fire detection alarm system, it comprises at least one detector (6) and power supply, it is characterized in that: described detector (6) is the two waveband image detector; Detector (6) output terminal is connected with video distributor (7), video distributor (7) one tunnel outputs to conventional closed monitor system (8), another road output is connected with image capture device (9), image capture device (9) is connected with microcomputer (10), microcomputer (10) then is connected with linkage controller (11), and linkage controller (11) is connected with warning, rescue aid respectively.
2, two waveband image recognition fire detection alarm system according to claim 1, it is characterized in that: described two waveband image detector comprises the camera head of a pair of employing CCD photo-sensitive cell (3) of arranging side by side, the preceding installation infrared fileter of CCD photo-sensitive cell (3) (4) of one of them camera head, the vision signal of formation infrared image and two kinds of different-wavebands of coloured image.
3, two waveband image recognition fire detection alarm system according to claim 2 is characterized in that: described infrared fileter (4) is installed in before the preceding or camera head lens of CCD photo-sensitive cell (3) or before the detector window or between camera and the CCD photo-sensitive cell (3); Or be attached on the CCD photo-sensitive cell (3).
4, two waveband image recognition fire detection alarm system according to claim 1 is characterized in that: be provided with video switcher (12) between described video distributor (7) and image capture device (9).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNU2007200261613U CN201117049Y (en) | 2007-08-10 | 2007-08-10 | Double-waveband image identification fire hazard detecting alarming system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNU2007200261613U CN201117049Y (en) | 2007-08-10 | 2007-08-10 | Double-waveband image identification fire hazard detecting alarming system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN201117049Y true CN201117049Y (en) | 2008-09-17 |
Family
ID=39991625
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNU2007200261613U Expired - Fee Related CN201117049Y (en) | 2007-08-10 | 2007-08-10 | Double-waveband image identification fire hazard detecting alarming system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN201117049Y (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101574567B (en) * | 2009-06-08 | 2011-08-31 | 南京航空航天大学 | Computer vision technique based method and system for detecting and extinguishing fire disaster intelligently |
| CN104318700A (en) * | 2014-11-07 | 2015-01-28 | 莆田学院 | Dual-waveband video dual-core fire detector and front end device thereof |
| CN106530569A (en) * | 2016-10-17 | 2017-03-22 | 北京小米移动软件有限公司 | Method and device for fire monitoring |
| CN108629940A (en) * | 2018-05-02 | 2018-10-09 | 北京准视科技有限公司 | A kind of image-type fire detection alarm system |
| CN108765860A (en) * | 2018-05-02 | 2018-11-06 | 北京准视科技有限公司 | A kind of image-type fire detecting and alarm method |
-
2007
- 2007-08-10 CN CNU2007200261613U patent/CN201117049Y/en not_active Expired - Fee Related
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101574567B (en) * | 2009-06-08 | 2011-08-31 | 南京航空航天大学 | Computer vision technique based method and system for detecting and extinguishing fire disaster intelligently |
| CN104318700A (en) * | 2014-11-07 | 2015-01-28 | 莆田学院 | Dual-waveband video dual-core fire detector and front end device thereof |
| CN106530569A (en) * | 2016-10-17 | 2017-03-22 | 北京小米移动软件有限公司 | Method and device for fire monitoring |
| CN108629940A (en) * | 2018-05-02 | 2018-10-09 | 北京准视科技有限公司 | A kind of image-type fire detection alarm system |
| CN108765860A (en) * | 2018-05-02 | 2018-11-06 | 北京准视科技有限公司 | A kind of image-type fire detecting and alarm method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100511305C (en) | Double wave band image identification fire hazard detection alarming system and monitoring method | |
| CN101458865B (en) | Fire disaster probe system and method | |
| KR101635000B1 (en) | Fire detector and system using plural cameras | |
| US6937743B2 (en) | Process and device for detecting fires based on image analysis | |
| US4999614A (en) | Monitoring system using infrared image processing | |
| EP1275094B1 (en) | Early fire detection method and apparatus | |
| CN101527073B (en) | Fire detection system and method thereof | |
| CN201259717Y (en) | Fire disaster probe system | |
| CN107483889A (en) | The tunnel monitoring system of wisdom building site control platform | |
| CN201117049Y (en) | Double-waveband image identification fire hazard detecting alarming system | |
| KR102195706B1 (en) | Method and Apparatus for Detecting Intruder | |
| KR101835552B1 (en) | Control center system of working environment for smart factory | |
| CN108432232B (en) | Safety camera system | |
| CN110240036B (en) | Device and method for detecting electric bicycle trying to enter elevator | |
| CN203630902U (en) | Improved image fire-hazard detector | |
| CN101764999B (en) | Sub-camera video capture device | |
| CN109040711A (en) | A kind of image-pickup method, monitoring method and imaging sensor | |
| KR102113527B1 (en) | Thermal ip cctv system and method for detecting fire by using the same | |
| CN201091014Y (en) | Fire detecting device | |
| CN104284070A (en) | Infrared waterproof high-resolution security camera | |
| CN202306757U (en) | Embedded flame detection device based on double-waveband | |
| CN201435128Y (en) | Visible perimeter alarm invasion detector | |
| CN212966489U (en) | Elevator early warning system on electric motor car based on intelligent analysis technique | |
| CN111486537B (en) | Air conditioner with security and monitoring functions | |
| CN101106727A (en) | Method for fire detection based on flame color template |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080917 Termination date: 20100810 |