CN106845410A - A kind of flame identification method based on deep learning model - Google Patents
A kind of flame identification method based on deep learning model Download PDFInfo
- Publication number
- CN106845410A CN106845410A CN201710047239.8A CN201710047239A CN106845410A CN 106845410 A CN106845410 A CN 106845410A CN 201710047239 A CN201710047239 A CN 201710047239A CN 106845410 A CN106845410 A CN 106845410A
- Authority
- CN
- China
- Prior art keywords
- fish eye
- flame
- eye images
- images
- measurement model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000013136 deep learning model Methods 0.000 title claims abstract description 13
- 241000251468 Actinopterygii Species 0.000 claims abstract description 33
- 238000002594 fluoroscopy Methods 0.000 claims abstract description 31
- 238000005259 measurement Methods 0.000 claims abstract description 26
- 238000010586 diagram Methods 0.000 claims abstract description 18
- 238000001914 filtration Methods 0.000 claims abstract description 5
- 239000000779 smoke Substances 0.000 claims abstract description 5
- 238000013527 convolutional neural network Methods 0.000 claims abstract 2
- 230000000007 visual effect Effects 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 4
- 230000008859 change Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 239000000443 aerosol Substances 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- G06T3/08—
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20224—Image subtraction
Abstract
The present invention relates to a kind of flame identification method based on deep learning model, it includes that (1) gathers video information, reads each two field picture, then carries out gaussian filtering and obtains filtered fish eye images;(2) fish eye images inner parameter is corrected;(3) fish eye images external parameter is corrected;(4) Sphere Measurement Model is built, the fish eye images after correction is projected on Sphere Measurement Model, then removal is projected in the repeat region on Sphere Measurement Model, forms spherical diagram picture;(5) by going the interference information that smog is recognized to flame portion in smoke model removal spherical diagram picture;(6) dynamic area on spherical diagram picture is obtained;(7) dynamic area part is generated into fluoroscopy images;(8) regular fluoroscopy images, using fluoroscopy images as the input for having trained seven layer architecture convolutional neural networks, whether identification dynamic area is flame, if dynamic area is flame, into step (9), otherwise terminates this operation;(9) show recognition result and produce warning message.
Description
Technical field
The invention belongs to the communications field, and in particular to a kind of flame identification method based on deep learning model.
Background technology
With China's industrialization and the continuous improvement of urban and town level, modern installations large public building towards space it is big, enter
The complicated diversification direction of deep and broad function is developed, this for anti-pyrotechnics towards space it is big, enter the complicated diversification side of deep and broad function
To development, proposed also for Smoke prevention, fire prevention and the reliable, stable of fire-fighting safety system, high precision design and operation higher
Requirement.
At present, security against fire is effectively lifted for live perception by the correlation technique of message area, effectively lifting
Security against fire is asked for the important research that rapidity that scene of fire early warning is disposed, levels of accuracy are safety engineering fields
Topic.
At present, Chen Wenhui etc. devises a Metro Train Fire detection alarm based on fire hazard aerosol fog image-obscuring properties
System, the method detects the generation of fire using the image spatial domain differential method.Li Shiwei etc. is obtained after being filtered using infrared fileter
Filtering image realizes that flame object is extracted.Wang great writers etc. carry out flame by steps such as image segmentation, image enhaucament, feature extractions
Identification.The behavioral characteristics and static nature of the analysis flame such as Wang Zulong include edge shake, area change and shape, color, text
Reason etc..But, the design system of above-mentioned flame identification system there is a problem of following some common:
(1) traditional method flase drop or the likelihood ratio of missing inspection are higher;
(2) consider that live dynamic area whether there is flame characteristic, the complexity of algorithm is high;
(3) region of detection, may not be flame, or compare alike with the textural characteristics of flame, then be likely to appearance
The situation of flase drop;
(4) the fire scope visual field of observable is small.
The content of the invention
Goal of the invention:The present invention makes improvement for the problem that above-mentioned prior art is present, i.e., the invention discloses one kind
Flame identification method based on deep learning model.
Technical scheme:A kind of flame identification method based on deep learning model, comprises the following steps:
(1) video information is gathered by back-to-back two fisheye cameras, then reads each frame of video information of collection
The fish eye images of acquisition are then carried out gaussian filtering by image, obtain filtered fish eye images;
(2) aligning step (1) obtains the inner parameter of fish eye images, and subsequently into step (3), inner parameter includes tangential
Error, radial error and the light heart error margin;
(3) external parameter of fish eye images is corrected, into step (4);
(4) Sphere Measurement Model is built, the fish eye images after step (3) is corrected are projected on Sphere Measurement Model, then removal is thrown
Repeat region of the shadow on Sphere Measurement Model, forms spherical diagram picture, subsequently into step (5);
(5) interference recognized to flame portion by smog in the spherical diagram picture that goes smoke model removal step (4) to obtain
Information, subsequently into step (6);
(6) using dynamic area on Codebook methods acquisition spherical diagram picture is improved, subsequently into step (7);
(7) the dynamic area part for obtaining step (6) generates fluoroscopy images;
(8) fluoroscopy images that normalisation step (7) is obtained, using fluoroscopy images as having trained seven layer architecture convolutional Neural nets
The input of network, whether identification dynamic area is flame, if dynamic area is flame, into step (9), otherwise terminates this behaviour
Make;
(9) recognition result is transferred to slave computer, shows recognition result in slave computer and produce warning message.
Further, when the fish eye images in step (3) are converted to the figure of 640 × 480 pixels, if being examined on fish eye images
When measuring angle point number more than 300, step (3) is comprised the following steps:
(31) normalization that perspective view is certain, size determines is generated centered on the characteristic point for detecting on the Sphere Measurement Model
Image block;
(32) relevance degree of characteristic point pair is calculated within the scope of the determination in adjacent two field pictures,
(33) best match pair is finally given, optimizes the relative pose of camera.
Further, when the fish eye images in step (3) are converted to the figure of 640 × 480 pixels, if being examined on fish eye images
When measuring angle point number less than 300, step (3) is comprised the following steps:
(31) using the side and geometry detected on image, end point pair is obtained on Sphere Measurement Model,
(32) transformation relation of structure on image is obtained based on end point pair, so as to calculate and optimize the relative of camera
Posture, so that Optimal Parameters.
Further, when the angle in the repetition visual field of big field-of-view image in step (3) is more than 15 degree, step (3) includes
Following steps:
(31) fluoroscopy images of wide-angle are generated by Sphere Measurement Model,
(32) repeat to calculate its degree of correlation on the visual field in fluoroscopy images, so that feedback adjustment camera posture, after adjustment again
Secondary generation fluoroscopy images are optimized.
Beneficial effect:A kind of flame identification method based on deep learning model disclosed by the invention has following beneficial effect
Really:
1st, false drop rate or loss are low;
2nd, the fire scope visual field of observable is big.
Brief description of the drawings
Fig. 1 is a kind of flow chart of flame identification method based on deep learning model disclosed by the invention;
Fig. 2 is Sphere Measurement Model schematic diagram;
Fig. 3 is the fluoroscopy images schematic diagram based on spherical diagram picture.
Specific embodiment:
Specific embodiment of the invention is described in detail below.
As shown in Figures 1 to 3, a kind of flame identification method based on deep learning model, comprises the following steps:
(1) video information is gathered by back-to-back two fisheye cameras, then reads each frame of video information of collection
The fish eye images of acquisition are then carried out gaussian filtering by image, obtain filtered fish eye images;
(2) aligning step (1) obtains the inner parameter of fish eye images, and subsequently into step (3), inner parameter includes tangential
Error, radial error and the light heart error margin;
(3) external parameter of fish eye images is corrected, into step (4);
(4) Sphere Measurement Model is built, the fish eye images after step (3) is corrected are projected on Sphere Measurement Model, then removal is thrown
Repeat region of the shadow on Sphere Measurement Model, forms spherical diagram picture, subsequently into step (5);
(5) interference recognized to flame portion by smog in the spherical diagram picture that goes smoke model removal step (4) to obtain
Information, subsequently into step (6);
(6) using dynamic area on Codebook methods acquisition spherical diagram picture is improved, subsequently into step (7);
(7) the dynamic area part for obtaining step (6) generates fluoroscopy images;
(8) fluoroscopy images that normalisation step (7) is obtained, using fluoroscopy images as having trained seven layer architecture convolutional Neural nets
The input of network, whether identification dynamic area is flame, if dynamic area is flame, into step (9), otherwise terminates this behaviour
Make;
(9) recognition result is transferred to slave computer, shows recognition result in slave computer and produce warning message.
Further, when the fish eye images in step (3) are converted to the figure of 640 × 480 pixels, if being examined on fish eye images
When measuring angle point number more than 300, step (3) is comprised the following steps:
(31) normalization that perspective view is certain, size determines is generated centered on the characteristic point for detecting on the Sphere Measurement Model
Image block;
(32) relevance degree of characteristic point pair is calculated within the scope of the determination in adjacent two field pictures,
(33) best match pair is finally given, optimizes the relative pose of camera.
Further, when the fish eye images in step (3) are converted to the figure of 640 × 480 pixels, if being examined on fish eye images
When measuring angle point number less than 300, step (3) is comprised the following steps:
(31) using the side and geometry detected on image, end point pair is obtained on Sphere Measurement Model,
(32) transformation relation of structure on image is obtained based on end point pair, so as to calculate and optimize the relative of camera
Posture, so that Optimal Parameters.
Further, when the angle in the repetition visual field of big field-of-view image in step (3) is more than 15 degree, step (3) includes
Following steps:
(31) fluoroscopy images of wide-angle are generated by Sphere Measurement Model,
(32) repeat to calculate its degree of correlation on the visual field in fluoroscopy images, so that feedback adjustment camera posture, after adjustment again
Secondary generation fluoroscopy images are optimized.
Fig. 2 is Sphere Measurement Model schematic diagram.Each pixel on fish eye images is projected into Sphere Measurement Model in calculating process
On azimuthAnd elevation angle theta.As shown in Fig. 2 spatial point P projects to the p points on Sphere Measurement Model.Position according to subpoint can
To calculate azimuthAnd elevation angle theta.Any point on sphereCalculate the coordinate (x of three dimensionsp,yp,zp).
The three-dimensional point on two width fish eye images is calculated in processing procedure respectively.Directly calculated for preceding image and project to space coordinates (xp,
yp,zp), such as shown in formula (1), the space coordinate transformation such as formula (2) for rear image projection to sphere is shown, will coordinate
90 degree of rotate counterclockwise.
It follows that the θ repeated on two images,Scope, if obtain repeat θ,Scope, then only take
Pixel wherein on piece image is projected.There is repeat region during projection can effectively be eliminated in the method.
As shown in figure 3, the fluoroscopy images schematic diagram based on spherical diagram picture.O is the center of circle, in being by any point p on sphere
The heart, the fluoroscopy images block of the fixed length and width of generation.P is the central point on fluoroscopy images block, and OpP is conllinear.The radius of the spheroid is rs,
Ordinary circumstance rs=1, as unit sphere.lsFor fluoroscopy images border projects to the intersection point of the centre of sphere and sphere to the arc length of p points,
φ is the angle of fluoroscopy images border projection line and OpP lines.If fluoroscopy images size is fixed, φ is bigger, and fluoroscopy images are regarded
Wild also bigger, fluoroscopy images are closer to sphere, conversely, fluoroscopy images are away from sphere.So change can be reached by adjusting φ angles
The effect of focal length.The foundation of the Sphere Measurement Model can be obtained at 360 degree without any viewpoint direction in dead angle can be with the perspective of varifocal
Image block.For follow-up flame identification is laid a good foundation.
Embodiments of the present invention are elaborated above.But the present invention is not limited to above-mentioned implementation method,
In the ken that art those of ordinary skill possesses, can also be done on the premise of present inventive concept is not departed from
Go out various change.
Claims (4)
1. a kind of flame identification method based on deep learning model, it is characterised in that comprise the following steps:
(1) video information is gathered by back-to-back two fisheye cameras, then reads each two field picture of video information of collection,
Then the fish eye images of acquisition are carried out into gaussian filtering, obtains filtered fish eye images;
(2) aligning step (1) obtains the inner parameter of fish eye images, and subsequently into step (3), inner parameter includes tangential mistake
Difference, radial error and the light heart error margin;
(3) external parameter of fish eye images is corrected, into step (4);
(4) Sphere Measurement Model is built, the fish eye images after step (3) is corrected are projected on Sphere Measurement Model, then removal is projected in
Repeat region on Sphere Measurement Model, forms spherical diagram picture, subsequently into step (5);
(5) interference information recognized to flame portion by smog in the spherical diagram picture that goes smoke model removal step (4) to obtain,
Subsequently into step (6);
(6) using dynamic area on Codebook methods acquisition spherical diagram picture is improved, subsequently into step (7);
(7) the dynamic area part for obtaining step (6) generates fluoroscopy images;
(8) fluoroscopy images that normalisation step (7) is obtained, using fluoroscopy images as having trained seven layer architecture convolutional neural networks
Input, whether identification dynamic area is flame, if dynamic area is flame, into step (9), otherwise terminates this operation;
(9) recognition result is transferred to slave computer, shows recognition result in slave computer and produce warning message.
2. a kind of flame identification method based on deep learning model according to claim 1, it is characterised in that work as step
(3) when the fish eye images in are converted to the figure of 640 × 480 pixels, if angle point number is detected on fish eye images more than 300, step
Suddenly (3) comprise the following steps:
(31) normalized image that perspective view is certain, size determines is generated centered on the characteristic point for detecting on the Sphere Measurement Model
Block;
(32) relevance degree of characteristic point pair is calculated within the scope of the determination in adjacent two field pictures,
(33) best match pair is finally given, optimizes the relative pose of camera.
3. a kind of flame identification method based on deep learning model according to claim 1, it is characterised in that work as step
(3) when the fish eye images in are converted to the figure of 640 × 480 pixels, if angle point number is detected on fish eye images less than 300, step
Suddenly (3) comprise the following steps:
(31) using the side and geometry detected on image, end point pair is obtained on Sphere Measurement Model,
(32) transformation relation of structure on image is obtained based on end point pair, so as to calculate and optimize the relative appearance of camera
Gesture, so that Optimal Parameters.
4. a kind of flame identification method based on deep learning model according to claim 1, it is characterised in that work as step
(3) when the angle in the repetition visual field of big field-of-view image is more than 15 degree in, step (3) is comprised the following steps:
(31) fluoroscopy images of wide-angle are generated by Sphere Measurement Model,
(32) repeat to calculate its degree of correlation on the visual field in fluoroscopy images, so that feedback adjustment camera posture, secondary again after adjustment
Optimized into fluoroscopy images.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710047239.8A CN106845410B (en) | 2017-01-22 | 2017-01-22 | Flame identification method based on deep learning model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710047239.8A CN106845410B (en) | 2017-01-22 | 2017-01-22 | Flame identification method based on deep learning model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106845410A true CN106845410A (en) | 2017-06-13 |
CN106845410B CN106845410B (en) | 2020-08-25 |
Family
ID=59120999
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710047239.8A Expired - Fee Related CN106845410B (en) | 2017-01-22 | 2017-01-22 | Flame identification method based on deep learning model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106845410B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108830143A (en) * | 2018-05-03 | 2018-11-16 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of video analytic system based on deep learning |
CN108985374A (en) * | 2018-07-12 | 2018-12-11 | 天津艾思科尔科技有限公司 | A kind of flame detecting method based on dynamic information model |
CN109165575A (en) * | 2018-08-06 | 2019-01-08 | 天津艾思科尔科技有限公司 | A kind of pyrotechnics recognizer based on SSD frame |
CN111310662A (en) * | 2020-02-17 | 2020-06-19 | 淮阴工学院 | Flame detection and identification method and system based on integrated deep network |
WO2020247358A1 (en) * | 2019-06-07 | 2020-12-10 | Honeywell International Inc. | Method and system for connected advanced flare analytics |
CN113450530A (en) * | 2021-06-03 | 2021-09-28 | 河北华电石家庄热电有限公司 | Safety early warning system based on intelligent video analysis algorithm |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060017578A1 (en) * | 2004-07-20 | 2006-01-26 | Shubinsky Gary D | Flame detection system |
CN101441712A (en) * | 2008-12-25 | 2009-05-27 | 北京中星微电子有限公司 | Flame video recognition method and fire hazard monitoring method and system |
CN101625789A (en) * | 2008-07-07 | 2010-01-13 | 北京东方泰坦科技有限公司 | Method for monitoring forest fire in real time based on intelligent identification of smoke and fire |
CN102163358A (en) * | 2011-04-11 | 2011-08-24 | 杭州电子科技大学 | Smoke/flame detection method based on video image analysis |
CN104200468A (en) * | 2014-08-28 | 2014-12-10 | 中国矿业大学 | Method for obtaining correction parameter of spherical perspective projection model |
CN104581076A (en) * | 2015-01-14 | 2015-04-29 | 国网四川省电力公司电力科学研究院 | Mountain fire monitoring and recognizing method and device based on 360-degree panoramic infrared fisheye camera |
-
2017
- 2017-01-22 CN CN201710047239.8A patent/CN106845410B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060017578A1 (en) * | 2004-07-20 | 2006-01-26 | Shubinsky Gary D | Flame detection system |
CN101625789A (en) * | 2008-07-07 | 2010-01-13 | 北京东方泰坦科技有限公司 | Method for monitoring forest fire in real time based on intelligent identification of smoke and fire |
CN101441712A (en) * | 2008-12-25 | 2009-05-27 | 北京中星微电子有限公司 | Flame video recognition method and fire hazard monitoring method and system |
CN102163358A (en) * | 2011-04-11 | 2011-08-24 | 杭州电子科技大学 | Smoke/flame detection method based on video image analysis |
CN104200468A (en) * | 2014-08-28 | 2014-12-10 | 中国矿业大学 | Method for obtaining correction parameter of spherical perspective projection model |
CN104581076A (en) * | 2015-01-14 | 2015-04-29 | 国网四川省电力公司电力科学研究院 | Mountain fire monitoring and recognizing method and device based on 360-degree panoramic infrared fisheye camera |
Non-Patent Citations (2)
Title |
---|
B. UGˇUR TO¨REYIN 等: "Computer vision based method for real-time fire and flame detection", 《PATTERN RECOGNITION LETTERS 27》 * |
丁飞: "基于视觉传感器的火焰识别与高压细水雾喷淋实验研究", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108830143A (en) * | 2018-05-03 | 2018-11-16 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of video analytic system based on deep learning |
CN108985374A (en) * | 2018-07-12 | 2018-12-11 | 天津艾思科尔科技有限公司 | A kind of flame detecting method based on dynamic information model |
CN109165575A (en) * | 2018-08-06 | 2019-01-08 | 天津艾思科尔科技有限公司 | A kind of pyrotechnics recognizer based on SSD frame |
CN109165575B (en) * | 2018-08-06 | 2024-02-20 | 天津艾思科尔科技有限公司 | Pyrotechnic recognition algorithm based on SSD frame |
WO2020247358A1 (en) * | 2019-06-07 | 2020-12-10 | Honeywell International Inc. | Method and system for connected advanced flare analytics |
US11927944B2 (en) | 2019-06-07 | 2024-03-12 | Honeywell International, Inc. | Method and system for connected advanced flare analytics |
CN111310662A (en) * | 2020-02-17 | 2020-06-19 | 淮阴工学院 | Flame detection and identification method and system based on integrated deep network |
CN113450530A (en) * | 2021-06-03 | 2021-09-28 | 河北华电石家庄热电有限公司 | Safety early warning system based on intelligent video analysis algorithm |
Also Published As
Publication number | Publication date |
---|---|
CN106845410B (en) | 2020-08-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106845410A (en) | A kind of flame identification method based on deep learning model | |
CN103824070B (en) | A kind of rapid pedestrian detection method based on computer vision | |
CN106600888B (en) | Automatic forest fire detection method and system | |
CN104123544B (en) | Anomaly detection method and system based on video analysis | |
CN111680588A (en) | Human face gate living body detection method based on visible light and infrared light | |
CN104992452B (en) | Airbound target automatic tracking method based on thermal imaging video | |
CN103761514A (en) | System and method for achieving face recognition based on wide-angle gun camera and multiple dome cameras | |
CN106707296A (en) | Dual-aperture photoelectric imaging system-based unmanned aerial vehicle detection and recognition method | |
CN104301669A (en) | Suspicious target detection tracking and recognition method based on dual-camera cooperation | |
CN111652082B (en) | Face living body detection method and device | |
CN111985365A (en) | Straw burning monitoring method and system based on target detection technology | |
CN102819847A (en) | Method for extracting movement track based on PTZ mobile camera | |
CN108038866A (en) | A kind of moving target detecting method based on Vibe and disparity map Background difference | |
TWI759767B (en) | Motion control method, equipment and storage medium of the intelligent vehicle | |
CN109063625A (en) | A kind of face critical point detection method based on cascade deep network | |
EP2813973A1 (en) | Method and system for processing video image | |
WO2022161139A1 (en) | Driving direction test method and apparatus, computer device, and storage medium | |
TWI726278B (en) | Driving detection method, vehicle and driving processing device | |
CN106915303A (en) | Automobile A-column blind area perspective method based on depth data and fish eye images | |
CN103544478A (en) | All-dimensional face detection method and system | |
CN114140745A (en) | Method, system, device and medium for detecting personnel attributes of construction site | |
CN110703760A (en) | Newly-increased suspicious object detection method for security inspection robot | |
CN114998737A (en) | Remote smoke detection method, system, electronic equipment and medium | |
CN106815567A (en) | A kind of flame detecting method and device based on video | |
WO2021248564A1 (en) | Panoramic big data application monitoring and control system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200825 Termination date: 20210122 |
|
CF01 | Termination of patent right due to non-payment of annual fee |