CN104463904A - High-voltage line foreign matter invasion target detection method - Google Patents
High-voltage line foreign matter invasion target detection method Download PDFInfo
- Publication number
- CN104463904A CN104463904A CN201410484914.XA CN201410484914A CN104463904A CN 104463904 A CN104463904 A CN 104463904A CN 201410484914 A CN201410484914 A CN 201410484914A CN 104463904 A CN104463904 A CN 104463904A
- Authority
- CN
- China
- Prior art keywords
- target
- image
- foreign body
- body intrusion
- background
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
- G06T7/251—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments involving models
-
- 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
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
A high-voltage line foreign matter invasion target detection method includes the steps that an adjacent-frame image is extracted from a video monitoring terminal or a video, the extracted adjacent-frame image is preprocessed, the movement characteristic of the adjacent-frame image is extracted, foreign matter or an invasion target is detected, the foreign matter invasion target is marked through a morphological method, the foreign matter or the invasion target is tracked, whether a power transmission line is affected or not is judged, and if yes, an alarm is given. Through the method, the foreign matter invasion target is fast and accurately detected, a high-voltage line is monitored, and an actual application requirement can be met.
Description
Technical field
The invention belongs to electrical network, power domain, particularly relate to a kind of method of hi-line foreign body intrusion target detection.
Background technology
Electric power is the important foundation industry of national economy, and safety, stable electric power supply are the prerequisites ensureing the development of national economy fast and stable.In national grid, the environment residing for ultra-high-tension power transmission line very complicated and be vulnerable to infringement, therefore guarantee that hi-line security of operation just seems particularly important.Traditional video monitoring system needs manually to supervise video recording, and monitoring performance is subject to the restriction of the physiologic factor of monitor staff itself.Along with the quick growth of China's ultra-high-tension power transmission line scale, power department assume responsibility for increasing hi-line and patrols and examines and maintenance work.
The data processing function that intelligent video treatment technology computer is powerful, high speed analysis is carried out to the mass data in video monitoring picture, filter out the information that monitor staff does not pay close attention to, be only the key message that monitor staff provides, greatly can alleviate the working strength of video monitoring personnel, wrong report can be reduced simultaneously, fail to report, improve the promptness of alert process.
Therefore, current in the urgent need to designing a kind of hi-line foreign body intrusion object detection method, to improve detection speed and to detect degree of accuracy, circuit monitoring maintainer is helped to increase work efficiency, ensure the safety of hi-line, and provide technology model for hi-line foreign body intrusion target automatic checkout system.
Summary of the invention
The object of the present invention is to provide a kind of method of hi-line foreign body intrusion target detection, the method realize to foreign body intrusion target quick, accurately detect, in real time hi-line is monitored, practical application needs can be met.
In order to achieve the above object, the invention provides a kind of hi-line foreign body intrusion object detection method, it is characterized in that comprising the following steps:
1, from Video Monitoring Terminal or video record, consecutive frame image is extracted;
2, the consecutive frame image extracted is carried out to the pre-service of image;
3, extract adjacent two two field picture motion features, detect foreign body intrusion target;
4, morphological method is adopted to mark foreign body intrusion target;
5, foreign body intrusion target is followed the tracks of, judge whether to affect transmission line of electricity, as impact, report to the police.
Described pre-service extraction consecutive frame image being carried out to image, first detects critical area relevant with transmission line of electricity in image, then the image of critical area is carried out gray processing, obtain gray-scale map;
Wherein
for pixel coordinate in image,
, m is picture altitude,
, n is picture traverse,
being respectively pixel coordinate is
the pixel value of red, green, blue passage,
for pixel coordinate
gray-scale value;
Carry out image smoothing, filtered noise to the image of critical area again after gray level image, image smoothing adopts the method for medium filtering.
The extraction of described two field picture motion feature is carried out between consecutive frame image, is the point of the difference extracting local motion and global motion, respectively adjacent two two field pictures is carried out to the filtering of different scale, can obtain image
, wherein
if current frame image is
, previous frame image is
, wherein filter template
meet:
Wherein
, for
dimensional vector,
, and meet:
The then motion feature of image
for:
Wherein
,
be two consecutive frame images
with
each respective pixel do difference.
Then each width motion feature figure is normalized, then merges and just define final target detection image:
for each pixel correspondence of different scale images is added.
Described to carry out mark to foreign body intrusion target be after foreign body intrusion target being detected, obtains the position coordinates of target, then position mark with the empty frame of rectangle.
Described following the tracks of foreign body intrusion target adopts the MeanShift target tracking algorism improved, by pixel value
uas eigenwert, respectively normalization histogram is set up to the eigenwert of target and background, the discrimination of objective definition and background, introduce weight coefficient
, object module is weighted, is used for judging object feature value and background characteristics value discrimination.
for eigenwert in target
uvalue in histogram,
for eigenwert in background
uvalue in histogram, background is the region of outer 20 pixel of target window here,
for fixed value, get
, then
larger, object and background discrimination is higher, and what show that this eigenwert describes is target,
less, object and background discrimination is lower, and what show that this eigenwert describes is background, and
span is
, will
normalization can obtain
,
The formula redefinable of MeanShift algorithm is:
In formula
, and
, be respectively normalization constant, meet
,
, other steps of algorithm are consistent with traditional MeanShift algorithm.
The present invention proposes a kind of method of hi-line foreign body intrusion target detection, can effectively carry out the detection of hi-line foreign body intrusion.Near monitoring hi-line, the motion state of the on-the-spot heavy mechanical equipment such as Construction traffic, loop wheel machine, digging machine, behavior and the Harm to circuit thereof carry out quick, intelligent judgement, can monitor that foreign matter is close or hang on wire, early warning information is sent to an innings power transmission state monitoring center simultaneously.The method can detect in real time, and accuracy of detection is high, for hi-line foreign body intrusion target automatic checkout system provides technology model, substitutes manual inspection.
Accompanying drawing explanation
Fig. 1 is a kind of FB(flow block) of hi-line foreign body intrusion object detection method;
Fig. 2 is a kind of hi-line foreign body intrusion target following process flow diagram.
Embodiment
As shown in Figure 1, concrete detection method is as follows:
In step S101, extract consecutive frame image from (high-definition camera) Video Monitoring Terminal or video record;
In step S102, according to the image of step S101 input, extraction consecutive frame image is carried out to the pre-service of image, first detects critical area relevant with transmission line of electricity in image, then the image of critical area is carried out gray processing, obtain gray-scale map:
Wherein
for pixel coordinate in image,
, m is picture altitude,
, n is picture traverse,
being respectively pixel coordinate is
the pixel value of red, green, blue passage,
for pixel coordinate
gray-scale value;
Carry out image smoothing, filtered noise to the image of critical area again after gray level image, image smoothing adopts the method for medium filtering.
In step S103, according to the Image semantic classification result that step S102 carries out, extract adjacent two two field picture motion features, detect foreign body intrusion target, the extraction of motion feature is carried out between consecutive frame image, be the point of the difference extracting local motion and global motion, respectively adjacent two two field pictures carried out to the filtering of different scale, can image be obtained
, wherein
if current frame image is
, previous frame image is
, wherein filter template
meet:
Wherein
, for
dimensional vector,
, and meet:
The then motion feature of image
for:
Wherein
,
be two consecutive frame images
with
each respective pixel do difference.
Then each width motion feature figure is normalized, then merges and just define final target detection image:
for each pixel correspondence of different scale images is added;
In step S104, according to the object detection results that step S103 carries out, morphological method is adopted to mark foreign body intrusion target.After foreign body intrusion target being detected, need to mark target, for target following is ready.In the present invention, according to the foreign body intrusion target detected, obtain the position coordinates of target, then position mark with the empty frame of rectangle;
In step S105, according to the target label result that step S104 carries out, foreign body intrusion target is followed the tracks of, judge whether to affect transmission line of electricity, as impact, report to the police.The present invention adopts the MeanShift target tracking algorism of improvement.Because target region color histogram can be subject to the impact of background pixel, when comprising background pixel in object module and being less, tradition MeanShift algorithm can obtain good tracking effect, but when comprising a large amount of background pixels in object module, tradition MeanShift algorithm can cause the deviation of target localization, in order to reduce the impact of background pixel on object module, the method of the invention using pixel value u as eigenwert, respectively normalization histogram is set up to the eigenwert of target and background, the discrimination of objective definition and background, introduces weight coefficient
, object module is weighted, is used for judging object feature value and background characteristics value discrimination.
for the value of eigenwert u in target in histogram,
for the value of eigenwert u in background in histogram, background is the region of outer 20 pixel of target window here,
for fixed value, get
, then
larger, object and background discrimination is higher, and what show that this eigenwert describes is target;
less, object and background discrimination is lower, and what show that this eigenwert more can describe is background.And
span is
, will
normalization can obtain
,
The formula redefinable of MeanShift algorithm is:
In formula
, and
, be respectively normalization constant, meet
,
.Other steps of algorithm are consistent with traditional MeanShift algorithm.
Then judge whether the target of following the tracks of is foreign matter or intrusion target, target setting is demarcated rectangle frame height threshold and is
h, when tracked target aloft, and move to hi-line region, and when resting on hi-line region, can judge have foreign matter to be suspended on wire, give the alarm; And tracked target is when ground moving, and when object height is greater than threshold value
h, can judge there is intrusion target close to hi-line, give the alarm.
As shown in Figure 2, tracing process is as follows:
Flow process starts from S201;
In step S202, filtering detects foreign body intrusion target: first reads image data from video flowing, and start the second frame read, by carrying out filtering to current frame image and previous frame image, difference carries out target detection, obtains target location.
In step S203, according to the target that step S202 detects, carry out morphology mark, position mark with the empty frame of rectangle.
In step S204, according to the target of step S203 mark, carry out foreign matter target or intrusion target judgement, target setting is demarcated rectangle frame height threshold and is
h, when tracked target aloft, and move to hi-line region, and when resting on hi-line region, can be judged to be that tracked target is foreign matter target; And tracked target is when ground moving, and when object height is greater than
h, can judge that tracked target is as intrusion target.
In step S205, according to the result that step S204 judges, the target's center to be tracked of initialization improvement of the present invention MeanShift track algorithm, calculates target signature and background characteristics weight coefficient, and normalization.
In step S206, according to the initialized result of step S205, carry out Mean Shift iteration and follow the tracks of, the destination probability distribution of calculation window internal object initial position and the probability distribution of next frame candidate target.
In step S207, according to the result that step S206 calculates.Calculate MeanShift vector iteration to convergence point.In each Mean Shift vector iterative process, calculate similarity measurements flow function d.
In step S208, for similarity judges, if result is "Yes", enter step S209
In step S209, if similarity measurement is
.If
, illustrate that target is mated with candidate family, follow the tracks of effectively.The convergence point obtained is set to initial position, and lost frames counting resets, and continues to calculate target initial distribution, have updated object module, carry out Mean Shift iteration next time.
In step S208, if result is "No", enter step S210.
In step S210, if
, illustrate that target following frame departs from target, follow the tracks of and lost efficacy.In order to correct the description of destination probability distribution, making kernel function act in target's center, retaining target signature distribution (namely not upgrading object module) now, conversion Kalman filter carries out tracking prediction.
In step S211, upgrade object module, for next frame tracking image target is prepared.
In step S212, according to step S210 and S211, target location is demarcated.
In step S213, foreign body intrusion target alarm decision, judges that whether target is close or adhere to hi-line, if result is "Yes", enters step S214.
In step S213, if result is "No", enter step S206.
In step S214, give the alarm.
In step S215, flow process terminates.
Claims (5)
1. a hi-line foreign body intrusion object detection method, is characterized in that comprising the following steps:
(1) from Video Monitoring Terminal or video record, consecutive frame image is extracted;
(2) the consecutive frame image extracted is carried out to the pre-service of image;
(3) extract adjacent two two field picture motion features, detect foreign body intrusion target;
(4) morphological method is adopted to mark foreign body intrusion target;
(5) foreign body intrusion target is followed the tracks of, judge whether to affect transmission line of electricity, as impact, report to the police.
2. a kind of hi-line foreign body intrusion object detection method according to claim 1, it is characterized in that: the described pre-treatment step to extracting consecutive frame image and carry out image is: first detect critical area relevant with transmission line of electricity in image, then the image of critical area is carried out gray processing, obtain gray-scale map;
Wherein
for pixel coordinate in image,
, m is picture altitude,
, n is picture traverse,
being respectively pixel coordinate is
the pixel value of red, green, blue passage,
for pixel coordinate
gray-scale value;
Carry out image smoothing, filtered noise to the image of critical area again after gray level image, image smoothing adopts the method for medium filtering.
3. a kind of hi-line foreign body intrusion object detection method according to claim 1, it is characterized in that: the extraction of described two field picture motion feature is carried out between consecutive frame image, it is the point of the difference extracting local motion and global motion, respectively adjacent two two field pictures are carried out to the filtering of different scale, can image be obtained
, wherein
if current frame image is
, previous frame image is
, wherein filter template
meet:
Wherein
, for
dimensional vector,
, and meet:
The then motion feature of image
for:
Wherein
,
be two consecutive frame images
with
each respective pixel do difference, then each width motion feature figure is normalized, then merges and just define final target detection image:
for each pixel correspondence of different scale images is added.
4. a kind of hi-line foreign body intrusion object detection method according to claim 1, it is characterized in that: described to carry out mark to foreign body intrusion target be after foreign body intrusion target being detected, obtain the position coordinates of target, then position mark with the empty frame of rectangle.
5. a kind of hi-line foreign body intrusion object detection method according to claim 1, is characterized in that: described following the tracks of foreign body intrusion target adopts the MeanShift target tracking algorism improved, by pixel value
uas eigenwert, respectively normalization histogram is set up to the eigenwert of target and background, the discrimination of objective definition and background, introduce weight coefficient
, object module is weighted, is used for judging object feature value and background characteristics value discrimination;
for eigenwert in target
uvalue in histogram,
for eigenwert in background
uvalue in histogram, background is the region of outer 20 pixel of target window here,
for fixed value, get
, then
larger, object and background discrimination is higher, and what show that this eigenwert describes is target,
less, object and background discrimination is lower, and what show that this eigenwert describes is background, and
span is
, will
normalization can obtain
,
The formula redefinable of MeanShift algorithm is:
In formula
, and
, be respectively normalization constant, meet
,
, other steps of algorithm are consistent with traditional MeanShift algorithm.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410484914.XA CN104463904A (en) | 2014-09-22 | 2014-09-22 | High-voltage line foreign matter invasion target detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410484914.XA CN104463904A (en) | 2014-09-22 | 2014-09-22 | High-voltage line foreign matter invasion target detection method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN104463904A true CN104463904A (en) | 2015-03-25 |
Family
ID=52909886
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410484914.XA Pending CN104463904A (en) | 2014-09-22 | 2014-09-22 | High-voltage line foreign matter invasion target detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104463904A (en) |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105491354A (en) * | 2016-01-18 | 2016-04-13 | 广州爱九游信息技术有限公司 | Video monitoring method and video monitoring platform |
CN105825619A (en) * | 2016-03-25 | 2016-08-03 | 南京第五十五所技术开发有限公司 | High-tension cable floater alarm method based on image processing |
CN105931420A (en) * | 2016-06-14 | 2016-09-07 | 国家电网公司 | Early warning system and method against hidden safety troubles of line channel by utilizing dynamic laser fence technology |
CN106101658A (en) * | 2016-08-13 | 2016-11-09 | 哈尔滨理工大学 | Shaft tower foreign body and disappearance intelligent video on-line monitoring system |
CN106131501A (en) * | 2016-08-13 | 2016-11-16 | 哈尔滨理工大学 | Electric line foreign matter and disappearance intelligent video on-line monitoring system |
CN106429910A (en) * | 2016-06-24 | 2017-02-22 | 国网山东省电力公司寿光市供电公司 | Onsite work high-voltage line anti-touch early warning apparatus |
CN108257152A (en) * | 2017-12-28 | 2018-07-06 | 清华大学苏州汽车研究院(吴江) | A kind of road intrusion detection method and system based on video |
CN109272535A (en) * | 2018-09-07 | 2019-01-25 | 广东中粤电力科技有限公司 | A kind of power distribution room safety zone method for early warning based on image recognition |
CN109785361A (en) * | 2018-12-22 | 2019-05-21 | 国网内蒙古东部电力有限公司 | Substation's foreign body intrusion detection system based on CNN and MOG |
CN110458090A (en) * | 2019-08-08 | 2019-11-15 | 成都睿云物联科技有限公司 | Working state of excavator detection method, device, equipment and storage medium |
CN110956614A (en) * | 2019-11-11 | 2020-04-03 | 国网山东省电力公司电力科学研究院 | Ground wire foreign matter detection method and device based on iterative search and projection method |
CN111538103A (en) * | 2020-06-04 | 2020-08-14 | 广东电网有限责任公司 | Detection device for hanging object |
CN111626204A (en) * | 2020-05-27 | 2020-09-04 | 北京伟杰东博信息科技有限公司 | Railway foreign matter invasion monitoring method and system |
CN112991318A (en) * | 2021-03-31 | 2021-06-18 | 中车青岛四方机车车辆股份有限公司 | Motor train unit pantograph fault detection method and device and storage medium |
CN113076899A (en) * | 2021-04-12 | 2021-07-06 | 华南理工大学 | High-voltage transmission line foreign matter detection method based on target tracking algorithm |
CN113487819A (en) * | 2021-05-11 | 2021-10-08 | 重庆金专新晟科技有限公司 | External damage prevention monitoring system for tower power transmission line |
CN114723678A (en) * | 2022-03-21 | 2022-07-08 | 盛视科技股份有限公司 | High-voltage wire foreign matter detection method and detection system based on video image |
CN115114466A (en) * | 2022-08-30 | 2022-09-27 | 成都实时技术股份有限公司 | Method, system, medium and electronic device for searching target information image |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070237359A1 (en) * | 2006-04-05 | 2007-10-11 | Zehang Sun | Method and apparatus for adaptive mean shift tracking |
CN103745203A (en) * | 2014-01-15 | 2014-04-23 | 南京理工大学 | Visual attention and mean shift-based target detection and tracking method |
-
2014
- 2014-09-22 CN CN201410484914.XA patent/CN104463904A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070237359A1 (en) * | 2006-04-05 | 2007-10-11 | Zehang Sun | Method and apparatus for adaptive mean shift tracking |
CN103745203A (en) * | 2014-01-15 | 2014-04-23 | 南京理工大学 | Visual attention and mean shift-based target detection and tracking method |
Non-Patent Citations (3)
Title |
---|
刘斌: "基于双目立体视觉的大锻件尺寸在位测量关键技术研究", 《中国博士学位论文全文数据库 工程科技I辑》 * |
杜超: "基于均值漂移与粒子滤波的目标跟踪算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
薛丽霞等: "基于帧间差分的自适应运动目标检测方法", 《计算机应用研究》 * |
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105491354A (en) * | 2016-01-18 | 2016-04-13 | 广州爱九游信息技术有限公司 | Video monitoring method and video monitoring platform |
CN105825619A (en) * | 2016-03-25 | 2016-08-03 | 南京第五十五所技术开发有限公司 | High-tension cable floater alarm method based on image processing |
CN105931420A (en) * | 2016-06-14 | 2016-09-07 | 国家电网公司 | Early warning system and method against hidden safety troubles of line channel by utilizing dynamic laser fence technology |
CN106429910A (en) * | 2016-06-24 | 2017-02-22 | 国网山东省电力公司寿光市供电公司 | Onsite work high-voltage line anti-touch early warning apparatus |
CN106101658A (en) * | 2016-08-13 | 2016-11-09 | 哈尔滨理工大学 | Shaft tower foreign body and disappearance intelligent video on-line monitoring system |
CN106131501A (en) * | 2016-08-13 | 2016-11-16 | 哈尔滨理工大学 | Electric line foreign matter and disappearance intelligent video on-line monitoring system |
CN108257152B (en) * | 2017-12-28 | 2022-04-08 | 清华大学苏州汽车研究院(吴江) | Road intrusion detection method and system based on video |
CN108257152A (en) * | 2017-12-28 | 2018-07-06 | 清华大学苏州汽车研究院(吴江) | A kind of road intrusion detection method and system based on video |
CN109272535A (en) * | 2018-09-07 | 2019-01-25 | 广东中粤电力科技有限公司 | A kind of power distribution room safety zone method for early warning based on image recognition |
CN109785361A (en) * | 2018-12-22 | 2019-05-21 | 国网内蒙古东部电力有限公司 | Substation's foreign body intrusion detection system based on CNN and MOG |
CN110458090A (en) * | 2019-08-08 | 2019-11-15 | 成都睿云物联科技有限公司 | Working state of excavator detection method, device, equipment and storage medium |
CN110956614A (en) * | 2019-11-11 | 2020-04-03 | 国网山东省电力公司电力科学研究院 | Ground wire foreign matter detection method and device based on iterative search and projection method |
CN110956614B (en) * | 2019-11-11 | 2023-04-07 | 国网山东省电力公司电力科学研究院 | Ground wire foreign matter detection method and device based on iterative search and projection method |
CN111626204A (en) * | 2020-05-27 | 2020-09-04 | 北京伟杰东博信息科技有限公司 | Railway foreign matter invasion monitoring method and system |
CN111538103A (en) * | 2020-06-04 | 2020-08-14 | 广东电网有限责任公司 | Detection device for hanging object |
CN112991318A (en) * | 2021-03-31 | 2021-06-18 | 中车青岛四方机车车辆股份有限公司 | Motor train unit pantograph fault detection method and device and storage medium |
CN113076899A (en) * | 2021-04-12 | 2021-07-06 | 华南理工大学 | High-voltage transmission line foreign matter detection method based on target tracking algorithm |
CN113487819A (en) * | 2021-05-11 | 2021-10-08 | 重庆金专新晟科技有限公司 | External damage prevention monitoring system for tower power transmission line |
CN114723678A (en) * | 2022-03-21 | 2022-07-08 | 盛视科技股份有限公司 | High-voltage wire foreign matter detection method and detection system based on video image |
CN115114466A (en) * | 2022-08-30 | 2022-09-27 | 成都实时技术股份有限公司 | Method, system, medium and electronic device for searching target information image |
CN115114466B (en) * | 2022-08-30 | 2022-12-13 | 成都实时技术股份有限公司 | Method, system, medium and electronic device for searching target practice information image |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104463904A (en) | High-voltage line foreign matter invasion target detection method | |
CN108307146B (en) | System and method for detecting potential safety hazard of high-voltage transmission line | |
CN102509078B (en) | Fire detection device based on video analysis | |
CN103069434B (en) | For the method and system of multi-mode video case index | |
KR101237089B1 (en) | Forest smoke detection method using random forest classifier method | |
CN112016414A (en) | Method and device for detecting high-altitude parabolic event and intelligent floor monitoring system | |
KR102122859B1 (en) | Method for tracking multi target in traffic image-monitoring-system | |
CN103246896B (en) | A kind of real-time detection and tracking method of robustness vehicle | |
CN104123544A (en) | Video analysis based abnormal behavior detection method and system | |
WO2022078182A1 (en) | Throwing position acquisition method and apparatus, computer device and storage medium | |
CN105611244A (en) | Method for detecting airport foreign object debris based on monitoring video of dome camera | |
CN108765453B (en) | Expressway agglomerate fog identification method based on video stream data | |
CN105678730A (en) | Camera movement self-detecting method on the basis of image identification | |
CN102955940A (en) | System and method for detecting power transmission line object | |
CN105554462A (en) | Remnant detection method | |
CN105184812A (en) | Target tracking-based pedestrian loitering detection algorithm | |
Xu et al. | Segmentation and tracking of multiple moving objects for intelligent video analysis | |
CN210072642U (en) | Crowd abnormal behavior detection system based on video monitoring | |
CN108898098A (en) | Early stage video smoke detection method based on monitor supervision platform | |
CN112287823A (en) | Facial mask identification method based on video monitoring | |
CN110390784A (en) | A kind of transmission line of electricity external force damage prevention monitoring system based on deep learning | |
CN110349172B (en) | Power transmission line external damage prevention early warning method based on image processing and binocular stereo ranging | |
CN103093197A (en) | Monitoring method and system for recognizing hanging behavior | |
Li et al. | Intelligent transportation video tracking technology based on computer and image processing technology | |
CN112668478B (en) | Well lid monitoring method and device, electronic equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150325 |