CN105915645A - Substation video monitoring automatic alarm system - Google Patents
Substation video monitoring automatic alarm system Download PDFInfo
- Publication number
- CN105915645A CN105915645A CN201610443380.5A CN201610443380A CN105915645A CN 105915645 A CN105915645 A CN 105915645A CN 201610443380 A CN201610443380 A CN 201610443380A CN 105915645 A CN105915645 A CN 105915645A
- Authority
- CN
- China
- Prior art keywords
- video
- image
- monitoring
- algorithm
- operator
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
- H04L67/025—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP] for remote control or remote monitoring of applications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
- H04N19/103—Selection of coding mode or of prediction mode
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/157—Assigned coding mode, i.e. the coding mode being predefined or preselected to be further used for selection of another element or parameter
- H04N19/159—Prediction type, e.g. intra-frame, inter-frame or bidirectional frame prediction
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/70—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by syntax aspects related to video coding, e.g. related to compression standards
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Abstract
The invention discloses a system for carrying out video monitoring automatic alarm on switch cabinet single indicator lamps, switches and knob states of a substation. Through application of various image recognition algorithms of color, texture, shape and space relationship features, functions of carrying out automatic analysis and abnormal alarm on substation monitoring are realized. The system comprises video image recognition algorithms, video transmission and compression algorithms, and research of automatic analysis and abnormal alarm methods of the video monitoring. According to the system, the ultimate purpose is that the automatic analysis and abnormal alarm are carried out on the substation monitoring by the video image recognition algorithms and the video transmission and compression algorithms. The defect of delayed reaction resulting from manual slack is thoroughly solved. Online automatic recognition is carried out on the substation video monitoring contents, thus discovering various abnormal conditions timely, and the operation security level and automatic level are improved.
Description
Technical field
Native system is by multiple image recognition algorithms such as color, texture, shape, spatial relationship features, it is achieved
Switch cabinet of converting station condition monitoring is automatically analyzed, the function of abnormal alarm, belong to power industry and supervise safely
Control technical field.
Background technology
Intelligent Video Surveillance Technology application in public security personnel field brings the dramatic change of monitoring system.
Now, many scientific research personnel are being able to wide variety of moving object detection and tracking in military affairs are guided at first
Technology is successfully introduced in monitoring system the intelligent monitoring system having constructed a new generation.This kind of system is only
Monitoring personnel provide " eyes ", it is still necessary to monitoring personnel's moment according to scenery control The Cloud Terrace pursuit movement target,
Its own only has simple autonomous monitoring capacity and can examine the moving target of monitoring scene in real time
Survey and follow the tracks of.
In order to reach both to provide " eyes " to monitoring personnel, provide " brain " also to monitoring personnel, provided herein is
Multiple characteristics of image identification technology being incorporated in monitoring device, system design can substitute monitoring personnel completely,
Can be used for the high-intelligentization monitoring in the case of unmanned, even if occurring extremely to send alarm signal.
Summary of the invention
Present invention seek to address that transformer station realizes unmanned, switch cabinet of converting station image is monitored and
Abnormal conditions automatic alarm.For achieving the above object, technical scheme is as follows:
Online automatic identification is carried out for transformer substation video monitoring content, to find various unusual conditions in time,
Improve and run level of security and automatization level.System BROAD SUMMARY includes:
(1) video image identification algorithm
Analyze transformer substation video monitoring image feature, select rational image recognition algorithm that characteristics of image is clicked on
Row is analyzed；The video image of collection carries out before characteristics of image identification pretreatment (noise reduction, conversion), feature is known
Not, the step such as analyzing and processing, feature extraction involved in the present invention includes following a few class: color characteristic, stricture of vagina
Reason feature, shape facility, spatial relationship feature.
(2) transmission of video compression algorithm
The present invention uses the JVT (Joint Video Expert Team) constituted jointly with ITU-T and MPEG
H.264 video compression coding standard, propose three kinds reduce the highest computation complexities optimizations calculate
Method---intraframe prediction algorithm, motion estimation algorithm and fast motion estimation algorithm, be effectively improved video and pass
Transmission quality and efficiency.
(3) the automatically analyzing of video monitoring, abnormal alarm method
The present invention realizes supervisory control of substation video image is carried out feature analysis, and feature object is to open in transformer station
Close the most each signal lights and on off state, feature object is automatically analyzed, system when abnormal conditions occur
Automatically alarm signal is sent.
The present invention passes through video image identification algorithm, transmission of video compression algorithm, and video monitoring is automatic
Analysis, the automatic alert of abnormal conditions, it is achieved that transformer substation video monitoring automatic alarm system, thorough
The end, solves the artificial slack reaction caused drawback not in time.
Accompanying drawing explanation
Fig. 1 is the major function FB(flow block) used in an embodiment of the present invention.
Detailed description of the invention
The invention will be further described with detailed description of the invention below in conjunction with the accompanying drawings.
Embodiment:
Step 1, video image compression is transmitted, the video data of collection is carried out Video coding, is regarded by minimizing
Dependency between frequency sequence, reduces the redundancy in video content, represents video content with less bit number,
Thus realize the compression to video.
Video compress not only removes video spatial domain and statistical redundancy, it is often more important that reduce the redundancy of time domain, i.e.
The determination information that can deduce is removed.The present invention uses huffman coding to carry out the compression of statistical redundancy degree；Logical
Cross frequency domain transform, the coefficient of original image signal DC component and minority low frequency AC components is represented, makes
By orthogonal cosine transform DCT method, the spatial redundancies of image is effectively compressed；Use difference is compiled
Code DPCM, effectively compresses the temporal redundancy of image；Low by after discrete cosine transform
Frequency component carries out fine quantization, and the visual redundancy degree of image is carried out by the method slightly quantified high fdrequency component
Effective compression, by statistical redundancy, spatial redundancy, time redundancy and the compression of visual redundancy degree, improves
Picture compression efficiency.
Step 2, carries out Image semantic classification, including image noise reduction and rim detection by the video image gathering transmission.
Effectively suppressing noise, the application for image has vital effect.In order to reach more preferable
Denoising effect, the present invention, according to different noise sources and impact, have employed filter in spatial domain, transform domain filtering
With Denoising Algorithm such as partial differential equation.Filter in spatial domain directly carries out data operation on original image, to pixel
Gray value process；The present invention uses Walsh-Hadanjard Transform and wavelet transformation by image from spatial domain
It is transformed into the transform domain filtering algorithm of transform domain；The present invention uses the partial differential equation of Perona and Malik,
The method has the biggest selection space when determining diffusion coefficient, to diffusion after having while forward direction spreads
Function, so, there is smoothed image and by edge sharpening ability, partial differential equation are close at low noise
The image procossing of degree achieves preferable effect.
Image semantic classification rim detection of the present invention, mainly for detection of switch cubicle image border, carries for subsequent treatment
For basic data.Different images gray scale is different, and boundary typically has obvious edge, utilizes this feature to divide
Cut image.According to different video image characteristics, the present invention mainly use Roberts operator, Prewitt operator,
Sobel operator, Isotropic Sobel operator and the Image Edge-Detection operator of Laplacian operator.Roberts
Operator, location, edge standard, but to noise-sensitive, and noise less image segmentation obvious for edge；
Prewitt operator, has inhibitory action to noise, and the principle of suppression noise is average by pixel, but pixel
Averagely be equivalent to the low-pass filtering to image；Sobel operator and Prewitt operator are all weighted averages, but
It is that the impact that current pixel is produced by the pixel of neighborhood is not of equal value that Sobel operator processes, so distance is not
Same pixel has different weights, and the impact producing operator result is the most different；Isotropic Sobel operator,
Weights are inversely proportional to the distance of adjoint point and central point, and when along different directions detection edge, gradient amplitude is consistent, just
The isotropism being known as；Laplacian operator, is Second Order Differential Operator.It has isotropism,
I.e. unrelated with change in coordinate axis direction, after coordinate axes rotates, gradient result is constant.But, it is more sensitive to noise ratio,
So, image typically first passes through smoothing processing.
Step 3, is identified out by pretreated image, mainly include image characteristics extraction and target with
Track processes, and the present invention mainly carries out feature extraction to switch cabinet of converting station signal lights and panel-switch state, right
The characteristic target extracted carries out image recognition, mainly includes color characteristic, textural characteristics, shape facility and sky
Between relationship characteristic comprehensively identify, switch cubicle running status is more recognized accurately, to the target figure after feature extraction
As being tracked processing.
The present invention uses color histogram method to express color characteristic, and its advantage is not become by image rotation and translation
The impact changed, also can not be changed by graphical rule by normalization further and be affected；Textural characteristics is utilized to examine
Rigging has the texture image of the aspect bigger difference such as thickness, density, such as: when on off state changes, opens
The texture thickness of pass, density etc. are prone to differentiate, the number detail change of primitive object, combined with texture in region
Feature can effectively identify that this type of on off state changes；The present invention uses contours extract algorithm process shape facility,
The method complex image being used to first rim detection contours extract again goes out the contours extract of target
Come；Mutual locus between the multiple targets split in image or relative direction relation, these
Relation also can be divided into connection/syntopy, overlapping/overlapping relation and comprise/containment relationship etc., the present invention use by
Continuous constraint and the method combined based on region segmentation process image space relationship characteristic.
After monitored video image is carried out feature extraction, it is tracked target image processing, this project
The motion target tracking monitoring used mainly is used: frame differential method, background subtraction and optical flow method continuously.
Frame differential method is for well adapting to property of dynamic environment continuously, is a kind of motion based on pixel inspection
Survey method, it obtains by two or three images adjacent in sequence of video images are carried out calculus of differences
Moving object contours；Background subtraction is compared by input picture and background image thus is partitioned into motion
Target, but the dynamic scene change causing illumination and external condition is excessively sensitive；Optical flow method can detect
Go out the object of self-movement, it is not necessary to be known a priori by any information of scene.
Owing to motion detection is in the bottom that video motion is analyzed, it is widely applied occasion and makes motion detection calculate
Method can process the situation of various complexity, is difficult to have a kind of algorithm can be suitable for all of application scenario, this
The bright video image according to different scene capture, enters the target of motion in conjunction with above-mentioned three kinds of method for tracking target
Row locating and tracking.
Step 4, the automatically analyzing of video monitoring, abnormal alarm, the present invention is primarily directed to image and carries out intelligence
Fractional analysis, is analyzed characteristic point processing, analyzes method such as step 3, the characteristic point that system will identify every time
In state data-in storehouse, the analysis result after every time patrolling and examining and the video analysis results contrast patrolling and examining survey last time,
If twice testing result is inconsistent, system sends alarm as abnormal conditions, simultaneously under abnormal conditions
Switch cubicle signal lamp, switch and knob state preserve, when facilitating maintainer inquiry fault to occur
Relevant information.
1) signal lamp accident analysis
Break down before and after's signal lamp color significant change, judge signal lights by color change feature therefore
Barrier.Record each signal lamp state given corresponding fault type, the signal lamp figure that will gather back
Sheet carries out one_to_one corresponding with corresponding quantized data, sets up and constantly improves signal lamp failure and analyze storehouse, finally
Reach the function of quantitative analysis.
For the detection project that different signal lamp definition is different, it is carried out accompanying drawing corresponding, often simultaneously
The different word of individual project definition describes and gives different grades.The quantization detection project correspondence that will define
Scene is detected, and carries out unifying input system by the information detected back, carries out the analysis of fault.According to
Different descriptive grades assesses different fault degrees, and the present invention sends phase automatically according to different faults type
Answer alarm signal.
2) switch and knob state analysis
The comprehensive characteristics failure judgement such as switch and knob malfunction color to be passed through, texture, shape, space
Type.Record each switch and knob state given corresponding fault type, by the switch gathered back and knob
Picture carries out one_to_one corresponding with corresponding quantized data, sets up and constantly improves switch and knob accident analysis storehouse,
It is finally reached the function of quantitative analysis.
Define different detection projects for different switches and knob, it is carried out accompanying drawing corresponding, often simultaneously
The different word of individual project definition describes and gives different grades.The quantization detection project correspondence that will define
Scene is detected, and carries out unifying input system by the information detected back, carries out the analysis of fault.According to
Different descriptive grades assesses different fault degrees, and the present invention sends phase automatically according to different faults type
Answer alarm signal.
Although the above-mentioned detailed description of the invention to native system invention is described, but not protects the present invention
The restriction of scope, the technological development personnel of art should be understood that the basis in technical scheme
On, the technological development personnel of this area and association area need not to pay that creative work can make is various
Amendment or deformation, still within protection scope of the present invention.
Claims (5)
1. the system carrying out video monitoring automatic alarm for switch cabinet of converting station signal lamp, switch and knob state, this system possesses unusual condition detection automatic alarm, it is characterized in that, the video signal of collection is transmitted compression algorithm and is transferred to backstage, first the video image of transmission is carried out pretreatment, then being identified processing to characteristic point by video image video algorithm, automatically analyze the content of monitoring, abnormal conditions send alarm signal.
2. according to the video image Preprocessing Algorithm described in claim 1, it is characterized in that, video image is first carried out denoising, Denoising Algorithm includes filter in spatial domain, transform domain filtering and partial differential equation, then image being carried out rim detection, edge detection operator has Roberts operator, Prewitt operator, Sobel operator, Isotropic Sobel operator and Laplacian operator.
3. according to the video image identification algorithm described in claim 1, it is characterized in that, analyze switch cabinet of converting station video monitoring image feature, different Image denoising algorithms is used according to feature of image, it is analyzed by color, texture, shape and spatial relationship feature, is identified switch cubicle running status processing.
4. according to the transmission of video compression algorithm described in claim 1；System have employed ITU-T and ISO and jointly formulated video encoding standard of new generation and H.264 encode, and uses intraframe prediction algorithm, motion estimation algorithm and fast motion estimation algorithm to carry out code optimization, improves video transmission efficiency.
5. according to the automatically analyzing of the video monitoring described in claim 1, abnormal alarm method；Switch cabinet of converting station monitored picture is analyzed, in conjunction with the detection analysis of concrete part, provides the hidden danger identification of each equipment, the abnormality checked is sent corresponding alarm signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610443380.5A CN105915645A (en) | 2016-06-20 | 2016-06-20 | Substation video monitoring automatic alarm system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610443380.5A CN105915645A (en) | 2016-06-20 | 2016-06-20 | Substation video monitoring automatic alarm system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105915645A true CN105915645A (en) | 2016-08-31 |
Family
ID=56758017
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610443380.5A Pending CN105915645A (en) | 2016-06-20 | 2016-06-20 | Substation video monitoring automatic alarm system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105915645A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106709466A (en) * | 2016-12-30 | 2017-05-24 | 国网山西省电力公司电力科学研究院 | MPEG7 standard-based electric equipment image recognition method and system |
| CN106894806A (en) * | 2016-11-29 | 2017-06-27 | 攀枝花市九鼎智远知识产权运营有限公司 | A kind of intelligent rig electrohydraulic control system and method |
| CN106980863A (en) * | 2017-03-24 | 2017-07-25 | 哈尔滨理工大学 | A kind of unit exception diagnostic model in transformer substation video monitoring |
| CN107807572A (en) * | 2017-11-01 | 2018-03-16 | 炜呈智能电力科技（杭州）有限公司 | River course lock station machine room monitoring system |
| CN107807573A (en) * | 2017-11-01 | 2018-03-16 | 炜呈智能电力科技（杭州）有限公司 | River course lock station center monitoring method and computer-readable storage medium |
| CN108932808A (en) * | 2018-07-02 | 2018-12-04 | 太仓市友达电气技术有限公司 | A kind of safety defense monitoring system for low-voltage distribution room |
| CN109102669A (en) * | 2018-09-06 | 2018-12-28 | 广东电网有限责任公司 | A kind of transformer substation auxiliary facility detection control method and its device |
| CN109412069A (en) * | 2018-11-24 | 2019-03-01 | 四川都睿感控科技有限公司 | The state monitoring method and device of switchgear |
| CN109553149A (en) * | 2017-09-25 | 2019-04-02 | 神华集团有限责任公司 | Strengthen natural evaporation device and the evaporation pond with the reinforcing natural evaporation device |
| CN109724647A (en) * | 2019-01-21 | 2019-05-07 | 青岛科技大学 | EMU car rises pressure case condition monitoring system and monitoring method |
| CN114101078A (en) * | 2021-11-23 | 2022-03-01 | 博众精工科技股份有限公司 | Detection device |
-
2016
- 2016-06-20 CN CN201610443380.5A patent/CN105915645A/en active Pending
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106894806A (en) * | 2016-11-29 | 2017-06-27 | 攀枝花市九鼎智远知识产权运营有限公司 | A kind of intelligent rig electrohydraulic control system and method |
| CN106709466A (en) * | 2016-12-30 | 2017-05-24 | 国网山西省电力公司电力科学研究院 | MPEG7 standard-based electric equipment image recognition method and system |
| CN106980863A (en) * | 2017-03-24 | 2017-07-25 | 哈尔滨理工大学 | A kind of unit exception diagnostic model in transformer substation video monitoring |
| CN109553149A (en) * | 2017-09-25 | 2019-04-02 | 神华集团有限责任公司 | Strengthen natural evaporation device and the evaporation pond with the reinforcing natural evaporation device |
| CN109553149B (en) * | 2017-09-25 | 2024-03-29 | 国家能源投资集团有限责任公司 | Enhanced natural evaporation device and evaporation pond with same |
| CN107807572A (en) * | 2017-11-01 | 2018-03-16 | 炜呈智能电力科技（杭州）有限公司 | River course lock station machine room monitoring system |
| CN107807573A (en) * | 2017-11-01 | 2018-03-16 | 炜呈智能电力科技（杭州）有限公司 | River course lock station center monitoring method and computer-readable storage medium |
| CN108932808A (en) * | 2018-07-02 | 2018-12-04 | 太仓市友达电气技术有限公司 | A kind of safety defense monitoring system for low-voltage distribution room |
| CN109102669A (en) * | 2018-09-06 | 2018-12-28 | 广东电网有限责任公司 | A kind of transformer substation auxiliary facility detection control method and its device |
| CN109412069A (en) * | 2018-11-24 | 2019-03-01 | 四川都睿感控科技有限公司 | The state monitoring method and device of switchgear |
| CN109724647A (en) * | 2019-01-21 | 2019-05-07 | 青岛科技大学 | EMU car rises pressure case condition monitoring system and monitoring method |
| CN114101078A (en) * | 2021-11-23 | 2022-03-01 | 博众精工科技股份有限公司 | Detection device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105915645A (en) | Substation video monitoring automatic alarm system | |
| Jiang et al. | Scene change detection techniques for video database systems | |
| EP0706155A1 (en) | Method and device for encoding image signal and image signal decoding device | |
| CN105976383A (en) | Power transmission equipment fault diagnosis method based on limit learning machine image recognition | |
| US20190371141A1 (en) | Automatic threat detection based on video frame delta information in compressed video streams | |
| CN107679495B (en) | Detection method for movable engineering vehicles around power transmission line | |
| Mohan et al. | Video image processing for moving object detection and segmentation using background subtraction | |
| CN115294117A (en) | Defect detection method and related device for LED lamp beads | |
| CA3194402A1 (en) | A line clearance system | |
| KR102261669B1 (en) | Artificial Neural Network Based Object Region Detection Method, Device and Computer Program Thereof | |
| Parihar et al. | Graph theory based approach for image segmentation using wavelet transform | |
| Liu et al. | Automatic video activity detection using compressed domain motion trajectories for H. 264 videos | |
| Warhade et al. | Shot boundary detection in the presence of illumination and motion | |
| Abdollahzadeh et al. | Multi-focus image fusion for visual sensor networks | |
| KR101581162B1 (en) | Automatic detection method, apparatus and system of flame, smoke and object movement based on real time images | |
| Hamida et al. | Spatio-temporal video filtering for video surveillance applications | |
| CN104125430A (en) | Method and device for detecting video moving objects as well as video monitoring system | |
| Maksimović et al. | Experimental analysis of wavelet decomposition on edge detection | |
| Boufares et al. | Adaptive threshold for background subtraction in moving object detection using stationary wavelet transforms 2D | |
| Marceline et al. | Cloud storage optimization for video surveillance applications | |
| Cardaillac et al. | Marine snow detection for real time feature detection | |
| US6944343B2 (en) | Image analyzing method for detecting significant changes in a time sequence of images | |
| CN112149683A (en) | Method and device for detecting living objects in night vision environment | |
| Su et al. | Shadow removal using background reconstruction | |
| Mohammed et al. | On the selection of the proper wavelet for moving object detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160831 |