CN109509169A - The pressing plate condition checkout gear and method of learning distance metric based on nearest neighbour classification - Google Patents
The pressing plate condition checkout gear and method of learning distance metric based on nearest neighbour classification Download PDFInfo
- Publication number
- CN109509169A CN109509169A CN201811034011.6A CN201811034011A CN109509169A CN 109509169 A CN109509169 A CN 109509169A CN 201811034011 A CN201811034011 A CN 201811034011A CN 109509169 A CN109509169 A CN 109509169A
- Authority
- CN
- China
- Prior art keywords
- pressing plate
- distance metric
- nearest neighbour
- processing unit
- feature
- 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
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
- G06V10/507—Summing image-intensity values; Histogram projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- 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
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
Abstract
A kind of the pressing plate condition checkout gear and method of the learning distance metric based on nearest neighbour classification, including central processing unit, data information transfer route, storage hard disk, camera system, data connection route, integrated plate, wherein, the central processing unit is connected with the storage hard disk by data information transfer route, the camera system is installed at the face position of pressing plate switch, it is connected in the integrated plate by data connection route, the integrated plate is connected with central processing unit.The present invention overcomes the deficiencies in the prior art, and intelligent measurement identification can be carried out to pressing plate switch state in real time by providing one kind, to further decrease the apparatus and method of traditional artificial detection identification erroneous judgement that may be present and caused security risk.
Description
Technical field
The present invention relates to vision automatic recognition technical fields, more particularly, to a kind of distance degree based on nearest neighbour classification
Measure the pressing plate condition checkout gear and method of study.
Background technique
Power Secondary pressing plate be to ensure that relay protection and automatic device safe operation important leverage means, effect be
Artificial one breakpoint of manufacture in trip protection circuit provides maintenance convenience and safety supports.It can by providing that the pressing plate must be
Depending on obvious cut-off point, the throwing of pressing plate determines the on-off of secondary circuit whether moving back, therefore the throwing state of moving back of pressing plate must be correct,
Otherwise consequence is hardly imaginable.The status monitoring demand of secondary pressure plate: secondary pressure plate needs specific location information prompt, i.e., without
It is to put into or exit by pressing plate, whether completely in place without clear to operator in virtual connection state after manual operation
Information alert；Secondary pressure plate needs to realize that a distant place monitors in real time, and anti-error host can conjugate in real time after pressing plate state changes, i.e.,
Pressing plate state after Shi Chengxian displacement；Pressing plate detection mode specifically includes that the pressing plate inspection of hardware detecting circuit, view-based access control model at present
Survey technology and contactless pressure plate status monitoring system.As machine learning, computer vision etc. flourish, view-based access control model
Pressing plate detection technique has been more and more widely used.Image Classfication Technology is broadly divided into three phases: image pre-processing phase,
Image characteristics extraction stage and classifier construction phase, image pre-processing phase, image characteristics extraction stage are with classifier structure
Make the stage.For given original image, need to pre-process it i.e. digitlization, normalization, smooth, recovery and enhancing
Etc. a variety of processing, so that the irrelevant information eliminated in image retains effective information, it is real by simplifying image data to greatest extent
Now preferably improve the validity of next feature extraction and image classification.Image characteristics extraction typically refers to pass through calculating
The processing of machine handles the invariant feature extraction of image, and one of the important link as Image Classfication Technology, feature extraction can
Image is expressed as to computer in turn can identify the vector form of union, and it is high to meet easily extraction, easily differentiation and robustness
Etc. performances.And the classifier for constructing superperformance is also one of important link of Image Classfication Technology, it mainly passes through logarithm
Learnt according to the great amount of images data sample of concentration.
Summary of the invention
The present invention in order to overcome at least one of the drawbacks of the prior art described above, provides a kind of distance based on nearest neighbour classification
The pressing plate condition checkout gear and method of metric learning.
The present invention is directed to solve above-mentioned technical problem at least to a certain extent.
Primary and foremost purpose of the invention is to meet to carry out intelligent measurement identification to pressing plate switch state in real time, with into one
Step reduces traditional artificial detection identification erroneous judgement that may be present and caused security risk.
In order to solve the above technical problems, technical scheme is as follows:
A kind of the pressing plate condition checkout gear and method of the learning distance metric based on nearest neighbour classification, including pressing plate switch, it is special
Sign is, further includes central processing unit, data information transfer route, storage hard disk, camera system, data connection route, collection
At plate, in which:
The central processing unit is connected with the storage hard disk by data information transfer route；
The camera system is installed at the face position of pressing plate switch, passes through data connection route phase in the integrated plate
Even；
The integrated plate is connected with central processing unit.
Workflow of the invention is as follows:
Obtain pressing plate switch image in real time by camera system, the pressing plate switch image transmitting that camera system is got is extremely
In central processing unit, central processing unit carries out image preprocessing to pressing plate switch image, reduces illumination and other adverse effects, base
In the learning distance metric method training pattern of big spacing nearest neighbour classification, whole pressing plate characteristics of image is carried out to pressing plate switch image
Detection, after detection system detects all pressing plate switchs, the pressing plate switch that detects is by information identification system, from before
The feature of current pressing plate switch is obtained in the characteristics of image of preservation, then using trained identification model to each pressing plate switch
State identified, finally export result.
Preferably, the data of storage hard disk storage are the property data base of pressing plate switch, are mentioned for detection pressing plate switch state
For contrast sample.
Preferably, integrated column includes AD conversion module, GPS module, reset controller, wireless transmission/receiving module,
In:
Camera system is connected with AD conversion module；
AD conversion module is connected by data connection route with wireless transmission/receiving module；
Wireless transmission/receiving module is connected by signal transmission line with reset controller；
GPS module is connected by information transmission line with wireless transmission/receiving module.
Camera system obtain pressing plate switch image become digital signal after AD conversion module, by wireless transmission/
Receiving module sends the signal to central processing unit, handles subsequent step by central processing unit；Relevant staff can pass through
GPS module is accurately positioned the pressing plate switch position broken down, and reset controller reverts to the working condition of integrated plate just
Beginning state, in order to avoid wireless transmission/receiving module error of performance.
Preferably, the learning distance metric method training pattern of nearest neighbour classification, comprising the following steps:
S1: for statistical analysis to the pixel distribution of three Color Channels H, S, V to extract hsv color histogram feature；
S2: Lab color characteristic is extracted；
S3: extracting the textural characteristics of LBPs operator, and the feature mode number that uniform LBPs texture pattern generates is P (P-1)+3,
In, P is the number of sampled point；
S4: utilizing k nearest neighbor classification method, realizes pressing plate state-detection to the learning distance metric based on nearest neighbour classification, passes through instruction
Same class keeps its k nearest neighbor after white silk, and significantly separates from different types of holding, in learning framework, by utilizing sample
This class label defines metric function, optimizes metric function using Semidefinite Programming.
This method extracts the color characteristics of pressing plate switch image of video camera shooting, in textural characteristics, with storage hard disk
Database is compared, by debugging algorithm, so that the pressing plate switch image of same state keeps k nearest neighbor, the pressure of different conditions
Switching plate image is then kept away from, and is realized and is judged pressing plate switch state automation and intelligentification.
Compared with prior art, the beneficial effect of technical solution of the present invention is:
The present invention carries out real-time monitoring acquisition to pressing plate switch state by camera system, and will be corresponding by integrated plate
The image information and location information of pressing plate switch are transmitted to central processing unit, pass through the number of central processing unit and property data base
It is effectively analyzed it is believed that breath combines, pressing plate state-detection, energy is realized by a kind of learning distance metric based on nearest neighbour classification
It is enough that intelligent measurement identification is carried out to pressing plate switch state in real time, it is that may be present to further decrease traditional artificial detection identification
Erroneous judgement and caused security risk make alarm to some emergencies convenient for relevant device, staff are prompted to pacify
Complete effective processing, improves power grid intelligent operation efficiency.
Detailed description of the invention
Fig. 1 is pressing plate state-detection process of the invention
Fig. 2 is structural schematic diagram of the invention.
Specific embodiment
The attached figures are only used for illustrative purposes and cannot be understood as limitating the patent；
In order to better illustrate this embodiment, the certain components of attached drawing have omission, zoom in or out, and do not represent the ruler of actual product
It is very little；
To those skilled in the art, the omitting of some known structures and their instructions in the attached drawings are understandable.
The following further describes the technical solution of the present invention with reference to the accompanying drawings and examples.
Embodiment 1
A kind of the pressing plate condition checkout gear and method of the learning distance metric based on nearest neighbour classification, including pressing plate switch, it is special
Sign is, further includes central processing unit, data information transfer route, storage hard disk, camera system, data connection route, collection
At plate, in which:
The central processing unit is connected with the storage hard disk by data information transfer route；
The camera system is installed at the face position of pressing plate switch, passes through data connection route phase in the integrated plate
Even；
The integrated plate is connected with central processing unit.
The data of storage hard disk storage are the property data base of pressing plate switch, provide contrast sample.
In the specific implementation process, it obtains pressing plate switch image in real time by camera system, camera system is obtained
In the pressing plate switch image transmitting to central processing unit arrived, central processing unit carries out image preprocessing, drop to pressing plate switch image
Low illumination and other adverse effects, the learning distance metric method training pattern based on nearest neighbour classification, to pressing plate switch image into
The detection of row entirety pressing plate characteristics of image is for statistical analysis to the pixel distribution of three Color Channels H, S, V to extract HSV
Color histogram feature；Extract Lab color characteristic；Extract the textural characteristics of LBPs operator.Generated feature mode number is P
(P-1)+3, make always to be to maintain its neighbour by the measurement after training and always belongs to same class, and protected from different types of sample
It holds and significantly separates, by utilizing sample class tag definition metric function, metric function is optimized using Semidefinite Programming.
After detection system detects all pressing plate switchs, the pressing plate switch that detects is by information identification system, from previously stored
The feature of current pressing plate switch is obtained in characteristics of image, then using trained identification model to the state of each pressing plate switch
It is identified, finally exports result.
The same or similar label correspond to the same or similar components；
The terms describing the positional relationship in the drawings are only for illustration, should not be understood as the limitation to this patent；
Obviously, the above embodiment of the present invention be only to clearly illustrate example of the present invention, and not be to this hair
The restriction of bright embodiment.For those of ordinary skill in the art, it can also do on the basis of the above description
Other various forms of variations or variation out.There is no necessity and possibility to exhaust all the enbodiments.It is all in the present invention
Spirit and principle within made any modifications, equivalent replacements, and improvements etc., should be included in the guarantor of the claims in the present invention
Within the scope of shield.
Claims (6)
1. a kind of pressing plate condition checkout gear of the learning distance metric based on nearest neighbour classification, including pressing plate switch, feature exist
In further including central processing unit, data information transfer route, storage hard disk, camera system, data connection route, circuit board
Block, in which:
The central processing unit is connected with the storage hard disk by data information transfer route；
The camera system is installed at the face position of the pressing plate switch, passes through data connecting line in the integrated plate
Road is connected；
The integrated plate is connected with the central processing unit.
2. the pressing plate condition checkout gear for the learning distance metric based on nearest neighbour classification told according to claim 1, feature
It is, the data of the storage hard disk storage are the property data base of pressing plate switch.
3. the pressing plate condition checkout gear of the learning distance metric according to claim 1 based on nearest neighbour classification, feature
It is, the integrated column includes AD conversion module, GPS module, reset controller, wireless transmission/receiving module, in which:
The camera system is connected with the AD conversion module；
The AD conversion module is connected by data connection route with told wireless transmission/receiving module；
Wireless transmission/the receiving module is connected by signal transmission line with the reset controller；
The GPS module is connected by information transmission line with the wireless transmission/receiving module.
4. a kind of application side of the pressing plate condition checkout gear of the learning distance metric described in claim 1 based on nearest neighbour classification
Method, which is characterized in that its step are as follows:
S1: pressing plate switch image is obtained by the camera system in real time, and the pressing plate that the camera system obtains is opened
Image transmitting is closed to central processing unit processing；
S2: the central processing unit, which carries out image preprocessing to pressing plate switch image, reduces illumination and other adverse effects;
S3: the learning distance metric method training pattern based on nearest neighbour classification carries out whole pressing plate image to pressing plate switch image
The detection of feature;
S4: after detecting all pressing plate switchs, detect pressing plate switch by information identify, deposited from the storage hardboard
The feature that current switch is obtained in the property data base of the pressing plate switch of storage, then using identification model to each pressing plate switch
State is identified, result is finally exported.
5. the pressing plate condition detection method of the learning distance metric according to claim 4 based on nearest neighbour classification, feature
It is, the learning distance metric method training pattern of the nearest neighbour classification, comprising the following steps:
S1: for statistical analysis to the pixel distribution of three Color Channels H, S, V to extract hsv color histogram feature；
S2: Lab color characteristic is extracted；
S3: extracting the textural characteristics of LBPs operator, and the feature mode number that uniform LBPs texture pattern generates is P (P-1)+3,
In, P is the number of sampled point；
S4: utilizing k nearest neighbor classification method, realizes pressing plate state-detection to the learning distance metric based on nearest neighbour classification, passes through instruction
Same class keeps its k nearest neighbor after white silk, and significantly separates from different types of holding.
6. the pressing plate condition detection method of the learning distance metric according to claim 5 based on nearest neighbour classification, feature
It is, the learning distance metric method training pattern of the nearest neighbour classification, in learning framework, by utilizing sample class mark
Adopted metric function is signed, metric function is optimized using Semidefinite Programming.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811034011.6A CN109509169A (en) | 2018-09-05 | 2018-09-05 | The pressing plate condition checkout gear and method of learning distance metric based on nearest neighbour classification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811034011.6A CN109509169A (en) | 2018-09-05 | 2018-09-05 | The pressing plate condition checkout gear and method of learning distance metric based on nearest neighbour classification |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109509169A true CN109509169A (en) | 2019-03-22 |
Family
ID=65745691
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811034011.6A Pending CN109509169A (en) | 2018-09-05 | 2018-09-05 | The pressing plate condition checkout gear and method of learning distance metric based on nearest neighbour classification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109509169A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111289853A (en) * | 2020-02-25 | 2020-06-16 | 安徽炬视科技有限公司 | Channel-space attention mechanism-based insulator detection system and algorithm |
| CN112508940A (en) * | 2020-12-22 | 2021-03-16 | 三峡大学 | Method for identifying switching state of functional protection pressing plate of transformer substation |
| CN115795276A (en) * | 2022-11-22 | 2023-03-14 | 南京电力设计研究院有限公司 | Secondary circuit pressing plate state evaluation method based on wavelet analysis and machine learning |
| CN115795276B (en) * | 2022-11-22 | 2024-04-26 | 南京电力设计研究院有限公司 | Secondary circuit pressing plate state evaluation method based on wavelet analysis and machine learning |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103413130A (en) * | 2013-07-25 | 2013-11-27 | 中国农业大学 | Recognition method of protection pressing plate state |
| CN104408473A (en) * | 2014-12-08 | 2015-03-11 | 中华人民共和国山东出入境检验检疫局 | Distance metric learning-based cotton grading method and device |
| CN105205493A (en) * | 2015-08-29 | 2015-12-30 | 电子科技大学 | Video stream-based automobile logo classification method |
| CN107527344A (en) * | 2017-09-18 | 2017-12-29 | 广东电网有限责任公司珠海供电局 | A kind of relay-protection pressing plate condition detecting system and detection method |
| CN107680093A (en) * | 2017-10-12 | 2018-02-09 | 广东电网有限责任公司电力科学研究院 | The discrimination method and device of state are moved back in a kind of throwing of the interconnection system switch of protection pressing plate |
-
2018
- 2018-09-05 CN CN201811034011.6A patent/CN109509169A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103413130A (en) * | 2013-07-25 | 2013-11-27 | 中国农业大学 | Recognition method of protection pressing plate state |
| CN104408473A (en) * | 2014-12-08 | 2015-03-11 | 中华人民共和国山东出入境检验检疫局 | Distance metric learning-based cotton grading method and device |
| CN105205493A (en) * | 2015-08-29 | 2015-12-30 | 电子科技大学 | Video stream-based automobile logo classification method |
| CN107527344A (en) * | 2017-09-18 | 2017-12-29 | 广东电网有限责任公司珠海供电局 | A kind of relay-protection pressing plate condition detecting system and detection method |
| CN107680093A (en) * | 2017-10-12 | 2018-02-09 | 广东电网有限责任公司电力科学研究院 | The discrimination method and device of state are moved back in a kind of throwing of the interconnection system switch of protection pressing plate |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111289853A (en) * | 2020-02-25 | 2020-06-16 | 安徽炬视科技有限公司 | Channel-space attention mechanism-based insulator detection system and algorithm |
| CN112508940A (en) * | 2020-12-22 | 2021-03-16 | 三峡大学 | Method for identifying switching state of functional protection pressing plate of transformer substation |
| CN115795276A (en) * | 2022-11-22 | 2023-03-14 | 南京电力设计研究院有限公司 | Secondary circuit pressing plate state evaluation method based on wavelet analysis and machine learning |
| CN115795276B (en) * | 2022-11-22 | 2024-04-26 | 南京电力设计研究院有限公司 | Secondary circuit pressing plate state evaluation method based on wavelet analysis and machine learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110826538B (en) | Abnormal off-duty identification system for electric power business hall | |
| CN102509100B (en) | Knife switch connecting-disconnecting reliability recognition method based on image pattern recognition | |
| CN108537154A (en) | Transmission line of electricity Bird's Nest recognition methods based on HOG features and machine learning | |
| CN110414462A (en) | A kind of unsupervised cross-domain pedestrian recognition methods and system again | |
| CN108055501A (en) | A kind of target detection and the video monitoring system and method for tracking | |
| CN102521565A (en) | Garment identification method and system for low-resolution video | |
| CN109271872B (en) | Device and method for judging on-off state and diagnosing fault of high-voltage isolating switch | |
| CN105404867B (en) | A kind of substation isolating-switch state identification method of view-based access control model | |
| CN109271844B (en) | Electrical cabinet electrical symbol recognition method based on OpenCV | |
| CN109583324A (en) | A kind of pointer meters reading automatic identifying method based on the more box detectors of single-point | |
| CN106778609A (en) | A kind of electric power construction field personnel uniform wears recognition methods | |
| CN109344842A (en) | A kind of pedestrian's recognition methods again based on semantic region expression | |
| CN106803248B (en) | Fuzzy license plate image blur evaluation method | |
| CN102609730B (en) | Method based on contact proportion of dynamic contact and static contact and used for distinguishing switching-on and switching-off reliability of disconnecting link | |
| CN110175571A (en) | The intellectual monitoring of substation equipment state and recognition methods | |
| CN109583295A (en) | A kind of notch of switch machine automatic testing method based on convolutional neural networks | |
| CN116824517B (en) | Substation operation and maintenance safety control system based on visualization | |
| CN109509169A (en) | The pressing plate condition checkout gear and method of learning distance metric based on nearest neighbour classification | |
| CN111639568A (en) | Navigation patrol method for intelligent robot of transformer substation | |
| CN109166293A (en) | Remote assistant method for early warning based on the detection of power transformation stand body | |
| CN110188624A (en) | A kind of substation's wind drift recognition methods and system based on deep learning | |
| CN110008828A (en) | Pairs of constraint ingredient assay measures optimization method based on difference regularization | |
| CN106803259B (en) | A kind of continuous productive process platform plume Automatic Visual Inspection and method of counting | |
| CN111199250A (en) | Transformer substation air switch state checking method and device based on machine learning | |
| CN111597919A (en) | Human body tracking method in video monitoring scene |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190322 |