CN109685051A - A kind of infrared image fault diagnosis system based on network system - Google Patents
A kind of infrared image fault diagnosis system based on network system Download PDFInfo
- Publication number
- CN109685051A CN109685051A CN201811353591.5A CN201811353591A CN109685051A CN 109685051 A CN109685051 A CN 109685051A CN 201811353591 A CN201811353591 A CN 201811353591A CN 109685051 A CN109685051 A CN 109685051A
- Authority
- CN
- China
- Prior art keywords
- infrared image
- feature
- module
- temperature
- management server
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/12—Details of acquisition arrangements; Constructional details thereof
- G06V10/14—Optical characteristics of the device performing the acquisition or on the illumination arrangements
- G06V10/143—Sensing or illuminating at different wavelengths
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
- G06F18/2193—Validation; Performance evaluation; Active pattern learning techniques based on specific statistical tests
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The present invention discloses a kind of infrared image fault diagnosis system based on network system, module, image processing module, characteristic extracting module, management server and display terminal are obtained including infrared image, infrared image obtains module and is connect by image processing module with characteristic extracting module, and management server device is connect with characteristic extracting module and display terminal respectively.The present invention obtains module, characteristic extracting module by infrared image and combines management server, the feature of the infrared image of grid equipment can be extracted and be analyzed, and the importance of each feature is analyzed in conjunction with random forests algorithm, to filter out the feature for influencing infrared image fault detection, improve the accuracy and convenience of fault detection, with intelligent characteristic, and substantially increase the ability of failure supervision and detection, so that detection is more convenient, financial resources and material resources are greatlyd save simultaneously, there is far-reaching influence in future.
Description
Technical field
The invention belongs to electric network failure diagnosis technical field, it is related to a kind of infrared image failure based on network system and examines
Disconnected system.
Background technique
The power transformation of the various voltages entirety that electric line forms in one's power, referred to as power network in electric system.It includes to become
Electricity, three transmission of electricity, distribution units, the task of power network are conveying and distribution electric energy, change voltage.Power grid in existing network system
Device fails make staff that can not accurately and easily obtain electric network fault, cause the accuracy of electric network failure diagnosis
Difference, supervision and detectability are insufficient, and then the material resources that waste financial resources significantly, and seriously affect the performance of power grid, more than solving
Problem now designs a kind of infrared image fault diagnosis system based on network system.
Summary of the invention
The purpose of the present invention is to provide the infrared image fault diagnosis system based on network system, solve existing electricity
During net fault detection, there is a problem of the accuracy of fault detection is poor, convenience difference, in turn result in failure supervision and inspection
The scarce capacity of survey, meanwhile, financial resource and material resource are wasted significantly.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of infrared image fault diagnosis system based on network system, including infrared image obtain module, image procossing
Module, characteristic extracting module, management server and display terminal, infrared image obtain module and pass through image processing module and feature
Extraction module connection, management server device are connect with characteristic extracting module and display terminal respectively；
Infrared image obtains module and uses FLIR video camera, for carrying out infrared image acquisition to grid equipment, and will adopt
The infrared image of collection is sent to image processing module, and the temperature of grid equipment and the temperature of place environment can be improved in infrared image
Degree；
Image processing module is filter, the infrared image that module is sent is obtained for receiving infrared image, to received
Noise in infrared image is filtered, and to reduce the noise jamming in infrared image, and filtered infrared image is sent
To characteristic extracting module；
Characteristic extracting module is used to receive the infrared image of image processing module transmission after filtering, in infrared image
Each feature is chosen, and the feature of extraction is sent to management server；
Management server receives the feature that characteristic extracting module is sent, and is carried out to received feature using random forests algorithm
Significance Analysis, and the importance of each feature of analysis is sent to display terminal；
Display terminal is used to receive the importance of each feature of management server transmission and is shown.
Further, the feature for the infrared image that the characteristic extracting module is chosen include infrared image resolution sizes,
Infrared image maximum temperature, minimum temperature, the temperature difference of maximum temperature and minimum temperature, infrared image 95%, 75%,
50%, 25%, 5% point is number, the mean value of image temperature, standard deviation, entropy, kurtosis, several temperature amplitude gradients and several temperature
Direction gradient.
Further, in the temperature amplitude gradient temperature amplitude range be 0-100, etc. amplitudes spacing be divided into 20
It is spaced, the range of temperature direction is -180 ° -180 ° in temperature direction gradient, and carries out temperature direction using every 15 ° as interval
It divides.
Further, the management server is during carrying out different degree to each feature, the knot given for one
Point t, its associated data set is D and Gini coefficient is such as given a definition:
Wherein, p (k | D), k=1 ..., Q are the ratios of k kind sample in D, and associated data set is that characteristic extracting module is extracted
All features constitute set；
For two points of classification problems, Q, which is equal to 2, G (D), to be indicated to randomly select the tool sample different there are two classification from set D
This probability, therefore, the value of G (D) is smaller, and purity, that is, purity will be bigger, when to Category Attributes j=1 ..., m, measurement base
Buddhist nun's coefficient is as follows:Wherein, | Dv| represent the quantity that j-th of attribute takes the sample of j-th of probable value, D
It is expressed as the number of element in D set.For connection attribute j, the mean value of k tree in Gini coefficientIndicate different degree:
Further, the random forests algorithm based on feature in the management server, specifically includes the following steps:
Step 1: initializing and number is set
For i=1to N1do；
Step 2: in null data set S, having using bootstrap method randomly select K new self-service samples with putting back to
Collection, and by secondary K post-class processing of building, the data not being pumped to constitute the outer data (OOB) of bag；
Step 3: being equipped with m feature, randomly select m from each node of every one treetryA feature generates feature
Collection, and calculates the gini coefficient of each feature, and the feature of minimum gini coefficient will be selected as point in character subset
Split rule；
Step 4: each tree increases to greatest extent, does not do any cutting；
Step 5: equalizing the importance of each feature, be arranged in decreasing order, obtain importance rate, and select preceding p%
As character subset
end for
For i=1to N1do；
Step 6: random division data set, and it is used as training set by 80%, 20% is used as test set；
Step 7: K tree of generation constitutes random forest, and classification results are voted by decision tree classifier and obtained；
Step 8: calculating the importance of each feature of test set and the accuracy and AUC value of model；
Step 9: the mean value and standard deviation and AUC value of computational accuracy；
end for
Where mtry=floor (log2(m)+1)。
Beneficial effects of the present invention:
Infrared image fault diagnosis system provided by the invention based on network system, by infrared image obtain module,
Characteristic extracting module simultaneously combines management server, can the feature of the infrared image of grid equipment be extracted and be analyzed, and
It is analyzed in conjunction with importance of the random forests algorithm to each feature, to filter out the feature for influencing infrared image fault detection,
The accuracy and convenience of fault detection are improved, there is intelligent characteristic, and substantially increase the energy of failure supervision and detection
Power so that detection is more convenient, while greatling save financial resources and material resources, will have far-reaching influence in future.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, will be described below to embodiment required
Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for ability
For the those of ordinary skill of domain, without creative efforts, it can also be obtained according to these attached drawings other attached
Figure.
Fig. 1 is a kind of schematic diagram of the infrared image fault diagnosis system based on network system in the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts all other
Embodiment shall fall within the protection scope of the present invention.
Refering to Figure 1, a kind of infrared image fault diagnosis system based on network system, including infrared image obtain
Module, image processing module, characteristic extracting module, management server and display terminal, infrared image obtain module and pass through image
Processing module is connect with characteristic extracting module, and management server device is connect with characteristic extracting module and display terminal respectively；
Infrared image obtains module and uses FLIR video camera, for carrying out infrared image acquisition to grid equipment, and will adopt
The infrared image of collection is sent to image processing module, and the temperature of grid equipment and the temperature of place environment can be improved in infrared image
Degree；
Image processing module is filter, the infrared image that module is sent is obtained for receiving infrared image, to received
Noise in infrared image is filtered, and to reduce the noise jamming in infrared image, and filtered infrared image is sent
To characteristic extracting module；
Characteristic extracting module is used to receive the infrared image of image processing module transmission after filtering, in infrared image
Each feature is chosen, and the feature of extraction is sent to management server, and the feature of the infrared image of selection includes infrared figure
As the temperature difference of resolution sizes, infrared image maximum temperature, minimum temperature, maximum temperature and minimum temperature, infrared image
95%, 75%, 50%, 25%, 5% point is number, the mean value of image temperature, standard deviation, entropy, kurtosis, several temperature amplitude gradients
With several temperature direction gradients, wherein in temperature amplitude gradient temperature amplitude range be 0-100, etc. amplitudes spacing be divided into 20
A interval, the range of temperature direction is -180 ° -180 ° in temperature direction gradient, and carries out temperature direction using every 15 ° as interval
Division；
Management server receives the feature that characteristic extracting module is sent, and is carried out to received feature using random forests algorithm
Significance Analysis, and the importance of each feature of analysis is sent to display terminal；
Display terminal is used to receive the importance of each feature of management server transmission and is shown.
The random forests algorithm is a kind of machine learning algorithm based on decision tree, and random forests algorithm is in missing data
With in the case where unbalanced data have good robustness.In random forest, Gini coefficient and base based on node purity
In the feature selection approach of the classifier accuracy rate of OOB data be the most common feature selection approach used.
Management server is during carrying out different degree to each feature, the node t given for one, its dependency number
According to collection be D and Gini coefficient is such as given a definition:
Wherein, p (k | D), k=1 ..., Q are the ratios of k kind sample in D, and associated data set is that characteristic extracting module is extracted
All features constitute set.
For two points of classification problems, Q, which is equal to 2, G (D), to be indicated to randomly select the tool sample different there are two classification from set D
This probability.Therefore, the value of G (D) is smaller, and purity, that is, purity will be bigger.When to Category Attributes j=1 ..., m, measurement base
Buddhist nun's coefficient is as follows:
Wherein, | Dv| the quantity that j-th of attribute takes the sample of j-th of probable value is represented, D is expressed as element in D set
Number.For connection attribute j, the mean value of k tree in Gini coefficientIt indicates different degree Δ (Gj):
Random forests algorithm based on feature selecting.In random forest the characteristic of the different degree of variable by with to each spy
In the descending arrangement of sign.Wherein, the meeting of preceding p% is chosen as character subset.Random division training set and test set, wherein 80% makees
Test set is used as training set, 20%.After bat has been calculated, by multiple averaging, the mean value and mark of AUC has been obtained
It is quasi- poor.
Demand needed for random forests algorithm is an empty data set S={ x1,x2,…,xn, based on the random of feature
Forest algorithm, specifically includes the following steps:
Step 1: initializing and number is set
For i=1to N1Do, indicates circulation N1 times；
Step 2: in null data set S, having using bootstrap method randomly select K new self-service samples with putting back to
Collection, and by secondary K post-class processing of building, the data not being pumped to constitute the outer data (OOB) of bag；
Step 3: being equipped with m feature, randomly select m from each node of every one treetryA feature generates feature
Collection, and calculates the gini coefficient of each feature, and the feature of minimum gini coefficient will be selected as point in character subset
Split rule；
Step 4: each tree increases to greatest extent, does not do any cutting；
Step 5: equalizing the importance of each feature, be arranged in decreasing order, obtain importance rate, and select preceding p%
As character subset
end for
For i=1to N1Do, expression terminate upper for circulation；
Step 6: random division data set, and it is used as training set by 80%, 20% is used as test set；
Step 7: K tree of generation constitutes random forest, and classification results are voted by decision tree classifier and obtained；
Step 8: calculating the importance of each feature of test set and the accuracy and AUC value of model；
Step 9: the mean value and standard deviation and AUC value of computational accuracy；
end for
Where mtry=floor (log2(m)+1)；
Management server carries out ACC, TPR and TNR in the AUC value of acquisition pair with the threshold value of ACC, TPR and TNR respectively
Than, and the threshold value setting of ACC, TPR and TNR are respectively 270/19,633 ≈ 0.01375, ntree=500, mtry=7.
Wherein, ACC indicates accuracy, ACC=(TP+TN)/(TN+FP+FN+TP), and TPR table is shown as real class rate,TNR is expressed as very negative class rate, the very negative class rate of TNR=TN/ (TN+FP), True negative
(TN) it is known as Kidney-Yin rate, shows that really sample number of the negative sample prediction at negative sample, False positive (FP) are referred to as false
Positive rate shows really negative sample prediction into the sample number of positive sample, and False negative (FN), referred to as False-Negative Rate show
At the sample number of negative sample, True positive (TP), referred to as kidney-Yang rate show really positive sample for really positive sample prediction
This prediction at positive sample sample number.
The region area that AUC value is covered by ROC curve, i.e. AUC are bigger, and classifier classifying quality is better.ROC
(Receiver Operating Characteristic) is translated as " recipient's operating characteristic curve ".Curve is by two variables
FPR and TPR is drawn, and TPR reflects positive class level of coverage.
Infrared image fault diagnosis system provided by the invention based on network system, by infrared image obtain module,
Characteristic extracting module simultaneously combines management server, can the feature of the infrared image of grid equipment be extracted and be analyzed, and
It is analyzed in conjunction with importance of the random forests algorithm to each feature, to filter out the feature for influencing infrared image fault detection,
The accuracy and convenience of fault detection are improved, there is intelligent characteristic, and substantially increase the energy of failure supervision and detection
Power so that detection is more convenient, while greatling save financial resources and material resources, will have far-reaching influence in future.
The above content is just an example and description of the concept of the present invention, affiliated those skilled in the art
It makes various modifications or additions to the described embodiments or is substituted in a similar manner, without departing from invention
Design or beyond the scope defined by this claim, be within the scope of protection of the invention.
Claims (5)
1. a kind of infrared image fault diagnosis system based on network system, it is characterised in that: including infrared image obtain module,
Image processing module, characteristic extracting module, management server and display terminal, infrared image obtain module and pass through image procossing mould
Block is connect with characteristic extracting module, and management server device is connect with characteristic extracting module and display terminal respectively；
Infrared image obtains module and uses FL I R video camera, for carrying out infrared image acquisition to grid equipment, and will acquisition
Infrared image be sent to image processing module, and the temperature of grid equipment and the temperature of place environment can be improved in infrared image；
Image processing module is filter, the infrared image that module is sent is obtained for receiving infrared image, to received infrared
Noise in image is filtered, and to reduce the noise jamming in infrared image, and filtered infrared image is sent to spy
Levy extraction module；
Characteristic extracting module is used to receive the infrared image of image processing module transmission after filtering, to each spy in infrared image
Sign is chosen, and the feature of extraction is sent to management server；
Management server receives the feature that characteristic extracting module is sent, important using random forests algorithm progress to received feature
Degree analysis, and the importance of each feature of analysis is sent to display terminal；
Display terminal is used to receive the importance of each feature of management server transmission and is shown.
2. a kind of infrared image fault diagnosis system based on network system according to claim 1, it is characterised in that: institute
The feature for stating the infrared image of characteristic extracting module selection includes infrared image resolution sizes, infrared image maximum temperature, most
Low temperature, the temperature difference of maximum temperature and minimum temperature, infrared image 95%, 75%, 50%, 25%, 5% point for number, figure
As the mean value of temperature, standard deviation, entropy, kurtosis, several temperature amplitude gradients and several temperature direction gradients.
3. a kind of infrared image fault diagnosis system based on network system according to claim 2, it is characterised in that: institute
The range for stating temperature amplitude in temperature amplitude gradient is 0-100, etc. amplitudes spacing be divided into 20 intervals, in temperature direction gradient
The range of temperature direction is -180 ° -180 °, and the division of temperature direction is carried out using every 15 ° as interval.
4. a kind of infrared image fault diagnosis system based on network system according to claim 1, it is characterised in that: institute
Management server is stated during carrying out different degree to each feature, the node t given for one, its associated data set is
D and Gini coefficient is such as given a definition:
Wherein, p (k | D), k=1 ..., Q are the ratios of k kind sample in D, and associated data set is the institute that characteristic extracting module is extracted
The set for thering is feature to constitute；
For two points of classification problems, Q, which is equal to 2, G (D), indicates to randomly select the different sample of classification there are two tools from set D
Probability, therefore, the value of G (D) is smaller, and purity, that is, purity will be bigger, when to Category Attributes j=1 ..., m, measurement Geordie system
Number is as follows:
Wherein, | Dυ| the quantity that j-th of attribute takes the sample of j-th of probable value is represented, D is expressed as the number of element in D set,
For connection attribute j, the mean value of k tree in Gini coefficientIndicate different degree:
5. a kind of infrared image fault diagnosis system based on network system according to claim 1, it is characterised in that: institute
The random forests algorithm based on feature in management server is stated, specifically includes the following steps:
Step 1: initializing and number is set
For i=1 to N1do；
Step 2: in null data set S, has using bootstrap method randomly select K new self-service sample sets with putting back to, and
By secondary K post-class processing of building, the data not being pumped to constitute the outer data (OOB) of bag；
Step 3: being equipped with m feature, randomly select m from each node of every one treetryA feature generates character subset, and
And the gini coefficient of each feature is calculated, the feature of minimum gini coefficient will be selected as division rule in character subset
Then；
Step 4: each tree increases to greatest extent, does not do any cutting；
Step 5: equalize the importance of each feature, be arranged in decreasing order, obtain importance rate, and select preceding p% as
As character subset
end for
For i=1 to N1do；
Step 6: random division data set, and it is used as training set by 80%, 20% is used as test set；
Step 7: K tree of generation constitutes random forest, and classification results are voted by decision tree classifier and obtained；
Step 8: calculating the importance of each feature of test set and the accuracy and AUC value of model；
Step 9: the mean value and standard deviation and AUC value of computational accuracy；
end for
Where mtry=floor (log2(m)+1)。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811353591.5A CN109685051A (en) | 2018-11-14 | 2018-11-14 | A kind of infrared image fault diagnosis system based on network system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811353591.5A CN109685051A (en) | 2018-11-14 | 2018-11-14 | A kind of infrared image fault diagnosis system based on network system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109685051A true CN109685051A (en) | 2019-04-26 |
Family
ID=66184616
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811353591.5A Pending CN109685051A (en) | 2018-11-14 | 2018-11-14 | A kind of infrared image fault diagnosis system based on network system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109685051A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110222087A (en) * | 2019-05-15 | 2019-09-10 | 平安科技（深圳）有限公司 | Feature extracting method, device and computer readable storage medium |
| CN112364928A (en) * | 2020-11-18 | 2021-02-12 | 浙江工业大学 | Random forest classification method in transformer substation fault data diagnosis |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107818320A (en) * | 2017-10-27 | 2018-03-20 | 国网四川省电力公司德阳供电公司 | Recognition methods based on OCR technique transformer infrared image numerical value of increasing income |
-
2018
- 2018-11-14 CN CN201811353591.5A patent/CN109685051A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107818320A (en) * | 2017-10-27 | 2018-03-20 | 国网四川省电力公司德阳供电公司 | Recognition methods based on OCR technique transformer infrared image numerical value of increasing income |
Non-Patent Citations (6)
| Title |
|---|
| 何弘等: "《无盘站与终端技术从入门到精通》", 31 January 2003 * |
| 方立华: "基于特征振动的风电机组传动系统机械故障诊断研究", 《中国优秀博硕士学位论文全文数据库（硕士）工程科技Ⅱ辑》 * |
| 李涛: "《数字图像处理之红外弱目标分割方法研究》", 30 June 2016 * |
| 李鑫等: "电力设备IＲ图像特征提取及故障诊断方法研究", 《激光与红外》 * |
| 葛致磊等: "《导弹导引系统原理》", 31 March 2016 * |
| 邢素霞: "《红外热成像与信号处理》", 31 January 2011 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110222087A (en) * | 2019-05-15 | 2019-09-10 | 平安科技（深圳）有限公司 | Feature extracting method, device and computer readable storage medium |
| WO2020228283A1 (en) * | 2019-05-15 | 2020-11-19 | 平安科技（深圳）有限公司 | Feature extraction method and apparatus, and computer readable storage medium |
| CN112364928A (en) * | 2020-11-18 | 2021-02-12 | 浙江工业大学 | Random forest classification method in transformer substation fault data diagnosis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104699755B (en) | A kind of intelligent multiple target integrated recognition method based on data mining | |
| CN107209853A (en) | Positioning and map constructing method | |
| EP4276487A2 (en) | Generating wireless network access point models using clustering techniques | |
| CN102110122A (en) | Method and device for establishing sample picture index table, method and device for filtering pictures and method and device for searching pictures | |
| CN106980827A (en) | A kind of method of Bird's Nest in identification transmission line of electricity based on Aerial Images | |
| CN110533116A (en) | Based on the adaptive set of Euclidean distance at unbalanced data classification method | |
| CN105631404B (en) | The method and device that photo is clustered | |
| CN110688888B (en) | Pedestrian attribute identification method and system based on deep learning | |
| Zhang et al. | Automatic recognition of oil industry facilities based on deep learning | |
| CN109685051A (en) | A kind of infrared image fault diagnosis system based on network system | |
| CN104317946A (en) | Multi-key image-based image content retrieval method | |
| CN109410190A (en) | Shaft tower based on High Resolution Remote Sensing Satellites image falls disconnected detection model training method | |
| CN101251896A (en) | Object detecting system and method based on multiple classifiers | |
| CN110458432B (en) | Cloud model-based reliability diagnosis method for electric power optical transmission network | |
| Gueguen et al. | Mapping human settlements and population at country scale from VHR images | |
| CN104933178B (en) | Official website determines method and system and the sort method of official website | |
| CN104680118B (en) | A kind of face character detection model generation method and system | |
| CN105868242A (en) | Sorting method and system for labels in network recommendation | |
| CN113313110A (en) | License plate type recognition model construction and license plate type recognition method | |
| CN109167673B (en) | Novel cloud service screening method integrating abnormal Qos data detection | |
| CN108268478A (en) | A kind of unbalanced dataset feature selection approach and device based on ur-CAIM algorithms | |
| CN110751201A (en) | SAR equipment task failure cause reasoning method based on textural feature transformation | |
| CN108615056A (en) | A kind of tree enhancing Naive Bayes Classification method based on decomposable asymmetric choice net score function | |
| CN105291104B (en) | Robot remote control method based on cloud platform | |
| CN108647275A (en) | The recognition methods of isolated island detecting state and device, storage medium, processor |
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: 20190426 |