CN110415236A - A kind of method for detecting abnormality of the complicated underground piping based on double-current neural network - Google Patents
A kind of method for detecting abnormality of the complicated underground piping based on double-current neural network Download PDFInfo
- Publication number
- CN110415236A CN110415236A CN201910695703.3A CN201910695703A CN110415236A CN 110415236 A CN110415236 A CN 110415236A CN 201910695703 A CN201910695703 A CN 201910695703A CN 110415236 A CN110415236 A CN 110415236A
- Authority
- CN
- China
- Prior art keywords
- image
- rgbdifference
- feature
- video
- rgb image
- 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
- 238000000034 method Methods 0.000 title claims abstract description 57
- 230000005856 abnormality Effects 0.000 title claims abstract description 35
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 27
- 230000002159 abnormal effect Effects 0.000 claims abstract description 34
- 238000001514 detection method Methods 0.000 claims abstract description 26
- 238000011946 reduction process Methods 0.000 claims abstract description 23
- 230000004927 fusion Effects 0.000 claims description 15
- 238000005070 sampling Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000009434 installation Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims 1
- 230000002547 anomalous effect Effects 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 230000003321 amplification Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000003199 nucleic acid amplification method Methods 0.000 description 2
- 230000035807 sensation Effects 0.000 description 2
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000000155 melt Substances 0.000 description 1
- 238000005086 pumping Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Multimedia (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to pipeline abnormality detection technical fields, disclose a kind of method for detecting abnormality of complicated underground piping based on double-current neural network, include: that data prediction is carried out to input video, obtains RGB image and RGBDifference image, respectively with growth data and fixed video duration；Noise reduction process is carried out to RGB image and RGBDifference image；According to the RGB image and RGBDifference image after noise reduction process, RGB image feature and RGBDifference characteristics of image are obtained respectively；According to RGB image feature and RGBDifference characteristics of image, image co-registration feature is obtained, to promote the image-capable of the method for detecting abnormality of complicated underground piping；Image co-registration feature is inputted into single task network, single task network includes a full articulamentum, video frame level score is obtained, to distinguish the normal or abnormal information of video；Image co-registration feature is inputted into multitask network, multitask network includes a full articulamentum, abnormal class score is obtained, to distinguish the abnormal class of video.Anomalous video recognition accuracy original text, is greatly saved manpower.
Description
Technical field
The present invention relates to pipeline abnormality detection technical field more particularly to it is a kind of based on double-current neural network intricately under
The method for detecting abnormality of pipeline.
Background technique
As China's economy is grown rapidly, traffic is further convenient, and a large amount of populations pour in city, mentions for the booming of city
Endlessly blood is supplied.With the expansion of city size, matching complicated Urban Buried Pipeline Network Systems have been propped up just
The townie every aspect of benefit.And every year to the maintenance of underground pipe network, need to put into huge manpower and material resources.In order to reduce money
Source loss ensures staff's life security of going into the well, when robot progress pipeline shooting+algorithm identification pipeline has become extremely
Main maintenance model down.
To complicated urban Underground pipeline method for detecting abnormality, it is concentrated mainly on artificial mark analysis, and utilize traditional machine
Device visible sensation method detects in the two aspects.Artificial mark analysis needs to arrange the staff for largely having received special training, by
While a pipe video of traversal frame by frame, it is also noted that whether distinguishing pipe video normally.And use conventional machines visible sensation method
Detection technique, calculated mostly using features descriptions such as HOG (gradient orientation histogram), Edgelet (edge feature) in image
Son, it is abnormal in conjunction with the machine learning classifications methods such as SVM (support vector machines), Adaboost detection pipeline.It is more rare to be based on
Deep learning method usually goes simply to be classified by the method for single frames picture classification.
In the prior art, the method for artificial mark analysis needs to expend a large amount of manpowers, and staff one by one manage frame by frame by traversal
Road video gives normal and abnormal label.In addition, this method is particularly susceptible to factor and individual subjective factor influence, it is same
Video may have different class labels in the hand of different people；Traditional machine vision method can save part manpower money
Source, but this method passes through the classification methods such as extraction characteristics of the underlying image+SVM mostly and is detected.Due to extracting aspect ratio
Rougher, recognition accuracy is not usually high；And it is existing a little based on deep learning method, it is all based on single-frame images and is divided
Class.Neural network designed by them is all than shallower, it is difficult to acquire preferable feature, accuracy of identification can be also pulled low.In addition to this,
Video pumping frame is processed into single-frame images, then carry out classification frame by frame to drag slow algorithm detection frame per second significantly.
Therefore, how to improve becomes skill urgently to be resolved to the detection accuracy of complicated urban Underground pipeline method for detecting abnormality
Art problem.
Summary of the invention
The technical problem to be solved in the present invention is that how to improve the inspection to complicated urban Underground pipeline method for detecting abnormality
Survey precision.
For this purpose, according in a first aspect, the embodiment of the invention discloses a kind of complicated buried pipes based on double-current neural network
The method for detecting abnormality in road, comprising the following steps: S10, to input video carry out data prediction, respectively obtain RGB image and
RGBDifference image, with growth data and fixed video duration；S20 schemes the RGB image and RGBDifference
As carrying out noise reduction process；S30 obtains RGB according to the RGB image and RGBDifference image after noise reduction process respectively
Characteristics of image and RGBDifference characteristics of image；S40, it is special according to the RGB image feature and RGBDifference image
Sign obtains image co-registration feature, to promote the image-capable of the method for detecting abnormality of complicated underground piping；S50, will be described
Image co-registration feature inputs single task network, and the single task network includes a full articulamentum, obtains video frame level score,
To distinguish the normal or abnormal information of video；Described image fusion feature is inputted multitask network, the multitask net by S60
Network includes a full articulamentum, abnormal class score is obtained, to distinguish the abnormal class of video.
Optionally, the step S10 includes: S110, and input video is equidistantly divided into n segment；S120, according to institute
N segment is stated, in each segment stochastical sampling RGB image；S130, according to the RGB image, near sample frame and its
Image difference generates RGBDifference image, with growth data and fixed video duration.
Optionally, the step S120 includes: according to the n segment, and stochastical sampling single frames RGB schemes in each segment
Picture or multiframe RGB image.
Optionally, the step S30 includes: S310, and the RGB image is handled using depth convolutional network, is obtained
RGB image feature；S320, the RGBDifference image are handled using depth convolutional network, are obtained
RGBDifference characteristics of image.
Optionally, the step S40 includes: S410, special according to the RGB image feature and RGBDifference image
Sign carries out corresponding video frame fusion to input video；S420, the corresponding fused feature of video frame are passed through at a full articulamentum
After reason, fusion feature is obtained.
According to second aspect, the embodiment of the invention discloses a kind of the different of complicated underground piping based on double-current neural network
Normal detection device, comprising: preprocessing module, for input video carry out data prediction, respectively obtain RGB image and
RGBDifference image, with growth data and fixed video duration；Noise reduction process module, for the RGB image and
RGBDifference image carries out noise reduction process；Characteristics of image module, for according to after noise reduction process the RGB image and
RGBDifference image obtains RGB image feature and RGBDifference characteristics of image respectively；Image co-registration module is used
According to the RGB image feature and RGBDifference characteristics of image, image co-registration feature is obtained, under being promoted intricately
The image-capable of the method for detecting abnormality of pipeline；Single task detection module, it is single for inputting described image fusion feature
Task Network, the single task network include a full articulamentum, obtain video frame level score, with distinguish the normal of video or
Exception information；Multitask detection module, for described image fusion feature to be inputted multitask network, the multitask network packet
Containing a full articulamentum, abnormal class score is obtained, to distinguish the abnormal class of video.
According to the third aspect, the embodiment of the invention discloses a kind of computer installation, including processor, processor is for holding
The computer program stored in line storage realize any one of above-mentioned first aspect based on double-current neural network intricately under
The method for detecting abnormality of pipeline.
According to fourth aspect, the embodiment of the invention discloses a kind of computer readable storage mediums, are stored thereon with calculating
Machine program, processor be used for execute the computer program stored in storage medium realize any one of above-mentioned first aspect based on double
Flow the method for detecting abnormality of the complicated underground piping of neural network.
The invention has the following advantages: technical solution disclosed by the embodiments of the present invention is pre- by carrying out to input video
Processing obtains RGB image and RGBDifference image respectively, and drops to RGB image and RGBDifference image
It makes an uproar processing, the data after noise reduction obtain fusion feature by biserial neural network, pass through single task network and detect the normal of video
Or exception information, the abnormal class of video is detected by multitask network；Anomalous video recognition accuracy is high, is greatly saved people
Power, and then the detection accuracy to complicated urban Underground pipeline method for detecting abnormality is improved, by large scale training data, increase
The detection performance of detection model, network structure, trains better detection model after facilitating amplification training data end to end.
Detailed description of the invention
It, below will be to specific in order to illustrate more clearly of the specific embodiment of the invention or technical solution in the prior art
Embodiment or attached drawing needed to be used in the description of the prior art be briefly described, it should be apparent that, it is described below
Attached drawing is some embodiments of the present invention, for those of ordinary skill in the art, before not making the creative labor
It puts, is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of method for detecting abnormality of the complicated underground piping based on double-current neural network disclosed in the present embodiment
Flow chart；
Fig. 2 is a kind of abnormal detector of the complicated underground piping based on double-current neural network disclosed in the present embodiment
Structural schematic diagram；
Fig. 3 is a kind of method for detecting abnormality of the complicated underground piping based on double-current neural network disclosed in the present embodiment
Jump the flow chart sampled；
Fig. 4 is a kind of method for detecting abnormality of the complicated underground piping based on double-current neural network disclosed in the present embodiment
The flow chart of circulating sampling；
Fig. 5 is a kind of method for detecting abnormality of the complicated underground piping based on double-current neural network disclosed in the present embodiment
Image co-registration figure；
Fig. 6 is a kind of method for detecting abnormality of the complicated underground piping based on double-current neural network disclosed in the present embodiment
Algorithm flow chart.
Appended drawing reference: 701, preprocessing module；702, noise reduction process module；703, characteristics of image module；704, image melts
Mold block；705, single task detection module；706, multitask detection module.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
The embodiment of the invention discloses a kind of method for detecting abnormality of complicated underground piping based on double-current neural network, such as
Shown in Fig. 1 and Fig. 6, comprising the following steps:
S10 carries out data prediction to input video, obtains RGB image and RGBDifference image respectively, to expand
Open up data and fixed video duration；In the present embodiment, as shown in Figure 3 and Figure 4, average video frame is calculated on entire data set
It is long, remember a length of Avg of average frameframes, video totalframes is Numframes, sampling interval Ngap.To NumframesGreater than Avgframes
Video carry out Skip Sample, sequentially form new video.If NumframesLess than Avgframes, then Recurrent is carried out
Sample obtains new video.
S20 carries out noise reduction process to RGB image and RGBDifference image；
S30, according to the RGB image and RGBDifference image after noise reduction process, obtain respectively RGB image feature and
RGBDifference characteristics of image；
S40 obtains image co-registration feature according to RGB image feature and RGBDifference characteristics of image, multiple to be promoted
The image-capable of the method for detecting abnormality of miscellaneous underground piping；
Image co-registration feature is inputted single task network by S50, and single task network includes a full articulamentum, obtains video
Frame grade scores, to distinguish the normal or abnormal information of video；In the specific implementation process, single task network is to image procossing
The other prediction of frame level；
Image co-registration feature is inputted multitask network by S60, and multitask network includes a full articulamentum, is obtained abnormal
Category score, to distinguish the abnormal class of video.In the specific implementation process, multitask network is that frame level is other to image procossing
Prediction.
It should be noted that obtaining RGB image and RGBDifference respectively by pre-processing to input video
Image, and noise reduction process is carried out to RGB image and RGBDifference image, the data after noise reduction pass through biserial neural network,
Fusion feature is obtained, the normal or abnormal information of video is detected by single task network, video is detected by multitask network
Abnormal class；Anomalous video recognition accuracy original text, is greatly saved manpower, and then improves and examine extremely to complicated urban Underground pipeline
The detection accuracy of survey method increases the detection performance of detection model, end to end network structure by large scale training data,
Better detection model is trained after facilitating amplification training data.
In the specific implementation process, as shown in Figure 3 and Figure 4, step S10 includes:
Input video is equidistantly divided into n segment by S110；
S120, according to n segment, stochastical sampling RGB image in each segment；In the present embodiment, step S120 packet
It includes: according to n segment, stochastical sampling single frames RGB image or multiframe RGB image in each segment.
S130 generates RGBDifference image by sample frame and its neighbouring image difference, to expand according to RGB image
Open up data and fixed video duration.
In the specific implementation process, step S30 includes:
S310, RGB image are handled using depth convolutional network, obtain RGB image feature；
S320, RGBDifference image are handled using depth convolutional network, obtain RGBDifference image
Feature.
In the specific implementation process, as shown in figure 5, step S40 includes:
S410 carries out corresponding video frame to input video according to RGB image feature and RGBDifference characteristics of image
Fusion；
S420, the corresponding fused feature of video frame obtain fusion feature after a full articulamentum processing.
The embodiment of the invention discloses a kind of abnormal detectors of complicated underground piping based on double-current neural network, such as
Shown in Fig. 2, comprising: preprocessing module 701, for input video carry out data prediction, respectively obtain RGB image and
RGBDifference image, with growth data and fixed video duration；Noise reduction process module 702, for RGB image and
RGBDifference image carries out noise reduction process；Characteristics of image module 703, for according to after noise reduction process RGB image and
RGBDifference image obtains RGB image feature and RGBDifference characteristics of image respectively；Image co-registration module 704,
For image co-registration feature being obtained, to promote complicated buried pipe according to RGB image feature and RGBDifference characteristics of image
The image-capable of the method for detecting abnormality in road；Single task detection module 705, for image co-registration feature to be inputted single task
Network, single task network include a full articulamentum, video frame level score are obtained, to distinguish the normal or abnormal letter of video
Breath；Multitask detection module 706, for image co-registration feature to be inputted multitask network, multitask network includes one and connects entirely
Layer is connect, abnormal class score is obtained, to distinguish the abnormal class of video.
In addition, also providing a kind of computer installation in the embodiment of the present invention, processor passes through computer instructions, thus
Realize following methods:
Data prediction is carried out to input video, obtains RGB image and RGBDifference image, respectively with spreading number
According to and fixed video duration；Noise reduction process is carried out to RGB image and RGBDifference image；According to the RGB after noise reduction process
Image and RGBDifference image obtain RGB image feature and RGBDifference characteristics of image respectively；Schemed according to RGB
As feature and RGBDifference characteristics of image, image co-registration feature is obtained, to promote the abnormality detection side of complicated underground piping
The image-capable of method；Image co-registration feature is inputted into single task network, single task network includes a full articulamentum, is obtained
Video frame level score, to distinguish the normal or abnormal information of video；Image co-registration feature is inputted into multitask network, multitask
Network includes a full articulamentum, abnormal class score is obtained, to distinguish the abnormal class of video.
It is that can lead to it will be understood by those skilled in the art that realizing all or part of the process in above-described embodiment method
Computer program is crossed to instruct relevant hardware and complete, which can be stored in a computer-readable storage medium, should
Program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, the storage medium can for magnetic disk, CD, only
Read storage memory (ROM) or random access memory (RAM) etc..Computer processor is used to execute to store in storage medium
Computer program realizes following methods:
Data prediction is carried out to input video, obtains RGB image and RGBDifference image, respectively with spreading number
According to and fixed video duration；Noise reduction process is carried out to RGB image and RGBDifference image；According to the RGB after noise reduction process
Image and RGBDifference image obtain RGB image feature and RGBDifference characteristics of image respectively；Schemed according to RGB
As feature and RGBDifference characteristics of image, image co-registration feature is obtained, to promote the abnormality detection side of complicated underground piping
The image-capable of method；Image co-registration feature is inputted into single task network, single task network includes a full articulamentum, is obtained
Video frame level score, to distinguish the normal or abnormal information of video；Image co-registration feature is inputted into multitask network, multitask
Network includes a full articulamentum, abnormal class score is obtained, to distinguish the abnormal class of video.
Obviously, the above embodiments are merely examples for clarifying the description, and does not limit the embodiments.It is right
For those of ordinary skill in the art, can also make on the basis of the above description it is other it is various forms of variation or
It changes.These also should be considered as protection scope of the present invention, these all will not influence the reality of effect and patent that the present invention is implemented
The property used.There is no necessity and possibility to exhaust all the enbodiments.This application claims protection scope should be with its right
It is required that content subject to, the records such as specific embodiment in specification can be used for explaining the content of claim.And thus
The obvious changes or variations extended out are still within the protection scope of the invention.
Claims (8)
1. a kind of method for detecting abnormality of the complicated underground piping based on double-current neural network, which is characterized in that including following step
It is rapid:
S10 carries out data prediction to input video, obtains RGB image and RGBDifference image, respectively with spreading number
According to and fixed video duration；
S20 carries out noise reduction process to the RGB image and RGBDifference image；
S30, according to the RGB image and RGBDifference image after noise reduction process, obtain respectively RGB image feature and
RGBDifference characteristics of image；
S40 obtains image co-registration feature according to the RGB image feature and RGBDifference characteristics of image, multiple to be promoted
The image-capable of the method for detecting abnormality of miscellaneous underground piping；
Described image fusion feature is inputted single task network by S50, and the single task network includes a full articulamentum, is obtained
Video frame level score, to distinguish the normal or abnormal information of video；
Described image fusion feature is inputted multitask network by S60, and the multitask network includes a full articulamentum, is obtained
Abnormal class score, to distinguish the abnormal class of video.
2. the method for detecting abnormality of the complicated underground piping according to claim 1 based on double-current neural network, feature
It is, the step S10 includes:
Input video is equidistantly divided into n segment by S110；
S120, according to the n segment, stochastical sampling RGB image in each segment；
S130 generates RGBDifference image by sample frame and its neighbouring image difference, to expand according to the RGB image
Open up data and fixed video duration.
3. the method for detecting abnormality of the complicated underground piping according to claim 2 based on double-current neural network, feature
It is, the step S120 includes:
According to the n segment, stochastical sampling single frames RGB image or multiframe RGB image in each segment.
4. the method for detecting abnormality of the complicated underground piping according to claim 1 based on double-current neural network, feature
It is, the step S30 includes:
S310, the RGB image are handled using depth convolutional network, obtain RGB image feature；
S320, the RGBDifference image are handled using depth convolutional network, obtain RGBDifference image
Feature.
5. the method for detecting abnormality of the complicated underground piping according to claim 1 based on double-current neural network, feature
It is, the step S40 includes:
S410 carries out corresponding video frame to input video according to the RGB image feature and RGBDifference characteristics of image
Fusion；
S420, the corresponding fused feature of video frame obtain fusion feature after a full articulamentum processing.
6. a kind of abnormal detector of the complicated underground piping based on double-current neural network characterized by comprising
Preprocessing module obtains RGB image and RGBDifference figure for carrying out data prediction to input video respectively
Picture, with growth data and fixed video duration；
Noise reduction process module, for carrying out noise reduction process to the RGB image and RGBDifference image；
Characteristics of image module, for being obtained respectively according to the RGB image and RGBDifference image after noise reduction process
RGB image feature and RGBDifference characteristics of image；
Image co-registration module, for obtaining image co-registration according to the RGB image feature and RGBDifference characteristics of image
Feature, to promote the image-capable of the method for detecting abnormality of complicated underground piping；
Single task detection module, for described image fusion feature to be inputted single task network, the single task network includes one
A full articulamentum obtains video frame level score, to distinguish the normal or abnormal information of video；
Multitask detection module, for described image fusion feature to be inputted multitask network, the multitask network includes one
A full articulamentum obtains abnormal class score, to distinguish the abnormal class of video.
7. a kind of computer installation, which is characterized in that including processor, the processor is based on executing and storing in memory
Calculation machine program realizes the abnormality detection of the complicated underground piping as described in any one in claim 1-5 based on double-current neural network
Method.
8. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that processor is deposited for executing
The computer program stored in storage media realizes the complexity based on double-current neural network as described in claim 1-5 any one
The method for detecting abnormality of underground piping.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910695703.3A CN110415236A (en) | 2019-07-30 | 2019-07-30 | A kind of method for detecting abnormality of the complicated underground piping based on double-current neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910695703.3A CN110415236A (en) | 2019-07-30 | 2019-07-30 | A kind of method for detecting abnormality of the complicated underground piping based on double-current neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110415236A true CN110415236A (en) | 2019-11-05 |
Family
ID=68364320
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910695703.3A Pending CN110415236A (en) | 2019-07-30 | 2019-07-30 | A kind of method for detecting abnormality of the complicated underground piping based on double-current neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110415236A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110907749A (en) * | 2019-11-19 | 2020-03-24 | 湖南国奥电力设备有限公司 | Method and device for positioning fault underground cable |
| CN113762007A (en) * | 2020-11-12 | 2021-12-07 | 四川大学 | Abnormal behavior detection method based on appearance and action characteristic double prediction |
| CN116486273A (en) * | 2023-06-20 | 2023-07-25 | 南昌工程学院 | Method for extracting water body information of small sample remote sensing image |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101299233A (en) * | 2008-04-08 | 2008-11-05 | 西安交通大学 | Device and method for realizing moving object identification and track based on FPGA |
| CN106598226A (en) * | 2016-11-16 | 2017-04-26 | 天津大学 | UAV (Unmanned Aerial Vehicle) man-machine interaction method based on binocular vision and deep learning |
| CN108038850A (en) * | 2017-12-08 | 2018-05-15 | 天津大学 | A kind of drainage pipeline Exception Type automatic testing method based on deep learning |
| CN108615230A (en) * | 2018-03-16 | 2018-10-02 | 北京邮电大学 | A kind of hub surface method for detecting abnormality and system |
| CN109058771A (en) * | 2018-10-09 | 2018-12-21 | 东北大学 | The pipeline method for detecting abnormality of Markov feature is generated and is spaced based on sample |
| CN109492129A (en) * | 2018-10-26 | 2019-03-19 | 武汉理工大学 | A kind of similar video searching method and system based on double-current neural network |
| CN109919031A (en) * | 2019-01-31 | 2019-06-21 | 厦门大学 | A kind of Human bodys' response method based on deep neural network |
-
2019
- 2019-07-30 CN CN201910695703.3A patent/CN110415236A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101299233A (en) * | 2008-04-08 | 2008-11-05 | 西安交通大学 | Device and method for realizing moving object identification and track based on FPGA |
| CN106598226A (en) * | 2016-11-16 | 2017-04-26 | 天津大学 | UAV (Unmanned Aerial Vehicle) man-machine interaction method based on binocular vision and deep learning |
| CN108038850A (en) * | 2017-12-08 | 2018-05-15 | 天津大学 | A kind of drainage pipeline Exception Type automatic testing method based on deep learning |
| CN108615230A (en) * | 2018-03-16 | 2018-10-02 | 北京邮电大学 | A kind of hub surface method for detecting abnormality and system |
| CN109058771A (en) * | 2018-10-09 | 2018-12-21 | 东北大学 | The pipeline method for detecting abnormality of Markov feature is generated and is spaced based on sample |
| CN109492129A (en) * | 2018-10-26 | 2019-03-19 | 武汉理工大学 | A kind of similar video searching method and system based on double-current neural network |
| CN109919031A (en) * | 2019-01-31 | 2019-06-21 | 厦门大学 | A kind of Human bodys' response method based on deep neural network |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110907749A (en) * | 2019-11-19 | 2020-03-24 | 湖南国奥电力设备有限公司 | Method and device for positioning fault underground cable |
| CN113762007A (en) * | 2020-11-12 | 2021-12-07 | 四川大学 | Abnormal behavior detection method based on appearance and action characteristic double prediction |
| CN113762007B (en) * | 2020-11-12 | 2023-08-01 | 四川大学 | Abnormal behavior detection method based on appearance and action feature double prediction |
| CN116486273A (en) * | 2023-06-20 | 2023-07-25 | 南昌工程学院 | Method for extracting water body information of small sample remote sensing image |
| CN116486273B (en) * | 2023-06-20 | 2023-09-01 | 南昌工程学院 | Method for extracting water body information of small sample remote sensing image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105868758B (en) | method and device for detecting text area in image and electronic equipment | |
| CN110415236A (en) | A kind of method for detecting abnormality of the complicated underground piping based on double-current neural network | |
| CN110059694A (en) | The intelligent identification Method of lteral data under power industry complex scene | |
| CN108921089A (en) | Method for detecting lane lines, device and system and storage medium | |
| CN107909033A (en) | Suspect's fast track method based on monitor video | |
| CN104244073B (en) | Automatic detecting and recognizing method of scroll captions in videos | |
| CN101833664A (en) | Video image character detecting method based on sparse expression | |
| CN108416774A (en) | A kind of fabric types recognition methods based on fine granularity neural network | |
| CN110189333B (en) | Semi-automatic marking method and device for semantic segmentation of picture | |
| CN109919060A (en) | A kind of identity card content identifying system and method based on characteristic matching | |
| CN107948586B (en) | Trans-regional moving target detecting method and device based on video-splicing | |
| CN113688665A (en) | Remote sensing image target detection method and system based on semi-supervised iterative learning | |
| CN111460927A (en) | Method for extracting structured information of house property certificate image | |
| CN108664970A (en) | A kind of fast target detection method, electronic equipment, storage medium and system | |
| CN112287884B (en) | Examination abnormal behavior detection method and device and computer readable storage medium | |
| CN110599453A (en) | Panel defect detection method and device based on image fusion and equipment terminal | |
| CN108509950A (en) | Railway contact line pillar number plate based on probability characteristics Weighted Fusion detects method of identification | |
| CN112465854A (en) | Unmanned aerial vehicle tracking method based on anchor-free detection algorithm | |
| CN115861738A (en) | Category semantic information guided remote sensing target detection active sampling method | |
| CN113989683A (en) | Ship detection method for synthesizing synchronous orbit sequence optical image space-time information | |
| CN111159167B (en) | Labeling quality detection device and method | |
| CN107451120B (en) | Content conflict detection method and system for open text information | |
| CN109766892A (en) | A kind of ray detection image tagged information character dividing method based on edge detection | |
| CN116486296A (en) | Target detection method, device and computer readable storage medium | |
| CN114694130A (en) | Method and device for detecting telegraph poles and pole numbers along railway based on deep learning |
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: 20191105 |