CN112508944A - Leakage detection method applied to underground water supply pipeline of coal mine - Google Patents
Leakage detection method applied to underground water supply pipeline of coal mine Download PDFInfo
- Publication number
- CN112508944A CN112508944A CN202011571561.9A CN202011571561A CN112508944A CN 112508944 A CN112508944 A CN 112508944A CN 202011571561 A CN202011571561 A CN 202011571561A CN 112508944 A CN112508944 A CN 112508944A
- Authority
- CN
- China
- Prior art keywords
- leakage
- value
- image
- candidate area
- water supply
- 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
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 17
- 239000003245 coal Substances 0.000 title claims abstract description 15
- 238000001514 detection method Methods 0.000 title claims abstract description 13
- 238000000034 method Methods 0.000 claims abstract description 21
- 238000012544 monitoring process Methods 0.000 claims abstract description 19
- 238000003062 neural network model Methods 0.000 claims abstract description 7
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 238000012545 processing Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000001914 filtration Methods 0.000 claims description 7
- 230000000877 morphologic effect Effects 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 7
- 238000002372 labelling Methods 0.000 claims description 5
- 230000003044 adaptive effect Effects 0.000 claims description 4
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000013145 classification model Methods 0.000 claims description 3
- 230000001186 cumulative effect Effects 0.000 claims description 2
- 239000000463 material Substances 0.000 claims description 2
- 230000000007 visual effect Effects 0.000 claims 1
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
Images
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
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M3/00—Investigating fluid-tightness of structures
- G01M3/38—Investigating fluid-tightness of structures by using light
-
- 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
- G06T2207/10004—Still image; Photographic image
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a leakage detection method applied to a coal mine underground water supply pipeline, which utilizes a machine vision technology, firstly acquires a video image of a leakage monitoring point through a camera, then carries out preprocessing, obtains a suspected leakage candidate area map based on a frame difference method, then takes the leakage candidate area map as input, judges whether the leakage candidate area map has a leakage phenomenon or not by utilizing a neural network model, finally secondarily confirms whether the leakage situation of the monitoring point really occurs or not in a time period recorded by the video by integrating leakage frequency, area and position information, and outputs a final judgment result.
Description
Technical Field
The invention belongs to the field of coal mine safety, and particularly relates to a leakage detection method applied to an underground water supply pipeline of a coal mine.
Background
The water supply pipeline system is an important component in a mine production system, and because the underground production condition of the coal mine is complex, the leakage problem of the water supply pipeline can be caused by environmental deformation, pipeline aging and joint loosening, the normal operation of underground equipment of the coal mine is influenced, and even safety accidents can be caused to harm the life safety of underground workers. Therefore, monitoring the working state of the water supply pipeline system is important for safe production of the coal mine. At present, the monitoring of the working state of a water supply pipeline mainly adopts a manual inspection mode, the inspection efficiency is low, potential safety hazards exist, partial pipelines are not easy to approach, and the working condition of the pipelines can not be grasped in time.
Disclosure of Invention
In order to solve the technical problems, the invention provides a leakage detection method applied to an underground water supply pipeline of a coal mine.
The technical scheme adopted by the invention is as follows: a method for detecting leakage of underground water supply pipeline in coal mine includes collecting video image of leakage monitoring point by camera, preprocessing, obtaining suspected leakage candidate region map based on frame difference method, judging whether leakage phenomenon exists in leakage candidate region map by using neural network model with leakage candidate region map as input, confirming whether leakage condition actually appears in monitoring point in time period of video recording by integrating leakage frequency, area and position information for two times and outputting final judgment result.
The preprocessing mainly comprises operations such as a frame difference method, filtering and morphological processing based on an adaptive threshold, and the suspected leakage candidate area image is obtained through the preprocessing, so that the influence of illumination change, noise and a complex background on a subsequent detection result is reduced.
In order to reduce the influence of background interference on the classification result, a convolutional neural network is adopted, a leakage candidate area graph is used as input, and a neural network model is used for predicting a preliminary judgment result.
Compared with the existing manual detection method, the method has the following beneficial effects:
1) compared with manual detection, the invention can continuously detect for 24 hours, and can feed back the running condition of the water supply pipeline in time.
2) The invention can realize remote automatic monitoring and effectively ensure the personal safety of workers.
3) The invention has simple equipment and is economical and practical.
4) The method is free from the interference of complex background, and has strong robustness and high portability.
Drawings
FIG. 1 is a flow chart of the overall algorithm of the present invention.
FIG. 2 is a flow chart of the adaptive threshold calculation according to the present invention.
FIG. 3 is a flowchart illustrating candidate region map extraction according to the present invention.
Fig. 4 is a diagram of the original video with leakage according to the present invention.
FIG. 5 is a diagram of candidate areas for suspected leaks in accordance with the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings;
as shown in fig. 1 to 5, a leakage detection method applied to a water supply pipeline under a coal mine well first collects historical videos of leakage monitoring points, including a leakage video and a non-leakage video; preprocessing each video segment to obtain a suspected leakage candidate area map set; then, classifying and labeling the suspected leakage candidate region graph set according to the actual situation, and dividing the graph set into a training set and a testing set according to the proportion; and taking the training set as input, and training the neural network to obtain a classification model. In the leakage detection stage, a video camera is used for collecting video images of leakage monitoring points; obtaining a candidate region map of suspected leakage through a frame difference method, high-pass filtering and morphological processing operation based on an adaptive threshold; taking the candidate area map of suspected leakage as input, and predicting and judging whether leakage occurs in the way by utilizing a neural network model; and finally, confirming whether the monitoring point really generates the leakage phenomenon in the video recording time period by integrating the leakage frequency, area and position for the second time, and outputting the final judgment result. The specific process comprises the following steps:
the method comprises the following steps: extracting a current frame image fc and a previous frame image fp of an original video, and acquiring a difference image d of the two images by using a frame difference method, wherein the calculation formula is as follows:
wherein fc (x, y) is the gray value of the pixel (x, y) of the image fc, fp (x, y) is the gray value of the pixel (x, y) of the image fp, and d (x, y) is the gray value of the pixel (x, y) of the difference map d;
step two: selecting a reference region ror from the difference map d, and calculating the filtering threshold T of the difference map d according to the gray values of all pixel points rorsubAnd requiring that the selected area is not all white and does not have a leakage phenomenon, wherein the specific calculation method of the threshold value is as follows:
step 2.1: acquiring gray level histograms of all pixel points in the reference area ror, acquiring a gray level value highest _ value _ idx of the highest peak value and a maximum gray level value max _ value _ idx with a non-zero pixel number;
step 2.2: taking highest _ value _ idx as the initial value of ror binarization threshold t;
step 2.3: taking t as a threshold value to carry out binarization on ror, then eliminating noise interference through morphological processing, and counting the number whites _ num of pixel points with the gray value of 255;
step 2.4: if white _ num/(ror. cols. ror. rows)>rario _ thresh, t value increased (max _ value _ idx-highest _ value _ idx) ratio, repeat step 2.3 until white _ num/(r<= ratio _ thresh, and assigns the current T to the filtering threshold Tsub。
Step three: by TsubObtaining a high pass filtered image dhp for the threshold value, the calculation formula is as follows;
step four: repeating the first step to the third step to obtain m dhp images each time, and accumulating the m images to obtain a cumulative image dsum, wherein the calculation formula is as follows:
step five: performing morphological processing on the accumulated graph dsum to eliminate noise interference to obtain a suspected leakage candidate area graph dst;
step six: the suspected leakage candidate area graph dst is used as input, whether leakage phenomenon exists at a monitoring point in a time period recorded by the current accumulated graph dsum is judged, and the method specifically comprises the following steps:
step 6.1: collecting historical videos of the leakage monitoring points, wherein the historical videos comprise videos with leakage phenomena and videos without leakage phenomena, the historical videos are used as original materials, and the steps one to five are repeated to obtain a suspected leakage candidate area atlas;
step 6.2: classifying and labeling the image set obtained in the step 6.1, labeling the image set as leakage and non-leakage according to actual conditions, and dividing the image set into a training set and a test set according to proportion;
step seven: and taking the images of the training set as input to train the neural network model, starting to train and test when the loss function of the model tends to be convergent, and taking the weight parameter with the highest accuracy on the test set as a final classification model.
Step eight: and (4) acquiring a video image of the leakage monitoring point by using a camera, repeating the steps from the first step to the fifth step, and processing a section of video to obtain k suspected leakage candidate area maps dst.
Step nine: and taking the k suspected leaked candidate area graphs dst as input, predicting by using the network model obtained in the step seven, and preliminarily judging whether each candidate area graph has a leakage phenomenon.
Step ten: and integrating the position, frequency and area information of the k candidate region maps dst, secondarily confirming whether the monitoring points are actually leaked in the time period of the video recording, and outputting the final judgment result.
Claims (2)
1. A leakage detection method applied to a coal mine underground water supply pipeline is characterized by comprising the following steps: the method is based on a machine vision technology, a camera is used for collecting visual images of leakage monitoring points of the underground coal mine water supply pipeline, and leakage detection is carried out on the underground coal mine water supply pipeline by using a method combining image processing and neural network detection; firstly, a suspected leakage candidate area map is obtained through preprocessing such as a frame difference method, filtering, morphological processing and the like based on an adaptive threshold, then a neural network classification method is adopted, a neural network model is used for predicting and judging whether the candidate area map has a leakage phenomenon, finally, information of frequency, position and leakage area of leakage in a video within a period of time is synthesized, whether a water supply pipeline really leaks within the video recording time period is secondarily confirmed, and a final judgment result is output.
2. The method for detecting the leakage of the underground water supply pipeline of the coal mine according to claim 1, wherein the method comprises the following steps: the method specifically comprises the following steps:
the method comprises the following steps: extracting a current frame image f of an original videocAnd the previous frame image fpAnd acquiring a difference image d of the two images by using a frame difference method, wherein the calculation formula is as follows:
wherein fc (x, y) is the gray value of the pixel (x, y) of the image fc, fp (x, y) is the gray value of the pixel (x, y) of the image fp, and d (x, y) is the gray value of the pixel (x, y) of the difference map d;
step two: selecting a reference region ror in the difference map d based onror calculating the filtering threshold T of the difference image d by the gray values of all the pixel pointssubAnd requiring that the selected area is not all white and does not have a leakage phenomenon, wherein the specific calculation method of the threshold value is as follows:
step 2.1: acquiring gray level histograms of all pixel points in the reference area ror, acquiring a gray level value highest _ value _ idx of the highest peak value and a maximum gray level value max _ value _ idx with a non-zero pixel number;
step 2.2: taking highest _ value _ idx as the initial value of ror binarization threshold t;
step 2.3: taking t as a threshold value to carry out binarization on ror, then eliminating noise interference through morphological processing, and counting the number whites _ num of pixel points with the gray value of 255;
step 2.4: if white _ num/(ror. cols. ror. rows)>rario _ thresh, t value increased (max _ value _ idx-highest _ value _ idx) ratio, repeat step 2.3 until white _ num/(r<= ratio _ thresh, and assigns the current T to the filtering threshold Tsub。
Step three: by TsubObtaining a high pass filtered image dhp for the threshold value, the calculation formula is as follows;
step four: repeating the first step to the third step to obtain m dhp images each time, and accumulating the m images to obtain a cumulative image dsum, wherein the calculation formula is as follows:
step five: performing morphological processing on the accumulated graph dsum to eliminate noise interference to obtain a suspected leakage candidate area graph dst;
step six: the suspected leakage candidate area graph dst is used as input, whether leakage phenomenon exists at a monitoring point in a time period recorded by the current accumulated graph dsum is judged, and the method specifically comprises the following steps:
step 6.1: collecting historical videos of the leakage monitoring points, wherein the historical videos comprise videos with leakage phenomena and videos without leakage phenomena, the historical videos are used as original materials, and the steps one to five are repeated to obtain a suspected leakage candidate area atlas;
step 6.2: classifying and labeling the image set obtained in the step 6.1, labeling the image set as leakage and non-leakage according to actual conditions, and dividing the image set into a training set and a test set according to proportion;
step seven: and taking the images of the training set as input to train the neural network model, starting to train and test when the loss function of the model tends to be convergent, and taking the weight parameter with the highest accuracy on the test set as a final classification model.
Step eight: and (4) acquiring a video image of the leakage monitoring point by using a camera, repeating the steps from the first step to the fifth step, and processing a section of video to obtain k suspected leakage candidate area maps dst.
Step nine: and taking the k suspected leaked candidate area graphs dst as input, predicting by using the network model obtained in the step seven, and preliminarily judging whether each candidate area graph has a leakage phenomenon.
Step ten: and integrating the position, frequency and area information of the k candidate region maps dst, secondarily confirming whether the monitoring points are actually leaked in the time period of the video recording, and outputting the final judgment result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011571561.9A CN112508944A (en) | 2020-12-27 | 2020-12-27 | Leakage detection method applied to underground water supply pipeline of coal mine |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011571561.9A CN112508944A (en) | 2020-12-27 | 2020-12-27 | Leakage detection method applied to underground water supply pipeline of coal mine |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112508944A true CN112508944A (en) | 2021-03-16 |
Family
ID=74923585
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011571561.9A Pending CN112508944A (en) | 2020-12-27 | 2020-12-27 | Leakage detection method applied to underground water supply pipeline of coal mine |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112508944A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113670524A (en) * | 2021-07-13 | 2021-11-19 | 江铃汽车股份有限公司 | Detection method and detection system for fuel leakage in automobile collision |
CN113781513A (en) * | 2021-08-19 | 2021-12-10 | 广东能源集团科学技术研究院有限公司 | Method and system for detecting leakage of water supply pipeline of power plant |
CN114001867A (en) * | 2021-10-19 | 2022-02-01 | 北京伟瑞迪科技有限公司 | Monitoring and coping method for gas leakage of park |
CN117237362A (en) * | 2023-11-16 | 2023-12-15 | 山东嘉源复合材料有限公司 | Vision-based propylene glycol diacetate canning sealing detection method |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110335294A (en) * | 2019-07-11 | 2019-10-15 | 中国矿业大学 | Mine water pump house leakage detection method based on frame difference method Yu 3D convolutional neural networks |
US20190331301A1 (en) * | 2016-12-30 | 2019-10-31 | Du Yuchuan | Method for leakage detection of underground pipeline corridor based on dynamic infrared thermal image processing |
-
2020
- 2020-12-27 CN CN202011571561.9A patent/CN112508944A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190331301A1 (en) * | 2016-12-30 | 2019-10-31 | Du Yuchuan | Method for leakage detection of underground pipeline corridor based on dynamic infrared thermal image processing |
CN110335294A (en) * | 2019-07-11 | 2019-10-15 | 中国矿业大学 | Mine water pump house leakage detection method based on frame difference method Yu 3D convolutional neural networks |
Non-Patent Citations (1)
Title |
---|
青春: "基于支持向量机的供水管道泄漏检测算法研究", 《 基于支持向量机的供水管道泄漏检测算法研究》,硕士电子期刊出版, 15 January 2019 (2019-01-15) * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113670524A (en) * | 2021-07-13 | 2021-11-19 | 江铃汽车股份有限公司 | Detection method and detection system for fuel leakage in automobile collision |
CN113670524B (en) * | 2021-07-13 | 2023-12-19 | 江铃汽车股份有限公司 | Detection method and detection system for automobile collision fuel leakage |
CN113781513A (en) * | 2021-08-19 | 2021-12-10 | 广东能源集团科学技术研究院有限公司 | Method and system for detecting leakage of water supply pipeline of power plant |
CN113781513B (en) * | 2021-08-19 | 2023-11-28 | 广东能源集团科学技术研究院有限公司 | Leakage detection method and system for water supply pipeline of power plant |
CN114001867A (en) * | 2021-10-19 | 2022-02-01 | 北京伟瑞迪科技有限公司 | Monitoring and coping method for gas leakage of park |
CN117237362A (en) * | 2023-11-16 | 2023-12-15 | 山东嘉源复合材料有限公司 | Vision-based propylene glycol diacetate canning sealing detection method |
CN117237362B (en) * | 2023-11-16 | 2024-01-26 | 山东嘉源复合材料有限公司 | Vision-based propylene glycol diacetate canning sealing detection method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112508944A (en) | Leakage detection method applied to underground water supply pipeline of coal mine | |
CN114937055B (en) | Image self-adaptive segmentation method and system based on artificial intelligence | |
CN109870461B (en) | Electronic components quality detection system | |
CN109359594B (en) | Mine belt abnormity detection method and device based on video processing | |
CN103279765B (en) | Steel wire rope surface damage detection method based on images match | |
Dubey et al. | Maximally stable extremal region marking-based railway track surface defect sensing | |
CN103442209A (en) | Video monitoring method of electric transmission line | |
WO2023082418A1 (en) | Power utility tunnel settlement crack identification method based on artificial intelligence technology | |
CN101738394A (en) | Method and system for detecting indoor smog | |
CN111814678A (en) | Video monitoring-based method and system for identifying coal flow in conveyor belt | |
CN106303158A (en) | A kind of striped method for detecting abnormality in video image | |
CN108009574B (en) | Track fastener detection method | |
CN111860143B (en) | Real-time flame detection method for inspection robot | |
CN111212291A (en) | DFL-CNN network-based video intra-frame object removal tamper detection method | |
CN111127448A (en) | Model and method for detecting air spring fault based on isolated forest | |
Fu et al. | Research on image-based detection and recognition technologies for cracks on rail surface | |
CN115035182A (en) | Landslide disaster early warning method and system in mountainous area | |
CN109767426B (en) | Shield tunnel water leakage detection method based on image feature recognition | |
Hashmi et al. | Computer-vision based visual inspection and crack detection of railroad tracks | |
Saranya et al. | Comparison of segmentation techniques for detection of defects in non-destructive testing images | |
CN112419336B (en) | Motor car gearbox oil leakage fault identification method and system based on wavelet transformation | |
CN103391441A (en) | Surveillance video object deleting, tampering and detecting algorithm based on capacity variance | |
CN105913008A (en) | Crowd exceptional event detection method based on hypothesis examination | |
CN109859141B (en) | Deep vertical shaft well wall image denoising method | |
CN111598042A (en) | Visual statistical method for underground drill rod counting |
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 |