A natural gas pipeline leakage monitoring system based on infrared thermal imager

Abstract

The present invention relates to the field of safety monitoring technology, and specifically to a natural gas pipeline leakage monitoring system based on an infrared thermal imager. It includes: a regional thermal imaging image acquisition module, a gas leakage area acquisition module, a danger level assessment module, and a leakage point positioning module. The regional thermal imaging image acquisition module is used to acquire a regional thermal imaging image based on a thermal imaging image and a regional image; the gas leakage area acquisition module is used to acquire a gas annotation image based on a segmentation neural network according to the regional thermal imaging image, and then acquire a gas leakage area; the danger level assessment module is used to record the dynamic change process of the gas leakage area and acquire the diffusion rate; acquire the temperature change during the gas leakage process, and evaluate the danger level of gas diffusion according to the diffusion rate and temperature change; the leakage point positioning module is used to determine the position of the leakage point according to the pixel value of each pixel point in the superimposed image. The present invention can monitor the natural gas pipeline in real time and accurately acquire the leakage location.

Description

Technical Field

[0001] The invention relates to the technical field of safety monitoring, and in particular to a natural gas pipeline leakage monitoring system based on an infrared thermal imager. Background Art

[0002] Natural gas is an indispensable energy source in the lives of urban residents. Currently, natural gas companies basically provide gas supply services through pipeline transportation. Natural gas pipelines can be said to be one of the urban infrastructures that have a great impact on the daily lives of urban residents. Urban natural gas pipeline projects are all buried underground. Under the influence of various reasons such as corrosion, natural gas pipelines may be damaged and leak after long-term use, causing accidents and huge economic and energy losses. Therefore, for the maintenance of natural gas pipelines, the most critical issue is to promptly detect the leakage of natural gas pipelines and accurately locate the leakage points of natural gas pipelines.

[0003] Currently, natural gas pipeline leaks are mostly detected through chemical gas sensors. However, as time goes by, the sensitivity of chemical gas sensors will decrease, and they will not be able to monitor natural gas leaks in real time and accurately. In the long run, this will lead to great safety hazards. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a natural gas pipeline leakage monitoring system based on an infrared thermal imager. The technical solution adopted is as follows:

[0005] An embodiment of the present invention provides a natural gas pipeline leakage monitoring system based on an infrared thermal imager, the system comprising the following modules:

[0006] A regional thermal imaging image acquisition module is used to collect thermal imaging images of the natural gas pipeline, obtain a regional image of the pipeline area by performing threshold segmentation on the thermal imaging image, and obtain a regional thermal imaging image based on the thermal imaging image and the regional image;

[0007] A gas leakage area acquisition module is used to acquire a gas annotation image based on a segmentation neural network according to the regional thermal imaging image, and then acquire the gas leakage area; the segmentation neural network extracts the first feature vector of the regional thermal imaging image through an encoder, and then uses a time convolution network to obtain the time series change characteristics of the first feature vector as a second feature vector, and then acquires the gas annotation image through a decoder;

[0008] The hazard level assessment module is used to record the dynamic changes in the gas leakage area, obtain the diffusion rate according to the area change of the gas leakage area and the leakage time; obtain the temperature change during the gas leakage process, and assess the hazard level of gas diffusion according to the diffusion rate and temperature change;

[0009] The leakage point positioning module is used to fit the leakage direction of the gas leakage area at each moment, obtain the fitted straight line image, superimpose all the fitted straight line images to obtain the superimposed image, and determine the position of the leakage point according to the pixel value of each pixel point in the superimposed image.

[0010] Preferably, the regional thermal imaging image acquisition module includes:

[0011] The region division unit is used to extract the pipeline connected domain as the pipeline internal region from the segmented image after the threshold segmentation; and to perform buffer zone analysis on the pipeline connected domain to obtain the pipeline external region.

[0012] Preferably, the regional thermal imaging image acquisition module includes:

[0013] The regional thermal imaging image acquisition unit is used to form the regional image from the internal area of ​​the pipeline and the external area of ​​the pipeline, and multiply the thermal imaging image and the corresponding pixels of the regional image to obtain the regional thermal imaging image.

[0014] Preferably, the gas leakage area acquisition module includes:

[0015] The segmentation neural network training unit is used to use the multi-channel regional thermal imaging image within a preset time period as the input of the neural network, extract features through the encoder to obtain the first feature image and flatten it into a first feature vector, use it as the input of the temporal convolutional network, extract the time-series change characteristics of the first feature vector as the second feature vector, reshape the second feature vector into a second feature image, and then output the gas annotation image through the decoder.

[0016] Preferably, the risk level assessment module includes:

[0017] A diffusion rate acquisition unit is used to acquire a first time when a gas leakage area first appears, a second time when it intersects with the internal area of ​​the pipeline, and a third time when it intersects with the external area of ​​the pipeline; to acquire an internal diffusion rate according to an internal diffusion area of ​​the gas leakage area between the first time and the second time and an internal diffusion time; to acquire an external diffusion rate according to an external diffusion area of ​​the gas leakage area between the second time and the third time and an external diffusion time; and to acquire a diffusion rate of the leaked gas based on the internal diffusion rate and the external diffusion rate.

[0018] Preferably, the risk level assessment module includes:

[0019] The temperature change acquisition unit is used to acquire the average temperature of the internal diffusion area as the internal area temperature, acquire the average temperature of the external diffusion area as the external area temperature, and take the temperature difference between the internal area temperature and the external area temperature as the temperature change.

[0020] Preferably, the leakage point locating module includes:

[0021] The leakage direction fitting unit is used to perform straight line fitting according to the gas leakage area at each moment, and take the longest line segment in the gas leakage area as the gas leakage direction at that moment.

[0022] Preferably, the leakage point locating module includes:

[0023] The fitted straight line image acquisition unit is used to assign a first pixel value to the straight line where the gas leakage direction is located at each moment, and assign a second pixel value to other pixel points, to obtain a binary image at each moment, and select the endpoint of the straight line in the internal area of ​​the pipeline as a reference point to linearly attenuate the binary image to obtain a fitted straight line image at each moment.

[0024] Preferably, the leakage point locating module includes:

[0025] The superimposed image acquisition unit is used to add the corresponding pixels of the fitted straight line image at each moment, and the added pixel values ​​decay in time sequence to obtain the superimposed image.

[0026] Preferably, the leakage point locating module includes:

[0027] The leakage point judgment unit is used to assign weights to each pixel value in the superimposed image based on a Gaussian kernel function, obtain a leakage point distribution image according to the pixel points of the superimposed image and the corresponding weights, and take the point with the largest pixel value as the leakage point.

[0028] The embodiments of the present invention have at least the following beneficial effects:

[0029] The embodiment of the present invention obtains the gas leakage area through thermal imaging images, and obtains the danger level of the gas leakage according to the changes in the gas leakage area in time sequence and the temperature changes at the same time; the connected domain image of the gas leakage is linearly fitted to obtain the gas leakage position. The embodiment of the present invention can monitor the leakage status of the natural gas pipeline in real time, and issue different degrees of alarms according to the degree of danger, accurately locate the leakage point, help relevant personnel to assist in determining the location of the natural gas pipeline leakage, and improve the efficiency of emergency repairs. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0031] Figure 1A system block diagram of a natural gas pipeline leakage monitoring system based on an infrared thermal imager is provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of a natural gas pipeline leakage monitoring system based on an infrared thermal imager proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments may be combined in any suitable form.

[0033] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0034] The following is a detailed description of a specific scheme of a natural gas pipeline leakage monitoring system based on an infrared thermal imager provided by the present invention in conjunction with the accompanying drawings.

[0035] See also Figure 1 , which shows a system block diagram of a natural gas pipeline leakage monitoring system based on an infrared thermal imager provided by an embodiment of the present invention, the system includes the following modules:

[0036] A regional thermal imaging image acquisition module 100 , a gas leakage region acquisition module 200 , a danger level assessment module 300 and a leakage point positioning module 400 .

[0037] The regional thermal imaging image acquisition module 100 is used to collect thermal imaging images of the natural gas pipeline, obtain regional images of the pipeline area by performing threshold segmentation on the thermal imaging image, and obtain regional thermal imaging images based on the thermal imaging image and the regional image.

[0038] Specifically, the regional thermal imaging image acquisition module 100 includes a thermal imaging image acquisition unit 110 , a region division unit 120 and a regional thermal imaging image acquisition unit 130 .

[0039] The thermal imaging image acquisition unit 110 is used to acquire thermal imaging images of the natural gas pipeline.

[0040] Thermal imaging technology refers to the use of infrared detectors and optical imaging lenses to receive the infrared radiation energy distribution pattern of the target to be measured and reflect it on the photosensitive element of the infrared detector, thereby obtaining an infrared thermal image, which corresponds to the heat distribution field on the surface of the object. In layman's terms, an infrared thermal imager converts the invisible infrared energy emitted by an object into a visible thermal image. The different colors on the thermal image represent the different temperatures of the object being measured.

[0041] Natural gas is mainly composed of methane and ethane, and medium-wave infrared thermal imagers can clearly detect these two gases. They can then detect information such as corrosion, rupture, thinning, blockage, and leakage of natural gas pipelines, and quickly and accurately obtain the two-dimensional temperature distribution on the surface of equipment and materials. For gas leaks in natural gas pipelines, they can quickly and accurately find the leak point in the early stage, which is very effective in preventing accidents and reducing energy consumption.

[0042] Deploy an infrared thermal imager and aim the camera lens at the natural gas pipeline that needs to be monitored to obtain temperature information from the on-site scene and ultimately obtain a real-time thermal imaging image of the natural gas pipeline.

[0043] The region division unit 120 is used to extract the pipeline connected domain as the pipeline internal region from the segmented image after the threshold segmentation; and perform buffer zone analysis on the pipeline connected domain to obtain the pipeline external region.

[0044] Due to the material and properties of the natural gas pipeline, its thermal distribution field is different from that of the background. Therefore, the maximum between-class variance method (OTSU) is used to segment the thermal imaging image to obtain a natural gas pipeline segmentation image. In this image, the pixel value of the natural gas pipeline is 1, and the pixel value of the non-natural gas pipeline is 0.

[0045] Then, the connected domain of the natural gas pipeline segmentation image is extracted to obtain the natural gas pipeline segmentation connected domain image. The connected domain extraction is used to reduce the cluttered image noise. The natural gas pipeline connected domain is used as the internal area of ​​the natural gas pipeline to analyze the area where the leakage occurs.

[0046] By performing a buffer analysis on the connected domain of the natural gas pipeline, the buffer analysis in the embodiment of the present invention is a buffer based on the polygonal boundary of the face element, which is expanded outward by a certain distance d to generate a new polygon to form a buffer as the external area of ​​the pipeline. The external area of ​​the pipeline is the area through which the gas passes when the natural gas leaks.

[0047] As an example, in the embodiment of the present invention, the certain distance d is set to 30.

[0048] The regional thermal imaging image acquisition unit 130 is used to form a regional image from the internal area of ​​the pipeline and the external area of ​​the pipeline, and multiply the thermal imaging image and the corresponding pixels of the regional image to obtain the regional thermal imaging image.

[0049] The outer area of ​​the natural gas pipeline and the inner area of ​​the natural gas pipeline together constitute the regional image of the natural gas pipeline. Then, the regional image of the natural gas pipeline is multiplied by corresponding pixels with the thermal imaging image to obtain the regional thermal imaging image of the natural gas pipeline.

[0050] The gas leakage area acquisition module 200 is used to obtain a gas annotation image based on a segmentation neural network according to the regional thermal imaging image, and then obtain the gas leakage area; the segmentation neural network extracts the first eigenvector of the regional thermal imaging image through an encoder, and then uses a time convolution network to obtain the time series change characteristics of the first eigenvector as the second eigenvector, and then obtains the gas annotation image through a decoder.

[0051] The gas leakage area acquisition module 200 includes a segmentation neural network training unit 210 , a gas annotation image acquisition unit 220 and a gas leakage area acquisition unit 230 .

[0052] The segmentation neural network training unit 210 is used to use the multi-channel regional thermal imaging image within a preset time period as the input of the neural network, extract features through the encoder to obtain a first feature image and flatten it into a first feature vector, use it as the input of the temporal convolutional network, extract the time-series change characteristics of the first feature vector as the second feature vector, reshape the second feature vector into a second feature image, and then output the gas annotation image through the decoder.

[0053] A segmentation neural network model is established. This neural network model can use existing semantic segmentation models, such as Enet, Unet, Deeplab, etc. Taking Unet as an example, Unet includes a two-dimensional image encoder and a two-dimensional image decoder.

[0054] The input of the two-dimensional image encoder is a multi-channel regional thermal imaging image within a preset time period. For example, the infrared thermal imager collects an image every 0.2 seconds, and the images collected within 1 second, that is, five thermal imaging images of the natural gas pipeline area, are input. After the concatenation operation, the final input is a five-channel natural gas pipeline area thermal imaging image.

[0055] The function of the two-dimensional image encoder is feature extraction. It extracts the features of the five-channel regional thermal imaging image and outputs it as the first feature map. The first feature map is flattened to become the first feature vector and then input into the temporal encoder. The temporal encoder uses a temporal convolutional network (TCN) to extract temporal features. It can avoid Unet only fitting the temperature in the thermal imaging. After adding TCN, it can learn the temperature change information of the region and extract the change characteristics of the first feature vector in time series as the second feature vector, that is, the output of TCN. The second feature vector is reshaped to become the second feature map and input into the two-dimensional image decoder to output the gas annotation image.

[0056] The input data of the constructed segmentation neural network model is a multi-channel regional thermal imaging image. When training the neural network, the gas area is manually marked based on the thermal imaging image of the natural gas pipeline area, and the gas pixels are marked as pixel value 1, and the non-gas pixels are marked as 0 as labels for network training; the loss function adopts cross entropy, and the network optimization method preferably adopts Adam optimizer. The output of the neural network is a gas labeled image, in which the gas pixels are marked as pixel value 1, and the non-gas pixels are marked as pixel value 0.

[0057] The gas annotation image acquisition unit 220 is used to input the regional thermal imaging image acquired in real time into the trained segmentation neural network and output the corresponding gas annotation image.

[0058] The gas leakage region acquisition unit 230 is used to take the region with a pixel value of 1 in the gas annotation image as the gas leakage region.

[0059] The segmentation neural network uses time-series regional thermal imaging images to effectively allow the network to learn temperature change information, thereby improving the accuracy of leaking gas segmentation.

[0060] The danger level assessment module 300 is used to record the dynamic changes of the gas leakage area, obtain the diffusion rate according to the area change of the gas leakage area and the leakage time; obtain the temperature change during the gas leakage process, and assess the danger level of gas diffusion according to the diffusion rate and temperature change.

[0061] Specifically, the danger level assessment module 300 includes a diffusion rate acquisition unit 310 , a temperature change acquisition unit 320 and a danger level assessment unit 330 .

[0062] The diffusion rate acquisition unit 310 acquires a first time when a gas leakage area first appears, a second time when it intersects with an internal area of ​​the pipeline, and a third time when it intersects with an external area of ​​the pipeline; acquires an internal diffusion rate based on an internal diffusion area of ​​the gas leakage area between the first time and the second time and an internal diffusion time; acquires an external diffusion rate based on an external diffusion area of ​​the gas leakage area between the second time and the third time and an external diffusion time; and acquires a diffusion rate of the leaked gas based on the internal diffusion rate and the external diffusion rate.

[0063] The connected domain of the gas leakage area in the regional thermal imaging image is extracted to eliminate the influence of noise and obtain the leaking gas connected domain. Each leaking gas connected domain is a gas leakage area of ​​the natural gas pipeline. Since gas leakage is a dynamic process, the dynamic change of each leaking gas connected domain is modeled.

[0064] The specific modeling process is: when the gas leakage area appears for the first time, record the current time as the first time Then obtain the second time when the gas leakage area intersects with the internal area of ​​the pipeline , the third time when the gas leakage area intersects with the external area of ​​the pipeline , that is, the leaked gas connection area intersects with the internal boundary line of the natural gas pipeline and the external boundary line of the pipeline. When there is an intersection, the time at this moment is recorded and recorded as and . At the same time, the location of the leakage gas connection area at the current two moments is recorded.

[0065] Calculate the time it takes for leaked gas to spread across regions Time: internal diffusion time , external diffusion time .

[0066] When a gas leak occurs, the area of ​​the leaking gas connection domain will gradually increase. Since the area of ​​the leaking gas connection domain will be affected by wind speed, etc., only the changes of the leaking gas in the natural gas pipeline area are analyzed and the second time is recorded. , Third Time The area of ​​the leaking gas connection domain has the second time The connected domain area under and the third time The connected domain area under the , external diffusion area .

[0067] Get the internal diffusion rate of gas diffusion in the internal region , and the external diffusion rate of gas diffusion in the external region , and then calculate the gas diffusion rate of the leaked gas .

[0068] in, The internal diffusion rate Weight, Represents the weight of the external diffusion rate.

[0069] As an example, in the embodiment of the present invention The value is 0.75. The value is 0.25.

[0070] The temperature change acquisition unit 320 is used to acquire the average temperature of the internal diffusion area as the internal area temperature, acquire the average temperature of the external diffusion area as the external area temperature, and take the temperature difference between the internal area temperature and the external area temperature as the temperature change.

[0071] The pixel value of each pixel in the thermal imaging image represents the temperature value. The average temperature of the internal diffusion area is calculated as the internal area temperature, and the average temperature of the external diffusion area is calculated as the external area temperature. , taking the temperature difference between the internal area temperature and the external area temperature as the temperature change T: .

[0072] Temperature changes It indicates the temperature difference between the internal area and the external area of ​​the leaking gas at a specified time. The larger the value, the faster the gas leakage rate is and the larger the leakage point of the natural gas pipeline may be.

[0073] The danger level assessment unit 330 is used to calculate the danger level U of gas leakage in the natural gas pipeline:

[0074]

[0075] The larger U is, the larger the leaked part is. At the same time, the faster the gas diffuses, and the higher the degree of danger is.

[0076] The leakage point positioning module 400 is used to fit the leakage direction of the gas leakage area at each moment, obtain a fitting straight line image, superimpose all the fitting straight line images to obtain a superimposed image, and determine the position of the leakage point according to the pixel value of each pixel point in the superimposed image.

[0077] A leakage direction fitting unit 410 , a fitting straight line image acquiring unit 420 , a superimposed image acquiring unit 430 , a leakage point judging unit 440 and an early warning unit 450 .

[0078] The leakage direction fitting unit 410 is used to perform straight line fitting according to the gas leakage area at each moment, and take the longest line segment in the gas leakage area as the gas leakage direction at that moment.

[0079] The gas leakage area at each moment between the first time and the third time of the time series is obtained, and a straight line fitting is performed on the leakage gas connection domain at each moment, and the longest line segment in the connection domain is used as the fitted gas leakage direction.

[0080] The fitted straight line image acquisition unit 420 is used to assign a first pixel value to the straight line where the gas leakage direction is located at each moment, and assign a second pixel value to other pixel points, to obtain a binary image at each moment, and select the endpoints of the straight line in the internal area of ​​the pipeline as reference points to linearly attenuate the binary image to obtain a fitted straight line image at each moment.

[0081] The pixel value of the straight line where the gas leakage direction is located at each moment is assigned to 1, and the other pixels are assigned to 0 to obtain a binary image. The end point of the straight line in the inner area of ​​the pipeline is selected as the reference point, and the distance between the other end point and the reference point is obtained, recorded as D. The gray value of the straight line is linearly attenuated using the distance D:

[0082]

[0083] in, Represents the pixel value after attenuation of each pixel point. It represents the pixel value of each pixel in the binary image, that is, the pixel value of the pixel where the straight line is located is 1, and the pixel value of other pixels is 0. x represents the linear attenuation coefficient.

[0084] As an example, in the embodiment of the present invention, the linear attenuation coefficient x is set to 0.9, so that the pixel value of the pixel in the straight line decays exponentially with the distance from the reference point. The greater the distance, the smaller the gray value.

[0085] For the binary image at each moment, the attenuated pixel value of each pixel is obtained to form a fitting straight line image.

[0086] The superimposed image acquisition unit 430 is used to add the corresponding pixels of the fitted straight line image at each moment, and the added pixel values ​​decay in time sequence to obtain a superimposed image.

[0087] Perform frequency superposition on the fitted straight line images at all times between the first time and the third time:

[0088]

[0089] in, is the pixel value of the superimposed pixel, is the gray value of the pixel at the current moment, is the gray value of the pixel at the next moment of the current moment, is the attenuation coefficient.

[0090] As an example, the attenuation coefficient in the embodiment of the present invention is The value is 0.3, which means that the more times the fitting straight line appears at the same pixel position, the larger its value.

[0091] Then, all pixel values ​​of the superimposed fitted straight line image are normalized to obtain a superimposed image. As an example, pixel value normalization adopts range normalization.

[0092] The leakage point judgment unit 440 is used to assign weights to each pixel value in the superimposed image based on the Gaussian kernel function, obtain a leakage point distribution image according to the pixel points of the superimposed image and the corresponding weights, and take the point with the largest pixel value as the leakage point.

[0093] The larger the pixel value in the superimposed image, the more likely it is a leakage point. However, since the fitted straight line can only represent the line segment where the leakage point may exist, it cannot accurately indicate the location of the leakage point. Therefore, a two-dimensional Gaussian distribution area is constructed based on the straight line distribution of the time series after the fitting straight line.

[0094] Based on the Gaussian kernel function, a weight is assigned to each pixel value greater than 0 in a 3*3 window range. The window range is a 3*3 window centered on the pixel point. The Gaussian kernel function is as follows:

[0095]

[0096] in, Indicates the coordinates of the center point of the window. express The weight of the pixel at .

[0097] As an example, the embodiment of the present invention sets The value of is 1.

[0098] Each window forms a Gaussian distribution and a weight is assigned to it.

[0099] The weight of each pixel is used as the pixel value to form a leakage point Gaussian distribution image, and then the leakage point Gaussian distribution image is multiplied by the corresponding pixel of the superimposed image to obtain the leakage point distribution image. The larger the pixel value in the image, the more likely it is the leakage point location, and the location with the largest pixel value is taken as the leakage point.

[0100] The early warning unit 450 is used to issue different degrees of alarms based on the danger level of the leakage point of the natural gas pipeline, and at the same time provide the location of the leakage point to relevant personnel as a data reference to assist emergency repair personnel in determining the location of the gas leakage point.

[0101] In summary, the embodiment of the present invention includes the following modules:

[0102] A regional thermal imaging image acquisition module 100 , a gas leakage region acquisition module 200 , a danger level assessment module 300 and a leakage point positioning module 400 .

[0103] Specifically, the regional thermal imaging image acquisition module is used to collect thermal imaging images of natural gas pipelines, obtain regional images of pipeline areas by threshold segmentation of thermal imaging images, and obtain regional thermal imaging images according to thermal imaging images and regional images; the gas leakage area acquisition module is used to obtain gas annotation images based on segmentation neural networks according to regional thermal imaging images, and then obtain gas leakage areas; the segmentation neural network extracts the first feature vector of the regional thermal imaging image through an encoder, and then uses a time convolution network to obtain the time series change characteristics of the first feature vector as the second feature vector, and then obtains the gas annotation image through a decoder; the danger level assessment module is used to record the dynamic change process of the gas leakage area, and obtain the diffusion rate according to the area change of the gas leakage area and the leakage time; obtain the temperature change during the gas leakage process, and assess the danger level of gas diffusion according to the diffusion rate and temperature change; the leakage point positioning module is used to fit the leakage direction of the gas leakage area at each moment, obtain the fitting straight line image, superimpose all the fitting straight line images to obtain the superimposed image, and judge the position of the leakage point according to the pixel value of each pixel in the superimposed image. The embodiment of the present invention can monitor the leakage state of the natural gas pipeline in real time, and make different degrees of alarms according to the degree of danger, accurately locate the leakage point, help relevant personnel to assist in judging the location of the natural gas pipeline leakage, and improve the repair efficiency.

[0104] It should be noted that the sequence of the embodiments of the present invention described above is for description only and does not represent the advantages and disadvantages of the embodiments. The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0105] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0106] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A natural gas pipeline leakage monitoring system based on infrared thermal imager, It is characterized in that The system includes the following modules: A regional thermal imaging image acquisition module is used to collect thermal imaging images of the natural gas pipeline, obtain a regional image of the pipeline area by performing threshold segmentation on the thermal imaging image, and obtain a regional thermal imaging image based on the thermal imaging image and the regional image; A gas leakage area acquisition module is used to acquire a gas annotation image based on a segmentation neural network according to the regional thermal imaging image, and then acquire the gas leakage area; the segmentation neural network extracts the first feature vector of the regional thermal imaging image through an encoder, and then uses a time convolution network to obtain the time series change characteristics of the first feature vector as a second feature vector, and then acquires the gas annotation image through a decoder; The hazard level assessment module is used to record the dynamic changes in the gas leakage area, obtain the diffusion rate according to the area change of the gas leakage area and the leakage time; obtain the temperature change during the gas leakage process, and assess the hazard level of gas diffusion according to the diffusion rate and temperature change; The leakage point positioning module is used to fit the leakage direction of the gas leakage area at each moment, obtain the fitted straight line image, superimpose all the fitted straight line images to obtain the superimposed image, and determine the position of the leakage point according to the pixel value of each pixel point in the superimposed image.

2. A natural gas pipeline leakage monitoring system based on infrared thermal imager according to claim 1, It is characterized in that The regional thermal imaging image acquisition module includes: The region division unit is used to extract the pipeline connected domain as the pipeline internal region from the segmented image after the threshold segmentation; and to perform buffer zone analysis on the pipeline connected domain to obtain the pipeline external region.

3. A natural gas pipeline leakage monitoring system based on infrared thermal imager according to claim 2, It is characterized in that The regional thermal imaging image acquisition module includes: The regional thermal imaging image acquisition unit is used to form the regional image from the internal area of ​​the pipeline and the external area of ​​the pipeline, and multiply the thermal imaging image and the corresponding pixels of the regional image to obtain the regional thermal imaging image.

4. A natural gas pipeline leakage monitoring system based on infrared thermal imager according to claim 1, It is characterized in that The gas leakage area acquisition module comprises: The segmentation neural network training unit is used to use the multi-channel regional thermal imaging image within a preset time period as the input of the neural network, extract features through the encoder to obtain the first feature image and flatten it into a first feature vector, use it as the input of the temporal convolutional network, extract the time-series change characteristics of the first feature vector as the second feature vector, reshape the second feature vector into a second feature image, and then output the gas annotation image through the decoder.

5. A natural gas pipeline leakage monitoring system based on infrared thermal imager according to claim 2, It is characterized in that The risk assessment module includes: A diffusion rate acquisition unit is used to acquire a first time when a gas leakage area first appears, a second time when it intersects with the internal area of ​​the pipeline, and a third time when it intersects with the external area of ​​the pipeline; to acquire an internal diffusion rate according to an internal diffusion area of ​​the gas leakage area between the first time and the second time and an internal diffusion time; to acquire an external diffusion rate according to an external diffusion area of ​​the gas leakage area between the second time and the third time and an external diffusion time; and to acquire a diffusion rate of the leaked gas based on the internal diffusion rate and the external diffusion rate.

6. A natural gas pipeline leakage monitoring system based on infrared thermal imager according to claim 5, It is characterized in that The risk assessment module includes: The temperature change acquisition unit is used to acquire the average temperature of the internal diffusion area as the internal area temperature, acquire the average temperature of the external diffusion area as the external area temperature, and take the temperature difference between the internal area temperature and the external area temperature as the temperature change.

7. A natural gas pipeline leakage monitoring system based on infrared thermal imager according to claim 1, It is characterized in that The leakage point positioning module includes: The leakage direction fitting unit is used to perform straight line fitting according to the gas leakage area at each moment, and take the longest line segment in the gas leakage area as the gas leakage direction at that moment.

8. A natural gas pipeline leakage monitoring system based on infrared thermal imager according to claim 7, It is characterized in that The leakage point positioning module includes: The fitted straight line image acquisition unit is used to assign a first pixel value to the straight line where the gas leakage direction is located at each moment, and assign a second pixel value to other pixel points, to obtain a binary image at each moment, and select the endpoint of the straight line in the internal area of ​​the pipeline as a reference point to linearly attenuate the binary image to obtain a fitted straight line image at each moment.

9. A natural gas pipeline leakage monitoring system based on infrared thermal imager according to claim 1, It is characterized in that The leakage point positioning module includes: The superimposed image acquisition unit is used to add the corresponding pixels of the fitted straight line image at each moment, and the added pixel values ​​decay in time sequence to obtain the superimposed image.

10. A natural gas pipeline leakage monitoring system based on infrared thermal imager according to claim 1, It is characterized in that The leakage point positioning module includes: The leakage point judgment unit is used to assign weights to each pixel value in the superimposed image based on a Gaussian kernel function, obtain a leakage point distribution image according to the pixel points of the superimposed image and the corresponding weights, and take the point with the largest pixel value as the leakage point.

Images