Insulating gas leakage imaging enhancement method and device based on GMM (Gaussian Mixture Model) and storage medium

By using the GMM model to identify and enhance insulating gas leak images, the problem of identifying trace gas leaks in complex backgrounds has been solved, achieving high-precision gas diffusion display and automatic identification.

CN119048548APending Publication Date: 2024-11-29STATE GRID ANHUI ELECTRIC POWER CO LTD ELECTRIC POWER SCI RES INST +1
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411072663.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-06
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

Under complex background interference, existing technologies struggle to effectively identify trace amounts of insulating gas leaks, especially in power scenarios where infrared imaging technology cannot meet practical application requirements.

Method used

An imaging enhancement method for insulating gas leakage based on Gaussian mixture model (GMM) is adopted. By initializing and updating the mean and variance matrices of the Gaussian mixture model, combined with the updating and normalization of weight parameters, gas diffusion images are identified and enhanced.

Benefits of technology

It effectively eliminates complex background interference, improves the accuracy and identification effect of insulating gas leak detection, automatically identifies gas diffusion patterns, and enhances the effect of infrared imaging detection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure BDA0004981313890000022
    Figure BDA0004981313890000022
  • Figure BDA0004981313890000023
    Figure BDA0004981313890000023
  • Figure BDA0004981313890000033
    Figure BDA0004981313890000033
Patent Text Reader

Abstract

The invention discloses a GMM-based insulating gas leakage imaging enhancement method, belongs to the technical field of insulating gas leakage detection, and solves the problem that trace insulating gas leakage is difficult to identify under complex background interference. Selecting continuous frame images of the infrared detector, initializing a mean value matrix and a variance matrix of each Gaussian mixture model, matching gray values obtained by pixel points with the current Gaussian mixture model, updating weight parameters, and updating a mean value and a variance of the Gaussian mixture model corresponding to the pixel points; performing normalization processing on the weight, re-sorting all Gaussian mixture models according to a sequence, selecting the Gaussian mixture model conforming to a weight threshold as a background, matching a current pixel point value with a background model, and judging whether the background information is background information or a moving foreground; according to the method, gas diffusion display is enhanced by deducting a complex dynamic background, the gas diffusion form is automatically identified while the detection precision is improved, and the infrared imaging detection effect is further improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of insulating gas leakage detection, and relates to an insulating gas leakage imaging enhancement method, device and storage medium based on GMM. Background Art

[0002] At present, power insulating gases mainly include sulfur hexafluoride, perfluoroisobutyronitrile, and mixed gases with carbon dioxide, nitrogen, etc. As insulating and arc-extinguishing media in equipment, gas-using equipment has the characteristics of compact structure, small size, electromagnetic environment friendliness, and high safety, and is widely used in power scenarios. Among them, sulfur hexafluoride gas has a strong greenhouse effect, and its greenhouse effect potential is more than 23,900 times that of carbon dioxide. Its half-life is as high as more than 3,200 years. Its decomposition products are severely corrosive and toxic, and therefore its use is restricted by many international conventions. Although perfluoroisobutyronitrile is an environmentally friendly gas, it is trace toxic. Large-scale leakage will cause casualties to surrounding personnel and bring serious social impacts. At the same time, once a large amount of insulating gas leaks, it will inevitably cause the insulation performance of the equipment to deteriorate and cause equipment operation failures. Therefore, the development of effective insulating gas leakage monitoring technology is of great significance for power production safety and ecological environmental protection.

[0003] Insulating gas leakage monitoring in the power industry mainly relies on infrared imaging, which realizes gas diffusion imaging display by measuring the regional temperature difference caused by the selective radiation absorption of insulating gas. Insulating gas infrared imaging technology is an important means of wide application in actual production. Compared with sensor monitoring, it can locate gas leakage points intuitively and quickly. Combining this monitoring technology with artificial intelligence algorithms can realize intelligent monitoring and greatly improve work efficiency.

[0004] In the infrared imaging monitoring of power insulating gas, there are problems such as difficulty in manual identification of trace gases under complex background interference and short effective detection distance. Currently, relying solely on conventional thermal infrared image differential enhancement methods still cannot meet practical applications. Summary of the invention

[0005] The technical solution of the present invention is used to solve the problem of difficulty in identifying trace insulating gas leakage under complex background interference.

[0006] The present invention solves the above technical problems through the following technical solutions:

[0007] A method for enhancing insulating gas leakage imaging based on GMM, comprising the following steps:

[0008] Step 1: Data collection;

[0009] Conduct an insulating gas leakage simulation experiment and use an infrared detector to collect infrared image data of the continuous diffusion of insulating gas leakage under natural background;

[0010] Step 2: Initialize model parameters;

[0011] Select continuous frame images of the infrared detector to initialize the mean matrix and variance matrix of each Gaussian mixture model;

[0012] Step 3: Update model parameters;

[0013] Step 3.1: Get the video frame at time t from the real-time video of the infrared detector, and convert the gray value I obtained by the pixel point (x, y) into t (x, y) is matched with the current Gaussian mixture model. If I t If (x,y) matches the jth model, then mark M j,t = 1, if there is no match with the jth model, then mark M j,t =0;

[0014] Step 3.2: After matching, the weight parameter ω of the j-th Gaussian mixture model is calculated according to each tag. j,t (x,y) is updated;

[0015] Step 3.3: If there is a marker M j,t =1, then the mean and variance of the j-th Gaussian mixture model corresponding to the pixel are updated; if the pixel value I i (x, y) does not match any Gaussian mixture model, then use the pixel value I i (x, y) replaces the mean of the corresponding position of the last Gaussian mixture model and redistributes the variance and weight;

[0016] Step 4: Model motion detection;

[0017] Step 4.1, normalize the weight of the Gaussian mixture model of each pixel;

[0018] Step 4.2: Follow Rearrange all Gaussian mixture models in order of size, and select the first B Gaussian mixture models that are greater than or equal to the weight threshold as the background;

[0019] Step 4.3: Set the current pixel value I i (x, y) is matched with the B background models obtained in step 3.2 to determine whether it is background information or motion foreground, and the pixels of background information and motion foreground are classified, thereby enhancing the infrared image of insulating gas leakage.

[0020] Furthermore, the formulas for initializing the mean matrix and variance matrix of each Gaussian mixture model are as follows:

[0021]

[0022]

[0023] In the formula, μ j.0 represents the mean matrix of the j-th Gaussian mixture model, represents the variance matrix of the j-th Gaussian mixture model, N represents the number of selected continuous frame images, I(i) represents the video sequence, where i=1, 2, ..., N; during parameter initialization, the video sequence I(i) is cumulatively sampled, and J represents the number of mean matrices and variance matrices.

[0024] Furthermore, the gray value I obtained by the pixel point (x, y) t The formula for matching (x, y) with the current Gaussian mixture model is as follows:

[0025] |I t (x,y)-μ j,t-1 (x,y)|≤c·σ j,t-1 (x,y)

[0026] In the formula, μ j,t-1 (x,y) is the mean matrix of the j-th Gaussian mixture model at time t-1, σ j,t-1 (x,y) is the standard deviation matrix of the j-th Gaussian mixture model at time t-1, and c is the empirical parameter.

[0027] Furthermore, the weight parameter ω of the j-th Gaussian mixture model according to each tag is j,t The formula for updating (x,y) is as follows:

[0028] ω j,t (x,y)=(1-α)·ω j,t-1 (x,y)+M j,t α

[0029] Among them, α is the learning rate of the weight, α∈[0,1], ω j,t (x,y) is the weight parameter of the j-th Gaussian mixture model at time t, ω j,t-1 (x, y) is the weight parameter of the j-th Gaussian mixture model at time t-1, M j,t is the matching label of the j-th Gaussian mixture model.

[0030] Furthermore, the formula for updating the mean and variance of the j-th Gaussian mixture model corresponding to the pixel point is as follows:

[0031] μ j.t (x,y)=(1-β)·μ j,t-1 (x,y)+B·I i (x,y)

[0032]

[0033] Where β is the learning rate of the mean and variance, β∈[0,1], μ j,t-1 (x,y) is the mean matrix of the j-th Gaussian mixture model at time t-1, I i (x,y) is the grayscale value of the i-th frame image, is the standard deviation matrix of the j-th Gaussian mixture model at time t-1.

[0034] Furthermore, the formula for selecting the first B Gaussian mixture models that are greater than or equal to the weight threshold as the background model is as follows:

[0035]

[0036] Among them, τ is the weight threshold and argmin() is the minimum function.

[0037] Furthermore, the formula for performing matching judgment is as follows:

[0038] |I i (x,y)-μ j,t (x,y)|≤c·σ j,t (x,y)

[0039] Among them, μ j,t (x,y) is the mean matrix of the j-th Gaussian mixture model at time t.

[0040] An electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above-mentioned GMM-based insulating gas leakage imaging enhancement method, and the processor is configured to execute the program stored in the memory.

[0041] A storage medium stores a computer program, and when the computer program is run by a processor, the steps of the insulating gas leakage imaging enhancement method based on GMM are executed.

[0042] The advantages of the present invention are:

[0043] The present invention selects continuous frame images of an infrared detector, initializes the mean matrix and variance matrix of each Gaussian mixture model, matches the grayscale value obtained by the pixel point with the current Gaussian mixture model, updates the weight parameter, and updates the mean and variance of the Gaussian mixture model corresponding to the pixel point; normalizes the weight, re-sorts all Gaussian mixture models in order, selects a Gaussian mixture model that meets the weight threshold as the background, matches the current pixel point value with the background model, and determines whether it is background information or a motion foreground; the present invention enhances the gas diffusion display by deducting the complex dynamic background, improves the detection accuracy, and automatically identifies the gas diffusion morphology, thereby further improving the infrared imaging detection effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a schematic diagram of the insulating gas infrared remote sensing imaging principle of the present invention;

[0045] Figure 2 This is a scene diagram of an infrared image acquisition experiment of insulating gas leakage according to the present invention;

[0046] Figure 3 This is a diagram showing the infrared imaging result of the insulating gas leakage of the present invention. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments:

[0049] Embodiment 1

[0050] 1. Principle of Insulating Gas Infrared Imaging

[0051] The infrared radiation transmission process of insulating gas infrared imaging detection can usually be described by a "three-layer" model, in which the first layer is the atmosphere, the second layer is the insulating gas cloud, and the third layer is the background.

[0052] like Figure 1 As shown in the figure, when the infrared detector detects the insulating gas cloud, a three-layer model can be used to describe the measurement process. The radiation brightness L (W / m 2 ·Sr·cm -1 ) can be expressed as:

[0053] L=(1-τ1)B1+τ1((1-τ2)B2+τ2L3) (1)

[0054] In the formula, τ i is the transmittance of the i-th layer; B i is the blackbody radiation brightness at the i-th (i=1,2) layer temperature; L3 is the spectral radiation brightness of the background object.

[0055] If there is no insulating gas cloud in the detector pixel area, that is, τ2 = 1, the radiation brightness L0 incident on the pixel area is:

[0056] L0=(1-τ1)B1+τ1L3 (2)

[0057] Due to the heat exchange between the target gas and the surrounding atmosphere, its temperature quickly approaches the atmospheric temperature, and the two can be approximately considered equal. Then the radiation brightness of the pixel area with insulating gas cloud is:

[0058] L=(1-τ1τ2)B1+τ1τ2L3 (3)

[0059] The radiation brightness of the pixel area with or without gas is compared. When gas appears, due to the strong characteristic absorption of the gas, τ2 is smaller, so L is smaller than L0. The manifestation on the infrared image is: the grayscale value of the area with gas is lower, thus showing the gas diffusion morphology.

[0060] 2. Gaussian mixture model algorithm

[0061] 2.1 Algorithm Principle

[0062] Gaussian mixture model (GMM) is a background modeling method based on pixel sample statistics. Assuming that the color information between pixels is independent of each other, the change of the value of each pixel in the video image can be regarded as a random process that continuously generates new pixel values. In statistics, this random process presents the form of Gaussian distribution. In reality, due to changes in light and slight jitter of objects in flame video images, the distribution of each pixel usually obeys a multi-peak Gaussian distribution. The motion detection algorithm process includes the initialization of GMM parameters, the update of model parameters, and the detection of moving targets.

[0063] (1) Model parameter initialization

[0064] In motion detection, the video image information within a period of time is usually counted to initialize the parameters of each Gaussian model, as follows:

[0065] Select continuous frame images of the infrared detector and initialize the mean matrix and variance matrix of each Gaussian mixture model. The formula is as follows:

[0066]

[0067]

[0068] In the formula, μ j.0 represents the mean matrix of the j-th Gaussian mixture model, represents the variance matrix of the j-th Gaussian mixture model, N represents the number of selected continuous frame images, I(i) represents the video sequence, where i=1, 2, ..., N; during parameter initialization, the video sequence I(i) is cumulatively sampled, and J represents the number of mean matrices and variance matrices.

[0069] (2) Model parameter update

[0070] When the GMM model parameter initialization state is completed, the next step of moving target detection begins. The video frame at time t is obtained from the real-time video of the infrared detector, and the gray value I obtained by the pixel point (x, y) is t (x, y) is matched with the current Gaussian mixture model. If I t If (x,y) matches the jth model, then mark M j,t =1; if there is no match with the jth model, mark M j,t =0, the matching judgment formula is as follows:

[0071] |I t (x,y)-μ j,t-1 (x,y)|≤c·σ j,t-1 (x,y) (6)

[0072] In the formula, μ j,t-1 (x,y) is the mean matrix of the j-th Gaussian mixture model at time t-1, σ j,t-1 (x, y) is the standard deviation matrix of the j-th Gaussian mixture model at time t-1, c is an empirical parameter, and generally 2 to 3 is used to obtain ideal detection results.

[0073] After the matching is completed, the weight parameter ω of the j-th Gaussian mixture model is adjusted according to each mark j,t (x,y) is updated according to formula (7).

[0074] ω j,t (x,y)=(1-α)·ω j,t-1 (x,y)+M j,t α (7)

[0075] Among them, α is the learning rate of the weight, α∈[0,1], ω j,t (x,y) is the weight parameter of the j-th Gaussian mixture model at time t, ω j,t-1(x, y) is the weight parameter of the j-th Gaussian mixture model at time t-1, M j,t is the matching label of the j-th Gaussian mixture model.

[0076] If there is a mark M j,t =1, then the mean and variance of the j-th Gaussian mixture model corresponding to the pixel are updated according to formula (8) and formula (9); if the pixel value I i (x, y) does not match any Gaussian mixture model, then use the pixel value I i (x, y) replaces the mean of the corresponding position of the last Gaussian mixture model and redistributes the variance and weight.

[0077] μ j.t (x,y)=(1-β)·μ j,t-1 (x,y)+β·I i (x,y) (8)

[0078]

[0079] Where β is the learning rate of the mean and variance, β∈[0,1], μ j,t-1 (x,y) is the mean matrix of the j-th Gaussian mixture model at time t-1, I i (x,y) is the grayscale value of the i-th frame image, is the standard deviation matrix of the j-th Gaussian mixture model at time t-1.

[0080] (3) Model motion detection

[0081] In the mixed Gaussian model, in order to obtain the background model at time t, all j Gaussian models that have completed parameter update at the current time t must be adjusted. The specific method is as follows:

[0082] First, the weight of the Gaussian mixture model of each pixel is normalized;

[0083] Secondly, according to All Gaussian models are reordered in order of size, and the first B models that satisfy the following formula (10) are selected as the background.

[0084] B=argmin(∑ j=1 ω j,t (x,y)≥τ (10)

[0085] Among them, τ is the weight threshold, argmin() is the minimum value function;

[0086] Finally, the current pixel value I i(x, y) is matched with the obtained B backgrounds according to formula (11). If a model j is found that satisfies formula (11), it is considered that the pixel point presents background information at time t; otherwise, the point is judged to be a moving foreground; the pixels of background information and moving foreground are classified, thereby realizing the enhancement of the insulating gas leakage infrared image.

[0087] |I i (x,y)-μ j,t (x,y)|≤c·σ j,t (x,y) (11) where μ j,t (x,y) is the mean matrix of the j-th Gaussian mixture model at time t.

[0088] 3. Experiments and Results

[0089] 3.1 Data Collection

[0090] An insulating gas (sulfur hexafluoride) leakage simulation experiment was carried out, and more than 700 frames of continuous diffusion infrared image data of insulating gas leakage under natural background were collected. The experimental scene is shown in Figure 2 .

[0091] The experiment uses the Telops FAST V1k long-wave thermal imager, and the instrument parameters are shown in Table 1. A gas cylinder filled with sulfur hexafluoride is used to simulate gas leakage, and the detection distance is 50 meters.

[0092] Table 1 Instrument parameters

[0093] Bands range 7.5~11.5um Resolution 1μL / s Detector SLS NETD 30Mk(@25℃)

[0094] The experimental parameters are shown in Table 2.

[0095] Table 2 Experimental parameters

[0096]

[0097]

[0098] 3.2 Experimental Results and Discussion

[0099] According to the insulating gas leakage infrared recognition processing flow, the collected infrared images of simulated sulfur hexafluoride gas leakage are processed to achieve real-time update and subtraction of dynamic background, and obtain the insulating gas diffusion recognition results. The following selects the gas diffusion key frame (covering the start release stage, diffusion stage and end stage) data processing results for analysis.

[0100] like Figure 3As shown in the figure, the original infrared images of insulating gas leakage diffusion at each stage ((1) is the beginning of leakage diffusion, (2) is the middle of diffusion, and (3) is the end of diffusion) and their subsequent sequences (1-1, 2-1, 3-1) are processed by Gaussian mixture model. Through dynamic background recognition and updating, the foreground insulating gas diffusion results (1-2, 2-2, 3-2) are extracted. The identified gas leakage area is further processed by morphological filtering to remove noise interference and obtain the filtered gas diffusion results (1-3, 2-3, 3-3). Finally, the gas diffusion results are fused with the original infrared image and displayed with pseudo-color annotation to obtain the insulating gas leakage diffusion recognition results.

[0101] Judging from the results of insulating gas diffusion identification, the method of the present invention effectively subtracts the static background and larger dynamic targets (walking people and fluttering leaves, etc.), successfully captures the dynamic infrared signals caused by gas diffusion, and realizes the effective identification and enhanced display of insulating gas diffusion that cannot be distinguished by the naked eye.

[0102] Embodiment 2

[0103] An electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the GMM-based insulating gas leakage imaging enhancement method in embodiment 1, and the processor is configured to execute the program stored in the memory.

[0104] Embodiment 3

[0105] A storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the insulating gas leakage imaging enhancement method based on GMM in embodiment 1 are executed.

[0106] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A GMM-based insulating gas leakage imaging enhancement method, characterized in that: The following steps are involved: Step 1: Data collection; Conduct an insulating gas leakage simulation experiment and use an infrared detector to collect infrared image data of the continuous diffusion of insulating gas leakage under natural background; Step 2: Initialize model parameters; Select continuous frame images of the infrared detector to initialize the mean matrix and variance matrix of each Gaussian mixture model; Step 3: Update model parameters; Step 3.1: Get the video frame at time t from the real-time video of the infrared detector, and convert the gray value I obtained by the pixel point (x, y) into t (x, y) is matched with the current Gaussian mixture model. If I t If (x,y) matches the jth model, then mark M j,t = 1, if there is no match with the jth model, then mark M j,t =0; Step 3.2: After matching, the weight parameter ω of the j-th Gaussian mixture model is calculated according to each tag. j,t (x,y) is updated; Step 3.3: If there is a marker M j,t =1, then the mean and variance of the j-th Gaussian mixture model corresponding to the pixel are updated; if the pixel value I i (x, y) does not match any Gaussian mixture model, then use the pixel value I i (x, y) replaces the mean of the corresponding position of the last Gaussian mixture model and redistributes the variance and weight; Step 4: Model motion detection; Step 4.1, normalize the weight of the Gaussian mixture model of each pixel; Step 4.2: Follow Rearrange all Gaussian mixture models in order of size, and select the first B Gaussian mixture models that are greater than or equal to the weight threshold as the background; Step 4.3: Set the current pixel value I i (x, y) is matched with the B background models obtained in step 3.2 to determine whether it is background information or motion foreground, and the pixels of background information and motion foreground are classified, thereby enhancing the infrared image of insulating gas leakage.

2. The GMM-based insulating gas leakage imaging enhancement method according to claim 1, characterized in that: The formulas for initializing the mean matrix and variance matrix of each Gaussian mixture model are as follows: In the formula, μ j.0 represents the mean matrix of the j-th Gaussian mixture model, represents the variance matrix of the j-th Gaussian mixture model, N represents the number of selected continuous frame images, I(i) represents the video sequence, where i=1, 2, ..., N; during parameter initialization, the video sequence I(i) is cumulatively sampled, and J represents the number of mean matrices and variance matrices.

3. The GMM-based insulating gas leakage imaging enhancement method according to claim 2, characterized in that: The gray value I obtained by the pixel point (x, y) t The formula for matching (x, y) with the current Gaussian mixture model is as follows: |I t (x,y)-μ j,t-1 (x,y)|≤c·σ j,t-1 (x,y) In the formula, μ j,t-1 (x,y) is the mean matrix of the j-th Gaussian mixture model at time t-1, σ j,t-1 (x,y) is the standard deviation matrix of the j-th Gaussian mixture model at time t-1, and c is the empirical parameter.

4. The GMM-based insulating gas leakage imaging enhancement method according to claim 3, characterized in that: The weight parameter ω of the j-th Gaussian mixture model according to each mark j,t The formula for updating (x,y) is as follows: oh j,t (x,y)=(1-α)·ω j,t-1 (x,y)+M j,t ·a Among them, α is the learning rate of the weight, α∈[0,1], ω j,t (x,y) is the weight parameter of the j-th Gaussian mixture model at time t, ω j,t-1 (x, y) is the weight parameter of the j-th Gaussian mixture model at time t-1, M j,t is the matching label of the j-th Gaussian mixture model.

5. The GMM-based insulating gas leakage imaging enhancement method according to claim 4, characterized in that: The formula for updating the mean and variance of the j-th Gaussian mixture model corresponding to the pixel is as follows: m j.t (x,y)=(1-β)·μ j,t-1 (x,y)+β·I i (x,y) Where β is the learning rate of the mean and variance, β∈[0,1], μ j,t-1 (x,y) is the mean matrix of the j-th Gaussian mixture model at time t-1, I i (x,y) is the grayscale value of the i-th frame image, is the standard deviation matrix of the j-th Gaussian mixture model at time t-1.

6. The GMM-based insulating gas leakage imaging enhancement method according to claim 5, characterized in that: The formula for selecting the first B Gaussian mixture models that are greater than or equal to the weight threshold as the background model is as follows: Among them, τ is the weight threshold and argmin() is the minimum function.

7. The GMM-based insulating gas leakage imaging enhancement method according to claim 1, characterized in that: The formula for matching judgment is as follows: |I i (x,y)-μ j,t (x,y)|≤c·σ j,t (x,y) Among them, μ j,t (x,y) is the mean matrix of the j-th Gaussian mixture model at time t.

8. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store a program that supports the processor to execute the GMM-based insulating gas leakage imaging enhancement method according to any one of claims 1 to 7, and the processor is configured to execute the program stored in the memory.

9. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the GMM-based insulating gas leakage imaging enhancement method according to any one of claims 1 to 7 are executed.

Citation Information

Cited By

  • Gas leakage detection visual identification method based on motion feature fusion

    CN119229251A

  • Intelligent detection device and method for natural gas leakage

    CN120740863A

  • Water conservancy flood bank safety monitoring system based on image recognition

    CN121147692A