Novel environment-friendly gas non-contact leakage detection method and system and electronic equipment

Abstract

The invention relates to a novel environment-friendly gas non-contact leakage detection method and system and electronic equipment. The method comprises the following steps: firstly, acquiring multi-spectral infrared image data, wherein the infrared image data comprises background information and leaked mixed gas information; secondly, calculating a gas index based on the multi-spectral infrared image data to obtain multiple groups of gas index data; and finally, preprocessing the multiple groups of gas index data, carrying out binarization classification by using a method based on an empirical model to obtain a binary image, and determining a target area of a mixed gas leakage position according to the pixel size in the binary image. Compared with the prior art, the method has the advantages that the weak gas monitoring capability of infrared imaging equipment is improved, so that the detection requirement of leakage of C4F7N / CO2 / O2 mixed gas is met, and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of power system safety monitoring and industrial production safety, and in particular to a novel non-contact leakage detection method, system and electronic equipment for environmentally friendly gas. Background Art

[0002] As an excellent insulating medium, SF6 gas is widely used in the power industry, but its strong greenhouse effect always poses a threat to the global atmospheric environment. Perfluoroisobutyronitrile (C4F7N) performs well in terms of electrical strength and environmental protection. A certain ratio of perfluoroisobutyronitrile mixed with carbon dioxide (CO2) has a good insulation protection function and can be used as an insulating medium for electrical equipment. C4F7N / CO2 / O2 mixed gas is currently the most promising environmentally friendly insulating gas for high-voltage electrical equipment. With the promotion and application of C4F7N / CO2 / O2 mixed gas insulation equipment, it will inevitably face operation and maintenance problems such as equipment status assessment and fault diagnosis.

[0003] Due to the low degree of promotion and application of new environmentally friendly gases, there are not many targeted leak detection research and related products. In existing research, the detection method for C4F7N / CO2 / O2 mixed gas leakage continues the traditional SF6 gas leakage detection method, such as using a contact pump suction detection method; or installing an in-situ sensor and using an electrochemical method for online detection; or using a CO2 sensor to detect CO2 in the mixed gas. However, there is still a lack of an effective non-contact leak detection method for C4F7N mixed gas.

[0004] Passive infrared thermal imaging technology has become an important industrial gas detection method in the fields of industry and power due to its advantage of working all day, and has great development prospects in the field of gas leakage detection. However, traditional thermal imagers are limited by various factors and cannot meet the needs of C4F7N / CO2 / O2 mixed gas leakage detection. Therefore, how to improve the infrared imaging equipment's monitoring ability for weak gases to meet the detection needs of C4F7N / CO2 / O2 mixed gas leakage has become a problem that needs to be solved in this field. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a new type of environmentally friendly gas non-contact leakage detection method, system and electronic equipment, which can effectively improve the monitoring ability of infrared imaging equipment for weak gases to meet the detection needs of C4F7N / CO2 / O2 mixed gas leakage.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] According to a first aspect of the present invention, there is provided a novel environmentally friendly gas contactless leakage detection method, comprising the following steps: an infrared image acquisition step, acquiring multi-spectral infrared image data, wherein the infrared image data includes background information and leaked mixed gas information; a gas index calculation step, calculating the gas index based on the multi-spectral infrared image data to obtain multiple groups of gas index data; a leakage detection step, pre-processing the multiple groups of gas index data, and performing binary classification using an empirical model-based method to obtain a binary image, and determining a target area of ​​the mixed gas leakage location according to the pixel size in the binary image.

[0008] As a preferred technical solution, the method also includes a detection result visualization step: grayscale histogram equalization and filter enhancement processing are performed on the multi-spectral infrared image data in sequence to obtain a visualized infrared gas leakage grayscale image; after morphological closing operation and connected domain extraction processing are performed on the binary image, a threshold judgment is performed on the size of the connected domain and positions with an area smaller than a preset threshold are eliminated to obtain a target area of ​​the mixed gas leakage position; the infrared gas leakage grayscale image in the target area is subjected to pseudo-color transformation to obtain a final leakage detection result image.

[0009] As a preferred technical solution, the specific process of obtaining multi-spectral infrared image data includes: for the long-wave band, the infrared image data of the thermal radiation of the background and the leaked mixed gas in the first absorption band and the first non-absorption band are obtained in sequence; for the medium-wave band, the infrared image data of the thermal radiation of the background and the leaked mixed gas in the second absorption band and the second non-absorption band are obtained in sequence.

[0010] As a preferred technical solution, the first absorption band includes 7.5-8.1um, 8.3-8.7um, 9.1-9.5um and 10-10.4um, and the first non-absorption band includes 11-11.5um; the second absorption band includes 4.2-4.5um, and the second non-absorption band includes 3.5-3.8um.

[0011] As a preferred technical solution, the preprocessing process includes noise reduction processing and data fusion.

[0012] As a preferred technical solution, the data fusion process is expressed as follows:

[0013]

[0014] in,

[0015] α+β=1

[0016] Wherein, α and β are different fusion weight coefficients, v1 represents the gas index between the second absorption band and the second non-absorption band, and NGI2, NGI3, v4 and NGI5 represent the gas indices between different first absorption bands and the first non-absorption bands, respectively.

[0017] As a preferred technical solution, the expression of the gas index is:

[0018]

[0019] In the formula, I abs and I nabs They represent the normalized infrared image data obtained by the absorption band and the non-absorption wave respectively, and L is the background correction coefficient.

[0020] As a preferred technical solution, the method based on the empirical model includes regression analysis, partial least squares method, random forest method and support vector machine method.

[0021] According to a second aspect of the present invention, there is provided a novel environmentally friendly gas contactless leakage detection system, the system being used to implement the method described, the system comprising an infrared image acquisition module and an image processing module connected in sequence, the image processing module comprising a gas index calculation unit, a leakage detection unit and a detection result visualization unit; the infrared image acquisition module being used to acquire multi-spectral infrared image data, the infrared image data comprising background information and leakage mixed gas information; the gas index calculation unit being used to calculate the gas index based on the multi-spectral infrared image data to obtain multiple groups of gas index data; the leakage detection unit being used to pre-process the multiple groups of gas index data, and to perform binary classification using an empirical model-based method to obtain a binary image, and to determine a target area of ​​a mixed gas leakage location according to the pixel size in the binary image; the detection result visualization unit being used to output and display a final leakage detection result image.

[0022] According to a third aspect of the present invention, there is provided an electronic device, comprising a memory and a processor, wherein a computer program is stored in the memory, and the method described above is implemented when the processor executes the program.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] 1. After acquiring multi-spectral infrared image data, the present invention calculates the gas index and performs data preprocessing. On this basis, a binary image is obtained using a method based on an empirical model, and the target area of ​​the mixed gas leakage position is determined according to the pixel size in the binary image. The non-contact leakage detection and positioning of the C4F7N / CO2 / O2 mixed gas can be realized, and the monitoring capability of the infrared imaging equipment for weak gases can be improved to meet the detection requirements of the C4F7N / CO2 / O2 mixed gas leakage;

[0025] 2. The present invention enhances the gas absorption information of the characteristic spectrum of the C4F7N / CO2 / O2 mixed gas and weakens the influence of the non-absorption band by calculating the gas index, thereby improving the detection accuracy; further, the interference of noise in the binarization process is eliminated by fusing the indexes of multiple spectrum bands to reduce the false detection rate; the measurement of the selected 7 spectrum bands can avoid the high-cost high-spectrum measurement method based on the interference principle, and adopt a low-cost multi-spectrum measurement method based on spectroscopy, thereby greatly reducing the detection cost of the same type of products. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is the transmittance curve of the C4F7N / CO2 / O2 mixed gas in Example 1 of the present invention;

[0027] Figure 2 This is the infrared imaging process of C4F7N / CO2 / O2 mixed gas leakage in Example 1 of the present invention;

[0028] Figure 3 A schematic diagram of a process for providing a method in Example 1 of the present invention;

[0029] Figure 4 This is a schematic structural diagram of a non-contact leakage detection system for C4F7N / CO2 / O2 mixed gas in Example 1 of the present invention;

[0030] Figure 5 This is the binarization process based on multi-band images and gas index in Example 1 of the present invention;

[0031] Figure 6 This is the visualization output process of the C4F7N / CO2 / O2 mixed gas leakage detection result image in Example 1 of the present invention. DETAILED DESCRIPTION

[0032] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0033] Example 1

[0034] Since traditional thermal imagers are limited by many factors and cannot meet the needs of C4F7N / CO2 / O2 mixed gas leakage detection, how to maximize the innovative optimization and improvement of traditional thermal imaging equipment, improve the infrared imaging equipment's monitoring ability for weak gases, and meet the needs of C4F7N / CO2 / O2 mixed gas leakage detection is an important research direction of infrared thermal imaging technology in the power field.

[0035] The present invention aims to provide a new type of environmentally friendly gas non-contact leakage detection method, system and electronic equipment based on infrared imaging, which can quickly and accurately find gas leakage points through multi-spectral filtering imaging, calculation of gas index and combination of target extraction algorithm. The new type of environmentally friendly gas is C4F7N / CO2 / O2 mixed gas. Specifically, based on traditional wide-band thermal infrared imaging, the present invention selects the absorption peak of C4F7N / CO2 / O2 mixed gas in the medium-wave and long-wave bands, uses narrow-band filters to remove the influence of non-absorption bands, obtains multiple groups of infrared images, and enhances gas leakage infrared imaging by constructing gas index. It is further combined with background modeling, binarization processing and pseudo-color methods to achieve separation of weak gas leakage and background noise, obtain gas leakage location, and improve detection robustness.

[0036] The basic principle of infrared imaging gas detection is to use a thermal imager to detect the infrared band containing a characteristic absorption peak of the gas based on the infrared absorption spectrum characteristics of the gas and the difference in thermal radiation between the gas and the background, and to display the leaked gas, showing the radiation changes that are imperceptible to the human eye with grayscale differences. Infrared imaging gas detection technology does not require quantitative and accurate measurement of gas concentration. Its goal is to quickly image the leaked gas and accurately locate the source of the gas leak. It has the characteristics of long-distance non-contact and passive imaging.

[0037] C4F7N gas has multiple strong absorption peaks between 7.5 and 10.5 μm, CO2 gas has a strong absorption peak at 4.2-4.4 μm, and the transmittance curve of C4F7N / CO2 / O2 mixed gas is as follows: Figure 1 shown.

[0038] When the C4F7N / CO2 / O2 gas mixture leaks, the leaked gas will generally gradually drift into the atmosphere. When the infrared radiation containing a specific band emitted by the background object passes through the C4F7N / CO2 / O2 gas mixture and reaches the infrared thermal imager, the C4F7N / CO2 / O2 gas mixture will absorb part of its energy, causing it to attenuate to a certain extent. The infrared radiation received by the thermal imager on the gas path will form a difference with the background radiation that is not absorbed by the mixed gas. This radiation difference is captured by the thermal imaging system to form an infrared thermal radiation image that reflects the surface characteristics and temperature distribution of the object, that is, an infrared image containing the leaking gas cloud, from which the location information of the leak can be extracted. Infrared imaging of gas leaks such as Figure 2 shown.

[0039] Based on the aforementioned detection principle, the C4F7N / CO2 / O2 mixed gas non-contact leakage detection method and system provided in this embodiment are described in detail below.

[0040] like Figure 3 As shown, the C4F7N / CO2 / O2 mixed gas non-contact leakage detection method provided in this embodiment includes the following steps:

[0041] Infrared image acquisition step, acquiring multi-spectral infrared image data, the infrared image data including background information and leaked mixed gas information;

[0042] A gas index calculation step, based on the multi-spectral infrared image data, calculates the gas index to obtain multiple groups of gas index data;

[0043] A leak detection step, preprocessing the multiple sets of gas index data, and performing binary classification using an empirical model-based method to obtain a binary image, and determining a target area of ​​the mixed gas leakage location according to the pixel size in the binary image; and

[0044] Detection result visualization steps: grayscale histogram equalization and filtering enhancement processing are performed on the multi-spectral infrared image data in turn to obtain a visualized infrared gas leakage grayscale image; after morphological closing operation and connected domain extraction processing are performed on the binary image, a threshold judgment is made on the size of the connected domain and positions with an area smaller than the preset threshold are eliminated to obtain the target area of ​​the mixed gas leakage position; the infrared gas leakage grayscale image in the target area is subjected to pseudo-color transformation to obtain the final leakage detection result image.

[0045] like Figure 4As shown, the C4F7N / CO2 / O2 mixed gas non-contact leakage detection system for implementing the above steps includes an infrared image acquisition module 1 and an image processing module 2 connected in sequence, and the image processing module 2 includes a gas index calculation unit, a leakage detection unit, and a detection result visualization unit. Among them, the infrared image acquisition module 1 is used to implement the infrared image acquisition step of the above method; the gas index calculation unit is used to implement the gas index calculation step of the above method; the leakage detection unit is used to implement the leakage detection step of the above method; and the detection result visualization unit is used to implement the detection result visualization step of the above method.

[0046] In this embodiment, the C4F7N / CO2 / O2 mixed gas non-contact leakage detection system is mainly composed of four parts: a lens, a filter group, a detector and an image processing circuit. Specifically, the infrared image acquisition module 1 includes two groups of infrared lenses, filter groups and detectors arranged in sequence, and the image processing module 2 includes an image processing circuit. During detection, the background and the leaked mixed gas are located at one end close to the infrared lens. The infrared lens collects and converges the infrared radiation energy from the mixed gas and the background; the detector receives the infrared radiation after spectral filtering and converts the energy into an electrical signal; the image processing circuit converts the electrical signal output by the detector into an infrared image visible to the human eye, and uses multi-spectral image data to determine the leaking gas target in the scene.

[0047] According to the aforementioned detection principle and system composition, the specific implementation steps of the C4F7N / CO2 / O2 mixed gas non-contact leakage detection method provided in this embodiment are as follows:

[0048] (1) Infrared image acquisition steps

[0049] like Figure 4 As shown, for the long-wave band, corresponding to the combination of lens 1, filter set 1 and detector 1, the narrow-band filter is rotated to sequentially obtain the thermal radiation of the background and the leaking mixed gas in the four first absorption bands (7.5-8.1um, 8.3-8.7um, 9.1-9.5um, 10-10.4um, corresponding to Figure 1 band3-band6) and a first non-absorption band (11-11.5um, corresponding to Figure 1 Infrared image of band7).

[0050] For the medium wave band, corresponding to the combination of lens 2, filter set 2 and detector 2, the narrow band filter is rotated to obtain the thermal radiation of the background and the leaking mixed gas in a second absorption band (4.2-4.5um, corresponding to Figure 1 band2) and a second non-absorption band (3.5-3.8um, corresponding to Figure 1 Infrared image of band1 in .

[0051] Multi-band spectral imaging is achieved by a narrow-band filter plus a rotating mechanism. The detector sends the seven sets of image data collected from the medium-wave and long-wave band channels to the image processing circuit, which completes the gas index calculation step, the leak detection step, and the detection result visualization step.

[0052] Among them, the gas index calculation step and the leakage detection step constitute a binarization process based on multi-band images and gas index, such as Figure 5 As shown, the specific process is as follows:

[0053] (2) Gas index calculation steps

[0054] The gas index is calculated using images from multiple bands. The normalized gas index is defined as:

[0055]

[0056] In the formula, I abs and I nabs They represent the normalized infrared image data obtained by the absorption band and the non-absorption wave, respectively, and L is the background correction coefficient. According to the calculation method of the gas index, the gas index between band1 and band2, band3 and band7, band4 and band7, band5 and band7, and band6 and band7 are calculated, respectively, and recorded as NGI1, NGI2, NGI3, NGI4, and NGI5.

[0057] (3) Leak detection steps

[0058] The five groups of gas index data in the gas index calculation step are subjected to noise reduction processing to eliminate abnormal data points, and then data fusion is performed. The gas index fusion calculation method for the medium wave channel and the long wave channel is as follows.

[0059]

[0060] in,

[0061] α+β=1 (3)

[0062] In the above formula, α and β are different fusion weight coefficients.

[0063] On the fused gas index data, a binary classification method based on the empirical model is used to obtain a binary image B indicating the target area of ​​the mixed gas, which is expressed as:

[0064] B=Binarize(NGI) (4)

[0065] Specific classification methods include but are not limited to regression analysis, partial least squares, random forest, support vector machine, etc. In the binary image, the area with a pixel of 0 is the background, and the area with a pixel of 1 is the target area of ​​the mixed gas leakage location.

[0066] Based on the gas index calculation steps and the leakage detection steps, grayscale histogram equalization, filter enhancement, morphological closing operation, connected domain extraction, pseudo color transformation and other methods are used to visualize the mixed gas leakage detection result image. Figure 6 As shown, the specific process is as follows:

[0067] (4) Steps for visualizing test results

[0068] The detector data of band 3 to band 6 obtained in the infrared image acquisition step are fused, and an 8-bit grayscale image is obtained after grayscale histogram equalization, and then filtered and enhanced to obtain a visualized infrared gas leakage grayscale image I.

[0069] The binary image B obtained through the gas index calculation step and the leakage detection step is subjected to morphological closing operation and connected domain extraction processing, and the size of the connected domain is judged by a threshold value, and the positions with an area smaller than the preset threshold are eliminated to obtain the target frame coordinates of the gas leakage position, and determine the target area of ​​the mixed gas leakage position.

[0070] On the basis of the grayscale image I, the image within the target area is transformed into a pseudo-color image to replace the grayscale image of the target area, which is output as the final leakage detection result image.

[0071] Example 2

[0072] This embodiment provides an electronic device, including a memory and a processor. The memory stores a computer program. When the processor executes the program, the method in Embodiment 1 is implemented. The specific implementation steps are the same as the step flow in Embodiment 1 and are not repeated here.

[0073] It should be noted that during the implementation of the C4F7N / CO2 / O2 mixed gas non-contact leakage detection method provided in Example 1, the chip of the image processing circuit needs to have certain hardware computing and storage resource support, and the following data conditions are required during the implementation of the detection method:

[0074] 1. Radiation calibration between different filter channels. In the gas index calculation step, before calculating the gas index, it is necessary to obtain a set of radiation calibration coefficients using a standard blackbody radiation source with medium-wave band 1 or long-wave band 3 as the benchmark. The calibration coefficients are used to unify the responses of each filter band and improve detection accuracy.

[0075] 2. In the leak detection step, leak detection requires model training to improve detection accuracy. During the implementation of the method, it is necessary to establish a gas index data set and conduct model training according to different gas leakage conditions, select the optimal model and binary classification method, and reduce the false detection rate.

[0076] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A novel non-contact leakage detection method for environmentally friendly gas, characterized in that: The following steps are involved: Infrared image acquisition step, acquiring multi-spectral infrared image data, wherein the infrared image data includes background information and leaked mixed gas information; A gas index calculation step, based on the multi-spectral infrared image data, calculates the gas index to obtain multiple groups of gas index data; The leakage detection step preprocesses the multiple groups of gas index data and performs binary classification using an empirical model-based method to obtain a binary image, and determines the target area of ​​the mixed gas leakage location based on the pixel size in the binary image.

2. The novel environmentally friendly gas non-contact leakage detection method according to claim 1 is characterized in that: The method further comprises the step of visualizing the detection results: The multi-spectral infrared image data is processed in sequence by grayscale histogram equalization and filtering enhancement to obtain a visualized infrared gas leakage grayscale image; After performing morphological closing operation and connected domain extraction processing on the binary image, threshold judgment is performed on the size of the connected domain and positions with an area smaller than a preset threshold are eliminated to obtain a target area of ​​the mixed gas leakage position; The infrared gas leakage grayscale image in the target area is subjected to pseudo-color transformation to obtain a final leakage detection result image.

3. The novel environmentally friendly gas non-contact leakage detection method according to claim 1 is characterized in that: The specific process of obtaining multispectral infrared image data includes: For the long-wave band, infrared image data of thermal radiation of the background and the leaked mixed gas in the first absorption band and the first non-absorption band are obtained in sequence; For the medium wave band, infrared image data of the thermal radiation of the background and the leaking mixed gas in the second absorption band and the second non-absorption band are acquired in sequence.

4. The novel environmentally friendly gas non-contact leakage detection method according to claim 3 is characterized in that: The first absorption band includes 7.5-8.1um, 8.3-8.7um, 9.1-9.5um and 10-10.4um, and the first non-absorption band includes 11-11.5um; the second absorption band includes 4.2-4.5um, and the second non-absorption band includes 3.5-3.8um.

5. The novel environmentally friendly gas non-contact leakage detection method according to claim 3 is characterized in that: The preprocessing process includes noise reduction processing and data fusion.

6. The novel environmentally friendly gas non-contact leakage detection method according to claim 5 is characterized in that: The data fusion process is expressed as: in, α+β＝1 Wherein, α and β are different fusion weight coefficients, NGI1 represents the gas index between the second absorption band and the second non-absorption band, and NGI2, NGI3, NGI4 and NGI5 represent the gas indices between different first absorption bands and the first non-absorption bands, respectively.

7. The novel environmentally friendly gas non-contact leakage detection method according to claim 1 is characterized in that: The expression of the gas index is: In the formula, I abs and I nabs They represent the normalized infrared image data obtained by the absorption band and the non-absorption wave respectively, and L is the background correction coefficient.

8. The novel environmentally friendly gas non-contact leakage detection method according to claim 1 is characterized in that: The methods based on empirical models include regression analysis, partial least squares method, random forest method and support vector machine method.

9. A new type of environmentally friendly gas non-contact leakage detection system, characterized in that: The system is used to implement the method according to any one of claims 1 to 8, the system comprises an infrared image acquisition module and an image processing module connected in sequence, the image processing module comprises a gas index calculation unit, a leakage detection unit and a detection result visualization unit; The infrared image acquisition module is used to acquire multi-spectral infrared image data, wherein the infrared image data includes background information and leaked mixed gas information; The gas index calculation unit is used to calculate the gas index based on the multi-spectral infrared image data to obtain multiple groups of gas index data; The leakage detection unit is used to pre-process the multiple groups of gas index data, and perform binary classification using a method based on an empirical model to obtain a binary image, and determine the target area of ​​the mixed gas leakage position according to the pixel size in the binary image; The detection result visualization unit is used to output and display the final leakage detection result image.

10. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 8 is implemented.

Images