Switchgear structure point cloud generation method and system based on X-ray image grayscale

Abstract

The present invention discloses a method for generating a switchgear structure point cloud based on the grayscale of an X-ray image. The method performs X-ray shooting on the switchgear to be photographed, performs noise reduction and enhancement preprocessing and regional segmentation on the obtained X-ray detection image, calculates the X-ray attenuation coefficient, calculates the transmission thickness through the grayscale value, constructs a specific spatial point cloud structure, and obtains a spatial point cloud three-dimensional model. The present invention also discloses a switchgear structure point cloud generation system based on the grayscale of an X-ray image, including: an X-ray image acquisition module, an X-ray image preprocessing module, a switchgear attenuation coefficient calculation module, a switchgear structure point cloud generation module, and a structure point cloud rendering module. The present invention efficiently utilizes X-ray image information, and can discover potential defects in switchgear without manual experience judgment. It has a high degree of automation, intuitive results, high detection accuracy and efficiency, and good detection effect.

Description

Technical Field

[0001] The present invention relates to the technical field of switchgear diagnosis, and in particular to a method and system for generating a switchgear structure point cloud based on X-ray image grayscale. Background Art

[0002] X-ray technology is the representative of radiation detection technology. It has not only been widely used in the fields of machinery, aviation, health, etc., but also has made great progress in the miniaturization, mobility and image acquisition methods of radiation machines. In recent years, X-ray detection has been widely used in the field of non-destructive testing of power equipment due to its advantage of being able to know the location of internal defects without disassembly. However, X-ray detection images have the characteristics of low contrast, small grayscale range, and low brightness. It is easy to cause misjudgment and missed judgment by visual diagnosis of defects alone. In addition, X-ray images are different from ordinary color RGB images. The principle of conventional photo imaging is the reflection of visible light, and the imaging principle of X-ray images is the energy attenuation caused by X-rays penetrating the object to be tested. It is a transmission image. Most of the conventional image processing methods applied to reflection images are not applicable to X-ray transmission images.

[0003] Switchgear is a vital component in the power system and plays a decisive role in the safe and stable operation of the power system. Therefore, whether X-ray detection can be used to quickly, efficiently and intuitively diagnose the defect location and determine the defect type in the switchgear is of great significance to the safe and stable operation of the power grid.

[0004] At this stage, X-ray inspection technology still remains at the stage of simply taking images and relying on the experience of diagnostic personnel to manually judge defects. The inspection efficiency is low and it is difficult to find potential defects in the equipment. There is a lack of a defect diagnosis method that efficiently utilizes X-ray image information and intuitively and effectively utilizes X-rays. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for generating a switchgear structure point cloud based on X-ray image grayscale to solve the above defects.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A method for generating a switchgear structure point cloud based on X-ray image grayscale comprises the following steps:

[0008] S1. According to the structure and volume of the switchgear to be tested, select appropriate X-ray detection parameters, record the working parameters of the X-ray machine workstation, and obtain the X-ray detection image;

[0009] S2, performing noise reduction and enhancement preprocessing on the captured X-ray detection image, and then performing regional segmentation on each structure in the X-ray detection image;

[0010] S3, according to the segmentation results of different regions of the X-ray detection image, confirm the material of each part and calculate the X-ray attenuation coefficient of each region;

[0011] S4, combining the X-ray detection parameters and the working parameters of the X-ray machine workstation, using the calculated X-ray attenuation coefficient, calculating the transmission thickness through the gray value of the X-ray detection image, and constructing a complete spatial point cloud structure;

[0012] S5. Use three-dimensional space visualization technology to render the complete space point cloud structure and finally obtain a three-dimensional model of the space point cloud.

[0013] Preferably, in step S2, the noise reduction and enhancement preprocessing adopts the following method: according to the characteristics of the X-ray detection image, a noisy X-ray image data set is established, and the noise at different noise levels in the X-ray detection image is extracted by using a deep learning method to obtain image denoising results at different noise levels.

[0014] Preferably, in step S2, the region segmentation adopts a method of: selecting an X-ray detection image under one of the noise levels, and segmenting the grayscale region of the X-ray detection image into different parts according to the grayscale difference of the image in the X-ray.

[0015] Preferably, in step S3, the X-ray attenuation coefficient of each area is calculated, specifically: according to the X-ray detection image data taken by the current X-ray detection system, the grayscale values ​​corresponding to different structural materials under different X-ray detection parameters are calculated respectively, and the X-ray attenuation coefficients of different materials under the system are calculated based on the grayscale values; the detailed steps are as follows:

[0016] The relationship between the X-ray transmission intensity of the X-ray machine and the thickness of the structure of the object to be measured is as follows:

[0017] (1), where is the received ray intensity, is the initial transmission intensity of the X-ray machine, is the tube current, is the tube voltage, is the focal length of the X-ray machine;

[0018] The relationship between grayscale and photoelectric conversion process in the radiographic imaging system is as follows:

[0019] (2),

[0020] In the formula, is the gray value, , , They are the photoelectric conversion coefficient, digital-to-analog conversion coefficient and grayscale value bits of the X-ray imaging system. is the X-ray attenuation coefficient, is the transmission thickness;

[0021] From formula (1) and formula (2), we can get the relationship between the gray value in the image and the X-ray machine parameters:

[0022] (3),

[0023] In the formula, It is a constant that characterizes the conversion relationship between image grayscale and X-ray imaging system;

[0024] According to the X-ray detection images that have been taken in the current X-ray imaging system, the constants are calculated based on the known equipment thickness information, camera parameter information and gray value information. and the X-ray attenuation coefficient .

[0025] Preferably, the step S4 is specifically as follows:

[0026] The calculated X-ray attenuation coefficient is combined with the grayscale value of each pixel in each area of ​​the X-ray detection image to calculate the transmission thickness of each structure in the switchgear to be tested in the X-ray detection image; according to the calculated thickness of each part and the selected camera posture, a three-dimensional point cloud structure in the corresponding direction is generated according to the transmission thickness.

[0027] The switchgear structure point cloud generation system based on X-ray image grayscale includes:

[0028] X-ray image acquisition module, X-ray image preprocessing module, switchgear attenuation coefficient calculation module, switchgear structure point cloud generation module, structure point cloud rendering module;

[0029] The X-ray image acquisition module can automatically select the X-ray machine parameters suitable for the switchgear structure to be tested according to the preset switchgear structure, and obtain the X-ray images of the switchgear to be tested from multiple angles;

[0030] The X-ray image preprocessing module is used to preprocess the acquired X-ray image, including noise filtering, image quality enhancement and segmentation of device structures in the image;

[0031] The switchgear attenuation coefficient calculation module uses an X-ray detection image taken by a reference device with known structural dimensions and material under the X-ray image acquisition module as a reference image to obtain the transmission thickness of the reference device, and then calculates the X-ray attenuation coefficients under different structural materials according to the reference image, and then detects the switchgear to be tested to obtain an X-ray detection image and calculates the X-ray transmission thickness;

[0032] The switchgear structure point cloud generation module generates a point cloud structure in three-dimensional space according to the X-ray transmission thickness and the specified resolution;

[0033] The structural point cloud rendering module reconstructs the surface of the spatial point cloud of the switchgear structure to form an entity, and can perform texture mapping on the surface to generate a spatial point cloud three-dimensional model.

[0034] The beneficial effects of the present invention are:

[0035] The switchgear structure point cloud generation method and system based on the grayscale of X-ray images of the present invention perform X-ray shooting of the switchgear to be photographed, perform noise reduction and enhancement preprocessing and regional segmentation on the obtained X-ray detection image, calculate the X-ray attenuation coefficient, calculate the transmission thickness through the grayscale value, construct a specific spatial point cloud structure, and obtain a spatial point cloud three-dimensional model. The switchgear structure point cloud generation method and system based on the grayscale of X-ray images of the present invention efficiently utilize X-ray image information, and can find potential defects in the switchgear without manual experience judgment, with a high degree of automation, intuitive results, high detection accuracy and efficiency, and good detection effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a flow chart of a method for generating a switchgear structure point cloud based on X-ray image grayscale in an embodiment of the present invention;

[0037] Figure 2 It is a structural schematic diagram of a switchgear structure point cloud generation system based on X-ray image grayscale in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The present invention is further described below in conjunction with the embodiments. It should be noted that these are merely examples and illustrations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the claims, they should be deemed to fall within the protection scope of the present invention. Embodiment 1:

[0039] Figure 1 FIG. 1 is a flow chart of a method for generating a switchgear structure point cloud based on the grayscale of an X-ray image according to the present invention. Figure 1 As shown, the method for generating a switchgear structure point cloud based on the grayscale of an X-ray image comprises the following steps:

[0040] S1. Acquire an X-ray detection image.

[0041] According to the structure and volume of the switchgear to be tested, select appropriate X-ray detection parameters and obtain X-ray detection images. The details are as follows:

[0042] According to the structure, size, volume, material, etc. of the switchgear to be tested and the resolution required for the detection image, determine the X-ray detection parameters suitable for the switchgear to be tested, including: X-ray effective focus size, X-ray tube voltage, X-ray tube current, focal length and X-ray exposure. After selection, take a full-scale photo of the switchgear to be tested.

[0043] Image pre-inspection. First, according to the X-ray B-level sensitivity detection scheme, select the X-ray detection parameters, take a set of X-ray detection images, calculate the unsharpness and image signal-to-noise ratio of the X-ray detection images, and if the above X-ray detection parameters meet the detection conditions, the object to be tested can be formally photographed and a standard X-ray detection image can be obtained; if the above conditions are not met, the X-ray tube voltage and X-ray exposure of the X-ray machine should be adjusted to continue the detection until the X-ray detection image meets the detection requirements.

[0044] S2. X-ray detection image preprocessing and region segmentation.

[0045] The captured X-ray detection image is pre-processed for noise reduction and enhancement to remove noise points in the image that affect the image grayscale; then the various structures of the switchgear to be tested in the X-ray detection image are segmented into regions, and the X-ray detection image of the switchgear to be tested is segmented into different parts, which is conducive to subsequent separate calculations and combinations for each region.

[0046] The denoising and enhancement preprocessing method is as follows: X-ray detection images are grayscale transmission images. According to their characteristics, noise of different noise levels is added to the X-ray detection images, and a noisy X-ray image dataset is established. The DNCNN neural network model is constructed using the deep learning method to denoise the image and extract the noise at different noise levels in the X-ray image, and obtain the image denoising results at different noise levels.

[0047] The method used for region segmentation is: select an X-ray detection image under one of the noise levels, and segment the grayscale area of ​​the X-ray image into different parts according to the grayscale difference of the image in the X-ray. Since the materials and thicknesses of various components in the switchgear usually have obvious differences, the device image can be segmented into several parts by limiting the grayscale value range of the X-ray image.

[0048] S3. Calculation of X-ray attenuation coefficient.

[0049] Based on the segmentation results of different areas of the X-ray detection image, the material of each part is confirmed and the X-ray attenuation coefficient of each area is calculated.

[0050] According to the X-ray detection image data taken by the current X-ray detection system (including X-ray machine, imaging board, transmission line, data processing workstation and other X-ray detection related hardware facilities), the grayscale values ​​corresponding to different structural materials under different X-ray detection parameters are calculated respectively, and the X-ray attenuation coefficients of different materials under the X-ray detection system are calculated based on this. , the detailed steps are as follows:

[0051] The relationship between the X-ray transmission intensity of the X-ray machine and the thickness of the structure of the object to be measured is as follows:

[0052] (1),

[0053] In the formula, is the received ray intensity, is the initial transmission intensity of the X-ray machine, is the tube current, is the tube voltage, is the focal length of the X-ray machine. This formula shows that the X-ray intensity can be regarded as a function of the X-ray tube voltage, tube current and spacing.

[0054] The relationship between grayscale and photoelectric conversion process in the radiographic imaging system is as follows:

[0055] (2),

[0056] In the formula, is the gray value, , , They are the photoelectric conversion coefficient, digital-to-analog conversion coefficient and grayscale value bits of the X-ray imaging system. is the X-ray attenuation coefficient, is the transmission thickness.

[0057] From formula (1) and formula (2), we can get the relationship between the gray value in the image and the X-ray machine parameters:

[0058] (3),

[0059] In the formula, A constant that characterizes the conversion relationship between image grayscale and X-ray imaging system.

[0060] According to the X-ray detection images that have been taken in the current X-ray imaging system, the constants are calculated based on the known equipment thickness information, camera parameter information and gray value information. and the X-ray attenuation coefficient .

[0061] S4. Transmission thickness calculation and spatial point cloud generation.

[0062] Combined with the working parameters of the X-ray detection parameters, using the calculated X-ray attenuation coefficient, through the gray value of the X-ray detection image, the transmission thickness of each structure of the switchgear to be tested in the X-ray detection image is calculated.

[0063] According to the calculated transmission thickness of each structure of the switchgear to be tested, combined with the selected camera posture, a three-dimensional space point cloud structure in the corresponding direction is generated according to the transmission thickness. By combining the corresponding areas of each part of the space point cloud according to the segmentation of the ray image, a three-dimensional space point cloud of the overall structure can be formed.

[0064] S5. Generate spatial point cloud three-dimensional model.

[0065] The three-dimensional space point cloud of the overall structure is rendered using three-dimensional space visualization technology to finally obtain a three-dimensional model of the space point cloud.

[0066] Figure 2 Schematic diagram of the structure of the switchgear structure point cloud generation system based on X-ray image grayscale in an embodiment of the present invention. Figure 2 As shown, a switchgear structural point cloud generation system based on X-ray image grayscale includes: an X-ray image acquisition module, an X-ray image preprocessing module, a switchgear attenuation coefficient calculation module, a switchgear structural point cloud generation module, and a structural point cloud rendering module.

[0067] The X-ray image acquisition module can automatically select the X-ray machine parameters suitable for the switchgear structure to be tested according to the preset switchgear structure, and obtain the X-ray image of the switchgear to be tested from multiple angles. In step S1, the X-ray detection image is obtained by the X-ray image acquisition module.

[0068] The X-ray image preprocessing module is used to preprocess the acquired X-ray image, including noise filtering, image quality enhancement and segmentation of the device structure in the image. In step S2, the acquired X-ray detection image can be preprocessed and region segmented by the X-ray image preprocessing module.

[0069] The switchgear attenuation coefficient calculation module uses the X-ray detection image taken by the reference device with known structural dimensions and materials under the X-ray image acquisition module as the reference image to obtain the transmission thickness of the reference device; then the X-ray attenuation coefficients under different structural materials are calculated based on the reference image, and then the switchgear to be tested is tested to obtain the X-ray detection image and calculate the X-ray transmission thickness.

[0070] First, a reference device is set up, and an X-ray detection image of the reference device is taken as a reference image under the current X-ray detection system. When the structural size and material of the reference device are known, the switch device attenuation coefficient calculation module can calculate the transmission thickness of the reference device; then the X-ray attenuation coefficient under different structural materials is calculated according to the reference image, and then the switch device to be tested is tested to obtain the X-ray detection image and calculate the X-ray transmission thickness. In steps S3 and S4, the switch device attenuation coefficient calculation module is used to calculate the X-ray attenuation coefficient and transmission thickness of the pre-processed and region-segmented X-ray detection image.

[0071] The switchgear structure point cloud generation module generates a point cloud structure in three-dimensional space according to the X-ray transmission thickness and the specified resolution. In step S4, the switchgear structure point cloud generation module forms a three-dimensional point cloud of the overall structure according to the calculated transmission thickness of each structure of the switchgear to be tested.

[0072] In step S5, the spatial point cloud of the switchgear structure is surface reconstructed through the structural point cloud rendering module to form an entity, and the surface can be texture mapped to generate a spatial point cloud three-dimensional model.

[0073] The switchgear structural point cloud generation method and system based on X-ray image grayscale of the present invention efficiently utilize X-ray image information and can discover potential defects in the switchgear without manual experience judgment. It has a high degree of automation, intuitive results, high detection accuracy and efficiency, and good detection effect.

[0074] The above is an exemplary description of the invention. Obviously, the specific implementation of the present invention is not limited to the above-mentioned method. As long as such non-substantial improvements are made by adopting the method concept and technical solution of the present invention, or the concept and technical solution of the invention are directly applied to other occasions without improvement, they are all within the protection scope of the present invention.

Claims

1. A method for generating a switchgear structure point cloud based on X-ray image grayscale, characterized in that: The following steps are involved: S1. According to the structure and volume of the switchgear to be tested, select appropriate X-ray detection parameters, record the working parameters of the X-ray machine workstation, and obtain the X-ray detection image; S2, performing noise reduction and enhancement preprocessing on the captured X-ray detection image, and then performing regional segmentation on each structure in the X-ray detection image; S3, according to the segmentation results of different regions of the X-ray detection image, confirm the material of each part and calculate the X-ray attenuation coefficient of each region; S4, combining the X-ray detection parameters and the working parameters of the X-ray machine workstation, using the calculated X-ray attenuation coefficient, calculating the transmission thickness through the gray value of the X-ray detection image, and constructing a complete spatial point cloud structure; S5. Using three-dimensional space visualization technology, the complete space point cloud structure is surface rendered to finally obtain a three-dimensional model of the space point cloud; In step S3, the X-ray attenuation coefficient of each area is calculated, specifically: according to the X-ray detection image data taken by the current X-ray detection system, the grayscale values ​​corresponding to different structural materials under different X-ray detection parameters are calculated respectively, and the X-ray attenuation coefficients of different materials under the system are calculated based on the grayscale values. ; The detailed steps are as follows: The relationship between the X-ray transmission intensity of the X-ray machine and the thickness of the structure of the object to be measured is as follows: (1), In the formula, is the received ray intensity, is the initial transmission intensity of the X-ray machine, is the tube current, is the tube voltage, is the focal length of the X-ray machine; The relationship between grayscale and photoelectric conversion process in the radiographic imaging system is as follows: (2), In the formula, is the gray value, , , They are the photoelectric conversion coefficient, digital-to-analog conversion coefficient and grayscale value bits of the X-ray imaging system. is the X-ray attenuation coefficient, is the transmission thickness; From formula (1) and formula (2), we can get the relationship between the gray value in the image and the X-ray machine parameters: (3), In the formula, It is a constant that characterizes the conversion relationship between image grayscale and X-ray imaging system; According to the X-ray detection images that have been taken in the current X-ray imaging system, the constants are calculated based on the known equipment thickness information, camera parameter information and gray value information. and the X-ray attenuation coefficient .

2. The method for generating switchgear structure point cloud based on X-ray image grayscale according to claim 1, characterized in that: In step S2, the noise reduction and enhancement preprocessing method is as follows: according to the characteristics of the X-ray detection image, a noisy X-ray image dataset is established, and the noise at different noise levels in the X-ray detection image is extracted by using a deep learning method to obtain image denoising results at different noise levels.

3. The method for generating switchgear structure point cloud based on X-ray image grayscale according to claim 1, characterized in that: In step S2, the region segmentation adopts the method of: selecting an X-ray detection image under one of the noise levels, and segmenting the grayscale region of the X-ray detection image into different parts according to the grayscale difference of the image in the X-ray.

4. The method for generating switchgear structure point cloud based on X-ray image grayscale according to claim 1, characterized in that: The S4 step is specifically as follows: The calculated X-ray attenuation coefficient is combined with the grayscale value of each pixel in each area of ​​the X-ray detection image to calculate the transmission thickness of each structure in the switchgear to be tested in the X-ray detection image; according to the calculated thickness of each part and the selected camera posture, a three-dimensional point cloud structure in the corresponding direction is generated according to the transmission thickness.

5. A switchgear structure point cloud generation system based on X-ray image grayscale, applied to a switchgear structure point cloud generation method based on X-ray image grayscale as claimed in any one of claims 1 to 4, characterized in that: include: X-ray image acquisition module, X-ray image preprocessing module, switchgear attenuation coefficient calculation module, switchgear structure point cloud generation module, structure point cloud rendering module; The X-ray image acquisition module can automatically select the X-ray machine parameters suitable for the switchgear structure to be tested according to the preset switchgear structure, and obtain the X-ray images of the switchgear to be tested from multiple angles; The X-ray image preprocessing module is used to preprocess the acquired X-ray image, including noise filtering, image quality enhancement and segmentation of device structures in the image; The switchgear attenuation coefficient calculation module obtains the transmission thickness of the reference device by taking an X-ray detection image of a reference device with known structural dimensions and material under the X-ray image acquisition module as a reference image, and then calculates the X-ray attenuation coefficients under different structural materials according to the reference image, and then detects the switchgear to be tested to obtain an X-ray detection image and calculates the X-ray transmission thickness; The switchgear structure point cloud generation module generates a point cloud structure in three-dimensional space according to the X-ray transmission thickness and the specified resolution; The structural point cloud rendering module reconstructs the surface of the spatial point cloud of the switchgear structure to form an entity, and performs texture mapping on the surface to generate a spatial point cloud three-dimensional model.

Images