Laser speckle digital correlation method image quality evaluation and optical path adjustment system and method

Abstract

The application discloses a kind of laser speckle digital correlation method image quality evaluation and optical path adjustment system and method.Due to laser speckle being susceptible to material topography features, detection angle, light source parameter etc., make the digital image generated image quality deterioration, influence the accuracy of partial filter target plate thermal strain diagnosis.The present application utilizes gray scale image recognition and processing technology, by quantitative analysis of speckle image gray scale key features, to the validity of image and image quality are evaluated.The collection and calculation process of unqualified image is terminated, and the imaging optical path is optimized and adjusted in real time, so as to obtain laser speckle image that can be effectively used for DIC calculation.The present application can pre-judge the quality of speckle image, avoid the collection and calculation of invalid image, greatly improve the detection accuracy and diagnosis efficiency of partial filter target plate thermal strain.

Description

Technical Field

[0001] The present invention relates to a laser speckle digital image correlation method image quality evaluation system and method, and belongs to the cross-research field of advanced nuclear fusion energy diagnosis methods and image processing. Background Art

[0002] Fusion energy is one of the potential effective ways to permanently resolve humanity's energy needs. Thermonuclear fusion of deuterium and tritium in a fusion reactor generates radiant heat and high-energy neutrons (14.1 MeV). These neutrons enter the plasma-facing components of the reactor, including the breeder blanket, divertor, and vacuum chamber, generating high heat fluxes in the first wall and depositing high-power-density nuclear heat within the structure. Damage or destruction of the divertor structure can affect its proper function, potentially preventing the production of high-performance plasma. In severe cases, this can lead to inappropriate plasma discharge and directly impact the safe operation of the Tokamak device. Therefore, key technologies for diagnosing divertor target plates under operational conditions are a crucial research topic for fusion demonstration reactors.

[0003] Most traditional methods for measuring heat flux on the divertor target plate surface have certain drawbacks. Thermocouple signals are easily disturbed by the strong magnetic fields within the fusion reactor environment, and when the temperature measurement point at the tip is disturbed, poor contact can occur, resulting in inaccurate measurements. FBG fiber optic sensors have low spatial resolution, and the fiber is fragile and easily damaged. Infrared thermal imaging technology has low resolution and is affected by the emissivity of the material, resulting in low measurement accuracy. Considering these drawbacks and the requirements of engineering applications, it is imperative to develop a new method suitable for remote, non-contact, and high-precision measurement of the divertor target plate heat load.

[0004] Laser speckle digital image correlation (DIC) can address these issues. By generating active laser speckle patterns to measure thermal strain in divertor target plates, DIC not only enables high-precision, non-contact measurement but also enables online, in-situ monitoring of target plate status. However, DIC calculations require the acquisition of high-quality laser speckle images. Ensuring that these images are sufficiently qualified presents a challenge in applying this method to fusion reactors. Summary of the Invention

[0005] The technical problem solved by the present invention is: in response to the requirements of the laser speckle digital correlation diagnosis method for divertor target plates, a laser speckle digital image correlation method image quality evaluation system and method are provided, which solves the measurement errors caused by the influence of light source, detection surface morphology, detection angle, etc., and effectively improves the diagnostic efficiency of the method.

[0006] On the one hand, the technical solution adopted by the present invention is:

[0007] A laser speckle digital correlation method image quality evaluation and optical path adjustment system, the system comprising: an imaging part and a calculation part;

[0008] The imaging part is an optical system for acquiring a digital image of laser speckle, including a laser, a spatial filter, a bandpass filter and a camera;

[0009] The calculation part includes an image quality evaluation system and a DIC calculation module, wherein the image quality evaluation system includes: a speckle density analysis module, a speckle size analysis module, a grayscale feature analysis module, a grayscale gradient analysis module, and an image correlation analysis module;

[0010] The laser emitted by the laser is expanded by a spatial filter to illuminate the detection area of ​​the target plate, and then is reflected by the target plate and filtered by a bandpass filter to generate a speckle image on the CCD of the camera. The camera inputs the speckle image into the calculation part;

[0011] The calculation part is configured to: after the speckle image is transmitted to the calculation part, the speckle density analysis module analyzes the image density characteristics, and confirms whether the shooting angle is correct based on the speckle density. If the density is unqualified, the shooting angle is adjusted. If the density is qualified, the image is transmitted to the speckle size analysis module; the speckle size analysis module analyzes the size characteristics of the speckles in the image, and confirms whether the camera is correctly focused based on the size characteristics. If it is not focused, the focus plane is adjusted to the surface to be measured, and the image captured after the adjustment is qualified is input to the grayscale feature analysis module to complete grayscale feature analysis; if the image fails the grayscale feature analysis, the laser intensity is adjusted, and the image captured after the adjustment is qualified is input to the grayscale gradient analysis module for grayscale gradient analysis; if the image fails the grayscale gradient analysis, the camera aperture is adjusted, and the image captured after the adjustment is qualified is input to the image correlation analysis module. The image correlation analysis module compares whether images captured at different times are correlated. If there is no correlation, the camera shooting frequency and exposure time are adjusted. If the correlation is qualified, the camera starts capturing images and the captured images are input to the DIC calculation module to complete the calculation of the thermal strain of the target plate.

[0012] Furthermore, when the speckle density analysis module analyzes the image density feature, if the speckle density is qualified, the image is transmitted to the size analysis module; otherwise, the shooting angle is adjusted and the speckle density is checked again after the adjustment. If the speckle density is qualified after the re-check, the image is transmitted to the size analysis module; if it is still unqualified, it is checked whether the camera focus plane is on the target plate surface. If not, the camera focus plane is adjusted, and the adjusted captured image is input into the grayscale feature analysis module.

[0013] In the present invention, an image evaluation system is added before the DIC calculation module to adjust the optical parameters of images that are not suitable for DIC calculation. The calculation is performed after determining whether the collected images are qualified. It is suitable for online detection of fusion reactor divertor target plates and improves diagnostic efficiency. Among them, the speckle density analysis module and the speckle size analysis module are used to confirm whether the camera shooting angle and focus plane are accurate. If not, corresponding adjustments are made to quickly build a qualified optical path and improve diagnostic efficiency. The grayscale feature analysis module, the grayscale gradient analysis module, and the image correlation analysis module are used to analyze whether the image quality of the collected image can meet the requirements of DIC calculation, improve measurement accuracy, and avoid the collection, processing, and calculation of invalid images. The present invention obtains qualified speckle density and size by adjusting the basic optical path, obtains qualified grayscale features by adjusting the light source intensity, obtains qualified grayscale gradients by adjusting the camera aperture, and obtains qualified image correlation by adjusting the camera frequency, thereby obtaining a speckle image that meets the DIC requirements.

[0014] On the other hand, the present invention further provides a method for the above-mentioned laser speckle digital correlation method image quality evaluation and optical path adjustment system, the method comprising the following steps:

[0015] Step S1: Setting up an imaging optical path; wherein the laser emitted by the laser is expanded by a spatial filter to illuminate the detection area of ​​the target plate, and then the reflected light is filtered by a bandpass filter to generate a speckle image on the camera. The speckle image is collected in real time and input into the speckle density analysis module; then proceeding to step S2;

[0016] Step S2: The speckle density analysis module performs real-time analysis on the speckle density features of the image to check whether the speckle density is qualified. If qualified, the image is input into the speckle size analysis module. If not qualified, the camera shooting angle is checked to see if it is correct. If so, the process proceeds to step S3. If the shooting angle is incorrect, it is adjusted to the correct shooting angle and the speckle density is checked again to see if it is qualified. If so, the process proceeds to step S3. If still unqualified, the camera focus plane is checked to see if it is on the target plate surface. If not, the camera focus plane is adjusted and the process proceeds to step S4.

[0017] Step S3: The speckle size analysis module performs real-time analysis on the size of speckle particles in the image. If the size is qualified, the image is input into the grayscale feature analysis module. If it is not qualified, the camera focus plane is checked to see if it is on the target plate surface. The camera focus plane is adjusted. If it is correct, the image is input into the grayscale feature analysis module; then the process goes to step S4;

[0018] Step S4: The grayscale feature analysis module performs real-time analysis on the grayscale features of the image. If the grayscale balance is qualified, the image is input into the grayscale gradient analysis module. If the grayscale balance is unqualified, the light source intensity is adjusted to meet the requirements before the image is input into the grayscale gradient analysis module. Then, the process proceeds to step S5.

[0019] Step S5: The grayscale gradient analysis module performs real-time analysis on the grayscale gradient of the image. If the grayscale gradient is qualified, the image is input into the image correlation analysis module. If the grayscale gradient is unqualified, the camera aperture is adjusted until the image grayscale gradient is qualified and the image is then input into the image correlation analysis module; then the process proceeds to step S6;

[0020] Step S6: The image correlation analysis module performs real-time comparison on the continuously acquired images. If the correlation is satisfactory, the image is input into the DIC calculation module. If the correlation is insufficient, the camera exposure time and acquisition frame rate are adjusted until the correlation is satisfactory. The image is then input into the DIC calculation module. The DIC calculation module performs calculation and analysis on the qualified image to obtain the thermal strain number of the divertor target plate. Through inversion, real-time diagnosis of the heat flow and health status of the divertor target plate is achieved.

[0021] The advantages of the present invention are:

[0022] 1. Optimize and adjust the optical path for unqualified speckle images. Because the quality of laser speckle images is easily affected by factors such as material topography, detection angle, and light source parameters, unqualified speckle images can lead to large errors when calculating material strain using DIC, rendering the method ineffective. Against this backdrop, the present invention proposes a method for quantitatively evaluating laser speckle image quality. The evaluation results are fed back to the imaging system and the optical path parameters are optimized, enabling the optimized speckle image to be effectively used in DIC calculations.

[0023] 2. It can achieve efficient and rapid adjustment of the imaging optical path. During the diagnosis of the divertor target plate, different imaging optical paths need to be built according to the space and window conditions of the fusion device. The image features acquired by different optical paths are different, which will cause different errors in the DIC calculation based on laser speckle. Adjusting the optical path during the measurement process is time-consuming and labor-intensive, and obtaining a qualified imaging optical path often requires a long period of optimization. The present invention uses grayscale image recognition technology to quantitatively analyze the image quality of various types of laser speckle, and uses statistical methods to obtain image quality parameter criteria that meet high-precision DIC calculations. Through the image quality criteria set by this system and method, the optical path can be adjusted to achieve the rapid generation of qualified images, greatly improving the diagnostic efficiency and accuracy of the divertor target plate. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a schematic diagram of an image quality evaluation and optical path adjustment system using a laser speckle digital correlation method according to the present invention;

[0025] Figure 2 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0026] The specific embodiments of the present invention are given below in conjunction with the accompanying drawings to illustrate the technical solutions of the present invention in detail.

[0027] Figure 1 Schematic diagram of a laser speckle digital correlation method image quality evaluation and optical path adjustment system of the present invention; Figure 1 As shown, the entire system consists of a laser 2, a spatial filter 3, a bandpass filter 4, a camera 5, a computer 6, a speckle density analysis module 7, a speckle size analysis module 8, a grayscale feature analysis module 9, a grayscale gradient analysis module 10, an image correlation analysis module 11, a DIC calculation module 12, etc.

[0028] The laser emitted by the laser 2 is expanded by the spatial filter 3 and then illuminates the divertor target plate 1 to generate speckle. After the stray light is filtered by the bandpass filter 4, the CCD of the camera 5 continuously collects the target plate speckle image. The collected speckle image is directly input into the evaluation system. The speckle density analysis module 7 performs real-time analysis on the grayscale characteristics of the image and adjusts the optical path parameters. The image is then input into the speckle size analysis module 8 or the grayscale feature analysis module 9. The speckle size analysis module 8 performs real-time analysis on the size of the speckle particles in the image and adjusts the optical path parameters. The image is then input into the grayscale feature analysis module. 9; the grayscale feature analysis module 9 performs real-time analysis on the grayscale features of the image and adjusts the optical path parameters, and then inputs the image into the grayscale gradient analysis module 10; the grayscale gradient analysis module 10 performs real-time analysis on the grayscale gradient of the image and adjusts the optical path, and then inputs the image into the image correlation analysis module 11; the image correlation analysis module 11 performs real-time comparison on the continuously acquired images and adjusts the optical path parameters, and then inputs the image into the DIC calculation module 12; the DIC calculation module 12 performs calculation and analysis on the qualified images to obtain real-time strain information, thereby realizing the health status diagnosis of the divertor target plate.

[0029] like Figure 2 As shown, the method of the present invention specifically includes the following steps:

[0030] Step S1: Build an imaging optical path that meets the basic requirements of the laser speckle DIC method. Laser 2 illuminates the target plate 1 after beam expansion through a spatial filter 3. Bandpass filter 4 filters stray light, generating a speckle image on the CCD of camera 5. This speckle image is collected in real time and input into the speckle density analysis module 7.

[0031] Step S2: The speckle density analysis module 7 performs real-time analysis on the speckle density characteristics of the image to check whether the speckle density is qualified. If qualified, the image is input to the speckle size analysis module 8. If unqualified, the camera shooting angle is checked to see if it is correct. If the shooting angle is incorrect, it is adjusted to the correct shooting angle and the speckle density is checked again to see if it is qualified. If it is still unqualified, the camera focus plane is checked to see if it is on the surface of the target plate 1. If not, the camera focus plane is adjusted, and then step S3 is skipped and the image is input to the grayscale feature analysis module 9.

[0032] Step S3: The speckle size analysis module 8 performs real-time analysis on the size of the speckle particles in the image. If the size is qualified, the image is input to the grayscale feature analysis module 9. If not, the camera focus plane is checked to see if it is on the surface of the target plate 1. The camera focus plane is adjusted and, if it is correct, the image is input to the grayscale feature analysis module 9.

[0033] Step S4: The grayscale feature analysis module 9 performs real-time analysis on the grayscale features of the image. If the grayscale balance is qualified, the image is input into the grayscale gradient analysis module 10. If the grayscale balance is unqualified, the light source intensity is adjusted until the grayscale is qualified and then input into the grayscale gradient analysis module 10.

[0034] Step S5: The grayscale gradient analysis module 10 performs real-time analysis on the grayscale gradient of the image. If the grayscale gradient is qualified, the image is input into the image correlation analysis module 11. If the grayscale gradient is unqualified, the camera aperture is adjusted until the image grayscale gradient is qualified and the image is then input into the image correlation analysis module 11.

[0035] Step S6: The image correlation analysis module 11 performs real-time comparison on the continuously acquired images. If the correlation is satisfactory, the image is input to the DIC calculation module 12. If the correlation is insufficient, the camera exposure time and acquisition frame rate are adjusted until the correlation is satisfactory. The image is then input to the DIC calculation module 12. The DIC calculation module 12 performs calculations and analysis on the qualified image to obtain the thermal strain number of the divertor target plate 1. Through inversion, real-time diagnosis of the heat flow and health status of the divertor target plate is achieved.

[0036] The preferred embodiments of the present invention are as follows:

[0037] In step S1, the initial light source intensity of the laser 2 is 20 mW, the wavelength is 532 nm, and the spatial filter 3 is a Newport MODELM-900; a 532 nm bandpass filter 4 is added in front of the camera to filter out interfering stray light, and the image resolution collected by the camera 5 is 2736×2192.

[0038] After the image is input into the computer, it needs to be binarized to separate the speckle from the background. The specific method is as follows:

[0039] 1. Take the grayscale average value G0 of the entire image;

[0040] 2. Take the grayscale average value G for all pixels in the image whose grayscale value is greater than G0 max , take the gray value G of the image whose gray value is less than G0 min ;

[0041] 3. Let G = (G min +G max ) / 2;

[0042] 4. If G is equal to G0, then G is the binarization threshold; otherwise, let G0=G, and repeat steps 2-3;

[0043] 5. Set the gray value of all pixel points in the image greater than G to 255, and set the gray value of all pixel points in the image less than G to 0, to complete the binarization.

[0044] All pixel points with a gray value of 255 in the binarized image are identified as speckle pixels, and connected speckle pixels are recorded as a speckle particle.

[0045] In the step S2, the speckle density analysis module 7 includes speckle particle number statistics and average density calculation. The speckle density is the ratio of the total number of speckle particles in the image to the total number of image pixel points, and is usually qualified at 35-50 per 10,000 pixels (30-50 / 10,000 pixels). Generally, a speckle image will not exceed this range, and if it does, it is mostly due to a camera shooting angle that is too large. The shooting angle is qualified at 0-5° to the detection surface, and exceeding this range will severely reduce the speckle density, resulting in a large DIC measurement error. Adjust the camera position to be perpendicular to the shooting, and if the speckle density is still unqualified, check whether the camera 5 focusing surface is on the target plate surface 1, and adjust the focusing surface.

[0046] In the step S3, the speckle size analysis module 8 includes speckle particle average size calculation, speckle image proportion calculation, and speckle particle size standard deviation analysis. The speckle particle average size is not required, and generally, the speckle particle average size of a qualified image is 40-80 pixels. The speckle image proportion is qualified at 20%-30%, and the speckle particle size standard deviation is qualified at 40 or more. The speckle particle size is the number of speckle pixels constituting a single speckle particle, and the arithmetic mean of all speckle particle sizes is the speckle particle average size. The speckle image proportion is the ratio of the total number of speckle pixels to the total number of image pixel points. The speckle particle size standard deviation is the statistical standard deviation of all speckle particle sizes in the image. A speckle that does not meet the above standards is mostly caused by a severe defocus or a lens that is too large, resulting in almost no speckle generation. After adjusting the camera angle and focusing surface, the above conditions can generally be met. Figure 1

[0047] In the step S4, the average gray value is qualified at 15 or more (gray scale is 0-255). The average gray value is the average value of all pixel gray values in the image, which is another form of representation of the overall brightness of the image. Too low an average gray value will cause the DIC method to be unable to accurately identify speckle particles in the image, and the average gray value can be increased by increasing the light source intensity when the gray value is too low.

[0048] ​In step S5, the average grayscale gradient algorithm treats the grayscale values ​​of the entire speckle image as a two-dimensional discrete function and averages the gradient modulus of all pixels. A value above 0.03 (on a grayscale scale of 0-255) is generally considered acceptable. The average grayscale gradient is another indicator of speckle pattern contrast; higher contrast indicates easier identification of speckle particles. If the average grayscale gradient is too low, it can be increased by increasing the camera aperture to adjust the amount of light entering.

[0049] In step S6, correlation analysis is performed on consecutive images collected from the same inspection area. A correlation of 85% or higher is generally considered acceptable. When the inspection platform is vibration-free, correlation is generally acceptable. However, when the inspection platform vibrates slightly, correlation decreases, requiring improvement by reducing exposure time and increasing acquisition frequency.

[0050] Although the specific implementation methods of the present invention are described above, those skilled in the art should understand that these are merely examples and that various changes or modifications may be made to these implementation methods without departing from the principles and implementations of the present invention. Therefore, the scope of protection of the present invention is limited by the appended claims.

Claims

1. A laser speckle digital correlation method image quality evaluation and optical path adjustment system, characterized by: The system includes: an imaging part and a computing part; The imaging part is an optical system for acquiring a laser speckle digital image, comprising a laser (2), a spatial filter (3), a bandpass filter (4) and a camera (5); The calculation part includes an image quality evaluation system and a DIC calculation module (12), wherein the image quality evaluation system includes: a speckle density analysis module (7), a speckle size analysis module (8), a grayscale feature analysis module (9), a grayscale gradient analysis module (10), and an image correlation analysis module (11); The laser light emitted by the laser (2) is expanded by a spatial filter (3) to illuminate the detection area of ​​the target plate (1), and then is reflected by the target plate and filtered by a bandpass filter (4) to generate a speckle image on the CCD of the camera (5). The camera (5) inputs the speckle image into the calculation part; The calculation part is configured as follows: after the speckle image is transmitted to the calculation part, the speckle density analysis module (7) analyzes the image density characteristics, confirms whether the shooting angle is correct according to the speckle density, adjusts the shooting angle if the density is unqualified, and transmits the image to the speckle size analysis module (8) if the density is qualified; the speckle size analysis module (8) analyzes the size characteristics of the speckle particles in the image, confirms whether the camera is correctly focused according to the size characteristics, adjusts the focus plane to the surface to be measured if it is not focused, and inputs the image shot after the adjustment to the grayscale feature analysis module (9) to complete the grayscale feature analysis; The laser intensity is adjusted for the unqualified image by feature analysis, and the image taken after the adjustment is qualified is input into the grayscale gradient analysis module (10) for grayscale gradient analysis. The camera aperture is adjusted for the unqualified image by grayscale gradient analysis, and the image taken after the adjustment is qualified is input into the image correlation analysis module (11). The image correlation analysis module (11) compares whether the images collected at different times are correlated. If there is no correlation, the camera shooting frequency and exposure time are adjusted. If the correlation is qualified, the camera starts to collect images and the collected images are input into the DIC calculation module (12) to complete the calculation of the target plate thermal strain.

2. The laser speckle digital correlation method image quality evaluation and optical path adjustment system according to claim 1, characterized in that: The speckle density analysis module (7) analyzes the image density feature. If the speckle density is qualified, the image is transmitted to the speckle size analysis module (8). Otherwise, the shooting angle is adjusted and the speckle density is checked again after the adjustment. If the speckle density is qualified after the re-check, the image is transmitted to the speckle size analysis module (8). If it is still unqualified, it is checked whether the camera focus plane is on the target plate (1) surface. If it is not on the target plate (1) surface, the camera focus plane is adjusted, and the adjusted captured image is input into the grayscale feature analysis module (9).

3. A method for the laser speckle digital correlation method image quality evaluation and optical path adjustment system according to claim 1 or 2, characterized in that: The method comprises the following steps: Step S1: constructing an imaging optical path; wherein the laser light emitted by the laser (2) is expanded by the spatial filter (3) to illuminate the detection area of ​​the target plate (1), and then the reflected light is filtered by the bandpass filter (4) to generate a speckle image on the camera (5), and the speckle image is collected in real time and input into the speckle density analysis module (7); then proceeding to step S2; Step S2: The speckle density analysis module (7) performs real-time analysis on the speckle density characteristics of the image to check whether the speckle density is qualified. If qualified, the image is input into the speckle size analysis module (8). If unqualified, the camera shooting angle is checked to see if it is correct. If qualified, the process proceeds to step S3. If the shooting angle is incorrect, the camera is adjusted to the correct shooting angle, and the speckle density is checked again to see if it is qualified. If qualified, the process proceeds to step S3. If still unqualified, the camera focus plane is checked to see if it is on the surface of the target plate (1). If not, the camera focus plane is adjusted, and the process proceeds to step S4. Step S3: The speckle size analysis module (8) performs real-time analysis on the size of the speckle particles in the image. If the size is qualified, the image is input into the grayscale feature analysis module (9). If it is unqualified, the camera focus plane is checked to see if it is on the surface of the target plate (1). The camera focus plane is adjusted. If it is correct, the image is input into the grayscale feature analysis module (9); then the process goes to step S4; Step S4: The grayscale feature analysis module (9) performs real-time analysis on the grayscale features of the image. If the grayscale balance is qualified, the image is input into the grayscale gradient analysis module (10). If the grayscale balance is unqualified, the light source intensity is adjusted to be qualified before inputting into the grayscale gradient analysis module (10); then the process proceeds to step S5; Step S5: The grayscale gradient analysis module (10) performs real-time analysis on the grayscale gradient of the image. If the grayscale gradient is qualified, the image is input into the image correlation analysis module (11). If the grayscale gradient is unqualified, the camera aperture is adjusted until the image grayscale gradient is qualified and the image is then input into the image correlation analysis module (11); then the process proceeds to step S6; Step S6: The image correlation analysis module (11) performs real-time comparison on the continuously collected images. If the correlation is qualified, the image is input into the DIC calculation module (12). If the correlation is not sufficient, the camera exposure time and the acquisition frame rate are adjusted until the correlation is qualified, and the image is then input into the DIC calculation module (12). The DIC calculation module (12) performs calculation and analysis on the qualified image to obtain the thermal strain number of the divertor target plate (1), and realizes real-time diagnosis of the heat flow and health status of the divertor target plate through inversion.

Images