Refractive index measuring method based on digital image correlation method and inversion method

Abstract

The invention relates to a refractive index measuring method based on a digital image correlation method and an inversion method, which comprises the following steps of: after synchronously acquiring a top view angle speckle image and a side view angle contour image of a liquid drop arranged on a speckle pattern, establishing consistency conversion of spatial scales between the top view angle speckle image and the side view angle contour image based on a geometrical optics principle; after overlook speckle displacement caused by refraction is extracted by adopting a segmentation auxiliary digital image correlation algorithm and a curvature parameter of the side wall contour of the liquid drop is extracted by adopting an edge detection algorithm, a comprehensive physical model is established to obtain a liquid drop refractive index corresponding to any refraction point; and finally, carrying out iterative optimization on the incidence angle and the refraction parameter of the liquid drop contour feature point by adopting a particle swarm optimization algorithm to obtain an optimal refraction solution. The method does not need other standard refractive index liquid as a calculation reference, avoids the dependence on the reference refractive index, and improves the dynamic variable interference response capability in actual engineering.

Description

Technical Field

[0001] The present invention relates to a technology in the field of optical measurement, specifically a refractive index measurement method based on digital image correlation method and inversion method. Background Art

[0002] Existing measurement technologies for the refractive index of trace transparent liquid droplets rely on highly precise optical instruments. Based on the assumption of an ideal steady-state environment, it is difficult to cope with the interference of dynamic variables such as temperature and concentration changes on the measurement results, and it is difficult to achieve real-time and effective monitoring of the dynamic change process of the refractive index of liquid droplets. Summary of the Invention

[0003] Aiming at the problems of the existing technology's dependence on precision optical devices for the accurate characterization of the refractive index of transparent liquid droplets and the insufficient ability to monitor dynamic processes, the present invention proposes a refractive index measurement method based on digital image correlation method and inversion method. By describing the imaging optical path in the form of a refraction model, it avoids the problems of measurement accuracy or failure caused by the inconsistency between the actual situation and the assumptions in the existing methods, and significantly improves the measurement accuracy and the practicality of the method. By using a physical model established based on geometric optics for calculation, only the air refractive index at the measurement temperature is taken as the calculation reference, and no other standard refractive index liquids are required as calculation references, avoiding the dependence on the reference refractive index and improving the ability to cope with dynamic variable interference in actual engineering.

[0004] The present invention is realized through the following technical solutions:

[0005] The present invention relates to a refractive index measurement method based on digital image correlation method and inversion method. After synchronously collecting the top-view speckle image and the side-view contour image of the liquid droplet placed on the speckle pattern, based on the principle of geometric optics, after establishing the consistency conversion of the spatial scale between the two, the segmentation-assisted digital image correlation algorithm (SA-DIC) is used to extract the top-view speckle displacement caused by refraction respectively. After using the edge detection algorithm to extract the curvature parameters of the liquid droplet sidewall contour, a comprehensive physical model is established to obtain the refractive index of the liquid droplet corresponding to any refraction point. Finally, the particle swarm optimization algorithm (PSO) is used to iteratively optimize the incident angle and refraction parameters of the liquid droplet contour feature points to obtain the optimal refraction solution.

[0006] The liquid droplet to be measured is a transparent and uniform liquid droplet, which may contain an appropriate color, but it should be ensured that there is no turbidity caused by crystal precipitation or undissolved substances inside the liquid droplet during measurement under dynamic environmental changes, so as to avoid reducing the clarity and recognition of the transmitted speckle pattern, thereby ensuring the accuracy of displacement calculation.

[0007] The contact angle of the droplet to be measured should be greater than 5° and less than 90°. Excessively large contact angles will cause secondary refraction in the speckle pattern captured from a top-down perspective, significantly deforming the speckle pattern in the image and increasing errors in the calculated results. However, if the contact angle is too small, the droplet's side profile loses its regular arc shape due to the interaction between the droplet's surface tension and the three-phase interface, making it difficult to accurately fit the droplet's profile in the side-view image with a smooth curve. This, in turn, significantly increases errors in the refractive index calculation. Therefore, appropriately controlling the contact angle range not only reduces the effects of refraction and deformation during image acquisition but also significantly improves the accuracy of refractive index and displacement measurements.

[0008] The speckle pattern is preferably prepared by inkjet printing on offset paper and polyvinyl chloride plastic sealing technology, which can provide good speckle clarity and recognition at a magnification that ensures clear droplet edges, thereby improving the accuracy of subsequent displacement calculations.

[0009] The top-view speckle image and side-view profile image are acquired via a dual-view coupled imaging system comprising a top-view camera and a side-view camera. The top-view camera utilizes a transmissive imaging mode. Transmitted light from the characteristic surface speckle pattern is incident from a planar backlight source through the bottom of the droplet. After penetrating the droplet and undergoing refraction at the liquid-air interface, the top-view camera ultimately records the difference in speckle image between the presence and absence of the droplet. The side-view camera captures backlit profile side-view images generated by double refraction from a vertical planar light source through the surface of the liquid being measured. This clearly records subtle changes in the droplet's side geometry under different conditions, thereby enabling precise capture and analysis of the droplet's dynamic behavior and physical parameters. To ensure the integrity and accuracy of droplet edge identification, the camera's top-view field of view is vertical, perpendicular to the horizontal speckle plane supporting the droplet, and covers the droplet's complete circular outline. The side-view field of view is horizontally aligned with the center of the speckle plane, covering the droplet's entire semicircular edge.

[0010] The top view image acquisition is configured with bottom light source transmission illumination to ensure the clarity and recognizability of the speckle pattern; the side view image acquisition adopts a planar backlight illumination method to avoid the characteristic surface reflection problem caused by projection lighting, which can enhance the side view contour contrast to improve the contour recognition accuracy.

[0011] The top-view camera and the side-view camera use the same frequency for image acquisition. That is, the two cameras use a unified sampling frequency and a precise trigger signal to synchronously record the changes in the droplet morphology and speckle pattern during the dynamic process, ensuring a strict correspondence between the recorded images on the time axis, thereby achieving accurate parameter calibration and high-fidelity restoration of the droplet dynamic morphology.

[0012] The geometrical optics principle specifically includes:

[0013] 1) Calibrate the overhead camera to obtain the internal parameters of the camera, including the focal length, the position of the optical center, and the lens distortion coefficient.

[0014] 2) According to the correspondence between the pixel coordinates of the calibration plate feature points extracted from the overhead image and the actual space points, accurately determine the actual physical size of the droplet contour in the overhead state, and establish a two-dimensional spatial scale mapping relationship for the overhead image.

[0015] 3) Analyze and fit the side-view image taken at the same moment, extract the geometric features of the droplet side contour, correspond the droplet diameter measured from the side-view image with the droplet diameter at the corresponding moment in the overhead image, establish the spatial correlation between the two cameras, and then realize the consistent conversion of the spatial scale between the two cameras, so that the image information obtained from different perspectives can be accurately unified into the same spatial coordinate system, effectively improving the accuracy and stability of the measurement of droplet geometric parameters.

[0016] The calibration mentioned above uses, but is not limited to, Zhang Zhengyou calibration method.

[0017] The described segmentation-assisted digital image correlation algorithm (SA-DIC) specifically includes:

[0018] A) Perform image segmentation processing on the overhead-view speckle image after consistent conversion, divide the image into multiple sub-regions with different characteristics, and achieve more accurate feature matching in each sub-region to handle the problem of discontinuous displacement that is easy to find when measuring the droplet boundary;

[0019] B) Compare the speckle patterns of the surface to be measured in different states with the reference image, and use gray-scale matching to find the corresponding relationship of each point in the sub-region of the image, so as to obtain the surface displacement information. Specifically: in the displacement field measurement, use the DIC algorithm based on multi-sub-region segmentation. After image segmentation processing, for a single sub-region, obtain the corresponding positions in the reference image and the target image through point-by-point matching.

[0020] For the described point-by-point matching, use the zero-mean normalized least square function (ZNSSD) to describe the matching degree of each sub-region in different images. Specifically: where: the average gray value of the reference sub-region the average gray value of the target sub-region f(x, y) represents the gray value of point P(x, y) in the reference sub-region during the matching of a single sub-region, g(x′, y′) represents the gray value of P′(x′, y′), and p is the deformation parameter vector from the reference sub-region to the target sub-region.

[0021] The smaller the value of the correlation coefficient C(p), the higher the reliability of the matching from the reference sub-region to the target sub-region. On this basis, by solving the deformation parameter vector p that minimizes the correlation coefficient C(p), a geometric mapping relationship is established to deduce the displacement value of the target region. For adjacent sub-regions, joint optimization is carried out in combination with the displacement continuity constraint, thereby effectively reducing the local matching error.

[0022] The edge detection algorithm mentioned above uses an improved Canny edge detection algorithm to construct an image edge detection model to accurately extract the edge information in the side-view image of the liquid droplet to be measured. It is implemented by the techniques described in "An improved Canny edge detection algorithm" published by RONG W et al. in 2014, but is not limited thereto.

[0023] The physical model mentioned above is established according to the actual refraction light path structure to establish the quantitative relationship between the refractive index of the liquid to be measured, the droplet profile, and the displacement of the speckle base image. To accurately quantify the refractive index of the transparent droplet, the accurate displacement of the speckle, the side-view droplet profile, and the refractive indices of the media through which the imaging light rays of the two perspectives pass are all considered in the refraction model, and more accurate model parameter values can be obtained through the calibration method, significantly improving the measurement accuracy.

[0024] The comprehensive physical model mentioned above means: taking the characteristic surface where the droplet is located as the reference plane, and selecting any maximum contour section of the droplet for analysis. Let the plane where the maximum contour section is located be the X-O-Z coordinate plane, and its origin is the projection coordinate position from the center of the industrial camera lens target surface in the top view to the base plane of the droplet. Then the coordinates of any refraction point R(x f ,z f ) satisfy: where: Z0(0,z0) is the optical center point of the lens in the pinhole imaging, and R(x f ,z f ) is the coordinate position of any refraction point on the maximum contour section; when a certain point on the imaging target surface has no droplet on the base, the point R corresponds to the object point X1(x1,0) on the X-axis in the actual space, and when there is a droplet, it corresponds to the image point X′1(x′1,0) in the space. The height z0 of the optical center point of the lens in the pinhole imaging is obtained by the Zhang-Zhengyou camera calibration method, the mathematical expression of the maximum contour f(x,z) is obtained by the improved Canny edge detection algorithm, and the corresponding relationship between the two points X1 and X′1 can be determined by the matching relationship of DIC; furthermore, the refractive index n sin γ = n α sin α of any refraction point is obtained, where: the refraction incident angle the refraction angle the object point X1, the image point X′1 and the refraction point R(x f ,z f)The included angle formed is θ, n α is the refractive index of air at a fixed temperature, vector vector vector

[0025] In the specific calculation process, a large number of feature points are extracted from the images before and after droplet refraction through the feature point matching algorithm, but only the feature points located on the maximum side view contour are selected to participate in the subsequent refractive index calculation. Since the displacements of each feature point conform to the refraction model, the refractive index value can be calculated separately for each feature point. Due to the fact that the refractive index calculation of a single feature point is susceptible to the combined influence of SA-DIC measurement error, edge detection error, and refraction point position error, there may be significant deviations. If the global average value is directly used for calculation, the data quality differences of each feature point will be ignored, resulting in the averaging of errors and reducing the final measurement accuracy.

[0026] The iterative optimization of the incident angle and refraction parameters of the droplet contour feature points using the particle swarm optimization algorithm (PSO) means: Select s feature points for optimizing the calculation of the refractive index. At the beginning of the calculation, initialize the particle positions and velocities and define the upper and lower bounds of the search space. Each time an iterative calculation is performed, the refractive indices of the s feature points are regarded as the positions of the particles to be optimized and updated, and then the refractive index is evaluated according to the fitness function. Specifically: The updated particle velocity v i (t + 1) = w(t)·v i (t) + c1·r1·(pBest i - x i ) + c2·r2·(gBest i - x i ), and the updated particle position x i (t + 1) = x i (t) + v i (t + 1), where: w(t) is the inertia weight that controls the influence of the current particle velocity on the update; c1, c2 are the individual and group learning factors; r1, r2 are the random factors that determine the randomness of the search; pBest i is the historical best position of the particle, that is, the refractive index calculated from a single refraction point; gBest i is the global best position, that is, the refractive index value obtained according to the weight; x i is the position of the currently calculated particle. In the optimization process, the goal of each particle is to minimize the fitness function F(x), which considers the differences between the incident angle and the exit angle related to the refractive index and assigns weights according to the situation of each point.

[0027] The described fitness function where: The weighting coefficient of each angle for the fitness maxw Parameters representing the weights of the incident angle and the refraction angle, sin in (i), sin out (i) are the incident cosine and the outgoing cosine of the i-th feature point respectively, and x is the refractive index value to be optimized.

[0028] The evaluated refractive index mentioned above refers to: continuously iterating the particle velocity and position through the particle swarm optimization algorithm. When the iterative change of the global fitness value F(x) is less than the set threshold, it is judged that the optimization result converges. When the optimization termination condition is satisfied, the iteration stops. The global optimal position gBest, that is, the best refractive index solution n, can be obtained by inverting the solution of v i (t + 1). Specifically: n = argmin[F(x i , sin in , sin out , α, γ, w(i), ……)], F(x) - F(x) prev < ε, where: F(x) prev is the result of the fitness function of the previous optimization; ε is the set threshold.

[0029] The threshold adopted in the present invention is ε = 10 -6 .

[0030] Preferably, during the setting process of the fitness weighting coefficient, in view of the large gradient of the refraction angle change in the region closer to the droplet center, a smaller weighting coefficient is given to it; conversely, in the far-field region, due to the enhanced stability of the refraction angle, a larger weighting coefficient is adopted. Technical effects

[0031] The present invention obtains the droplet top-view speckle image and the side-view contour image through a comprehensive refraction physical model based on geometric optics and the law of refraction, and uses the segmentation-assisted digital image correlation (SA-DIC) technology and the improved Canny edge detection algorithm to accurately quantify the speckle displacement field and the droplet contour curvature characteristics caused by refraction; further combines the particle swarm optimization (PSO) algorithm to iteratively optimize and solve the incident angle and refraction parameters in the physical model. Compared with the existing methods that rely on precision optical elements, static measurement conditions and reference refractive index liquids, the present invention can effectively improve the accuracy, stability and environmental adaptability of the refractive index measurement of transparent droplets. The measurement relative error does not exceed 0.217%, reaching the accuracy level of a laboratory-grade refractometer; realizing real-time and non-invasive monitoring of the dynamic change process of the refractive index of trace liquids. Brief description of the drawings

[0032] Figure 1 is the flow chart of the present invention;

[0033] Figure 2 is the schematic diagram of the embodiment scenario;

[0034] Figure 3 It is a schematic diagram of the embodiment system. Specific implementation manner

[0035] As Figure 2 shown, this embodiment relates to a refractive index measuring device based on digital image correlation method and inversion method, including: a calculation unit (not shown in the figure) and two industrial cameras 1 and 2, a speckle substrate 5, a planar light source 7, a backlight 8 and a stage 6 respectively connected thereto, wherein: the stage 6 is adjusted to be horizontal and is provided with a planar light source, and the speckle substrate 5 is placed on the stage and fixed; the liquid droplet 3 to be measured is placed on the plane of the speckle substrate by a pipette device in the hanging drop method. The two industrial cameras are respectively located directly above and horizontally on the side of the liquid droplet to be measured, wherein the planar light source 7 and the backlight 8 are respectively arranged behind the liquid droplet to be measured in the directions of the side-view and top-view camera lenses, so that the two industrial cameras can perform dual-view imaging on the liquid droplet to be measured; the calculation unit receives and processes the image data collected by the dual-view imaging system, and calculates the refractive index of the liquid droplet to be measured based on a pre-established physical model.

[0036] The described speckle substrate is made of a light-transmitting material, and is illuminated by the planar light source below, so that the industrial camera lens in the top-view direction can clearly collect the image information of the liquid droplet, thereby ensuring the measurement accuracy.

[0037] In this embodiment, the liquid droplet 3 to be measured is distilled water. During measurement, the measurement should be carried out after the liquid droplet to be measured contacts the substrate and its morphology is stable. The quartz substrate 4, the speckle substrate 5 and the stage 6 are all light-transmitting types, and are used in cooperation with the planar light source 7 installed on the back, so that the liquid droplet 3 to be measured can also be clearly imaged in a darker environment, thereby ensuring the measurement accuracy.

[0038] As Figure 1 shown, this embodiment is a method for measuring the dynamic change of the refractive index of a transparent liquid droplet based on the above device, specifically including:

[0039] S1. Build a measurement device.

[0040] S2. Clean the stage and fix the quartz substrate 4 and the speckle substrate 5, and place the liquid droplet 3 to be measured on the speckle substrate 5 in the hanging drop method.

[0041] S3. According to the measurement requirements, start the industrial cameras 1 and 2 to respectively capture the dual-view dynamic morphological changes of the liquid droplet to be measured during evaporation over time.

[0042] S4. After the dynamic continuous shooting is completed, keep the original camera parameters of industrial cameras 1 and 2 to shoot the original speckle pattern without the droplet to be measured, and use a calibration board to calibrate the overhead lens to obtain the internal parameters of industrial camera 1. In this embodiment, the camera calibration method used is the Zhang Zhengyou calibration method. Through multi-attitude acquisition of in-plane translation, lifting, and rotation of the calibration board, establish the mapping relationship between the external parameter translation matrix and the lens optical center height to determine the parameter z0.

[0043] S5. Compare and analyze the speckle pattern covered by the droplet to be measured continuously shot by industrial camera 1 with the original speckle pattern 5. Take the fully evaporated image as the DIC reference image and the images acquired at each moment as the target images, and use the SA-DIC method to calculate the speckle displacement. By calculating the image displacement of the characteristic surface points before and after refraction, obtain the displacement field in the area near the top-down maximum contour.

[0044] S6. Take the side-view images synchronously acquired by industrial camera 2 at each moment as the contour recognition target images, obtain a continuous and smooth arc-shaped contour through edge detection, and establish a complete geometric model and its mathematical expression including an arc segment (the side edge of the droplet) and a straight segment (the base plane).

[0045] S7. Use computer 9 to calculate the refractive index of the droplet to be measured, specifically including:

[0046] 1) Uniformly select an appropriate number of refraction points on the recognized edge contour according to the calculation accuracy and efficiency requirements.

[0047] 2) According to the numerical calculation framework constructed based on the established physical model, calculate the single-point refractive index value corresponding to the refraction points selected in step 1) point by point.

[0048] 3) Combine the particle swarm optimization (PSO) algorithm for parameter inversion to obtain the optimized best refractive index.

[0049] 4) According to the refractive index data of the droplet to be measured calculated at each moment, obtain the time-refractive index change relationship of the droplet to be measured during the dynamic change process.

[0050] Through specific experiments, a camera with a resolution of 3648 pixels × 5472 pixels was used, and a checkerboard calibration board with a single grid size of 0.25 mm × 0.25 mm (overall size of 5 mm × 5 mm) was selected for camera calibration. To verify the feasibility of the established physical model, 10 μL sodium chloride solutions with mass fractions of 0%, 5%, 10%, 15%, and 20% were used as experimental working fluids. Under the constant temperature condition of 20 ± 0.5 °C, the evaporation process was recorded at a synchronous sampling frequency of 10 seconds / frame until the working fluid was completely evaporated or crystallization occurred. Using the completely evaporated image as the DIC reference image and the images collected at each moment as the target images, a sampling interval of 20 pixels was set, and the displacement field in the area near the maximum contour was obtained by calculating the image displacement of the characteristic surface points before and after refraction. Using the side-view images collected at each moment as the contour recognition target images, the contours obtained by canny edge detection were combined with the physical model and the refractive index concentration relationship to analyze the dynamic changes of the refractive index and concentration of the target liquid.

[0051] As shown in Table 1, the maximum relative error between the refractive index measured by this method and the standard values of various classical methods is 2.196%, which proves that it has high practical accuracy in the concentration range of 0 - 20%, and achieves the purpose of refractive index measurement with low cost, high precision, and dynamic measurement.

[0052] Table 1

[0053] Compared with the prior art, the present invention does not require complex optical devices and can complete real-time online monitoring of micro-droplets only through conventional imaging equipment. Its non-invasive characteristic avoids interference with the sample and is applicable to long-term dynamic research of micro-liquids (microliter level), providing a convenient and low-cost measurement solution for in-situ analysis of the physical and chemical properties of solutions.

[0054] In summary, this method eliminates the defects in the traditional technology that the measurement accuracy is affected by dynamic environment interference and optical device dependence, and can complete real-time online monitoring of micro-droplets only through conventional imaging equipment. While ensuring the advantages of non-contact measurement, the relative error between the measurement result and the standard value of the classical measurement method is ≤0.217%, and its accuracy level is close to that of laboratory-grade refractometers, meeting the requirements of actual engineering measurements. This method can directly track the dynamic changes of the refractive index and concentration of the solution during evaporation, mixing, and chemical reaction processes. Its non-invasive characteristic avoids interference with the sample and is applicable to long-term dynamic research of micro-liquids (microliter level), providing a convenient and low-cost measurement solution for in-situ analysis of the physical and chemical properties of solutions.

[0055] The above specific implementation can be locally adjusted in different ways by those skilled in the art without departing from the principles and purposes of the present invention. The protection scope of the present invention is subject to the claims and is not limited by the above specific implementation, and all implementation solutions within its scope are subject to the constraints of the present invention.

Claims

1. A refractive index measurement method based on digital image correlation method and inversion method, characterized in that, After synchronously collecting the top-view speckle image and the side-view contour image of the droplet placed on the speckle pattern, and establishing the consistent conversion of the spatial scale between the two based on the geometric optical principle, the top-view speckle displacement caused by refraction is extracted by using the segmentation-assisted digital image correlation algorithm (SA-DIC), and the curvature parameter of the sidewall contour of the droplet is extracted by using the edge detection algorithm. Then, a comprehensive physical model is established to obtain the refractive index of the droplet corresponding to any refraction point. Finally, the particle swarm optimization algorithm (PSO) is used to iteratively optimize the incident angle and refraction parameters of the characteristic points of the droplet contour to obtain the optimal refraction solution; The comprehensive physical model mentioned above refers to: taking the characteristic surface where the droplet is located as the reference plane, selecting any maximum contour section of the droplet for analysis, setting the plane where the maximum contour section is located as the X-O-Z coordinate plane, and its origin as the projection coordinate position from the center of the industrial camera lens target plane in the top view to the base plane of the droplet. Then, the coordinates of any refraction point R(x f ,z f ) satisfy: Where: Z0(0,z0) is the optical center point of the lens in the pinhole imaging. R(x f ,z f ) is the coordinate position of any refraction point on the maximum contour section. When there is no droplet on the base at a certain point on the imaging target plane, the point R on the X-axis corresponds to the object point X1(x1,0) in the actual space. When there is a droplet, it corresponds to the image point X′1(x′1,0) in the space. The height z0 of the optical center point of the lens in the pinhole imaging is obtained by the Zhang Zhengyou camera calibration method. The mathematical expression of the maximum contour f(x,z) is obtained by the edge detection algorithm. The corresponding relationship between the two points X1 and X′1 is determined by the matching relationship of DIC (Digital Image Correlation). Furthermore, the refractive index of the droplet corresponding to any refraction point is obtained as nsinγ = n α sinα, where the refraction incident angle and the refraction angle The included angles formed by the object point X1, the image point X′1 and the refraction point R(x f ,z f ) are θ. n α is the refractive index of air at a fixed temperature. The vector vector vector 2. The refractive index measurement method based on the digital image correlation method and the inversion method according to claim 1, characterized in that, The top-view speckle image and the side-view contour image are collected by a dual-view coupled imaging system including a top-view camera and a side-view camera. Among them: the optical path of the top-view camera adopts a transmissive imaging mode. The transmitted light of the characteristic surface speckle pattern is incident from the bottom of the droplet by a plane backlight source, and after passing through the inside of the droplet and experiencing refraction at the liquid-gas interface, the top-view camera finally records the difference in the speckle image in the presence or absence of the droplet. The side-view camera collects the side-view imaging of the backlight contour generated by the vertical plane light source after two refractions on the surface of the liquid to be measured, and records the subtle changes in the side geometric shape of the droplet in different states, so as to accurately capture and analyze the dynamic behavior characteristics and physical parameters of the droplet.

3. The refractive index measurement method based on the digital image correlation method and the inversion method according to claim 1, characterized in that, The geometric optical principle mentioned above specifically includes: 1) Calibrate the top-view camera to obtain the internal parameters of the camera, including the focal length, the position of the optical center, and the lens distortion coefficient; 2) According to the correspondence between the pixel coordinates of the calibration plate feature points extracted from the top-view image and the actual spatial points, accurately determine the actual physical size of the droplet contour in the top-view state, and establish the two-dimensional spatial scale mapping relationship of the top-view image; 3) Analyze and fit the side-view image taken at the same moment, extract the geometric features of the side contour of the droplet, correspond the droplet diameter measured from the side-view image to the droplet diameter at the corresponding moment in the top-view image, establish the spatial correlation between the two cameras, and then realize the consistent conversion of the spatial scale between the two cameras, so that the image information obtained from different perspectives can be accurately unified into the same spatial coordinate system.

4. The refractive index measurement method based on the digital image correlation method and the inversion method according to claim 1, characterized in that, The segmentation-assisted digital image correlation algorithm mentioned above specifically includes: A) Perform image segmentation processing on the top-view speckle image after consistent conversion, divide the image into multiple sub-regions with different characteristics, and achieve more accurate feature matching in each sub-region according to its characteristics; B) Compare the speckle patterns of the surface to be measured in different states with the reference image, and use gray-scale matching to find the corresponding relationship of each point in the sub-region of the image, so as to obtain the surface displacement information. Specifically: in the displacement field measurement, the DIC algorithm based on multi-sub-region segmentation is adopted. After the image segmentation processing, for a single sub-region, the corresponding positions in the reference image and the target image are obtained through point-by-point matching.

5. The refractive index measurement method based on the digital image correlation method and the inversion method according to claim 4, characterized in that, The point-by-point matching adopts the zero-mean normalized least square function to describe the matching degree of each sub-region in different images, specifically as follows: Among them: the average gray value of the reference sub-region the average gray value of the target sub-region f(x, y) represents the gray value of the point P(x, y) in the reference sub-region during the matching of a single sub-region, g(x′, y′) represents the gray value of P′(x′, y′), and p is the deformation parameter vector from the reference sub-region to the target sub-region.

6. The refractive index measurement method based on the digital image correlation method and the inversion method according to claim 5, characterized in that, By solving the deformation parameter vector p that minimizes the correlation coefficient C(p), establish a geometric mapping relationship to calculate the displacement value of the target region; for adjacent sub-regions, combine the displacement continuity constraint for joint optimization, so as to effectively reduce the local matching error.

7. The refractive index measurement method based on the digital image correlation method and the inversion method according to claim 1, characterized in that, Among the feature points extracted from the images before and after droplet refraction through the feature point matching algorithm, the feature points located on the largest side view contour are involved in the subsequent refractive index calculation.

8. The refractive index measurement method based on the digital image correlation method and the inversion method according to claim 1, characterized in that, The iterative optimization of the incident angle and refraction parameters of the droplet contour feature points using the particle swarm optimization algorithm means: Select s feature points for optimizing the calculation of the refractive index. At the beginning of the calculation, initialize the particle positions and velocities and define the upper and lower bounds of the search space. In each iterative calculation, consider the refractive indices of the s feature points as the positions of the particles to be optimized and update them. Then, evaluate the refractive index according to the fitness function. Specifically: The updated particle velocity v i (t + 1) = w(t)·v i (t) + c1·r1·(pBest i - x i ) + c2·r2·(gBest i - x i ). The updated particle position x i (t + 1) = x i (t) + v i (t + 1), where: w(t) is the inertia weight that controls the influence of the current particle velocity on the update; c1, c2 are the individual and swarm learning factors; r1, r2 are the random factors that determine the randomness of the search; pBest i is the historical best position of the particle, that is, the refractive index calculated for a single refraction point; gBest i is the global best position, that is, the refractive index value obtained according to the weight; x i is the position of the currently calculated particle. In the optimization process, the goal of each particle is to minimize the fitness function F(x), which considers the differences in the incident angle and exit angle related to the refractive index and assigns weights according to the situation of each point.

9. The refractive index measurement method based on the digital image correlation method and the inversion method according to claim 8, characterized in that, The fitness function described Where: the weighting coefficient of each angle to the fitness max w Represents the parameter for controlling the weights of the incident angle and the refraction angle, sin in (i), sin out (i) are the incident cosine and the outgoing cosine of the i-th feature point respectively, and x is the refractive index value to be optimized; The so-called evaluated refractive index refers to: continuously iterating the particle velocity and position through the particle swarm optimization algorithm. When the iterative change of the global fitness value F(x) is less than the set threshold, it is judged that the optimization result converges, and the iteration stops when the optimization termination condition is satisfied. The global optimal position gBest, that is, the best refractive index solution n, can be obtained by inverting the solution of v i (t + 1), specifically: n = argmin[F(x i , sin in , sin out , α, γ, w(i), ……)], F(x) - F(x) prev < ε, where: F(x) prev is the result of the fitness function of the previous optimization; ε is the set threshold.

10. The refractive index measurement method based on the digital image correlation method and the inversion method according to claim 9, characterized in that, During the setting process of the fitness weighting coefficient, it is considered that the weighting coefficient of the refraction angle in the region closer to the droplet center is greater than that in the far-field region.

Images