Fluorescent microfilament dynamic flow spectrum separation and adhesion flow quantitative interpretation method

Abstract

The invention discloses a fluorescent microfilament dynamic flow spectrum separation and adhesion flow quantitative interpretation method, and aims to solve the problems of low interpretation precision and poor result repeatability caused by the fact that a fluorescent microfilament dynamic flow spectrum display method mainly depends on subjective experience and qualitative observation of operators at present. The method comprises the following steps: S01, preprocessing a fluorescent microfilament image acquired in an experiment; s02, calculating position coordinates of the robot; s03, the deflection angle of the fluorescent microfilament is calculated, and an angle value set is generated; s04, judging a flow state existing in the fluorescent microfilament numbered image; s05, dividing separation and attachment flow boundaries in the fluorescent microfilament numbered image; and S06, recording and storing the flow state judgment process and the boundary division process. According to the method, the quantitative criterion of the attached flow and the separated flow is provided by utilizing the physical and mathematical significance of the bimodal coefficient, the dynamic tracking of the evolution process is realized by distinguishing the transient flow structure, and the problem that the unsteady flow field structure is difficult to identify is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of aerodynamic flow visualization and identification, in particular to a fluorescent micro-filament dynamic flow spectrum separation and attached flow quantitative interpretation method. BACKGROUND

[0002] The fluorescent micro-filament dynamic flow spectrum display technology refers to arranging a fluorescent micro-filament array on the surface of an aircraft model, recording the movement state of the fluorescent micro-filament under the action of airflow by using high-speed photography, and thus realizing the visualization observation of flow structure. This method is an important means in the field of aircraft surface flow visualization, and is widely used due to its simple operation, rapid response, and real-time dynamic display of transient flow.

[0003] Accurate interpretation of the flow state revealed by the fluorescent micro-filament dynamic flow spectrum, especially accurate identification of the boundary of the separation and attachment regions, is crucial for determining the safe flight envelope of the aircraft and verifying the accuracy of the computational fluid dynamics (CFD) numerical simulation results, and is a key link for improving the reliability of aerodynamic characteristic analysis.

[0004] However, the current core problem faced by this technology is that the interpretation of separation flow and attached flow in the fluorescent micro-filament flow spectrum mainly relies on the subjective experience and qualitative observation of the operator. This method based on subjective judgment has significant limitations, with low interpretation accuracy and poor result repeatability. Specifically, it is easy to miss key information representing the transition of flow state (such as fine separation bubbles and reattachment points), resulting in unclear details of the flow field; it is difficult to objectively and quantitatively define the boundaries of the separation and attachment regions. These problems directly weaken the reliability and engineering practical value of the fluorescent micro-filament dynamic flow spectrum visualization results, which is not only not conducive to aerodynamic shape optimization design based on flow field details, but also affects the accuracy of safe flight envelope determination.

[0005] Therefore, there is an urgent need for a new method to solve the above problems. SUMMARY

[0006] The purpose of the present application is to provide a fluorescent micro-filament dynamic flow spectrum separation and attached flow quantitative interpretation method to solve the problem of low interpretation accuracy and poor result repeatability caused by the current reliance on subjective experience and qualitative observation of the operator in the fluorescent micro-filament dynamic flow spectrum display method.

[0007] To achieve the above purpose, the present application adopts the following technical solutions.

[0008] A fluorescent micro-filament dynamic flow spectrum separation and attached flow quantitative interpretation method, comprising the following steps: S01, pre-processing the fluorescent micro-filament images collected by the experiment to save the fluorescent micro-filament pre-processed images; S02, each fluorescent microfilament in the fluorescent microfilament pretreatment image is blocked and numbered, and its position coordinates are calculated, and a fluorescent microfilament number image is obtained; S03, according to the position coordinates and number of each fluorescent microfilament in the fluorescent microfilament number image, the deflection angle of the fluorescent microfilament is calculated, and an angle value set is generated; the angle value set includes the number, position coordinates and deflection angle of each fluorescent microfilament; S04, based on the angle value set, the flow state existing in the fluorescent microfilament number image is judged; the flow state includes adhesion flow, separation flow and mixed flow, and the mixed flow refers to the existence of adhesion flow and separation flow at the same time; S05, according to the angle polar coordinate distribution diagram and the size of the double peak coefficient, the separation and adhesion flow boundary in the fluorescent microfilament number image is divided; S06, the process of judging the flow state and boundary division is recorded and saved.

[0009] In the step S01, the fluorescent microfilament image collected by experiment is sequentially subjected to gray scale, binarization, noise reduction treatment, and the fluorescent microfilament pretreatment image is saved.

[0010] In the step S01, the operation of gray scale processing is as follows: based on the RGB three channels of each pixel point in the fluorescent microfilament image collected by experiment, R value, G value and B value are obtained respectively, and the gray value Gray of the corresponding pixel point is calculated by using the formula Gray = 0.299R + 0.587G + 0.114B; The operation of binarization processing is as follows: local adaptive threshold mean method is used for binarization processing; The operation of noise reduction processing is as follows: after the binarization processing is completed, the Hessian matrix eigenvector is used to detect the direction of the fluorescent microfilament, the gray value change is kept smooth along the direction of the fluorescent microfilament from the fixed end to the free end, and the gray value change is kept sharp perpendicular to the direction of the fluorescent microfilament, the noise reduction processing is completed, and the fluorescent microfilament pretreatment image is saved.

[0011] In the step S01, the operation of binarization processing is as follows: after the gray scale processing is completed, the average value mean(x, y) of all pixel gray scales in the neighborhood of each pixel point is calculated, then the local threshold T(x, y) of the pixel point is T(x, y) = mean(x, y)-C, and the gray value Gray(x, y) of the pixel point is compared with T(x, y), if Gray(x, y) ≥ T(x, y), then the gray value Gray(x, y) of the pixel point is 255; if Gray(x, y) < T(x, y), then the gray value Gray(x, y) of the pixel point is 0; Wherein, x is the horizontal coordinate of the pixel point, y is the vertical coordinate of the pixel point, mean(x, y) is the average value of the gray scale of all pixels in the neighborhood around the single pixel point, T(x, y) is the local threshold value of the pixel point, and C is a constant.

[0012] The average value of the gray scale of all pixels in the neighborhood around the single pixel point is mean(x, y); all pixels in the neighborhood around the single pixel point refer to adjacent pixels, for example, 9 grids, and the neighborhood around the middle point is the remaining eight points. C is a constant, and the range is 0-255.

[0013] In the step S01, the greater the noise, the greater the C value, and the greater the neighborhood range, the smaller the C value.

[0014] In the step S02, according to the position coordinates of each fluorescent microfilament, the endpoint coordinates (x1, y1) of the fixed end and the endpoint coordinates (x2, y2) of the free end are obtained; the direction vector v of the fluorescent microfilament is calculated through the formula v=(Δx / Δy)=(x2-x1) / (y2-y1); the angle value θ is calculated using the four-quadrant arctangent function arctan2(Δy, Δx), and the output range is [-180°, 180°]; finally, through angle normalization, it is converted to 0°-360° range.

[0015] In the step S04, the specific operation of judging the flow state existing in the fluorescent microfilament number image is as follows: According to the deflection angle of each fluorescent microfilament, the angle polar coordinate histogram corresponding to the fluorescent microfilament is drawn, and the double-peak coefficient of the angle polar coordinate histogram is calculated according to the standard double-peak coefficient formula, so as to identify the mixed flow state of the coexistence of separation and adhesion flow; wherein, b is the standard double-peak coefficient, and n is the sample capacity, g is the sample skewness, k is the sample kurtosis, is the sample mean; Standard double-peak coefficient b The calculation formula is: ; Sample skewness g The calculation formula is: ; Sample kurtosis k The calculation formula is: ; Sample mean The calculation formula is: .

[0016] In the formula, is the first iThe deflection angle of the fluorescent microfilament at the same position of the image.

[0017] In the step S05, the position range of the mixed flow state is determined according to the bimodal coefficient of the angle polar histogram of each fluorescent microfilament, and then the position range of the transition from the attachment flow to the separation flow is confirmed, and finally the boundary of the two flow states is determined.

[0018] In the step S06, the judgment of the flow state and the process of dividing the boundary are recorded and saved, and the deflection angle polar histogram of each fluorescent microfilament and the flow state are saved.

[0019] In view of the foregoing problems, the present application provides a fluorescent microfilament dynamic flow spectrum separation and attachment flow quantitative interpretation method. In one specific example, it includes the following steps.

[0020] S01, batch pre-processing is performed on the fluorescent microfilament images collected under different working conditions, including temperature, attack angle, Mach number, Reynolds number, pressure, etc. The batch pre-processing includes gray scale, binary, and noise reduction operations. Save the pre-processed images of the fluorescent microfilaments under different working conditions, and the file name includes the image name and the image format. The image name is determined by the corresponding working condition, and the image format is determined by the batch pre-processing result. Specifically, the picture format of the processed image is set in the batch pre-processing, which is intended to facilitate saving and viewing, and only helps to distinguish when analyzing data, without involving specific technical processing.

[0021] After the fluorescent microfilament images under different working conditions are collected, the images are pre-processed in the order of gray scale, binary, and noise reduction. The gray scale processing operation is as follows: using the weighted average gray scale method, according to the sensitivity of the human eye to different wavelengths of light, different weights are given to the RGB three channels of each pixel point in the fluorescent microfilament images collected under different working conditions, and then the sum is obtained to get the gray value of the corresponding pixel point; the weighted average formula is Gray = 0.299R + 0.587G + 0.114B; wherein R, G, B are the RGB three channel values of each pixel point, and 0.299, 0.587, 0.114 are the weight coefficients, which come from the ITU-R BT.709 standard. Green (G) has the greatest contribution to the brightness perception of the human eye, followed by red (R), and blue (B) is the lowest.

[0022] The operation of the binarization processing is as follows: a local adaptive threshold mean method is used, the average value mean(x, y) of the gray scale of all pixels in the neighborhood of each pixel point is calculated according to the gray scale result, and the local threshold T(x, y) of the pixel point is T(x, y) = mean(x, y)-C; the gray scale value Gray(x, y) of the pixel point is compared with T(x, y): if Gray(x, y) ≥ T(x, y), then the gray scale value Gray(x, y) of the pixel point is 255; otherwise, Gray(x, y) = 0. Wherein, the range of the pixel point neighborhood is determined by the resolution size of the fluorescent microfilament image collected by the experiment; C in the local threshold calculation formula is a constant, which is determined by the noise size of the fluorescent microfilament image collected by the experiment and the neighborhood range of the pixel point, wherein: the larger the noise, the larger the value of C, and the larger the neighborhood range, the smaller the value of C.

[0023] After the binarization processing is completed, the noise reduction processing is performed, and the specific operation is as follows: the line segment set feature is used for noise reduction, and the main principle is to use the direction-adjustable Gaussian derivative kernel, that is, the Hessian matrix feature vector to detect the direction of the fluorescent microfilament, to keep the gray scale value change smooth along the main direction and keep the gray scale value change sharp perpendicular to the direction of the fluorescent microfilament.

[0024] S02, based on the fluorescent microfilament pre-processing image under the corresponding working condition saved by S01, each fluorescent microfilament in the fluorescent microfilament pre-processing image is blocked and numbered, and its position coordinates are calculated, and a fluorescent microfilament numbered image is obtained.

[0025] Based on the fluorescent microfilament ordered grid pasting and coordinate conversion; a single fluorescent microfilament corresponds to a block, and adaptive blocking is performed according to the number of fluorescent microfilaments, and the order is from top to bottom and from left to right, and the numbering starts from 1; by using index conversion, one-dimensional block serial number is converted into two-dimensional grid coordinates, and coordinate offset calculation is performed. The offset principle is that the offset of each block is determined by the position of its upper left corner in the global image, and the calculation formula is: offset = [(column index-1) × block width, (row index-1) × block height], and finally the local coordinates are converted into global coordinates to determine the position coordinates of the fluorescent microfilament.

[0026] S03, according to the position coordinates and number of each fluorescent microfilament in the fluorescent microfilament numbered image, the deflection angle of the fluorescent microfilament is calculated, and an angle value set is generated; the angle value set includes the number, position coordinates and deflection angle of each fluorescent microfilament.

[0027] Specifically, based on the position coordinates of each fluorescent microfilament, the coordinates of the fixed end (x1, y1) and the coordinates of the free end (x2, y2) can be obtained. The direction vector v of the fluorescent microfilament is calculated using the formula v=(Δx / Δy)=(x2-x1) / (y2-y1). The angle value θ is calculated using the four-quadrant arctan2(Δy, Δx), with an output range of [-180°, 180°]. Finally, the angle is normalized to be converted to the range of 0°~360°.

[0028] S04. Based on the set of angle values, determine the flow state present in the fluorescent microfilament images under different working conditions, and determine the flow state present in the fluorescent microfilament numbered images; the flow state includes attached flow, separated flow, and mixed flow in which both flows exist.

[0029] Specifically, an angular polar histogram is plotted based on the deflection angle of each fluorescent microfilament, and the bimodal coefficient of the angular polar histogram is calculated using the standard bimodal coefficient formula to identify the mixed flow state in which separation and attached flow coexist.

[0030] The standard bimodal coefficients are given below ( b ), sample skewness ( g ), sample kurtosis ( k ) and sample mean The formula for calculating is as follows: where n is the sample size.

[0031] Standard bimodal coefficient b The calculation formula is: ; Sample skewness g The calculation formula is: ; Sample kurtosis k The calculation formula is: ; Sample mean The calculation formula is: .

[0032] S05. Based on the angular polar coordinate distribution diagram and the magnitude of the bimodal coefficient, delineate the boundary between the separation and attachment flows in the fluorescent microfilament numbering image.

[0033] In step S05, the location range of the mixed flow state is determined based on the bimodal coefficient of the polar coordinate histogram of each fluorescent microfilament, thereby confirming the location range of the transition from the attached flow to the separated flow, and finally determining the boundary between the two flow states.

[0034] S06, record and save the judgment flow state and boundary division process.

[0035] Record and save the judgment flow state and boundary division process; wherein, save the deflection angle polar coordinate histogram of each fluorescent microfilament and the flow state; wherein, the image name is determined by the different flow states and working conditions, and the image format is determined by the batch pre-processing result, only for distinguishing in the analysis data, and not involving specific technical processing.

[0036] Compared with the prior art, the present application has at least the following beneficial effects: the simple model near-wall surface complex flow structure visualization is realized by using the fluorescent microfilament and image processing technology, and has the characteristics of high sensitivity and high contrast; the quantitative criterion of attached flow and separated flow is proposed by using the physical and mathematical meaning of the double-peak coefficient, the dynamic tracking of the evolution process is realized by distinguishing the transient flow structure, and the problem of difficult identification of unsteady flow field structure is solved. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0038] The present application will be described by examples and with reference to the accompanying drawings, in which: Figure 1 Flowchart of Example 1; Figure 2 Preprocessing result schematic diagram of fluorescent microfilament image of Example 1; Figure 3 Fluorescent microfilament block diagram of Example 1; Figure 4 Deflection angle polar coordinate histogram of fluorescent microfilament of Example 1; Figure 5 Boundary division schematic diagram of different flow structures of Example 1. DETAILED DESCRIPTION

[0039] All features disclosed in this specification, or all steps of any method or process disclosed in this specification, can be combined in any manner, except where features or steps are mutually exclusive.

[0040] Any feature disclosed in this specification, unless specifically stated otherwise, can be replaced by an alternative feature or equivalent feature having the same or similar effect. That is, unless specifically stated otherwise, each feature is merely one example out of a range of equivalent or similar features.

[0041] Clearly, the described embodiments are merely a part of the embodiments of the present application, rather than all the embodiments of the present application.

[0042] Embodiment 1 The embodiment provides a fluorescent microfiber dynamic flow spectrum separation and attachment flow quantitative interpretation method. The method is used on a certain simple wing model, and includes the following steps.

[0043] S01, under certain wind tunnel experimental conditions (experimental conditions include Mach number, Reynolds number, temperature, pressure, angle, etc.), after gray scale, binarization and noise reduction preprocessing operations are performed on the fluorescent microfiber image, a result graph as shown in FIG. 1 is obtained. Figure 2

[0044] The weighted average gray scale method is adopted. Different weights are given to three channels R, G and B of each pixel point in the fluorescent microfiber image collected under different conditions according to the sensitivity of the human eye to light of different wavelengths, and then the gray value of the corresponding pixel point is obtained by summation. The weighted average formula is Gray = 0.299R + 0.587G + 0.114B. The weight coefficient comes from the ITU-R BT.709 standard. Green (G) has the largest contribution to the brightness perception of the human eye, red (R) is the second, and blue (B) is the lowest.

[0045] After the gray scale processing is completed, the local adaptive threshold mean value method is used for binarization processing. The operation of binarization processing is as follows: after the gray scale processing is completed, the average value mean(x, y) of the gray scale of all pixels in the neighborhood of each pixel point is calculated, the local threshold T(x, y) of the pixel point is T(x, y) = mean(x, y)-C, and the gray value Gray(x, y) of the pixel point is compared with T(x, y). If Gray(x, y) ≥ T(x, y), the gray value Gray(x, y) of the pixel point is 255; if Gray(x, y) < T(x, y), the gray value Gray(x, y) of the pixel point is 0. Wherein, x is the horizontal coordinate of the pixel point, y is the vertical coordinate of the pixel point, mean(x, y) is the average value of the gray scale of all pixels in the neighborhood of a single pixel point, T(x, y) is the local threshold of the pixel point, and C is a constant. The range of the pixel point neighborhood is determined by the resolution of the fluorescent microfiber image collected by the experiment. C in the local threshold calculation formula is a constant, which is determined by the image noise size and the pixel point neighborhood range; the larger the noise is, the larger the value of C is; the larger the neighborhood range is, the smaller the value of C is.

[0046] ​After the binarization processing is completed, noise reduction processing is performed, and the specific operation is as follows: noise reduction is performed using the line segment set feature, and the main principle is to use a direction-adjustable Gaussian derivative kernel, i.e., a Hessian matrix feature vector, to detect the direction of the fluorescent microfilament, keep the gray value change smooth along the main direction, and keep the gray value change sharp perpendicular to the fluorescent microfilament direction.

[0047] S02, each fluorescent microfilament in the fluorescent microfilament preprocessed image is blocked and numbered, and its position coordinates are calculated to obtain a fluorescent microfilament numbered image, and the specific operation is as follows.

[0048] Figure 3 The fluorescent microfilament numbering and blocking result schematic diagram is given. Based on the ordered grid pasting and coordinate conversion of the fluorescent microfilament, a single fluorescent microfilament corresponds to a block, and adaptive blocking is performed according to the number of fluorescent microfilaments, and is numbered from 1 in the order of from top to bottom and from left to right. Using index conversion, one-dimensional block serial number is converted into two-dimensional grid coordinates, and coordinate offset calculation is performed. The offset principle is that the offset of each block is determined by the position of its upper left corner in the global image, and the calculation formula is: offset=[(column index-1)×block width, (row index-1)×block height], and finally the local coordinates are converted into global coordinates to determine the position coordinates of the fluorescent microfilament.

[0049] S03, according to the position coordinates and number of each fluorescent microfilament in the fluorescent microfilament numbered image, the deflection angle of the fluorescent microfilament is calculated, and an angle value set is generated; the angle value set includes the number, position coordinates and deflection angle of each fluorescent microfilament.

[0050] Specifically, according to the position coordinates of each fluorescent microfilament, the endpoint coordinates (x1, y1) of the fixed end and the endpoint coordinates (x2, y2) of the free end can be obtained, and the direction vector v of the fluorescent microfilament is calculated by the formula v=(Δx / Δy)=(x2-x1) / (y2-y1); the angle value is calculated using the four-quadrant arctangent function arctan2(Δy, Δx), and the output range is [-180°, 180°], and finally the angle is normalized to the range of 0°~360°.

[0051] S04, based on the angle value set, the flow state existing in the fluorescent microfilament numbered image is judged; the flow state includes adhesion flow, separation flow, and mixed flow of both.

[0052] Specifically, Figure 4 the polar coordinate histogram of the deflection angle of each fluorescent microfilament, Figure 4The number in the upper left corner indicates the serial number of the fluorescent microfilament, i.e., which fluorescent microfilament is being represented. MATLAB is used to plot the polar histogram corresponding to the deflection angle of each fluorescent microfilament, and the bimodal coefficient of the angular polar histogram is calculated using the standard bimodal coefficient formula to identify the mixed flow state where separation and attached flow coexist.

[0053] The standard bimodal coefficients are given below ( b ), sample skewness ( g ), sample kurtosis ( k ) and sample mean The formula for calculating is as follows: where n is the sample size.

[0054] Standard bimodal coefficient b The calculation formula is: .

[0055] Sample skewness g The calculation formula is: .

[0056] Sample kurtosis k The calculation formula is: .

[0057] Sample mean The calculation formula is: .

[0058] S05. Based on the angular polar coordinate distribution diagram and the magnitude of the bimodal coefficient, delineate the separation and attachment flow boundaries in the fluorescent microfilament numbering image.

[0059] Specifically, Figure 5 A schematic diagram of the flow structure boundary of a fluorescent microfilament exhibiting mixed flow is presented. Figure 5 The serial numbers corresponding to the dots in the middle are 35, 96, 106, 133, 140, and 156. Figure 4 The serial numbers of the fluorescent microfilaments correspond one-to-one. Specifically, based on the magnitude of the bimodal coefficient of the polar histogram of each fluorescent microfilament, the location range of the mixed flow state is determined, thereby confirming the location range of the transition from the attached flow to the separated flow, and finally determining the boundary between the two flow states.

[0060] S06. Record and save the process of determining the flow state and boundary delineation.

[0061] In step S06, the process of determining the flow state and boundary delineation is recorded and saved; among which, the polar coordinate histogram of the deflection angle of each fluorescent microfilament and the flow state are saved.

[0062] Through calculation, as Figure 4The angle polar coordinate distribution of each fluorescent microfilament is shown, and the results are as follows Figure 5 The flow state change of each position is shown. Figure 5 The boundary between the attached flow and the detached flow is clearly shown (as shown by the green line in the lower right corner), and above the boundary is the attached flow and below the boundary is the strong detached flow; the flow near the green line has both the attached flow and the detached flow.

[0063] Finally, it should be noted that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the examples, those skilled in the art should understand that modifications or equivalent replacements of the technical solutions of the present application do not deviate from the spirit and scope of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A fluorescent microfilament dynamic flow spectrum separation and adhesion flow quantitative interpretation method, characterized in that, The method comprises the following steps: S01, pre-processing the fluorescence microfilament image collected in the experiment to obtain a fluorescence microfilament pre-processed image; S02, dividing and numbering each fluorescence microfilament in the fluorescence microfilament pre-processed image and calculating the position coordinates to obtain a fluorescence microfilament numbered image; S03, calculating the deflection angle of each fluorescence microfilament in the fluorescence microfilament numbered image according to the position coordinates and the number of each fluorescence microfilament and generating an angle value set; the angle value set comprises the number, position coordinates and deflection angle of each fluorescence microfilament; S04, judging the flow state existing in the fluorescence microfilament numbered image based on the angle value set; the flow state comprises an adherent flow, a separated flow and a mixed flow, and the mixed flow refers to the existence of the adherent flow and the separated flow at the same time; S05, dividing the boundary between the separated flow and the adherent flow in the fluorescence microfilament numbered image according to the angle polar coordinate distribution diagram and the size of the double-peak coefficient; S06, recording and saving the process of judging the flow state and the boundary division.

2. The method of claim 1, wherein, In the step S01, the fluorescence microfilament image collected in the experiment is sequentially subjected to grayscale processing, binarization processing and noise reduction processing to obtain the fluorescence microfilament pre-processed image.

3. The method of claim 2, wherein, In the step S01, the grayscale processing is performed as follows: based on the RGB three channels of each pixel point in the fluorescence microfilament image collected in the experiment, R value, G value and B value are obtained, and the gray value Gray of the corresponding pixel point is calculated by using the formula Gray = 0.299R + 0.587G + 0.114B; The binarization processing is performed as follows: the local adaptive threshold mean value method is used for binarization processing; The noise reduction processing is performed as follows: after the binarization processing is completed, the Hessian matrix feature vector is used to detect the direction of the fluorescence microfilament, the gray value change is kept smooth along the direction of the fluorescence microfilament from the fixed end to the free end, and the gray value change is kept sharp perpendicular to the direction of the fluorescence microfilament, the noise reduction processing is completed, and the fluorescence microfilament pre-processed image is saved.

4. The method of claim 3, wherein, In the step S01, the binarization processing is performed as follows: after the grayscale processing is completed, the average value mean(x, y) of the gray values of all pixels in the neighborhood of each pixel point is calculated, the local threshold T(x, y) of the pixel point is T(x, y) = mean(x, y)-C, and the gray value Gray(x, y) of the pixel point is compared with T(x, y); if Gray(x, y) ≥ T(x, y), the gray value Gray(x, y) of the pixel point is 255; if Gray(x, y) < T(x, y), the gray value Gray(x, y) of the pixel point is 0; wherein x is the horizontal coordinate of the pixel point, y is the vertical coordinate of the pixel point, mean(x, y) is the average value of the gray values of all pixels in the neighborhood of a single pixel point, T(x, y) is the local threshold of the pixel point, and C is a constant.

5. The method of claim 1, wherein, In the step S02, according to the position coordinates of each fluorescent microfiber, the end point coordinates (x1, y1) of the fixed end and the end point coordinates (x2, y2) of the free end are obtained; the direction vector v of the fluorescent microfiber is calculated by the formula v=(Δx / Δy)=(x2-x1) / (y2-y1); the angle value is calculated using the four-quadrant arctangent function arctan2(Δy, Δx), and the output range is [-180°, 180°]; finally, the angle is normalized to convert to the range of 0°~360°.

6. The method of claim 1, wherein, In the step S04, the specific operation of judging the flow state existing in the fluorescent microfiber number image is as follows: According to the deflection angle of each fluorescent microfiber, the angle polar coordinate histogram corresponding to the fluorescent microfiber is drawn, and the bimodal coefficient of the angle polar coordinate histogram is calculated according to a standard bimodal coefficient formula, so as to identify the mixed flow state coexisting with the separation and the attachment flow. b is a standard bimodal coefficient, n is a sample capacity, g is a sample skewness, k is a sample kurtosis, is a sample mean. Standard bimodality coefficient b The formula for calculating the standard bimodality coefficient is: ； Sample skewness g The formula for calculating the sample skewness is: ； Sample kurtosis k The formula for calculating the sample kurtosis is: ； sample mean The formula for calculating the sample mean is: 。 7. The method of claim 1, wherein, In the step S05, according to the size of the bimodal coefficient of the angle polar coordinate histogram of each fluorescent microfiber, the position range of the mixed flow state is determined, and then the position range of the transition from the attached flow to the separated flow is confirmed, and finally the boundaries of the two flow states are determined.

8. The method of claim 1, wherein, In the step S06, the process of judging the flow state and the boundary division is recorded and saved; wherein the deflection angle polar coordinate histogram of each fluorescent microfiber and the flow state are saved.

Images