Digital image correlation subset adaptive algorithm for cross-scale measurement

Abstract

The invention discloses a digital image related subset adaptive algorithm for cross-scale measurement, which comprises the following steps of: firstly, intercepting sub-region windows from small to large at the central position of a speckle image; and carrying out binaryzation on the intercepted sub-area window image, and projecting a pixel point of which the pixel value is 1 after binaryzation to an X axis and a Y axis. According to a calculation rule, calculating the variance of the projection of the pixel point on the X axis and the Y axis, and calculating the difference value of the two variances; and finding a position where the variance difference value has a minimum value and tends to zero for the first time, and adaptively outputting a subset of the position. According to the method, when a DIC system measures a test sample in a cross-scale manner and processes speckle images with different characteristics under different scales, subsets with proper sizes are selected in a self-adaptive manner, the influence of human factors is eliminated, and the calculation precision of a displacement field is ensured.

Description

Technical Field

[0001] The invention belongs to the technical field of digital image correlation subset adaptive algorithms, and in particular relates to a digital image correlation subset adaptive algorithm for cross-scale measurement. Background Art

[0002] Digital image correlation (DIC) is an advanced non-contact, non-destructive image detection technology that can efficiently and accurately capture and detect the movement, deformation, destruction and damage of objects or structures. The key to its technology is that it is necessary to capture the optical influence characteristics of the surface of the object being analyzed, and then use the relevant image processing algorithm for analysis and processing. When using DIC technology to measure the deformation field, the size of the subset is an important parameter, which will affect the calculation precision and accuracy of the displacement field. If the subset is too large, some local detail information will be lost, thereby reducing the matching accuracy, increasing the amount of calculation, and causing the matching algorithm to run slower. In the case of severe deformation, a subset that is too large may not accurately capture the deformation information of local details; if the subset is too small, the matching accuracy is insufficient and the pixel-level displacement cannot be accurately captured. In addition, a small subset may be affected by noise, thereby introducing uncertainty. Therefore, in DIC, selecting a subset of appropriate size is crucial to obtaining an accurate displacement field. In DIC cross-scale measurement, the measurement system needs to process speckle images with different characteristics at different scales. If the subset size is set manually, the accuracy cannot be guaranteed and is greatly affected by human factors.

[0003] Therefore, it is necessary to provide a subset adaptive algorithm for digital image correlation speckle images, which can adaptively provide appropriate subset sizes for DIC data processing of different types of speckle images, ensure the calculation accuracy of the displacement field, and improve the controllability and implementation efficiency of the experiment. Summary of the invention

[0004] In view of the above problems, the present invention provides a cross-scale measurement digital image correlation subset adaptive algorithm, which can adaptively provide a suitable subset size for different types of speckle images to ensure the calculation accuracy of the displacement field.

[0005] In order to achieve the above object, the present invention mainly includes the following core contents:

[0006] Step 1. Cross-scale subset adaptive selection mechanism: This algorithm first obtains the speckle image of the test sample through a CCD camera. During the image processing, the algorithm will successively intercept sub-area windows from the center of the image from small to large according to the preset size range. Each sub-area window represents a specific part of the image for subsequent analysis and comparison;

[0007] Step 2. Binarization and projection analysis: Then, the algorithm performs binarization on all intercepted sub-area windows, simplifying the pixel values ​​into two states: 0 and 1. Then, the pixel points with a binarized pixel value of 1 are projected onto the X-axis and Y-axis respectively, and subsequent statistical analysis is performed based on this.

[0008] Step 3. Calculate and compare variance differences: The algorithm calculates the variances of all pixels that are 1 on the X-axis and Y-axis projections, and calculates the difference between the two variances of the same subset on the X-axis and Y-axis. Then, compare the variance differences of subsets of different sizes to determine which subset is most suitable for subsequent measurement and analysis;

[0009] Step 4. Curve drawing and optimal subset determination: In order to more intuitively display the trend of variance difference, the algorithm sets a coordinate axis, takes the subset (pixel size) as the independent variable, takes the X-axis variance, Y-axis variance and variance difference as the dependent variable, and uses MATLAB and other tools to draw pixel line graphs of the three curves. When the variance difference of the speckle image subset first approaches a minimum value of zero, the subset at this position is output. The subset at this position is defined as the most appropriate subset size;

[0010] Step 5. Cross-scale measurement and precision improvement: This algorithm is particularly suitable for cross-scale test sample speckle image measurement. A two-dimensional DIC (digital image correlation) system can be used to accurately measure speckle images of different scales. In addition, by binarizing, projecting and performing variance analysis on the intercepted sub-area images, and combining the minimum value of the variance difference approaching zero as the evaluation criterion, this algorithm can automatically find the optimal subset size of different speckles, thereby ensuring the calculation accuracy of the displacement field and improving the controllability and implementation efficiency of the test;

[0011] Step 6. Adaptability and elimination of human factors: The core advantage of this adaptive algorithm is that it can adaptively select the optimal subset for different speckle images, thereby eliminating the influence of artificially set subset size on measurement accuracy. Whether in laboratory environments or in practical applications, this algorithm can provide stable and reliable measurement results.

[0012] The digital image correlation subset adaptive algorithm described above further illustrates S3, and the average value of the projection of the pixel point on the X-axis and the Y-axis is calculated as follows:

[0013]

[0014] The digital image correlation subset adaptive algorithm described above further illustrates S3, the variance calculation formula of the pixel point on the X-axis and Y-axis with the average value is:

[0015]

[0016] The digital image correlation subset adaptive algorithm described above further illustrates that by comparing the variance difference between the X-axis and the Y-axis, the validity of the projection of this point can be obtained, and the stable fluctuation size of the projection can be obtained.

[0017] The digital image related subset adaptive algorithm described above further illustrates that when the difference is an extreme value and approaches zero, the point is taken and the subset at the position is defined as the most appropriate subset size.

[0018] Advantages of the present invention

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] 1. Strong adaptability: The present invention can adaptively select the optimal subset size according to different speckle images without manual intervention or preset parameters, which greatly improves the flexibility and adaptability of measurement;

[0021] 2. High measurement accuracy: By accurately calculating the variance and variance difference of each subset and selecting the minimum value where the variance difference tends to zero as the optimal subset, the algorithm can ensure the calculation accuracy of the displacement field, thereby providing high-quality measurement results;

[0022] 3. Cross-scale measurement capability: The present invention is applicable to the measurement of speckle images of test samples of different scales, which can be achieved using a single DIC system without the need to replace equipment or adjust parameters, thus simplifying the measurement process and improving measurement efficiency;

[0023] 4. Reduce the impact of human factors: Since the algorithm can automatically select the optimal subset, the impact of human factors on measurement accuracy is eliminated, making the measurement results more objective and reliable;

[0024] 5. Visual analysis: By drawing a pixel line graph containing X-axis variance, Y-axis variance and variance difference, the algorithm provides an intuitive visual analysis tool, making it easier for users to understand and analyze measurement results;

[0025] 6. Improve the controllability and efficiency of the test: The algorithm automatically finds the optimal subset to ensure the controllability of the test and reduce the complexity and uncertainty of manual operation. At the same time, the efficiency of the algorithm also improves the efficiency of the test and shortens the measurement cycle. This algorithm can adaptively provide suitable subsets for different types of speckle images for subsequent displacement field calculations, ensuring the calculation accuracy of the displacement field and improving the controllability and efficiency of the test.

[0026] In summary, the digital image correlation subset adaptive algorithm for cross-scale measurement has the advantages of strong adaptability, high measurement accuracy, cross-scale measurement capability, reduction of human factors, visual analysis, and improved experimental controllability and implementation efficiency, providing an efficient and accurate measurement tool for the field of digital image processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0028] Attached Figure 1 The present invention is a flowchart of a cross-scale measurement digital image correlation subset adaptive algorithm.

[0029] Attached Figure 2 This is a schematic diagram of a DIC cross-scale measurement system of a digital image correlation subset adaptive algorithm for cross-scale measurement in the present invention.

[0030] Attached Figure 3 This is an algorithm principle diagram of a digital image correlation subset adaptive algorithm for cross-scale measurement of the present invention.

[0031] Attached Figure 4 , 5 6 is an output result diagram of a cross-scale measurement digital image correlation subset adaptive algorithm of the present invention when processing different types of speckle images.

[0032] Attached Figure 7 The binary image of the optimal subset when the cross-scale measurement digital image correlation subset adaptive algorithm processes different speckle images in the present invention DETAILED DESCRIPTION

[0033] In order to make the principles, features and advantages of the present invention more clearly understood, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. It should be emphasized that the specific embodiments described here are only used to explain the present invention and are not used to limit the present invention.

[0034] like Figure 1 As shown, a flowchart of a digital image correlation subset adaptive algorithm for cross-scale measurement provided by the present invention. The algorithm first intercepts sub-area windows from small to large at the center position of the captured speckle image, and then binarizes the sub-area windows. The pixel points with a pixel value of 1 after binarization are projected to the X-axis and Y-axis respectively, the variance of the projection of the point with a pixel value of 1 on the x-axis and y-axis is calculated, and the difference between the two variances is compared. If the difference between the two variances appears to be a minimum value for the first time and tends to zero, the subset at this position is output.

[0035] like Figure 2As shown, a schematic diagram of a DIC cross-scale measurement system of a digital image correlation subset adaptive algorithm for cross-scale measurement provided by the present invention is shown. The system can obtain speckle images with different characteristics for measurement objects at different distances and different scales.

[0036] like Figure 3 As shown, an algorithm principle diagram of a digital image correlation subset adaptive algorithm for cross-scale measurement provided by the present invention further explains the principle of the present invention.

[0037] like Figure 4 , 5 6, the output result diagram of a cross-scale measurement digital image correlation subset adaptive algorithm provided by the present invention when processing 6 groups of speckle images of different types. When the variance difference value first reaches a minimum value and tends to zero, the subset size at that position is output.

[0038] like Figure 7 As shown, a binary image of the optimal subset when a cross-scale measurement digital image correlation subset adaptive algorithm provided by the present invention processes different speckle images. For different types of speckles, the present invention can output subsets of different sizes.

[0039] Taking 6 groups of speckle images with different types as an example, the specific workflow of a digital image correlation subset adaptive algorithm for cross-scale measurement described in the present invention includes the following steps:

[0040] (1) Image acquisition and preprocessing

[0041] Use a CCD camera to obtain 6 groups of speckle images with different types. Perform necessary preprocessing on the acquired images, such as denoising and enhancement, to improve the accuracy of subsequent analysis;

[0042] (2) Sub-area window capture

[0043] The variation range of the subset window is set to 11 to 101, and the sub-area windows are successively intercepted from small to large in the central area of ​​the preprocessed speckle image according to the set size range;

[0044] (3) Binarization and projection analysis

[0045] Binarize the sub-area windows of different sizes obtained by interception, and convert the pixel values ​​into two states: 0 and 1. Project the pixel points with a pixel value of 1 after binarization to the X-axis and Y-axis respectively, and record the pixel distribution of each subset on the X-axis and Y-axis;

[0046] (4) Variance difference calculation

[0047] Calculate the variance of all pixels that are 1 on the X-axis and Y-axis projections. Calculate the difference between the two variances of the same subset on the X-axis and Y-axis to form a variance difference data set;

[0048] (5) Optimal subset selection

[0049] Compare the variance differences of subsets of different sizes and find the minimum point where the variance difference approaches zero for the first time. Output the subset corresponding to the minimum point as the optimal subset. The pixel size of this subset is the most appropriate subset size for measuring the speckle image. The subset sizes obtained by processing 6 groups of different types of speckle images are 19x19, 73x73, 65x65, 19x19, 25x25, and 21x21 respectively.

[0050] (6) Visual analysis and verification

[0051] Set a coordinate axis, take the subset (pixel size) as the independent variable, and take the X-axis variance, Y-axis variance, and variance difference as the dependent variable, and draw a pixel line graph of the three curves. By analyzing the image, verify the accuracy of the optimal subset selection.

[0052] The digital image correlation subset adaptive algorithm for cross-scale measurement of the present invention can be applied in the cross-scale measurement of the digital image correlation method.

[0053] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the principles and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A cross-scale measurement digital image correlation subset adaptive algorithm, characterized in that: The steps include: Step (1) in the speckle images of the test sample at different scales obtained by the CCD camera, starting from the central area of ​​the image, successively intercepting sub-area windows from small to large; Step (2) binarizes all the intercepted sub-area windows, and then projects the pixel points with a binarized pixel value of 1 onto the X-axis and Y-axis; Step (3) Calculate the variances of all pixels that are 1 on the X-axis and Y-axis projections, and calculate the difference between the two variances of the same subset on the X-axis and Y-axis, and compare the variance differences of subsets of different sizes; Step (4) Arrange and compare the variance differences of different sizes on the X-axis and Y-axis, set a coordinate axis, take the subset (pixel size) as the independent variable, take the X-axis variance, Y-axis variance and variance difference as the dependent variable, and use MATLAB to make a pixel line graph of the three curves; Step (5) The variation trend of the variance difference can be seen from the image. When the variance difference of the speckle image approaches zero for the first time, the subset at that position is output. The pixel size of the subset at that position is the most appropriate subset size when measuring the speckle image.

2. The digital image correlation subset adaptive algorithm for cross-scale measurement according to claim 1, characterized in that: A 2D DIC system can be used to measure the speckle image of an experimental sample across scales, such as the one shown in FIG2 .

3. The cross-scale measurement digital image correlation subset adaptive algorithm according to claim 1, characterized in that: The intercepted sub-area image is binarized and then projected onto the X-axis and Y-axis. The variances and variance differences are calculated. The minimum value of the variance difference approaching zero is used as the judgment criterion to find the optimal subset size of different speckles. This can ensure the calculation accuracy of the displacement field and improve the controllability and implementation efficiency of the experiment.

4. The digital image correlation subset adaptive algorithm for cross-scale measurement according to claim 1, characterized in that: The adaptive algorithm can adaptively select the optimal subset for different speckle images and eliminate the influence of human factors on the measurement accuracy.

Images