Soft material large-deformation high-precision stretch ratio measuring method based on monocular vision

Abstract

The invention provides a soft material large-deformation high-precision stretch ratio measurement method based on monocular vision. The method comprises the following steps: 1) preparing a speckle pattern on the surface of a soft material test piece; 2) fixing the soft material test piece with the prepared speckle pattern on a universal testing machine, applying a tensile load to the soft material test piece through the universal testing machine, and recording an image of the tensile process of the soft material test piece by adopting a single digital camera; and 3) carrying out image processing and data analysis on the image sequence recorded by the digital camera through a data processing device, and calculating to obtain the stretch ratio of the soft material test piece in the stretching process. According to the method, a single digital camera is adopted, and the SURF algorithm and adaptive image processing are combined, so that high-precision and low-cost stretch ratio measurement under the condition of large deformation of the soft material is realized; meanwhile, by adjusting a calculation model, the method is also suitable for strain calculation under a small deformation condition, and has good application flexibility.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photomechanics, and particularly relates to a large deformation high-precision tensile ratio measurement method for soft materials based on monocular vision. BACKGROUND

[0002] With the rapid development of material science and engineering technology, soft materials have become indispensable functional materials in the fields of biomedical engineering, flexible electronics, environmental engineering and energy storage due to their excellent biocompatibility, high water content and unique environmental response characteristics. Specifically, in biomedical engineering, soft materials are widely used to construct bionic tissue engineering scaffolds, effectively promoting the regeneration and repair of damaged tissues such as cartilage or skin; in the field of flexible electronics, conductive soft materials can be used as high-sensitivity strain sensors to monitor real-time joint movement or subtle expression changes; in environmental engineering, functionalized soft materials can efficiently adsorb heavy metal ions or organic pollutants in industrial wastewater, achieving the goal of green purification; in the field of energy storage, soft material electrolytes have become a key component of the new generation of flexible zinc ion batteries due to their good ion conductivity and mechanical flexibility.

[0003] Currently, the strain measurement of soft materials mainly relies on the displacement sensor equipped in the universal testing machine system. This method calculates the stress based on the force sensor reading of the universal testing machine and the cross-sectional area of the test piece, and then combines the data measured by the displacement sensor with the initial distance between the testing machine clamps to obtain the strain, and finally draws the stress-strain curve (see Chinese Patent Publication Nos. CN120078951 A, CN114752080 A, CN120192462 A, CN118702875 A; Xue, B., Gu, J., Li, L. et al., Nature Communication, 2021, 12, 7156; etc.). However, this method is limited by the measurement accuracy of the displacement sensor itself, and the test process involves super-elastic soft materials, which are prone to relative sliding between the clamps, resulting in significant deviations in strain measurement results. Especially under the condition of large deformation of soft materials, the error is more significant.

[0004] In addition, the existing optical extensometer method also has many limitations when measuring large deformation of soft materials. For example, some methods are difficult to achieve effective strain measurement under large deformation conditions (such as CN 118857935 A); some other schemes use a double-camera configuration, resulting in a complex system structure and strict requirements for camera synchronization (such as CN 110763146 A); and schemes based on traditional image algorithms such as gray center-of-gravity method (such as CN 102003945 A) have insufficient robustness and calculation accuracy; strain calculation methods based on digital image correlation method (DIC) have measurement blind areas at the edges of the image and long calculation time (such as Liu, M., Guo, J., Li, Z. et al., Mechanics of Materials, 2019, 138, 103158).

[0005] It is worth noting that the speeded up robust features (SURF) algorithm has been widely used in image stitching (such as CN 118134758 A, CN 111951162 A, CN 111861883 A) and three-dimensional reconstruction (such as CN 103413352 A) and other fields due to its strong robustness, high matching accuracy and other advantages. However, this algorithm has not been directly and effectively introduced into the field of large deformation tensile ratio measurement of soft materials to solve the problem of high-precision, non-contact measurement.

[0006] Therefore, developing a new tensile ratio measurement method that can balance measurement accuracy, system simplicity and large deformation adaptability has become a key problem to be solved in this technical field. SUMMARY

[0007] In view of the technical problems in the prior art that the existing soft material strain measurement methods generally have precision decline due to relative sliding between the test piece and the clamp, complex structure of the double-camera optical system, and insufficient robustness of traditional image algorithms when facing large deformation, the present application provides a soft material large deformation high precision tensile ratio measurement method based on monocular vision, which aims to realize low-cost, high-precision and reliable large deformation tensile ratio measurement by using a single digital camera and an optimized SURF image processing algorithm.

[0008] To solve the above technical problems, the soft material large deformation high precision tensile ratio measurement method based on monocular vision provided by the present application comprises the following steps:

[0009] 1) Prepare a speckle pattern on the surface of a soft material test piece;

[0010] 2) Fix the soft material test piece with the prepared speckle pattern on a universal testing machine, apply tensile load to the soft material test piece by the universal testing machine, and record the image of the tensile process of the soft material test piece by a single digital camera;

[0011] 3) The image sequence recorded by the digital camera is processed and analyzed by the data processing device to calculate the stretch ratio of the soft material specimen during the stretching process.

[0012] As a preferred embodiment of the present invention, step 2) specifically includes the following steps:

[0013] 2.1) Based on the pre-set soft material specimen shape parameters, determine the loading speed of the universal testing machine to be 10 mm / min;

[0014] 2.2) Adjust the position of the digital camera so that its optical axis is perpendicular to the surface of the soft material specimen, and ensure that the speckle pattern remains within the field of view of the digital camera throughout the tensile process;

[0015] 2.3) Adjust the focus and aperture parameters of the digital camera and imaging lens to make the speckle pattern clear and complete in the field of view;

[0016] 2.4) Set the image acquisition frequency of the digital camera according to the loading speed of the universal testing machine.

[0017] As a preferred embodiment of the present invention, step 3) specifically includes the following steps:

[0018] 3.1) Select the first frame image recorded by the digital camera as the reference image, define two regions of interest in the reference image, retain the grayscale information within the region, and set the grayscale value outside the region to zero;

[0019] 3.2) Use the accelerated robust feature algorithm to extract and match feature points in the image sequence. When the number of successfully matched feature point pairs exceeds the preset threshold, record the position coordinates of these feature points. If the number of matched point pairs does not reach the threshold, use the previous frame of the current image as the new reference image and repeat the feature extraction and matching process in this step until a match is successful.

[0020] 3.3) Based on the image feature points extracted and successfully matched in step 3.2), obtain the coordinates of the matching feature points in the reference image and the deformed image respectively, obtain the coordinates of all feature points extracted in the reference image and the deformed image, and calculate the tensile ratio of the soft material specimen during the tensile process based on the different coordinates of all feature points in the reference image and the deformed image.

[0021] As a preferred embodiment of the present invention, the specific process of feature point extraction and matching in step 3.2) is as follows:

[0022] 3.2.1) Establish a Cartesian coordinate system within the image. The sum of pixels within the rectangular region formed between point X and the origin is defined as I. Σ (X);

[0023] 3.2.2) Define the Hessian matrix of the image as:

[0024] ;

[0025] In the above formula, x and y represent the coordinates of each point in the image; σ represents the scale of the Gaussian filter; I(x,y) represents the gray level of each point in the image; * represents the convolution operation; g(x,y,σ) represents the two-dimensional Gaussian function.

[0026] 3.2.3) Feature points in the image are extracted using the approximation of the Hessian matrix determinant. The expression for the approximation of the Hessian matrix determinant is:

[0027] ;

[0028] In the above formula, D xx This represents the convolution result of the box filter in the x-direction with the original image; a point is identified as an extremum only when the determinant of its Hessian matrix exceeds a predefined threshold.

[0029] 3.2.4) Calculate the Haar wavelet responses in the x and y directions for the extracted feature points using the Haar wavelet template, and obtain the feature descriptors of the feature points. ;

[0030] 3.2.5) The similarity between the reference image and the deformed image is calculated using Euclidean distance, and the ratio of the shortest Euclidean distance to the second shortest Euclidean distance is used as the matching criterion. If this ratio is lower than a predefined threshold, the two feature points are considered a valid match. The Euclidean distance calculation formula is:

[0031] .

[0032] In a preferred embodiment of the present invention, the stretching ratio in step 3.3) is calculated according to the following formula:

[0033] ;

[0034] In the above formula, λ is the stretch ratio; Y ui and Y lj This represents the ordinate of the feature points extracted from the two regions of interest in the reference image; y ui and y lj , where are the ordinates of the corresponding points in the deformed image that match the above feature points; m and n are the number of feature points that are successfully matched in the two regions of interest, respectively.

[0035] As a preferred embodiment of the present invention: step 1) is to create a high-contrast random speckle pattern on the surface of a soft material specimen by spraying paint.

[0036] As a preferred embodiment of the present invention: before step 2), the two ends of the soft material specimen are firmly clamped in the upper and lower chucks of the universal testing machine, and then the digital camera with the imaging lens is fixed on the tripod. The position and angle of the digital camera are carefully adjusted to ensure that the optical axis of the imaging lens is perpendicular to the surface of the specimen.

[0037] As a preferred embodiment of the present invention: in step 2), the speckle pattern during the tensile load applied to the soft material specimen (1) by the universal testing machine is located at the center of the horizontal direction of the camera's field of view.

[0038] As a preferred embodiment of the present invention: after calculating the stretch ratio λ in step 3), the data processing device outputs and displays the curve of the calculated stretch ratio λ changing with time or load, thereby intuitively reflecting the mechanical behavior of soft materials during large deformation.

[0039] As a preferred embodiment of the present invention, the measurement method, by replacing the calculation formula with the engineering strain formula under the small deformation assumption, is also applicable to strain calculation of various specimens under small deformation conditions. The strain calculation formula under small deformation is:

[0040] ;

[0041] In the above formula, λ is the stretch ratio; Y ui and Y lj This represents the ordinate of the feature points extracted from the two regions of interest in the reference image; y ui and y lj , where are the ordinates of the corresponding points in the deformed image that match the above feature points; m and n are the number of feature points that are successfully matched in the two regions of interest, respectively.

[0042] By adopting the above technical solution, the present invention has the following beneficial effects:

[0043] This invention uses a single digital camera, combined with the SURF algorithm and adaptive image processing, to achieve high-precision and low-cost tensile ratio measurement of soft materials under large deformation conditions. At the same time, by adjusting the calculation formula, it is also applicable to strain calculation under small deformation conditions, and has good application flexibility.

[0044] The main features and advantages of this invention are reflected in the following aspects:

[0045] (1) The present invention achieves high-precision measurement of large deformation: The present invention takes the stretch ratio as the core measurement target and directly targets the large deformation scenario of soft materials. Through non-contact image measurement, it completely avoids the error introduced by the sliding of the specimen in the traditional method.

[0046] (2) The measuring equipment is simple and inexpensive: the use of a single digital camera greatly simplifies the structure of the measuring system and reduces the requirements for hardware synchronization and overall cost;

[0047] (3) Strong robustness: The SURF algorithm is used for feature matching, and combined with an adaptive reference map update strategy, which effectively ensures the success rate of feature matching and the continuity of measurement during large deformation.

[0048] (4) Wide range of applications: Although this invention is mainly aimed at the measurement of large deformation tensile ratio, its core image acquisition and displacement measurement mechanism is also applicable to small deformation scenarios. Strain measurement can be achieved by selecting engineering strain calculation formula, which shows good application flexibility. Attached Figure Description

[0049] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0050] Figure 1 A schematic diagram illustrating a high-precision tensile ratio measurement method for soft materials under large deformation, provided in an embodiment of the present invention;

[0051] Figure 2 A flowchart of a high-precision tensile ratio measurement method for soft materials under large deformation provided in an embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram illustrating the speckle pattern fabrication principle of a high-precision tensile ratio measurement method for large deformation of soft materials, provided in an embodiment of the present invention.

[0053] Figure 4 This is a schematic diagram of the strain calculation method for a high-precision tensile ratio measurement method for large deformation of soft materials provided in an embodiment of the present invention.

[0054] Figure 5 The speckle pattern produced by the speckle pattern production method provided in the embodiments of the present invention;

[0055] Figure 6 The strain measurement value and the actual strain value comparison curve and the strain measurement error curve are provided for the high-precision tensile ratio measurement method for large deformation of soft materials provided in the embodiments of the present invention.

[0056] Figure 7 The large deformation strain measurement results of polydimethylsiloxane provided in the embodiments of the present invention are applicable to the high-precision tensile ratio measurement method for large deformation of soft materials.

[0057] Figure 8 The results of large deformation strain measurement of hydrogels provided in the embodiments of the present invention are applicable to the high-precision tensile ratio measurement method for large deformation of soft materials.

[0058] In the figure: 1- Soft material specimen; 2- Speckle pattern on the surface of the soft material specimen; 3- Imaging lens; 4- Digital camera; 5- Data processing device; 6- Tripod. Detailed Implementation

[0059] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] The present invention will be further explained below with reference to specific embodiments.

[0061] like Figure 1 As shown in the figure, this embodiment provides a high-precision tensile ratio measurement method for large deformation of soft materials based on monocular vision, which includes the following steps:

[0062] S100. Prepare speckle pattern 2 on the surface of soft material specimen 1.

[0063] S200. Fix the soft material specimen 1 with the prepared speckle pattern 2 on a universal testing machine, apply a tensile load to the soft material specimen 1 through the universal testing machine, and simultaneously record the tensile process of the soft material specimen 1 using a single digital camera 4; specifically including the following steps:

[0064] S201. Based on the pre-set shape parameters of the soft material specimen 1, determine the loading speed of the universal testing machine to be 10 mm / min;

[0065] S202. Adjust the position of the digital camera 4 so that its optical axis is perpendicular to the surface of the soft material specimen 1, and ensure that the speckle pattern 2 is always within the field of view of the digital camera 4 throughout the entire stretching process.

[0066] S203. Adjust the focus and aperture parameters of the digital camera 4 and the imaging lens 3 to make the speckle pattern 2 clear and complete in the field of view.

[0067] S204. Set the image acquisition frequency of digital camera 4 according to the loading speed of the universal testing machine.

[0068] S300, The image sequence recorded by the digital camera 4 is processed and analyzed by the data processing device 5 to calculate the tensile ratio of the soft material specimen 1 during the tensile process; specifically including the following steps:

[0069] S301. Select the first frame image recorded by digital camera 4 as the reference image, define two regions of interest in the reference image, retain the grayscale information within the region, and set the grayscale value outside the region to zero.

[0070] S302. Use the Accelerated Robust Feature (SURF) algorithm to extract and match feature points in the image sequence. When the number of successfully matched feature point pairs exceeds a preset threshold, record the position coordinates of these feature points. If the number of matched point pairs does not reach the threshold, use the previous frame of the current image as the new reference image and repeat the feature extraction and matching process in this step until a match is successful.

[0071] The specific process of feature point extraction and matching in step S302 above is as follows:

[0072] S3021. Establish a Cartesian coordinate system within the image. The sum of pixels within the rectangular region formed between point X and the origin is defined as I. Σ (X);

[0073] S3022. Define the Hessian matrix of the image as follows:

[0074] ;

[0075] In the above formula, x and y represent the coordinates of each point in the image; σ represents the scale of the Gaussian filter; I(x,y) represents the gray level of each point in the image; * represents the convolution operation; g(x,y,σ) represents the two-dimensional Gaussian function; in order to speed up the computation, the SURF algorithm uses a box filter as an approximation of the second-order Gaussian filter;

[0076] S3023. Extract feature points from the image using the approximation of the Hessian matrix determinant. The expression for the approximation of the Hessian matrix determinant is:

[0077] ;

[0078] In the above formula, D xx This represents the convolution result of the box filter in the x-direction with the original image; a point is identified as an extremum only when the determinant of its Hessian matrix exceeds a predefined threshold.

[0079] S3024. Calculate the Haar wavelet responses in the x and y directions for the extracted feature points using the Haar wavelet template, and obtain the feature descriptors of the feature points. ;

[0080] S3025. Calculate the similarity between the reference image and the deformed image using Euclidean distance, and use the ratio of the shortest Euclidean distance to the second shortest Euclidean distance as the matching criterion; if this ratio is lower than a predefined threshold, the two feature points are considered a valid match; the Euclidean distance calculation formula is:

[0081] .

[0082] S303. Based on the image feature points extracted and successfully matched in step S302, obtain the coordinates of the corresponding feature points in the reference image and the deformed image. Obtain the coordinates of all feature points extracted in the reference image and the deformed image, and calculate the tensile ratio of the soft material specimen 1 during the tensile process based on the different coordinates of all feature points in the reference image and the deformed image. The tensile ratio λ is calculated according to the following formula:

[0083] ;

[0084] In the above formula, Y ui and Y lj This represents the ordinate of the feature points extracted from the two regions of interest in the reference image; y ui and y lj , where are the ordinates of the corresponding points in the deformed image that match the above feature points; m and n are the number of feature points that are successfully matched in the two regions of interest, respectively.

[0085] This invention replaces the calculation formula with an engineering strain formula under the assumption of small deformation, making it equally applicable to strain measurement of various specimens (such as metals and plastics) under small deformation conditions. The strain calculation formula for small deformation is as follows:

[0086] ;

[0087] In the above formula, λ is the stretch ratio; Y ui and Y lj This represents the ordinate of the feature points extracted from the two regions of interest in the reference image; y ui and y lj , where are the ordinates of the corresponding points in the deformed image that match the above feature points; m and n are the number of feature points that are successfully matched in the two regions of interest, respectively.

[0088] In this embodiment, the digital camera 4 is fixed by a tripod 6, and its optical axis is precisely adjusted to ensure that it is perpendicular to the surface of the soft material specimen 1 under test, thereby minimizing image distortion and ensuring clear imaging of the speckle pattern 2. The data processing device 5 is responsible for receiving and processing the image sequence recorded by the digital camera 4.

[0089] This embodiment takes the measurement of the tensile ratio of a typical soft material—hydrogel sample as an example. The specific steps are as follows:

[0090] S010, Specimen Preparation and System Setup

[0091] S011. A high-contrast random speckle pattern is created on the surface of the hydrogel sample by spraying paint.

[0092] S012. Securely clamp both ends of the specimen in the upper and lower chucks of the universal testing machine;

[0093] S013. Fix the digital camera 4 with the imaging lens 3 on the tripod 6, carefully adjust its position and angle to ensure that the optical axis of the lens is perpendicular to the sample surface, and that the speckle pattern 2 is located at the center of the field of view of the digital camera 4 throughout the stretching process.

[0094] S014. Adjust the focal length and aperture of the digital camera 4 to make the speckle pattern clear and uniform in brightness in the field of view; set the acquisition frequency (1fps) of the digital camera 4 according to the preset loading speed (10 mm / min) of the universal testing machine to ensure that the deformation process can be fully captured.

[0095] S020, Image Acquisition and Data Processing

[0096] S021. Start the universal testing machine and digital camera 4 to simultaneously perform tensile loading and image recording;

[0097] S022. After reading the image sequence, the data processing device 5 executes the following core algorithm process:

[0098] S0221. Image preprocessing and ROI setting: Set the first recorded frame image as the reference image and define two regions of interest (ROIs) on it; retain the grayscale information within the ROI and set the grayscale of pixels outside the region to zero to improve the efficiency of subsequent processing.

[0099] S0222, Feature Point Matching and Adaptive Reference Image Update: The Accelerated Robust Feature (SURF) algorithm is used to extract and match feature points in the image sequence; a threshold for the number of matching point pairs is set (300 pairs); when the number of successfully matched feature point pairs exceeds this threshold, the program records the position coordinates of these feature points; if the threshold is not reached, the system automatically uses the previous frame of the current image as the new reference image and re-executes the feature matching process; this adaptive mechanism effectively ensures the continuity of measurement under large deformation conditions;

[0100] S0223, Calculation of stretch ratio: Based on the coordinates of the successfully matched feature points, calculate the average displacement difference between the two ROIs in the stretch direction (Y direction); further, the formula for calculating the stretch ratio λ is:

[0101] ;

[0102] Among them, Y ui and Ylj This represents the ordinate of the feature points extracted from the two regions of interest in the reference image; y ui and y lj , where are the ordinates of the corresponding points in the deformed image that match the above feature points; m and n are the number of feature points that are successfully matched in the two regions of interest, respectively.

[0103] S030. Results Output and Analysis

[0104] The data processing device 5 outputs and displays the calculated stretch ratio λ as a function of time or load, thus intuitively reflecting the mechanical behavior of soft materials during large deformation processes.

[0105] Verification of the beneficial effects of the embodiments of the present invention:

[0106] Compared with existing technologies, this implementation method exhibits significant advantages:

[0107] Successful implementation of large deformation measurement: This method successfully solves the problems of excessive error caused by specimen slippage in traditional contact measurement of soft materials under large deformation conditions, as well as the insufficient robustness of traditional image algorithms (such as gray-scale centroid method).

[0108] Excellent accuracy and efficiency: The SURF algorithm is used for feature matching, which has a much higher accuracy than the gray-scale centroid method and a faster calculation speed than the complex digital image correlation (DIC) method. Testing shows that this method achieves an accuracy of 0.001 pixels in displacement measurement and a strain measurement error of less than 22 με, providing a reliable foundation for high-precision tensile ratio calculation.

[0109] Low system cost and complexity: Only a single camera is required, eliminating the need for a complex dual-camera synchronization system, which greatly reduces implementation costs and technical barriers.

[0110] In summary, this invention provides a practical and feasible high-precision method for measuring the tensile ratio of soft materials under large deformation, which has good prospects for industrial application.

[0111] This invention is simple to operate, low in cost, and highly accurate, making it particularly suitable for large deformation testing of soft materials. It uses a single digital camera, combined with the SURF algorithm and adaptive image processing, to achieve high-precision, low-cost tensile ratio measurement of soft materials under large deformation conditions. At the same time, by adjusting the calculation model, it is also suitable for strain measurement under small deformation conditions, demonstrating good application flexibility.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for high-precision measurement of the tensile ratio of soft materials under large deformation based on monocular vision, characterized in that, Includes the following steps: 1) Prepare speckle patterns (2) on the surface of the soft material specimen (1); 2) Fix the soft material specimen (1) with the prepared speckle pattern (2) on the universal testing machine, apply tensile load to the soft material specimen (1) through the universal testing machine, and at the same time use a single digital camera (4) to record the tensile process of the soft material specimen (1). 3) The image sequence recorded by the digital camera (4) is processed and analyzed by the data processing device (5) to calculate the stretching ratio of the soft material specimen (1) during the stretching process.

2. The method for high-precision tensile ratio measurement of soft materials with large deformation based on monocular vision as described in claim 1, characterized in that, Step 2) specifically includes the following steps: 2.1) Based on the pre-set shape parameters of the soft material specimen (1), the loading speed of the universal testing machine is determined to be 10 mm / min; 2.2) Adjust the position of the digital camera (4) so ​​that its optical axis is perpendicular to the surface of the soft material specimen (1) and ensure that the speckle pattern (2) is always within the field of view of the digital camera (4) throughout the stretching process; 2.3) Adjust the focus and aperture parameters of the digital camera (4) and the imaging lens (3) so that the speckle pattern (2) is clear and complete in the field of view; 2.4) Set the image acquisition frequency of the digital camera (4) according to the loading speed of the universal testing machine.

3. The method for high-precision tensile ratio measurement of soft materials with large deformation based on monocular vision as described in claim 1, characterized in that, Step 3) specifically includes the following steps: 3.1) Select the first frame image recorded by the digital camera (4) as the reference image, define two regions of interest in the reference image, retain the grayscale information in the region, and set the grayscale value outside the region to zero; 3.2) Use the accelerated robust feature algorithm to extract and match feature points in the image sequence. When the number of successfully matched feature point pairs exceeds the preset threshold, record the position coordinates of these feature points. If the number of matched point pairs does not reach the threshold, use the previous frame of the current image as the new reference image and repeat the feature extraction and matching process in this step until a match is successful. 3.3) Based on the image feature points extracted and matched in step 3.2), the coordinates of the matching feature points in the reference image and the deformed image are obtained respectively. The coordinates of all feature points extracted in the reference image and the deformed image are obtained. Based on the different coordinates of all feature points in the reference image and the deformed image, the tensile ratio of the soft material specimen (1) during the tensile process is calculated.

4. The method for high-precision tensile ratio measurement of soft materials with large deformation based on monocular vision as described in claim 3, characterized in that, The specific process of feature point extraction and matching in step 3.2) is as follows: 3.2.1) Establish a Cartesian coordinate system within the image. The sum of pixels within the rectangular region formed between point X and the origin is defined as I. Σ (X); 3.2.2) Define the Hessian matrix of the image as: ； In the above formula, x and y represent the coordinates of each point in the image; σ represents the scale of the Gaussian filter; I(x,y) represents the gray level of each point in the image; * represents the convolution operation; g(x,y,σ) represents the two-dimensional Gaussian function. 3.2.3) Feature points in the image are extracted using the approximation of the Hessian matrix determinant. The expression for the approximation of the Hessian matrix determinant is: ； In the above formula, D xx This represents the convolution result of the box filter in the x-direction with the original image; a point is identified as an extremum only when the determinant of its Hessian matrix exceeds a predefined threshold. 3.2.4) Calculate the Haar wavelet responses in the x and y directions for the extracted feature points using the Haar wavelet template, and obtain the feature descriptors of the feature points. ; 3.2.5) The similarity between the reference image and the deformed image is calculated using Euclidean distance, and the ratio of the shortest Euclidean distance to the second shortest Euclidean distance is used as the matching criterion. If this ratio is lower than a predefined threshold, the two feature points are considered a valid match. The Euclidean distance calculation formula is: 。 5. The method for high-precision tensile ratio measurement of large deformation of soft materials based on monocular vision as described in claim 3, characterized in that: The stretching ratio in step 3.3) is calculated according to the following formula: ； In the above formula, λ is the stretch ratio; Y ui and Y lj This represents the ordinate of the feature points extracted from the two regions of interest in the reference image; y ui and y lj , where are the ordinates of the corresponding points in the deformed image that match the above feature points; m and n are the number of feature points that are successfully matched in the two regions of interest, respectively.

6. The method for high-precision tensile ratio measurement of large deformation of soft materials based on monocular vision as described in claim 1, characterized in that: Step 1) involves creating a high-contrast random speckle pattern (2) on the surface of a soft material specimen (1) using a spray painting method.

7. The method for high-precision tensile ratio measurement of soft materials with large deformation based on monocular vision as described in claim 1, characterized in that: Before step 2), the soft material specimen (1) is firmly clamped at both ends in the upper and lower chucks of the universal testing machine. Then, the digital camera (4) with the imaging lens (3) is fixed on the tripod (6). The position and angle of the digital camera (4) are carefully adjusted to ensure that the optical axis of the imaging lens (3) is perpendicular to the surface of the specimen.

8. The method for high-precision tensile ratio measurement of soft materials with large deformation based on monocular vision as described in claim 1, characterized in that: In step 2), the speckle pattern (2) during the tensile load applied to the soft material specimen (1) by the universal testing machine is located at the center of the horizontal direction of the camera's field of view.

9. The method for high-precision tensile ratio measurement of soft materials with large deformation based on monocular vision as described in claim 1, characterized in that: After calculating the stretch ratio λ in step 3), the data processing device (5) outputs and displays the curve of the calculated stretch ratio λ changing with time or load, thereby intuitively reflecting the mechanical behavior of soft materials during large deformation.

10. The method for high-precision tensile ratio measurement of large deformation of soft materials based on monocular vision as described in claim 1, characterized in that: The measurement method, by replacing the calculation formula with the engineering strain formula under the small deformation assumption, is also applicable to strain calculation of various specimens under small deformation conditions. The strain calculation formula for small deformation is: ； In the above formula, λ is the stretch ratio; Y ui and Y lj This represents the ordinate of the feature points extracted from the two regions of interest in the reference image; y ui and y lj , where are the ordinates of the corresponding points in the deformed image that match the above feature points; m and n are the number of feature points that are successfully matched in the two regions of interest, respectively.

Images