A fracture image-assisted analysis software performance testing method and system

Abstract

The present invention provides a method and system for testing the performance of fracture image-assisted analysis software, and relates to the technical field of software testing. The method includes: obtaining fracture image test data; enhancing the fracture image test data to obtain fracture image enhancement data; determining the recognition accuracy and processing speed of the fracture image-assisted analysis software to be tested for the labeled information; extracting the region of interest of each fracture image in the fracture image enhancement data by combining the moment of inertia matrix and the rotating rectangular frame to obtain a set of regions of interest; inputting it into the software for repeated testing to obtain the secondary recognition accuracy; when the secondary recognition accuracy is greater than the preset secondary recognition accuracy, determining that the performance of the fracture image-assisted analysis software to be tested meets the standard, otherwise, outputting the abnormal category corresponding to the labeled information; determining that the performance of the fracture image-assisted analysis software to be tested does not meet the standard, and outputting the abnormal category. Improve the test coverage, test speed and test accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of software testing, and in particular to a method and system for testing the performance of fracture image-assisted analysis software. Background Art

[0002] Fracture image-assisted analysis software is an algorithmic tool specifically used to automatically process and analyze fracture areas in medical images (such as X-ray or CT images). Its functions include detection, classification and segmentation of fracture areas, aiming to assist physicians in diagnosis and evaluation and improve the efficiency and accuracy of fracture diagnosis.

[0003] The performance test of fracture image-assisted analysis software can verify whether the software performs stably, accurately and efficiently in the actual clinical environment. Through performance testing, it can be ensured that the software can stably provide high-quality results under a variety of different imaging conditions, patient characteristics and fracture types, ensuring its reliability and safety in medical diagnosis.

[0004] However, in the performance testing process of existing fracture image-assisted analysis software, it is difficult to obtain high-quality medical data for testing, the test coverage is insufficient during the testing process, and for important large-scale repetitive testing processes, the current testing process does not consider the impact of resolution and quality differences caused by different image acquisition devices. The injection of too much redundant data in the test data leads to too much computing resources being used to process redundant data during the direct repetitive testing process. The dispersion of computing resources affects the accuracy of the test, and leads to a long repetitive testing cycle and high test complexity, making it impossible to reliably test the software in a short period of time. Summary of the invention

[0005] In order to solve the technical problems that in the prior art, during the performance testing process of fracture image-assisted analysis software, it is difficult to obtain high-quality medical data for testing, the test coverage is insufficient during the testing process, and for a large number of important repetitive tests, the current testing process does not take into account the impact of resolution and quality differences caused by different image acquisition devices, and the injection of too much redundant data in the test data leads to too much computing resources being used to process redundant data during the direct repetitive test, and the dispersion of computing resources affects the accuracy of the test, and leads to too long a repetitive test cycle, high test complexity, and inability to reliably test the software in a short time, the present invention provides a fracture image-assisted analysis software performance testing method and system.

[0006] The technical solution provided by the embodiment of the present invention is as follows:

[0007] First aspect

[0008] An embodiment of the present invention provides a method for testing the performance of fracture image-assisted analysis software, comprising:

[0009] S1: Acquire fracture image test data with different annotation information, wherein the annotation information includes the fracture area contour segmentation line, the fracture area location and the fracture category;

[0010] S2: Enhance the fracture image test data by using a Laplace pyramid with an adaptive enhancement coefficient to obtain fracture image enhanced data;

[0011] S3: inputting the fracture image enhancement data into the fracture image auxiliary analysis software to be tested according to the annotated information category, and determining the recognition accuracy and processing speed of the fracture image auxiliary analysis software to be tested for the annotated information;

[0012] S4: if the recognition accuracy is greater than the preset recognition accuracy and the processing speed is greater than the preset processing speed, proceed to step S5; otherwise, output the abnormality category corresponding to the annotation information and proceed to step S8;

[0013] S5: extracting the region of interest of each fracture image in the fracture image enhancement data by combining the moment of inertia matrix and the rotating rectangular frame, and obtaining a set of regions of interest with annotation information;

[0014] S6: inputting the set of regions of interest into the image-assisted analysis software for the fracture to be tested according to the categories of the annotated information, performing repeated tests for a preset number of times, and obtaining the secondary recognition accuracy of the image-assisted analysis software for the fracture to be tested on the annotated information;

[0015] S7: if the secondary recognition accuracy is greater than the preset secondary recognition accuracy, it is determined that the performance of the fracture image-assisted analysis software to be tested meets the standard; otherwise, the abnormal category corresponding to the marked information is output, and the process proceeds to step S8;

[0016] S8: Determine that the performance of the fracture image-assisted analysis software to be tested does not meet the standards and output an abnormal category.

[0017] Second aspect

[0018] An embodiment of the present invention provides a fracture image-assisted analysis software performance testing system, comprising:

[0019] processor;

[0020] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method for testing the performance of fracture image-assisted analysis software as described in the first aspect is implemented.

[0021] The third aspect

[0022] An embodiment of the present invention provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method for testing the performance of fracture image-assisted analysis software as described in the first aspect is implemented.

[0023] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0024] In the present invention, the fracture image test data is enhanced by a Laplace pyramid with an adaptive enhancement coefficient to obtain fracture image enhancement data. The calculation method of the adaptive enhancement coefficient can realize intelligent regulation of details, edges and noise, has good balance and adaptability, is particularly suitable for enhancement processing in images with rich details or containing noise, can flexibly adjust the enhancement effect according to the local complexity and noise level of the image, ensure the overall quality of the image and the naturalness of the local details, complete the effective expansion of the test data, and then increase the test coverage breadth and test accuracy. In addition, the fracture image enhancement data is used for a test again. After the processing speed meets the requirements, the recognition accuracy of the software is tested for a second time with a large number of repetitive tests to further test the recognition effect of the software under different working environments. In the secondary test process, the inertia moment matrix and the rotating rectangular frame are used to extract the region of interest of each fracture image in the fracture image enhancement data, which is equivalent to performing secondary data enhancement on the test data. The secondary data test based on this data effectively removes a large amount of irrelevant redundant data, puts more computing resources in the recognition accuracy test process to verify the recognition effect of the software under high usage, effectively reduces the software test cycle, and can complete the accuracy test and reliability test of the software in a short time. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0026] Figure 1 A schematic diagram of a process for testing the performance of fracture image-assisted analysis software provided by an embodiment of the present invention;

[0027] Figure 2 A structural schematic diagram of a fracture image-assisted analysis software performance testing system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0029] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0030] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same. "of", "corresponding, relevant" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same.

[0031] In the embodiments of the present invention, sometimes a subscript such as W1 may be mistakenly written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are consistent.

[0032] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0033] Reference Manual Attached Figure 1 , showing a flow chart of a method for testing the performance of fracture image-assisted analysis software provided in an embodiment of the present invention.

[0034] The embodiment of the present invention provides a method for testing the performance of fracture image-assisted analysis software. The method can be implemented by a fracture image-assisted analysis software performance testing device, and the fracture image-assisted analysis software performance testing device can be a terminal or a server. The processing flow of the fracture image-assisted analysis software performance testing method can include the following steps:

[0035] S1: Obtain fracture image test data with different annotation information.

[0036] The annotation information includes the fracture area contour segmentation line, the fracture area location and the fracture category.

[0037] Among them, annotation information refers to the detailed marking of each fracture image, providing important information about the fracture area, including its shape, location and type. This information provides a reference standard for subsequent software recognition. The fracture area contour segmentation line refers to the boundary line of the fracture area, which identifies the precise range of the fracture. The fracture area position refers to the specific location coordinates of the fracture in the image, which identifies the exact location of the fracture area on the image. Fracture category refers to the type classification of fractures, such as simple fractures, comminuted fractures, etc. Collect image data containing fracture areas and provide detailed annotation information for each image. This information includes the contour segmentation, location and category of the fracture area, so that the analysis software can be accurately tested and evaluated in subsequent steps. This ensures the diversity of test data and helps verify the performance of the software when processing different types of fracture images.

[0038] Among them, the fracture image test data contains multiple fracture images, and each fracture image has corresponding annotation information.

[0039] In a possible implementation manner, after S1, the method further includes:

[0040] Data desensitization based on character masking is performed on fracture image test data.

[0041] Among them, character masking is a method of data desensitization, which is mainly used to remove or hide sensitive information in image data, such as patient names, ID numbers, medical records, etc., by covering relevant text in the image or replacing the original characters with unrecognizable symbols to protect privacy. The main purpose of data desensitization is to protect the privacy of patients or users and avoid leaking personal sensitive information in the test set during the test process, so as to remove patient identity information and medical information, while ensuring the integrity and security of the data.

[0042] S2: Enhance the fracture image test data through the Laplace pyramid with adaptive enhancement coefficient to obtain fracture image enhanced data.

[0043] Among them, the adaptive enhancement coefficient is a parameter that is dynamically adjusted according to the local characteristics of the image (such as noise, details, texture, etc.). It flexibly changes the intensity of image enhancement according to the complexity or noise level of different areas in the image to avoid excessive detail enhancement or noise amplification and ensure the balance of image processing. Laplacian pyramid is a multi-resolution image processing technology that retains high-frequency details of the image (such as edges and textures) by constructing image hierarchies of different resolutions. It generates an image structure that retains details by performing differential calculations on each layer of the Gaussian pyramid, and is used for image enhancement, compression or restoration. The adaptive enhancement coefficient is combined with the Laplacian pyramid technology to perform multi-resolution image enhancement processing on the fracture image data, ensuring that details are retained and noise is suppressed at different resolutions, and finally generating fracture image enhancement data that is more suitable for software analysis.

[0044] It should be noted that due to the different fracture image acquisition devices, the main difference in the fracture images collected is the difference in acquisition resolution. Therefore, in the enhancement process, the image details of the original image should be maintained as much as possible while adjusting the resolution to complete the data enhancement. Using the Laplace pyramid to enhance its test set can obtain more accurate test results to determine whether the fracture image auxiliary analysis software to be tested has sufficient auxiliary analysis capabilities for different fracture images. Fracture image test sets are difficult to obtain, and enhancing the test data can increase the accuracy of subsequent tests. This adaptive enhancement coefficient calculation method can realize intelligent control of details, edges and noise, has good balance and adaptability, and is particularly suitable for enhancement processing in images with rich details or containing noise. It can flexibly adjust the enhancement effect according to the local complexity and noise level of the image to ensure the overall quality of the image and the naturalness of local details.

[0045] In a possible implementation manner, the adaptive enhancement coefficient is specifically:

[0046]

[0047] Among them, α(x,y) represents the adaptive enhancement coefficient at the image pixel (x,y), represents the local variance at the image pixel (x, y), η 2 (x,y) represents the dynamic noise variance at the image pixel (x,y), ò(x,y) represents the enhancement intensity adjustment parameter at the image pixel (x,y), ò min =10 -3 represents a constant used to control the lower limit of the enhancement strength adjustment parameter, H(x,y) represents the local entropy value at the image pixel (x,y), log represents the logarithmic function, P i (x, y) represents the probability that the gray value of the image pixel (x, y) in the eight-neighborhood is i, and L represents the number of gray levels, that is, the number of gray value categories.

[0048] It should be noted that in image processing, the enhancement strength adjustment parameter is mainly used to control the strength of the enhancement to avoid instability caused by a too small denominator in the calculation. The key goal is to ensure that the enhancement operation can be properly adjusted in different areas (such as edges, details, and noise parts), rather than a simple constant or rough local statistic.

[0049] In actual application, the local variance in the formula reflects the grayscale change near the pixel (x, y). When the local variance is large, it means that the area contains more edge and detail information, and the enhancement coefficient will be relatively large, thereby retaining more details. When the local variance is small, it means that the area is relatively smooth, and the enhancement coefficient will be reduced, suppressing excessive enhancement of the smooth area and avoiding noise amplification.

[0050] The noise variance dynamically estimates the noise level in the image. In areas with greater noise, the enhancement coefficient will be reduced to prevent the noise from being further enhanced and reducing the image quality.

[0051] Local entropy reflects the complexity of the local area. The enhancement coefficient of high entropy areas (such as edges and parts with rich details) will be adjusted according to the entropy value, allowing more details to be enhanced. Low entropy areas (such as smooth areas) will reduce the enhancement to avoid artifacts or unnatural effects caused by over-enhancement. The entropy value provides a more accurate measure of image complexity, making the enhancement process more intelligent. The entropy-based enhancement intensity adjustment parameter ò(x,y) controls the intensity of the enhancement. The larger the entropy value, the higher the image complexity, and the enhancement coefficient will be adaptively reduced to keep the enhancement stable. In places with smaller entropy values, the enhancement coefficient will be larger and the enhancement effect will be more significant. This dynamic adjustment mechanism ensures that the enhancement coefficient is flexibly adjusted according to the local complexity of the image. Through the dynamic adjustment of ò(x,y), the denominator will not be over-enhanced due to local extreme conditions, avoiding over-enhancement or artifacts in some areas of the image. At the same time, the adjustment of ò(x,y) keeps the entire enhancement process computationally stable, ensuring that the global visual effect of the image is not disturbed by local details. Through multi-level control of local variance, noise estimation and local entropy, this method can better balance detail preservation, noise suppression and computational stability.

[0052] It should be noted that this fine control enables image enhancement to not only improve the clarity of edges and details, but also effectively suppress excessive noise amplification, and can adapt to different types of images and image areas of different complexity. The enhancement effect is more flexible and natural. This adaptive enhancement coefficient calculation method based on local statistical characteristics can process different areas in the image. Compared with global enhancement, it avoids using the same enhancement strength in different areas, has stronger adaptability, and can cope with complex image scenes.

[0053] In a possible implementation, S2 specifically includes:

[0054] S201: Smooth the fracture image test data through Gaussian filtering to obtain the base layer of the Gaussian pyramid:

[0055] G0(x,y)=I0(x,y)*Gaussian kernel

[0056] Among them, G0(x,y) represents the base layer, that is, the image after Gaussian blur processing, * represents the convolution operation, and Gaussian kernel represents the Gaussian kernel function, and I0(x,y) represents the fracture image test data.

[0057] S202: Downsampling is performed with the base layer as the starting point to obtain a Gaussian pyramid layer.

[0058] S203: The number of iterations is increased by 1, and the obtained Gaussian pyramid layer is downsampled until the resolution of the Gaussian pyramid layer is less than the preset resolution or the number of iterations is greater than the preset number of iterations.

[0059] It should be noted that those skilled in the art can set the size of the preset resolution according to actual needs, and the present invention is not limited thereto.

[0060] S204: The base layer is used as the first layer, and the obtained Gaussian pyramid layers are arranged in order of the number of iterations from small to large, to obtain a Gaussian pyramid.

[0061] S205: performing differentiation on adjacent layers of the Gaussian pyramid to obtain a Laplacian pyramid having multiple layers of Laplacian pyramid images.

[0062] It should be noted that the Laplacian pyramid is generated by the difference between two adjacent layers of the Gaussian pyramid, which can retain high-frequency details in the image, such as edge and texture information. These images can be used to enhance the test set. Specifically, the Laplacian pyramid can be generated by the difference between two adjacent layers of the Gaussian pyramid images, which retains the high-frequency details of the image (such as edge and texture information) and prepares for the subsequent application of the enhancement coefficient. First, each layer of the Gaussian pyramid is upsampled to restore to the resolution of the previous layer:

[0063]

[0064] in, Represents the k+1th level of the Gaussian pyramid G k+1 The image obtained after upsampling (x, y). Then, the difference between the k-th layer Gaussian pyramid and the upsampled image is calculated to generate the k-th layer Laplacian pyramid image:

[0065]

[0066] S206: Calculate the adaptive enhancement coefficient of each layer of the Laplacian pyramid.

[0067] S207: Input the fracture image test data into the Laplace pyramid, multiply the adaptive enhancement coefficient and the Laplace pyramid image of the corresponding layer number, and obtain fracture image enhancement data with multiple different resolutions:

[0068] L' k (x,y)=α k (x,y)×L k (x,y)

[0069] Among them, L' k (x, y) represents the k-th layer of fracture image enhancement data, i.e., the k-th layer of Laplace pyramid L k (x,y) enhanced image, α k (x, y) represents the adaptive enhancement coefficient corresponding to the k-th layer of the Laplacian pyramid.

[0070] Specifically, the process first smoothes the fracture image through Gaussian filtering and constructs a multi-layer Gaussian pyramid. Then, a Laplacian pyramid is generated through downsampling and differential operations to retain the high-frequency details in the image. Subsequently, the adaptive enhancement coefficient is calculated based on the characteristics of each layer of the image, and combined with the Laplacian pyramid to generate multi-resolution enhanced data. This process can retain image details and enhance important features at different resolutions, improve the clarity and analysis accuracy of fracture images, and ensure noise suppression effects.

[0071] S3: Input the fracture image enhancement data into the fracture image assisted analysis software to be tested according to the annotation information category, and determine the recognition accuracy and processing speed of the fracture image assisted analysis software to be tested for the annotation information.

[0072] It should be noted that the enhanced fracture image data are input into the fracture image auxiliary analysis software to be tested according to the categories of annotation information (such as fracture area, location, and category). Through this step, the recognition accuracy and processing speed of the software for different annotation information are measured, so as to evaluate its analysis ability and efficiency under different fracture image characteristics. This step can help verify the performance of the software in actual use.

[0073] In a possible implementation, S3 specifically includes:

[0074] When the annotation information is the fracture area contour segmentation line, the recognition accuracy is calculated by the chamfer distance:

[0075]

[0076] Among them, D charf (T,I) represents the chamfer distance between the fracture area contour segmentation line I of the fracture image-assisted analysis software to be tested and the fracture area contour segmentation line T corresponding to the annotation information, i.e., the recognition accuracy, d I (t) represents the shortest distance between any point t in T and I, and || represents the absolute value.

[0077] Here, the chamfer distance represents the overlap of the two dividing lines, which can be used as the recognition accuracy. The higher the overlap, the higher the recognition accuracy.

[0078] Optionally, when the annotation information is a fracture area contour segmentation line, the recognition accuracy may also be calculated using the Hausdorff distance.

[0079] When the labeled information is the location of the fracture area, the recognition accuracy is calculated using the Dice coefficient:

[0080]

[0081] Among them, Dice represents the Dice coefficient, that is, the recognition accuracy, Out represents the fracture segmentation area of ​​the fracture image-assisted analysis software to be tested, True represents the fracture segmentation area marked by the fracture area contour segmentation line, || represents taking the absolute value, and ∩ represents taking the intersection.

[0082] Optionally, the recognition accuracy can also be calculated using the Jaccard coefficient:

[0083]

[0084] Among them, Jaccard represents the Jaccard coefficient, and ∪ represents taking the union.

[0085] When the labeled information is fracture category, the ratio of the number of fracture categories correctly identified by the fracture image-assisted analysis software to be tested to the total number of fracture image test data to be identified is taken as the recognition accuracy.

[0086] It should be noted that in this step, different types of annotation information will use appropriate features and methods to calculate the recognition accuracy. Fracture area contour segmentation line: The chamfer distance is used to measure the difference between the fracture area contour segmentation line generated by the software to be tested and the annotated segmentation line. The chamfer distance evaluates the degree of overlap between the points between the two segmentation lines and the nearest point by calculating the distance between them. The higher the overlap, the higher the recognition accuracy. In addition, the Hausdorff distance can also be used to measure the maximum difference between the two boundaries to ensure segmentation accuracy. Fracture area position: The Dice coefficient or Jaccard coefficient is used to evaluate the degree of overlap between the fracture area generated by the software to be tested and the annotated fracture area. These coefficients are used to measure the overlap rate between the two areas to ensure the accuracy of the position annotation. Fracture category: The evaluation of the fracture category determines the recognition accuracy by calculating the ratio of the number of categories correctly identified by the software to the total number of annotations. In this way, the recognition ability of the software to be tested is comprehensively and accurately measured to ensure that accurate performance test results can be provided under different features.

[0087] S4: When the recognition accuracy is greater than the preset recognition accuracy and the processing speed is greater than the preset processing speed, proceed to step S5; otherwise, output the abnormal category corresponding to the annotation information and proceed to step S8.

[0088] It should be noted that the software performance is initially evaluated to detect whether the recognition accuracy and processing speed meet the preset standards. If the recognition accuracy and processing speed meet the requirements, the subsequent in-depth testing will continue. Otherwise, the corresponding exception category (such as recognition error or too slow speed) will be marked and the exception handling process will be entered. This step ensures the quality and reliability of the test and screens out situations that do not meet the standards.

[0089] In a possible implementation, the abnormality categories include fracture area segmentation abnormality, fracture position identification abnormality, fracture category identification abnormality, fracture area segmentation speed abnormality, fracture position identification speed abnormality, and fracture category identification speed abnormality.

[0090] It should be noted that those skilled in the art can set the preset recognition accuracy and the preset recognition speed according to actual needs, and the present invention does not limit this herein.

[0091] S5: The region of interest of each fracture image in the fracture image enhancement data is extracted by combining the moment of inertia matrix and the rotating rectangular frame to obtain a set of regions of interest with annotation information.

[0092] Among them, the moment of inertia matrix, also known as the second-order moment matrix, describes the geometric characteristics of the pixel distribution in the image. By analyzing the distribution of pixel points and calculating the main axis direction and secondary axis direction of the image, it helps to analyze and extract the shape more accurately. The rotating rectangular frame is a rectangular frame that can adaptively adjust the rotation angle according to the image content. It can better align with the main axis direction of the object or area, so as to accurately surround the target area, reduce background interference, and retain effective information. The region of interest is a specific part of the image that contains key target information. For fracture images, the region of interest refers to the fracture area or the part related to the analysis. By combining the moment of inertia matrix and the rotating rectangular frame technology, the region of interest in each fracture image is extracted and aligned with the main axis direction, and a set of regions of interest with annotation information is automatically and accurately generated, providing more accurate data for subsequent analysis and testing.

[0093] It should be noted that traditional labeling methods usually adopt fine-grained labeling, that is, fully labeling detailed information. This cannot be used for fracture contours and fracture area scenes with complex details that are difficult to accurately divide. Too detailed labeling information increases the image recognition cost, and when complex images that are difficult to label appear, it will directly cause a large deviation in the recognition results. Labeling by rotating the rectangular box can effectively reduce the labeling cost during the test process, and compress the image to be identified to verify the recognition accuracy of the software in different scenarios.

[0094] It can also effectively reduce the amount of image data processed in the analysis software. The purpose is to improve the test speed and reduce the test time when performing a large number of repetitive tests on the software, complete the repetitive test faster, and obtain the recognition accuracy of the image-assisted analysis software for different annotation information under a larger test base. Accordingly, more test cycles can be executed within a fixed time, thereby obtaining the recognition effect of the software under a large number of repeated tests. And the region of interest set is also an enhanced form of fracture image. This multi-level testing method not only greatly reduces the test time, but also verifies the recognition results of the image-assisted analysis software for fracture tests in different image forms, effectively improving the reliability, accuracy and long-term stability of the test results.

[0095] Compared with Laplace enhancement, the rotating rectangular frame can also be regarded as a way of image enhancement, which can reduce the amount of data processing in the test process. While ensuring that the processing speed meets the standard, the test data can be cropped. While ensuring the accuracy of the image test, it can greatly improve the test speed of the image-assisted analysis software for the fracture to be tested, help to cycle multiple cycles of testing, and improve the effectiveness and accuracy of the test results.

[0096] In a possible implementation, S5 specifically includes:

[0097] S501: Segment the fracture image and extract the fracture area:

[0098]

[0099] Among them, M(x,y) represents a binary mask indicating whether the pixel point (x,y) belongs to the fracture area.

[0100] Specifically, the fracture image can be segmented using a U-Net-based deep learning segmentation network or a region growing segmentation algorithm to automatically segment the fracture area. The segmentation result is a binary image in which the fracture area is the foreground and the background is black.

[0101] S502: Calculate the major axis direction and minor axis direction of the fracture area through the moment of inertia matrix.

[0102] The main axis direction is the rotation angle of the main axis direction relative to the x-axis, and the secondary axis direction is the rotation angle of the secondary axis direction relative to the x-axis.

[0103] In a possible implementation, S502 specifically includes:

[0104] S502A: Establish the moment of inertia matrix of the fracture area:

[0105]

[0106] Where M represents the inertia matrix, I xx and I yy Respectively represent the pixel distribution of the fracture area along the x direction and along the y direction, I xy It represents the mixed second-order moment reflecting the correlation between x and y, (x c ,y c ) represents the centroid coordinates of the fracture area, (x i ,y i ) represents the coordinates of the pixel point in the i-th fracture area, and N represents the total number of pixels in the fracture area.

[0107] S502B: Perform eigenvalue decomposition on the moment of inertia matrix to obtain the first eigenvalue corresponding to the main axis direction and the second eigenvalue corresponding to the secondary axis direction:

[0108]

[0109] Wherein, λ1 represents the first eigenvalue, and λ2 represents the second eigenvalue.

[0110] S502C: According to the first eigenvalue and the second eigenvalue, respectively calculate and obtain a first eigenvector corresponding to the main axis direction and a second eigenvector corresponding to the secondary axis direction:

[0111]

[0112] Wherein, v1 represents the first eigenvector, and v2 represents the second eigenvector.

[0113] S502D: Calculate the major axis direction and the minor axis direction according to the first eigenvector and the second eigenvector:

[0114]

[0115] Wherein, θ1 and θ2 represent the major axis direction and the minor axis direction respectively, and arctan represents the inverse tangent function.

[0116] Specifically, in steps S502A to S502D, firstly, by extracting the pixel distribution of the fracture area, an inertia matrix is ​​established, which contains the distribution information of the area in different directions. Then, the inertia matrix is ​​subjected to eigenvalue decomposition to obtain the eigenvalues ​​corresponding to the main axis direction and the secondary axis direction. The larger the eigenvalue, the greater the ductility of the pixel distribution in the direction. Next, the eigenvectors of the main axis direction and the secondary axis direction are calculated to describe the main extension direction of the fracture area. Finally, the rotation angle of the main axis direction and the secondary axis direction relative to the x-axis is calculated by the inverse tangent function, so as to achieve the accurate positioning of the fracture area direction. The pixel distribution of the fracture area is analyzed by the inertia matrix, combined with the eigenvalue decomposition, the main axis direction and the secondary axis direction of the area are accurately determined, and the eigenvector and direction calculation are used to not only retain the shape and structural information of the fracture area, but also ensure the optimal alignment of the rotating rectangular frame, reduce unnecessary background interference, thereby improving the validity of the test data and the accuracy of recognition.

[0117] S503: With the goal of aligning the rotating rectangular frame along the main axis direction, the rotating rectangular frame is rotated using the main axis direction as a rotation angle.

[0118] S504: Taking the projection length of the fracture area in the major axis direction as the width of the minimum rotation rectangular frame, and taking the projection length of the fracture area in the minor axis direction as the height of the minimum rotation rectangular frame, to obtain the minimum rotation rectangular frame surrounding the fracture area.

[0119] S505: Output the minimum rotated rectangular frame as the region of interest.

[0120] S506: Replace the fracture image, and repeat steps S501 to S505 to obtain a region of interest set having multiple regions of interest.

[0121] Specifically, in steps S501 to S506, the fracture image is first segmented to generate a binary mask image, and the fracture area is extracted as the foreground. Next, the main axis direction and the secondary axis direction of the fracture area are calculated by the moment of inertia matrix to ensure that the main extension direction of the area can be accurately determined according to the geometric distribution of pixels. Then, the rotating rectangular frame is aligned with the main axis direction as the rotation angle to ensure that the direction of the rectangular frame is consistent with the fracture area. The width and height of the rotating rectangular frame are determined according to the projection length of the main axis and the secondary axis to generate the smallest rotating rectangular frame. Finally, the rotating rectangular frame is output as the region of interest, and the process is repeated to obtain a set of regions of interest for multiple fracture images. By segmenting the fracture area, using the moment of inertia matrix to accurately calculate the main axis direction, and combining the alignment of the rotating rectangular frame, background interference can be effectively reduced, key details can be retained, and the amount of data processing during the test can be reduced, thereby improving recognition efficiency and accuracy.

[0122] S6: Inputting the set of regions of interest into the image-assisted analysis software for the fracture to be tested according to the categories of the annotation information, and performing repeated tests for a preset number of times to obtain the secondary recognition accuracy of the image-assisted analysis software for the fracture to be tested on the annotation information.

[0123] It should be noted that those skilled in the art can set the preset number of times according to actual needs, and the present invention does not limit this.

[0124] It should be noted that the extracted regions of interest are re-entered into the software to be tested according to the categories of annotation information for repeated testing. Through these tests, the software's secondary recognition accuracy for various types of annotation information is obtained. This step helps to verify the stability and accuracy of the software after multiple tests, ensuring that it has a higher recognition ability when processing different categories of annotation information and a large test base, so as to accurately verify the software performance.

[0125] S7: When the secondary recognition accuracy is greater than the preset secondary recognition accuracy, it is determined that the performance of the fracture image-assisted analysis software to be tested meets the standard; otherwise, the abnormal category corresponding to the labeled information is output and the process proceeds to step S8.

[0126] It should be noted that the performance of the software in multiple repeated tests is evaluated to check whether the secondary recognition accuracy meets the preset standard. If the secondary recognition accuracy meets the requirements, the software performance is determined to be up to standard. Otherwise, the corresponding exception category is marked and the exception handling step is entered. This step ensures that the software still has high performance and stability under complex and repeated test conditions.

[0127] It should be noted that those skilled in the art can set the preset secondary recognition accuracy rate according to actual needs, and the present invention is not limited thereto.

[0128] S8: Determine that the performance of the fracture image-assisted analysis software to be tested does not meet the standards and output an abnormal category.

[0129] In actual use, the process starts with obtaining the fracture image data with annotated information, performing multi-resolution enhancement on the data, and then inputting the enhanced data into the software to be tested for preliminary evaluation of recognition accuracy and processing speed. After extracting the region of interest, the software's secondary recognition accuracy is further evaluated through repeated tests, and finally it is determined whether the software performance meets the standards. This process can fully evaluate the accuracy, stability and processing efficiency of the software under different image conditions through multi-level testing methods, ensuring the efficiency and reliability of the software in actual use.

[0130] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0131] In the present invention, the fracture image test data is enhanced by a Laplace pyramid with an adaptive enhancement coefficient to obtain fracture image enhancement data. The calculation method of the adaptive enhancement coefficient can realize intelligent regulation of details, edges and noise, has good balance and adaptability, is particularly suitable for enhancement processing in images with rich details or containing noise, can flexibly adjust the enhancement effect according to the local complexity and noise level of the image, ensure the overall quality of the image and the naturalness of local details, complete the effective expansion of the test data, and then increase the test coverage breadth and test accuracy. In addition, the fracture image enhancement data is used for a test again. After the processing speed meets the requirements, the recognition accuracy of the software is tested for a second time with a large number of repetitive tests to further test the recognition effect of the software under different working environments. In the secondary test process, the inertia moment matrix and the rotating rectangular frame are used to extract the region of interest of each fracture image in the fracture image enhancement data, which is equivalent to performing secondary data enhancement on the test data. The secondary data test based on this data effectively removes a large amount of irrelevant redundant data, puts more computing resources in the recognition accuracy test process to verify the recognition effect of the software under high usage, effectively reduces the software test cycle, and can complete the accuracy test and reliability test of the software in a short time.

[0132] Reference Manual Attached Figure 2 , showing a structural schematic diagram of a fracture image-assisted analysis software performance testing system provided by the present invention.

[0133] The present invention also provides a fracture image-assisted analysis software performance testing system 20, which is applied to the fracture image-assisted analysis software performance testing method, comprising:

[0134] Processor 201.

[0135] The memory 202 stores computer-readable instructions. When the computer-readable instructions are executed by the processor 201, the method for testing the performance of the fracture image-assisted analysis software in the method embodiment is implemented.

[0136] The fracture image-assisted analysis software performance testing system 20 provided by the present invention can execute the above-mentioned fracture image-assisted analysis software performance testing method and achieve the same or similar technical effects. To avoid repetition, the present invention will not go into details.

[0137] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0138] In the present invention, the fracture image test data is enhanced by a Laplace pyramid with an adaptive enhancement coefficient to obtain fracture image enhancement data. The calculation method of the adaptive enhancement coefficient can realize intelligent regulation of details, edges and noise, has good balance and adaptability, is particularly suitable for enhancement processing in images with rich details or containing noise, can flexibly adjust the enhancement effect according to the local complexity and noise level of the image, ensure the overall quality of the image and the naturalness of the local details, complete the effective expansion of the test data, and then increase the test coverage breadth and test accuracy. In addition, the fracture image enhancement data is used for a test again. After the processing speed meets the requirements, the recognition accuracy of the software is tested for a second time with a large number of repetitive tests to further test the recognition effect of the software under different working environments. In the secondary test process, the inertia moment matrix and the rotating rectangular frame are used to extract the region of interest of each fracture image in the fracture image enhancement data, which is equivalent to performing secondary data enhancement on the test data. The secondary data test based on this data effectively removes a large amount of irrelevant redundant data, puts more computing resources in the recognition accuracy test process to verify the recognition effect of the software under high usage, effectively reduces the software test cycle, and can complete the accuracy test and reliability test of the software in a short time.

[0139] It should be understood that the processor in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0140] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0141] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0142] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0143] In the present invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0144] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0145] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0146] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0147] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0148] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0149] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0150] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0151] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method for testing the performance of fracture image-assisted analysis software as described in the method embodiment.

[0152] A computer-readable storage medium provided by the present invention can implement the steps and effects of the fracture image-assisted analysis software performance testing method of the above method embodiment. To avoid repetition, the present invention will not go into details.

[0153] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0154] In the present invention, the fracture image test data is enhanced by a Laplace pyramid with an adaptive enhancement coefficient to obtain fracture image enhancement data. The calculation method of the adaptive enhancement coefficient can realize intelligent regulation of details, edges and noise, has good balance and adaptability, is particularly suitable for enhancement processing in images with rich details or containing noise, can flexibly adjust the enhancement effect according to the local complexity and noise level of the image, ensure the overall quality of the image and the naturalness of the local details, complete the effective expansion of the test data, and then increase the test coverage breadth and test accuracy. In addition, the fracture image enhancement data is used for a test again. After the processing speed meets the requirements, the recognition accuracy of the software is tested for a second time with a large number of repetitive tests to further test the recognition effect of the software under different working environments. In the secondary test process, the inertia moment matrix and the rotating rectangular frame are used to extract the region of interest of each fracture image in the fracture image enhancement data, which is equivalent to performing secondary data enhancement on the test data. The secondary data test based on this data effectively removes a large amount of irrelevant redundant data, puts more computing resources in the recognition accuracy test process to verify the recognition effect of the software under high usage, effectively reduces the software test cycle, and can complete the accuracy test and reliability test of the software in a short time.

[0155] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0156] There are a few points to note:

[0157] (1) The drawings of the embodiments of the present invention only relate to the structures related to the embodiments of the present invention, and other structures may refer to the general design.

[0158] (2) For the sake of clarity, in the drawings used to describe the embodiments of the present invention, the thickness of the layers or regions is exaggerated or reduced, that is, these drawings are not drawn according to the actual scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or there may be intermediate elements.

[0159] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to obtain new embodiments.

[0160] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for testing the performance of fracture image-assisted analysis software, characterized in that: include: S1: Acquire fracture image test data with different annotation information, wherein the annotation information includes a fracture area contour segmentation line, a fracture area position, and a fracture category; S2: Enhance the fracture image test data by using a Laplace pyramid with an adaptive enhancement coefficient to obtain fracture image enhanced data; S3: inputting the fracture image enhancement data into the fracture image auxiliary analysis software to be tested according to the type of the annotation information, and determining the recognition accuracy and processing speed of the fracture image auxiliary analysis software to be tested for the annotation information; S4: if the recognition accuracy is greater than a preset recognition accuracy and the processing speed is greater than a preset processing speed, proceed to step S5; otherwise, output the abnormality category corresponding to the annotation information and proceed to step S8; S5: extracting the region of interest of each fracture image in the fracture image enhancement data by combining the moment of inertia matrix and the rotating rectangular frame to obtain a set of regions of interest with the annotation information; S6: inputting the set of regions of interest into the image-assisted analysis software for fracture to be tested according to the categories of the annotation information, performing repeated tests for a preset number of times, and obtaining the secondary recognition accuracy of the image-assisted analysis software for fracture to be tested on the annotation information; S7: if the secondary recognition accuracy rate is greater than a preset secondary recognition accuracy rate, determining that the performance of the fracture image-assisted analysis software to be tested meets the standard; otherwise, outputting the abnormality category corresponding to the annotation information and entering step S8; S8: Determine that the performance of the fracture image-assisted analysis software to be tested does not meet the standard, and output the abnormal category.

2. The fracture image-assisted analysis software performance testing method according to claim 1, characterized in that: After S1, the following is also included: The fracture image test data is desensitized based on character masking.

3. The fracture image-assisted analysis software performance testing method according to claim 1, characterized in that: The adaptive enhancement coefficient is specifically: Among them, α(x,y) represents the adaptive enhancement coefficient at the image pixel (x,y), represents the local variance at the image pixel (x, y), η 2 (x,y) represents the dynamic noise variance at the image pixel (x,y), ò(x,y) represents the enhancement intensity adjustment parameter at the image pixel (x,y), ò min =10 -3 represents a constant used to control the lower limit of the enhancement strength adjustment parameter, H(x,y) represents the local entropy value at the image pixel (x,y), log represents the logarithmic function, P i (x, y) represents the probability that the gray value of the image pixel (x, y) in the eight-neighborhood is i, and L represents the number of gray levels, that is, the number of gray value categories.

4. The method for testing the performance of fracture image-assisted analysis software according to claim 3, characterized in that: The S2 specifically includes: S201: Smooth the fracture image test data through Gaussian filtering to obtain the base layer of the Gaussian pyramid: G0(x,y)=I0(x,y)*Gaussian kernel Among them, G0(x,y) represents the base layer, that is, the image after Gaussian blur processing, * represents the convolution operation, and Gaussian kernel represents the Gaussian kernel function, I0(x,y) represents the fracture image test data; S202: downsampling is performed with the base layer as a starting point to obtain a Gaussian pyramid layer; S203: the number of iterations is increased by 1, and the obtained Gaussian pyramid layer is downsampled until the resolution of the Gaussian pyramid layer is less than the preset resolution or the number of iterations is greater than the preset number of iterations; S204: taking the base layer as the first layer, and arranging the obtained Gaussian pyramid layers in order of the number of iterations from small to large, to obtain a Gaussian pyramid; S205: performing differentiation on adjacent layers of the Gaussian pyramid to obtain a Laplacian pyramid having multiple layers of Laplacian pyramid images; S206: Calculate the adaptive enhancement coefficient of each layer of the Laplacian pyramid; S207: Input the fracture image test data into the Laplace pyramid, multiply the adaptive enhancement coefficient and the Laplace pyramid image of the corresponding layer number to obtain fracture image enhancement data with multiple different resolutions: The k (x,y)=α k (x,y)×L k (x,y) Among them, L' k (x, y) represents the k-th layer of fracture image enhancement data, i.e., the k-th layer of Laplace pyramid L k (x,y) enhanced image, α k (x, y) represents the adaptive enhancement coefficient corresponding to the k-th layer of the Laplacian pyramid.

5. The method for testing the performance of fracture image-assisted analysis software according to claim 1, characterized in that: The S3 specifically includes: When the annotation information is a fracture area contour segmentation line, the recognition accuracy is calculated by the chamfer distance: Among them, D charf (T,I) represents the chamfer distance between the fracture area contour segmentation line I of the fracture image-assisted analysis software to be tested and the fracture area contour segmentation line T corresponding to the annotation information, i.e., the recognition accuracy, d I (t) represents the shortest distance between any point t and I in T, and || represents the absolute value; When the annotation information is the location of the fracture area, the recognition accuracy is calculated by the Dice coefficient: Among them, Dice represents the Dice coefficient, i.e., the recognition accuracy rate; Out represents the fracture segmentation area of ​​the fracture image-assisted analysis software to be tested; True represents the fracture segmentation area marked by the fracture area contour segmentation line; || represents taking the absolute value; and ∩ represents taking the intersection; When the labeled information is fracture category, the ratio of the number of fracture categories correctly identified by the fracture image auxiliary analysis software to be tested to the total number of fracture image test data to be identified is used as the recognition accuracy rate.

6. The fracture image-assisted analysis software performance testing method according to claim 1, characterized in that: The abnormal categories include fracture area segmentation abnormality, fracture position identification abnormality, fracture category identification abnormality, fracture area segmentation speed abnormality, fracture position identification speed abnormality and fracture category identification speed abnormality.

7. The method for testing the performance of fracture image-assisted analysis software according to claim 1, characterized in that: The S5 specifically includes: S501: Segment the fracture image to extract the fracture area: Among them, M(x,y) represents a binary mask indicating whether the pixel point (x,y) belongs to the fracture area; S502: Calculating the major axis direction and the minor axis direction of the fracture area by using a moment of inertia matrix; S503: with the goal of aligning the rotating rectangular frame along the main axis direction, rotating the rotating rectangular frame using the main axis direction as a rotation angle; S504: taking the projection length of the fracture area in the direction of the major axis as the width of the minimum rotated rectangular frame, and taking the projection length of the fracture area in the direction of the minor axis as the height of the minimum rotated rectangular frame, to obtain the minimum rotated rectangular frame surrounding the fracture area; S505: Outputting the minimum rotated rectangular frame as the region of interest; S506: Replace the fracture image, and repeat steps S501 to S505 to obtain a region of interest set having multiple regions of interest.

8. The method for testing the performance of fracture image-assisted analysis software according to claim 7, characterized in that: The S502 specifically includes: S502A: Establish the moment of inertia matrix of the fracture area: Where M represents the inertia matrix, I xx and I yy Respectively represent the pixel distribution of the fracture area along the x direction and along the y direction, I xy It represents the mixed second-order moment reflecting the correlation between x and y, (x c ,y c ) represents the centroid coordinates of the fracture area, (x i ,y i ) represents the coordinates of the pixel point in the i-th fracture area, and N represents the total number of pixel points in the fracture area; S502B: Perform eigenvalue decomposition on the moment of inertia matrix to obtain a first eigenvalue corresponding to the main axis direction and a second eigenvalue corresponding to the secondary axis direction: Wherein, λ1 represents the first eigenvalue, and λ2 represents the second eigenvalue; S502C: According to the first eigenvalue and the second eigenvalue, respectively calculate and obtain a first eigenvector corresponding to the main axis direction and a second eigenvector corresponding to the secondary axis direction: Wherein, v1 represents the first eigenvector, and v2 represents the second eigenvector; S502D: Calculate the major axis direction and the minor axis direction according to the first eigenvector and the second eigenvector: Wherein, θ1 and θ2 represent the major axis direction and the minor axis direction respectively, and arctan represents the inverse tangent function.

9. A fracture image-assisted analysis software performance testing system, characterized in that: include: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method for testing the performance of fracture image-assisted analysis software according to any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for testing the performance of fracture image-assisted analysis software as described in any one of claims 1 to 8 is implemented.

Images