Intelligent nondestructive detection method and system for cylinder cover conduit defect and storage medium

Abstract

The invention discloses an intelligent nondestructive testing method and system for cylinder cover conduit defects and a storage medium, and relates to the technical field of cylinder cover nondestructive testing. The method comprises the specific steps that according to the form of an enclosed area, the crack possibility degree of the enclosed area is obtained; all the enclosed areas are screened according to the crack possible degree, and suspected crack areas are obtained; according to the gray difference condition of the edge pixel points of the enclosed area, the friction degree of the enclosed area is obtained; screening all the enclosed areas according to the friction degree to obtain a friction trace area; determining a damage area in the rod testing range through a friction trace area and a suspected crack area in the rod testing range; according to the crack possibility degree and the gray value of the damaged area and the distribution condition of other damaged areas, the damage degree of the damaged area is obtained; and performing defect detection on the cylinder cover conduit based on the damage degree. According to the invention, the accuracy of the cylinder cover conduit crack defect detection result is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cylinder head nondestructive testing, and in particular to an intelligent nondestructive testing method, system and storage medium for cylinder head conduit defects. Background Art

[0002] If the cracks in the cylinder head duct are not detected in time during quality inspection, it will have a huge impact and loss on the subsequent assembly and sale of the vehicle, so the detection of cracks in the cylinder head duct is particularly important; then, an adapted inspection rod can be made according to the size of the inner hole diameter of the duct. When the duct is bent and deformed, resulting in a narrowing of the inner hole diameter, the inspection rod cannot pass through the inner hole of the duct, thereby indicating that the duct has cracks; the existing technology mainly performs edge detection on the inspection rod image, and analyzes whether there are cracks in the inspection rod range in the edge detection image in the vertical direction of the inspection rod travel, but this method often cannot distinguish between slight cracks on the inspection rod caused by slight deformation of the duct, resulting in inaccurate cylinder head duct defect detection results. Summary of the invention

[0003] The purpose of the present invention is to provide a cylinder head duct defect intelligent non-destructive detection method, system and storage medium to solve the problems raised in the background technology.

[0004] To achieve the above object, the present invention provides the following solution: an intelligent nondestructive detection method for cylinder head duct defects based on data analysis, the specific steps of which include the following:

[0005] Acquire a rod inspection image of the cylinder head guide tube;

[0006] According to the shape of the enclosed area in the inspection stick detection image, obtaining the possible degree of cracks in the enclosed area;

[0007] The enclosed area is screened according to the possible degree of the crack to obtain all suspected crack areas within each test stick range;

[0008] Determine the damaged area within the test rod range according to the suspected crack area and the friction mark area within the test rod range;

[0009] Obtaining the damage degree of the damaged area through the possible degree of cracks in the damaged area, the gray value and the distribution with other damaged areas;

[0010] A defect detection is performed on the cylinder head guide based on the damage degree.

[0011] Preferably, the step of obtaining the friction mark area is:

[0012] Acquire all enclosed areas in the stick detection image;

[0013] Obtaining the friction degree of the enclosed area according to the grayscale difference of the edge pixels of the enclosed area;

[0014] The enclosed area is screened according to the friction degree to obtain the friction mark area within the range of each test stick.

[0015] Preferably, the specific steps of obtaining the enclosed area are:

[0016] Segmenting the stick inspection image to obtain all connected domains in the stick inspection image;

[0017] Each connected domain is used as a stick-checking range, edge detection is performed on each stick-checking range, and all edge pixel points in the stick-checking range are obtained;

[0018] All closed areas formed by the edge pixels within the range of each test stick are recorded as the enclosed area.

[0019] Preferably, the screening method for the suspected crack area is:

[0020] A crack threshold parameter is preset, and for any enclosed area within the range of any test stick, if the possible degree of cracks in the enclosed area is greater than or equal to the crack threshold parameter, the enclosed area is recorded as the suspected crack area.

[0021] Preferably, the specific steps of obtaining the damage degree of the damaged area are:

[0022] For any damaged area within the range of any test rod, the cumulative sum of the Euclidean distances between the centroid of the damaged area and the centroids of all other damaged areas is recorded as the distribution value of the damaged area;

[0023] The normalized value of the product of the distribution value of the damaged area, the possible degree of cracks in the damaged area and the average value of the grayscale values ​​of all pixels in the damaged area is used as the damage degree of the damaged area.

[0024] On the other hand, a cylinder head duct defect intelligent nondestructive detection system based on data analysis is provided, comprising an image acquisition module, a first data processing module, a screening module, a damage area determination module, a damage degree determination module, and a defect detection module; wherein,

[0025] The image acquisition module is used to acquire a test rod detection image of the cylinder head conduit;

[0026] The first data processing module is used to obtain the possible degree of cracks in the enclosed area according to the shape of the enclosed area in the inspection stick detection image;

[0027] The screening module is used to screen the enclosed area according to the possible degree of the crack to obtain all suspected crack areas within each test stick range;

[0028] The damage area determination module is used to determine the damage area within the test rod range according to the suspected crack area and the friction mark area within the test rod range;

[0029] The damage degree determination module is used to obtain the damage degree of the damaged area through the possible degree of cracks in the damaged area, the gray value and the distribution with other damaged areas;

[0030] The defect detection module is used to perform defect detection on the cylinder head conduit based on the damage degree.

[0031] Finally, a computer storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the intelligent non-destructive detection method for cylinder head duct defects based on data analysis are implemented.

[0032] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: obtaining the degree of friction of the enclosed area according to the grayscale difference of the edge pixels of the enclosed area; screening all the enclosed areas according to the degree of friction to obtain all the friction mark areas within each test stick range; determining the damaged area within the test stick range through the friction mark areas and the suspected crack areas within the test stick range; obtaining the degree of damage to the damaged area according to the possible degree of cracks in the damaged area, the grayscale value and the distribution with other damaged areas; performing defect detection on the cylinder head duct based on the degree of damage; thereby, not only the more obvious cracks in the cylinder head duct can be obtained, but also the fine cracks in the cylinder head duct can be detected, thereby improving the accuracy of the cylinder head duct defect detection results. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. 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 paying creative labor.

[0034] Figure 1 is a flow chart of the method of the present invention;

[0035] Figure 2 It is a system structure diagram of the present invention. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] The purpose of the present invention is to provide an intelligent nondestructive detection method for cylinder head duct defects based on data analysis, such as Figure 1 As shown, the specific steps include the following:

[0038] S1, obtaining a test rod detection image of the cylinder head guide tube;

[0039] S2. Obtaining the possible degree of cracks in the enclosed area according to the shape of the enclosed area in the inspection stick detection image;

[0040] S3. Screen the enclosed area according to the possible degree of cracks to obtain all suspected crack areas within each test stick range;

[0041] S4. Determine the damaged area within the test rod range based on the suspected crack area and friction mark area within the test rod range;

[0042] S5. Obtain the damage degree of the damaged area through the possible degree of cracks in the damaged area, the gray value, and the distribution with other damaged areas;

[0043] S6. Perform defect detection on the cylinder head guide tube based on the degree of damage.

[0044] It should be noted that since an adaptive inspection rod can be made according to the size of the inner hole diameter of the catheter, when the catheter bends and deforms, causing the inner hole diameter to narrow, the inspection rod cannot pass through the inner hole of the catheter, thereby indicating that there is a crack in the catheter; that is, when the inspection rod is inserted into the cylinder head catheter, when it encounters a crack, it continues to apply pressure, and the inspection rod will be blocked in its progress, resulting in cracks.

[0045] Further, in step S1, all grayscale images of the inspection rods that have been inspected on the cylinder head conduit are obtained, and all enclosed areas within the inspection rod range in the grayscale images of the inspection rods are obtained; the specific process is:

[0046] S11, arranging all the test rods for testing the cylinder head conduit in sequence, photographing them with a high-definition camera, and obtaining a grayscale image of the test rods for testing the cylinder head conduit;

[0047] S12, performing threshold segmentation on the stick-checking grayscale image using the Otsu threshold segmentation algorithm, obtaining all connected domains in the stick-checking grayscale image, and using each connected domain as a stick-checking range;

[0048] S13, using the lobel bidirectional edge detection algorithm to perform edge detection on each inspection stick range, and obtain all edge pixel points in each inspection stick range;

[0049] S14. All closed areas formed by edge pixels within the range of each test stick are recorded as enclosed areas.

[0050] So far, the grayscale images of all the test rods that have been inspected on the cylinder head duct are obtained through the above method.

[0051] Furthermore, in step S2, whether there is a crack in the cylinder head duct is detected by the deformation degree of the test rod. When the crack in the cylinder head duct is more serious, the duct is misaligned and the crack on the test rod is more obvious. Since the test rod is subjected to the force applied in its travel direction when cracks occur, the crack direction when the test rod breaks is generally perpendicular to the travel direction of the test rod. Edge detection is performed on the grayscale image. If there is a crack perpendicular to the travel direction of the test rod within the range of the test rod, the wider the crack, the more likely it is that the duct detected by the test rod has a crack. When the travel of the test rod is hindered and cracks are caused, deformation will occur, causing the longer edge of the test rod to change from a straight line to a curve. It can be concluded that when the longer edge of the obtained test rod range is closer to a straight line, the possibility of cracks on the test rod is smaller, and the possibility of cracks inside the cylinder head duct is smaller.

[0052] In some implementations of the embodiments of the present invention, a specific method for obtaining the possible degree of cracks in the enclosed area according to the shape of the enclosed area is:

[0053] S21. For any enclosed area within the range of any test stick, the length and width of the minimum circumscribed rectangle of the enclosed area are used as the length and width of the enclosed area;

[0054] S22, using PCA principal component analysis to analyze the inspection range, obtaining the principal component direction of the inspection range, and using it as the overall direction of the inspection range; using the least square method to perform linear fitting on all edge pixel points on both sides of the inspection range in the overall direction, and using the sum of two root mean square errors in the fitting process as the deformation degree of the inspection range;

[0055] S23. The ratio of the width to the length of the enclosed area is recorded as the crack factor; the normalized value of the product of the crack factor and the deformation degree of the test rod range is used as the possible degree of cracks in the enclosed area.

[0056] The specific calculation formula for the possible degree of cracks is:

[0057]

[0058] Among them, Q represents the possible degree of cracks in the enclosed area, W represents the width of the enclosed area; L represents the length of the enclosed area; α represents the deformation degree of the test rod range; norm() represents the linear normalization function.

[0059] It should be noted that When it is larger, it means that the crack width is larger, and the crack on the test rod is more obvious, so the detected cylinder head duct is more likely to have cracks; the greater the deformation degree of the test rod range, the worse the effect of fitting the edge pixel points on both sides of the test rod range into straight lines, which proves that the more curved the two sides of the test rod range are, the greater the deformation degree of the test rod, so the detected cylinder duct is more likely to have cracks; PCA principal component analysis method and least squares method are existing technologies, and this embodiment will not be described in detail here.

[0060] Further, in step S3, all enclosed areas are screened according to the possible degree of cracks, and the specific method for obtaining all suspected crack areas within each test rod range is:

[0061] S31, preset a crack threshold parameter T1, wherein this embodiment is described by taking T1=0.5 as an example, and this embodiment is not specifically limited, wherein T1 is determined according to specific implementation conditions;

[0062] S32. For any enclosed area within the range of any test stick, if the crack possibility of the enclosed area is greater than or equal to the crack threshold parameter T1, the enclosed area is recorded as a suspected crack area.

[0063] The analysis of suspected crack areas can only identify test rods with obvious deformation, but cannot identify slight deformation of the test rod caused by slight deformation of the cylinder head duct. Since the diameter of the test rod for testing the cylinder head duct is the same as that of the cylinder head duct, friction marks will be left during the test, resulting in slight deformation that is difficult to detect and will cause the friction marks to break. Therefore, any enclosed area within the test rod range can be screened to obtain the friction mark area. The damaged area within the test rod range is then determined through the friction mark area and the suspected crack area. The degree of damage to the damaged area is obtained based on the grayscale value of the damaged area and its distribution with other damaged areas.

[0064] Furthermore, in step S4, the steps of obtaining the friction mark area are:

[0065] S41, obtaining all enclosed areas in the stick inspection image;

[0066] S42, obtaining the friction degree of the enclosed area according to the grayscale difference of the edge pixels of the enclosed area;

[0067] S43. Screen the enclosed area by the degree of friction to obtain the friction mark area within the range of each test stick.

[0068] In step S42, according to the grayscale difference of the edge pixels of the enclosed area, the specific method for obtaining the friction degree of the enclosed area is:

[0069] S421, preset an edge parameter n, wherein this embodiment is described by taking n=5 as an example, and this embodiment is not specifically limited, wherein n is determined according to specific implementation conditions;

[0070] S422, for any enclosed area within the range of any test stick, the sine value of the angle between the overall direction of the enclosed area and the vertical direction is used as the offset factor of the enclosed area;

[0071] S423, for any edge pixel point of the enclosed area, select pixel points along the normal direction of the edge pixel point to jointly form an edge component sequence of the edge pixel point; and obtain the edge grayscale clarity of the enclosed area according to the grayscale difference of the pixel points in the edge component sequence of the edge pixel point of the enclosed area;

[0072] S424, taking the normalized value of the product of the inverse of the offset factor of the enclosed area, the crack factor of the enclosed area and the edge grayscale clarity of the enclosed area as the friction degree of the enclosed area.

[0073] The specific formula is:

[0074]

[0075] Among them, R represents the friction degree of the enclosed area; Z represents the offset factor of the enclosed area; W represents the width of the enclosed area; L represents the length of the enclosed area; B represents the grayscale clarity of the edge of the enclosed area; norm() table represents the linear normalization function.

[0076] It should be noted that The larger the value, the more elongated the enclosed area is in the moving direction of the test rod, and the more likely it is that the enclosed area is a trace of friction; the greater the grayscale clarity of the edge of the enclosed area, the clearer the edge of the enclosed area is, and the more likely it is a trace of friction.

[0077] The specific method for obtaining the grayscale clarity of the edge of the enclosed area in step S423 is:

[0078] For any edge pixel in the enclosed area, the absolute value of the difference in grayscale values ​​between the th pixel and the th pixel in the edge composition sequence of the edge pixel is recorded as the grayscale difference of the th pixel; the cumulative sum of the grayscale differences of all pixels in the edge composition sequence of the edge pixel is recorded as the grayscale difference value of the edge pixel; the reciprocal of the cumulative sum of the grayscale difference values ​​of all edge pixels in the enclosed area is taken as the edge grayscale clarity of the enclosed area.

[0079] The specific formula is:

[0080]

[0081] Where X represents the grayscale clarity of the edge of the enclosed area; Np represents the number of all edge pixels in the enclosed area; n represents the preset edge parameter; A p,m A represents the gray value of the mth pixel in the edge component sequence of the pth edge pixel in the enclosed area; p,m+1 Represents the grayscale value of the m+1th pixel in the edge component sequence of the pth edge pixel in the enclosed area; || means taking the absolute value.

[0082] In step S43, all enclosed areas are screened according to the degree of friction, and the specific method for obtaining all friction trace areas within the range of each test stick is:

[0083] S431, preset a friction threshold parameter T2, wherein this embodiment is described by taking T2=0.65 as an example, and this embodiment is not specifically limited, wherein T2 is determined according to specific implementation conditions;

[0084] S432. For any enclosed area within the range of any test rod, if the crack possibility of the enclosed area is greater than or equal to the friction threshold parameter T2, the enclosed area is recorded as a friction mark area.

[0085] When taking pictures of the inspection stick, due to the influence of environmental factors such as lighting, reflected light spots will be generated on the inspection stick, which will affect the identification of friction marks. The friction marks are generated when the inspection stick moves, and the friction is generated along the moving direction of the inspection stick, and have clear and obvious edges. However, the light spots are not necessarily along the moving direction of the inspection stick, and the edges of the light spot areas are often unclear. Therefore, except for the suspected crack areas in the inspection stick range, the enclosed area is analyzed to determine whether it is a friction mark based on the overall direction and edge clarity of the enclosed area, and all friction mark areas within the inspection stick range are obtained.

[0086] It should be noted that because the diameter of the test rod and the cylinder head duct are the same, during normal inspection, friction marks will be left on the surface of the test rod along the direction of movement of the test rod, and the friction marks should be continuous; but when there is a slight crack in the cylinder head duct causing a slight deformation of the test rod, but there is no friction mark in the area of ​​slight deformation, the continuous friction marks will be broken. Therefore, the friction mark area within the range of the test rod and the suspected crack area are determined to determine the damaged area within the range of the test rod.

[0087] The specific method of determining the damaged area within the test rod range is as follows: for any test rod range, the friction trace area between adjacent suspected crack areas is recorded as the damaged area. Thus, all damaged areas within each test rod range are obtained.

[0088] It should be noted that the test rod may have surface damage due to problems with operations or materials during production; such damage may also destroy the continuous friction marks on the test rod itself; since defects in the cylinder head duct will cause damage to the continuous friction mark area on the test rod, it is mostly manifested as an overall direction that is perpendicular to the direction of travel of the test rod, and the grayscale value is large; while the grayscale value of the damage to the test rod itself is small, and the shape is more random, and the distribution of the damaged area is relatively dense; therefore, the degree of damage to the damaged area can be quantified by the shape of the damaged area, the grayscale value of each pixel in the damaged area, and the distribution of multiple damaged areas.

[0089] Furthermore, the specific steps for obtaining the damage degree of the damaged area are:

[0090] S51. For any damaged area within the range of any test rod, the sum of the Euclidean distances between the centroid of the damaged area and the centroids of all other damaged areas is recorded as the distribution value of the damaged area;

[0091] S52, taking the normalized value of the product of the distribution value of the damaged area, the possible degree of cracks in the damaged area and the average of the gray values ​​of all pixels in the damaged area as the damage degree of the damaged area.

[0092] The specific formula is:

[0093] U = norm(Q × G × Lp);

[0094] Among them, U represents the damage degree of the damaged area; Q represents the possible degree of cracks in the damaged area; G represents the mean grayscale value of all pixels in the damaged area; Lp represents the distribution value of the damaged area; norm() represents the linear normalization function.

[0095] The greater the possible degree of crack, the closer the damaged area is to the test rod crack caused by the internal crack of the cylinder head duct, and the greater the degree of damage; the greater the average gray value of all pixels in the damaged area, the less consistent with the characteristic of the small gray value of the damaged area of ​​the test rod itself, the more likely the damaged area is caused by the internal crack of the cylinder head duct, and the greater the degree of damage; the farther the damaged area is from other damaged areas, the less consistent with the characteristic of the relatively concentrated damaged areas of the test rod itself, the more likely the damaged area is caused by the internal crack of the cylinder head duct, and the greater the degree of damage. Since the internal cracks and deformation of the cylinder head duct form obvious and clear cracks, the severity of the internal cracks of the cylinder head duct is judged by the degree of damage in the damaged area.

[0096] Furthermore, in step S6, a specific method for performing defect detection on the cylinder head conduit based on the degree of damage is as follows:

[0097] S61, preset a serious threshold parameter T3, wherein this embodiment is described by taking T3=0.75 as an example, and this embodiment is not specifically limited, wherein T3 is determined according to specific implementation conditions;

[0098] S62. For any damaged area within the range of any test rod, if the damage degree of the damaged area is greater than or equal to the friction threshold parameter T3, the cylinder head guide tube is recorded as severely damaged; it is necessary to arrange maintenance work on the cylinder head guide tube in time.

[0099] Another aspect of the present invention provides a cylinder head duct defect intelligent non-destructive detection system based on data analysis, such as Figure 2 As shown, it includes an image acquisition module, a first data processing module, a screening module, a damage area determination module, a damage degree determination module, and a defect detection module; wherein,

[0100] An image acquisition module, used to acquire a test rod detection image of the cylinder head guide tube;

[0101] The first data processing module is used to obtain the possible degree of cracks in the enclosed area according to the shape of the enclosed area in the inspection stick detection image;

[0102] A screening module, used to screen the enclosed area according to the possible degree of cracks, and obtain all suspected crack areas within the range of each test stick;

[0103] A damage area determination module is used to determine the damage area within the test rod range according to the suspected crack area and the friction mark area within the test rod range;

[0104] A damage degree determination module is used to obtain the damage degree of the damaged area through the possible degree of cracks in the damaged area, the gray value, and the distribution with other damaged areas;

[0105] The defect detection module is used to perform defect detection on the cylinder head guide tube based on the degree of damage.

[0106] Finally, a computer storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of an intelligent non-destructive detection method for cylinder head duct defects based on data analysis are implemented.

[0107] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. An intelligent nondestructive detection method for cylinder head duct defects based on data analysis, characterized in that: The specific steps include the following: Acquire a rod inspection image of the cylinder head guide tube; According to the shape of the enclosed area in the inspection stick detection image, obtaining the possible degree of cracks in the enclosed area; The enclosed area is screened according to the possible degree of the crack to obtain all suspected crack areas within each test stick range; Determine the damaged area within the test rod range according to the suspected crack area and the friction mark area within the test rod range; Obtaining the damage degree of the damaged area through the possible degree of cracks in the damaged area, the gray value and the distribution with other damaged areas; A defect detection is performed on the cylinder head guide tube based on the damage degree.

2. The method for intelligent nondestructive detection of cylinder head duct defects based on data analysis according to claim 1 is characterized in that: The steps for obtaining the friction trace area are: Acquire all enclosed areas in the stick detection image; Obtaining the friction degree of the enclosed area according to the grayscale difference of the edge pixels of the enclosed area; The enclosed area is screened according to the friction degree to obtain the friction mark area within the range of each test stick.

3. The method for intelligent nondestructive detection of cylinder head duct defects based on data analysis according to claim 2 is characterized in that: The specific steps of obtaining the enclosed area are: Segmenting the stick inspection image to obtain all connected domains in the stick inspection image; Each connected domain is used as a stick-checking range, edge detection is performed on each stick-checking range, and all edge pixel points in the stick-checking range are obtained; All closed areas formed by the edge pixels within the range of each test stick are recorded as the enclosed area.

4. The method for intelligent nondestructive testing of cylinder head duct defects based on data analysis according to claim 1 is characterized in that: The screening method of the suspected crack area is: A crack threshold parameter is preset, and for any enclosed area within the range of any test stick, if the possible degree of cracks in the enclosed area is greater than or equal to the crack threshold parameter, the enclosed area is recorded as the suspected crack area.

5. The method for intelligent nondestructive testing of cylinder head duct defects based on data analysis according to claim 1 is characterized in that: The specific steps of obtaining the damage degree of the damaged area are: For any damaged area within the range of any test rod, the cumulative sum of the Euclidean distances between the centroid of the damaged area and the centroids of all other damaged areas is recorded as the distribution value of the damaged area; The normalized value of the product of the distribution value of the damaged area, the possible degree of cracks in the damaged area and the average value of the grayscale values ​​of all pixels in the damaged area is used as the damage degree of the damaged area.

6. An intelligent nondestructive testing system for cylinder head duct defects based on data analysis, characterized in that: It includes an image acquisition module, a first data processing module, a screening module, a damage area determination module, a damage degree determination module, and a defect detection module; wherein, The image acquisition module is used to acquire a test rod detection image of the cylinder head conduit; The first data processing module is used to obtain the possible degree of cracks in the enclosed area according to the shape of the enclosed area in the inspection stick detection image; The screening module is used to screen the enclosed area according to the possible degree of the crack to obtain all suspected crack areas within each test stick range; The damage area determination module is used to determine the damage area within the test rod range according to the suspected crack area and the friction mark area within the test rod range; The damage degree determination module is used to obtain the damage degree of the damaged area through the possible degree of cracks in the damaged area, the gray value and the distribution with other damaged areas; The defect detection module is used to perform defect detection on the cylinder head conduit based on the damage degree.

7. A computer storage medium, characterized in that: The computer storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the intelligent non-destructive detection method for cylinder head duct defects based on data analysis as described in any one of claims 1 to 5 are implemented.

Images