A kind of steel cage welding quality intelligent detection system and method

Abstract

The present invention relates to the field of image data processing technology, and in particular to an intelligent detection system and method for the welding quality of a steel cage. The method comprises the steps of: obtaining the edge connected domain of each welding pattern of the steel cage, taking any edge connected domain as a target connected domain, and determining the abnormal amount of the target connected domain; respectively obtaining the absolute difference between the abnormal amount of the target connected domain and the adjacent edge connected domains on the left and right sides, and determining the morphological difference amount of the target connected domain; taking any pixel point in the target connected domain as a target pixel point, obtaining the neighboring pixel points of the target pixel point, and obtaining the crack degree of the target pixel point; obtaining the target gray value by multiplying the crack degree of the target pixel point and the gray value, adjusting the gray value of the target pixel point to the target gray value, and detecting and obtaining the quality detection result of the welding pattern image, which can effectively improve the efficiency of the welding quality detection of the steel cage.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and in particular to a steel cage welding quality intelligent detection system and method. Background Art

[0002] Steel cage is an important material commonly used in construction, and its welding quality is directly related to the overall performance of the structural system. Testing the welding quality can ensure that key indicators such as the strength, toughness and formability of the weld meet the standards, thereby ensuring the stability and safety of the building structure. At the same time, high-quality welding can also significantly improve the bearing capacity and stability of the steel cage, thereby improving the quality of the entire project. Traditional manual inspection of steel cage quality relies on the experience of the staff, which is prone to misjudgment and omission, and poses a major safety hazard.

[0003] The patent application document with publication number CN109034204A discloses a weld defect recognition method based on an improved convolutional neural network. The method establishes a pooling model that comprehensively considers the feature distribution of the pooling domain and the feature map in which it is located, constructs a deep convolutional neural network with the above pooling model and feature selection method, and trains a weld defect recognition neural network to realize the recognition of weld defect types, thereby avoiding the process of manual feature extraction required by traditional weld defect recognition methods.

[0004] Although the above-mentioned existing scheme provides a method for identifying the type of weld defects, in actual use, the welding process of the steel cage may be performed by multiple workers, and the welding method and effect may vary from person to person. If a crack defect recognition neural network is constructed for each worker's welding method and effect, the project workload is large and the cost-effectiveness is low.

[0005] Based on this, how to effectively improve the efficiency of steel cage crack quality detection is a problem that needs to be solved urgently by technical personnel in this field. Summary of the invention

[0006] In order to solve the technical problem of how to effectively improve the efficiency of steel cage crack quality detection, the present invention provides a steel cage welding quality intelligent detection system and method.

[0007] In a first aspect, the present invention provides a method for intelligently detecting the welding quality of a steel cage, which adopts the following technical solution:

[0008] A method for intelligently detecting the welding quality of a steel cage comprises the following steps:

[0009] The welding texture image of the steel cage is processed by an edge detection algorithm to obtain an edge connected domain of each welding texture, wherein the edge connected domain contains a plurality of pixel points; any edge connected domain is taken as a target connected domain, and the abnormal amount of the target connected domain is determined, and the abnormal amount is positively correlated with the ordinate of the pixel point in the target connected domain; the absolute difference between the abnormal amount of the target connected domain and the adjacent edge connected domains on the left and right sides is obtained respectively, and the morphological difference amount of the target connected domain is determined, and the morphological difference amount is positively correlated with the absolute difference; any pixel point in the target connected domain is taken as a target pixel point, and the neighboring pixel points of the target pixel point are obtained, and the sum of the grayscale differences between the target pixel point and each neighboring pixel point is determined; and the crack degree of the target pixel point is obtained, and the crack degree is positively correlated with the sum of the grayscale differences and the morphological difference amount; the target grayscale value is obtained by multiplying the crack degree of the target pixel point and the grayscale value, and after adjusting the grayscale value of the target pixel point to the target grayscale value, the quality detection result of the welding texture image is obtained by clustering algorithm detection.

[0010] The effect is that: the present invention takes into account that there may be large differences in morphology between crack pixels and other pixels, and therefore obtains the morphological difference amount of the target connected domain by analyzing the changes of each pixel in the target connected domain. On this basis, since the grayscale value between the crack pixel and other pixels may be different, the grayscale value of each pixel in the target connected domain is adjusted according to the morphological difference amount of each pixel in the target connected domain, and the gap between the crack pixel and the normal pixel is increased before detection, which can effectively improve the quality detection efficiency of the welding texture image.

[0011] According to a steel cage welding quality intelligent detection method provided by the present invention, the determination of the abnormal amount of the target connected domain includes: obtaining the minimum circumscribed rectangle of the target connected domain, determining the angle between the minimum circumscribed rectangle and the welding direction in the welding pattern image; calculating the abnormal amount of the target connected domain:

[0012] ;

[0013] In the formula, represents the abnormal amount of the i-th target connected domain, represents the angle between the minimum circumscribed rectangle of the i-th target connected domain and the welding direction in the welding pattern image, represents the number of pixels in the i-th target connected domain, represents the ordinate of the jth pixel in the i-th target connected domain, Represents the ordinate of the j+1th pixel in the i-th target connected domain.

[0014] According to a method for intelligently detecting welding quality of a steel cage provided by the present invention, the morphological difference amount of the target connected domain satisfies the relationship:

[0015] ;

[0016] In the formula, represents the morphological difference of the i-th target connected domain, represents the number of adjacent edge connected domains of the i-th target connected domain, represents the abnormal amount of the i-th target connected domain, represents the abnormal amount of the Lth adjacent edge connected domain on the left side of the i-th target connected domain, Represents the anomaly of the Lth adjacent edge connected domain to the right of the i-th target connected domain.

[0017] The effect is that the present invention provides an accurate method for calculating the morphological difference of a target connected domain, and the morphological difference of the target connected domain can be accurately obtained by analyzing the difference in the abnormal amount between the target connected domain and the adjacent edge connected domain.

[0018] According to an intelligent detection method for steel cage welding quality provided by the present invention, the crack degree of the target pixel point is obtained, including: taking the edge connected domain with the smallest difference in morphological difference with the target connected domain in the welding texture image as the similar connected domain of the target connected domain; in the similar connected domain, taking the pixel point with the shortest distance to the target pixel point as the similar pixel point of the target pixel point; taking the sum of the grayscale differences between the target pixel point and each neighboring pixel point as the first feature, taking the sum of the grayscale differences between the similar pixel point and each neighboring pixel point of the similar pixel point as the second feature, determining the crack degree of the target pixel point, and the crack degree is positively correlated with the deviation between the first feature and the second feature.

[0019] The effect is that: the present invention takes into account that the difference in morphological difference between the target connected domain and the similar connected domain is the smallest, and therefore combines the similar connected domain and the neighborhood pixel points for combined analysis from two aspects, thereby effectively improving the accuracy of the crack degree of each pixel point in the obtained target connected domain.

[0020] According to a method for intelligently detecting welding quality of a steel cage provided by the present invention, the crack degree of the target pixel point satisfies the relationship:

[0021] ;

[0022] In the formula, Indicates the crack degree of the pth target pixel in the i-th target connected domain, represents the morphological difference of the i-th target connected domain, Represents the morphological difference of the similar connected domain of the i-th target connected domain, Represents the number of neighborhood pixels, Represents the gray value of the pth target pixel in the i-th target connected domain, Represents the gray value of the pth similar pixel in the i-th target connected domain, The qth neighboring pixel of the pth target pixel in the i-th target connected domain, Represents the gray value of the qth neighboring pixel of the pth similar pixel in the i-th target connected domain.

[0023] The effect is that the present invention provides a method for calculating the crack degree of a pixel point, which can accurately obtain the crack degree of each pixel point. The higher the crack degree, the higher the possibility that the pixel point is a crack pixel point.

[0024] According to a method for intelligently detecting the welding quality of a steel cage provided by the present invention, the quality detection result of the welding texture image is obtained by detecting with a clustering algorithm, comprising: adjusting the target pixel points based on the target grayscale value, and then clustering all the pixel points in the target connected domain by a K-means clustering algorithm to obtain the quality detection result of the welding texture image.

[0025] According to a method for intelligently detecting the welding quality of a steel cage provided by the present invention, after obtaining the quality detection result of the welding texture image, it also includes: after marking the quality detection result, displaying the welding texture image and the corresponding quality detection result externally.

[0026] The effect is that in order to facilitate the staff to process the welding pattern images with cracks in time, the quality inspection results can be marked and fed back to the staff.

[0027] In a second aspect, the present invention provides an intelligent detection system for steel cage welding quality, which adopts the following technical solution:

[0028] A steel cage welding quality intelligent detection system comprises: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned steel cage welding quality intelligent detection method is implemented.

[0029] By adopting the above technical solution, the above-mentioned intelligent detection method for welding quality of steel cage is generated into a computer program and stored in a memory so as to be loaded and executed by a processor, thereby making a terminal device based on the memory and the processor for easy use.

[0030] The present invention has the following technical effects:

[0031] Based on the above scheme, when detecting the welding quality of the steel cage, the present invention takes into account that there will be a large difference in the morphology between the crack pixel point and other pixel points. Therefore, by analyzing the changes in each pixel point in the target connected domain, the abnormal amount of the target connected domain is obtained, and combined with the changes in the abnormal amount between the target connected domain and the adjacent edge connected domain, the morphological difference amount of the target connected domain is accurately obtained. On this basis, since the grayscale value between the crack pixel point and other pixel points will be different, the grayscale value of each pixel point is adjusted by the morphological difference amount of each pixel point in the target connected domain, and the gap between the crack pixel point and the normal pixel point is increased before detection, which can effectively improve the quality detection efficiency of the welding texture image. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0033] Figure 1 A schematic diagram of a flow chart of a method for intelligently detecting welding quality of a reinforcement cage provided by an embodiment of the present invention;

[0034] Figure 2 A schematic diagram of a welding pattern image provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0035] 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 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0036] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0037] It should be noted that cracks may exist at the welding points of the steel cage due to personal technical reasons. Although some cracks can be identified through the neural network model, the welding process of the steel cage may be performed by multiple workers, and the welding methods and effects vary from person to person. If a crack defect recognition neural network is constructed for each worker's welding method and effect, the project workload will be large and the cost-effectiveness will be low.

[0038] Based on this, the embodiment of the present invention discloses an intelligent detection method for steel cage welding quality, referring to Figure 1 , Figure 1 A flow chart of an intelligent detection method for steel cage welding quality provided in an embodiment of the present invention. In this scheme, the target grayscale value of each pixel in the welding texture image can be obtained by calculating the degree of cracks of each welding texture in the welding texture image, and then the pixel points are adjusted based on the target grayscale value to increase the difference between pixel points where cracks may exist and other pixel points, so as to facilitate subsequent clustering processing of the pixel points through a clustering algorithm, such as a K-means clustering algorithm, thereby obtaining the quality detection result of the steel cage welding, which can effectively improve the quality detection efficiency of the steel cage welding.

[0039] In order to implement the above solution, the following steps S1 to S5 may be performed:

[0040] S1: The welding pattern image of the steel cage is processed by edge detection algorithm to obtain the edge connected domain of each welding pattern.

[0041] For example, you can use a high-definition industrial camera to collect the welding texture image of the steel cage, and then process the welding texture image through semantic segmentation. Figure 2 As shown, Figure 2 A schematic diagram of a welding pattern image provided for an embodiment of the present invention shows that, in the absence of crack interference, the direction of the welding pattern is longitudinal, and the overall welding direction is horizontal from left to right; the minimum circumscribed rectangle of the edge connected domain can be obtained through algorithm processing, and there will be an angle between the minimum circumscribed rectangle and the welding direction, and the angle will vary with the change of the welding pattern. In order to make the range of the angle change in the interval of 0-90 degrees for subsequent analysis, a rectangular coordinate system can be constructed by the overall trend of the minimum circumscribed rectangle to obtain the angle: if the overall trend of the minimum circumscribed rectangle is tilted to the right, a rectangular coordinate system is constructed at the lower left corner of the minimum circumscribed rectangle; if the overall trend of the minimum circumscribed rectangle is tilted to the left, a rectangular coordinate system is constructed at the lower right corner of the minimum circumscribed rectangle.

[0042] Combination Figure 2 It can be seen from the example that the overall trend of the minimum circumscribed rectangle in the figure is tilted to the right, so a rectangular coordinate system is constructed at the lower left corner of the minimum circumscribed rectangle of the edge connected domain to obtain the angle between the minimum circumscribed rectangle and the welding direction.

[0043] Among them, the edge detection algorithm can be a canny algorithm, and the edge connected domain obtained after processing by the edge detection algorithm contains multiple pixel points; the minimum enclosing rectangle can be determined by rotating the search area of ​​the enclosing rectangle with the smallest area; the edge detection algorithm and the minimum enclosing rectangle acquisition algorithm can be specifically set according to actual needs, and the embodiment of the present invention does not impose too many restrictions on this.

[0044] After the edge connected domain of each welding pattern is obtained based on the above-mentioned method, the abnormality of each pixel point in the edge connected domain can be analyzed to obtain the crack degree of each pixel point, that is, the following steps are performed:

[0045] S2: Take any edge connected domain as the target connected domain, obtain the ordinate of the pixel point of the target connected domain, and determine the abnormal amount of the target connected domain.

[0046] It should be noted that since the welding lines are formed by stacking layers of solder, the overall trend of the welding lines in the obtained welding line image is that the changes are larger on the vertical axis and smaller on the horizontal axis. Based on this, the embodiment of the present invention analyzes the changes in the welding lines in the vertical direction to obtain the abnormal conditions of the edge connected domains of the welding lines.

[0047] For example, in an embodiment of the present invention, when determining the abnormal amount of the target connected domain, the minimum circumscribed rectangle of the target connected domain can be obtained, and the angle A between the minimum circumscribed rectangle and the welding direction in the welding pattern image can be determined; finally, the abnormal amount of the target connected domain is calculated, and the abnormal amount of the target connected domain is positively correlated with the vertical coordinate of the pixel points of the target connected domain.

[0048] For example, the abnormal amount of the target connected domain can be determined by referring to the following formula:

[0049] ;

[0050] In the formula, represents the abnormal amount of the i-th target connected domain, represents the angle between the minimum circumscribed rectangle of the i-th target connected domain and the welding direction in the welding pattern image, represents the number of pixels in the i-th target connected domain, represents the ordinate of the jth pixel in the i-th target connected domain, Represents the ordinate of the j+1th pixel in the i-th target connected domain.

[0051] Based on the above formula, the overall morphological abnormality of each edge connected domain can be obtained, where It can characterize the degree to which the edge connected domain in the welding texture image is affected by the crack. The larger the value, the greater the degree to which the edge connected domain is affected by the crack, and the greater the possibility of the existence of cracks in the edge connected domain. Conversely, the smaller the value, the smaller the degree to which the edge connected domain is affected by the crack, and the smaller the possibility of the existence of cracks in the edge connected domain.

[0052] It should be noted that if we only focus on the edge connected domain of a single welding pattern, there may be a situation where a crack runs vertically through the edge connected domain. In this way, the difference between the pixels in the edge connected domain will be small, but cracks actually exist. Based on this, the morphological difference of each edge connected domain can be obtained by analyzing the abnormal changes between the edge connected domain and the adjacent connected domain in the welding pattern image:

[0053] S3: Obtain the absolute difference between the abnormal amount of the target connected domain and the adjacent edge connected domains on the left and right sides respectively, and determine the morphological difference amount of the target connected domain.

[0054] Among them, the morphological difference is positively correlated with the absolute difference between the target connected domain and the abnormal amount of the adjacent edge connected domains on the left and right sides.

[0055] It should be noted that the welding patterns of the same technician are similar on the whole, so the difference in the anomaly between the edge connected domains will be smaller. Based on this, the morphological difference of the target connected domain can be obtained by analyzing the anomaly between the target connected domain and the adjacent edge connected domains on the left and right sides.

[0056] For example, in an embodiment of the present invention, the morphological difference amount of the target connected domain satisfies the relationship:

[0057] ;

[0058] In the formula, represents the morphological difference of the i-th target connected domain, represents the number of adjacent edge connected domains of the i-th target connected domain, represents the abnormal amount of the i-th target connected domain, represents the abnormal amount of the Lth adjacent edge connected domain on the left side of the i-th target connected domain, Represents the anomaly of the Lth adjacent edge connected domain to the right of the i-th target connected domain.

[0059] In the above formula, It can characterize the morphological difference between the edge connected domain and the adjacent edge connected domains on the left and right sides. The larger the value, the greater the possibility of cracks in the target connected domain; conversely, the smaller the value, the smaller the possibility of cracks in the target connected domain.

[0060] Based on the above method, by analyzing the difference between the edge connected domain and the adjacent edge connected domain in the welding pattern image, the overall morphological difference of each edge connected domain can be obtained. On this basis, in order to more accurately obtain the crack degree of each pixel point in the edge connected domain, the crack degree can be determined by combining the neighboring pixel points of each pixel point in the edge connected domain, that is, performing the following steps:

[0061] S4: taking any pixel point in the target connected domain as the target pixel point, obtaining the neighboring pixel points of the target pixel point, determining the sum of the grayscale differences between the target pixel point and each neighboring pixel point, and obtaining the crack degree of the target pixel point.

[0062] For example, in an embodiment of the present invention, when obtaining the degree of crack of a target pixel, the grayscale difference between the target pixel and the neighboring pixel can be analyzed to finally obtain the grayscale difference of the target pixel, wherein the degree of crack of the target pixel is positively correlated with the sum of the grayscale differences and is positively correlated with the morphological difference.

[0063] It should be noted that if there are cracks in the pixels of the edge connected domain, the grayscale value of the crack pixels will be different from that of the normal pixels. Therefore, the target connected domain and the similar connected domain can be combined by obtaining the similar connected domain with the smallest difference in morphological difference with the target connected domain in the welding texture image, and the crack degree can be further optimized through the grayscale value difference of the pixels to accurately obtain the crack degree of each pixel.

[0064] For example, in an embodiment of the present invention, when obtaining the degree of crack of a target pixel, the edge connected domain in the welding texture image with the smallest difference in morphological difference with the target connected domain can be used as a similar connected domain of the target connected domain; in the similar connected domain, the pixel with the shortest distance to the target pixel is used as a similar pixel of the target pixel; the sum of the grayscale differences between the target pixel and each neighboring pixel is used as the first feature, and the sum of the grayscale differences between the similar pixel and each neighboring pixel of the similar pixel is used as the second feature to determine the degree of crack of the target pixel, and the degree of crack is positively correlated with the deviation between the first feature and the second feature.

[0065] For example, in the embodiment of the present invention, the crack degree of the target pixel point is determined, and the specific formula may be referred to as follows:

[0066] ;

[0067] In the formula, Indicates the crack degree of the pth target pixel in the i-th target connected domain, represents the morphological difference of the i-th target connected domain, Represents the morphological difference of the similar connected domain of the i-th target connected domain, Represents the number of neighborhood pixels, Represents the gray value of the pth target pixel in the i-th target connected domain, Represents the gray value of the pth similar pixel in the i-th target connected domain, The qth neighboring pixel of the pth target pixel in the i-th target connected domain, Represents the gray value of the qth neighboring pixel of the pth similar pixel in the i-th target connected domain.

[0068] In the above formula, if the pth target pixel has 8 neighboring pixels, the number of its neighboring pixels is 8; if the pth target pixel has 4 neighboring pixels, the number of its neighboring pixels is 4, which can be set according to actual needs. In an embodiment of the invention, in order to obtain the accurate crack degree of each pixel, the number of neighboring pixels can be 8.

[0069] After the crack degree of each pixel point is accurately obtained based on the above steps S1 to S4, the grayscale value of the pixel point can be adjusted based on the crack degree of each pixel point, thereby increasing the difference between the pixel point where cracks may exist and other pixel points:

[0070] S5: The target gray value is obtained by multiplying the crack degree and the gray value of the target pixel point. After the gray value of the target pixel point is adjusted to the target gray value, the quality detection result of the welding texture image is obtained by clustering algorithm detection.

[0071] For example, in an embodiment of the present invention, when the quality inspection result of the welding texture image is obtained, the target pixel points can be adjusted based on the target grayscale value, and then all the pixel points in the target connected domain can be clustered through a clustering algorithm to obtain the quality inspection result of the welding texture image.

[0072] For example, the K-means clustering algorithm can be used to cluster the pixels. During the clustering process, the K value can be set to 2, and a grayscale threshold can be set. The grayscale values ​​of pixels greater than or equal to the grayscale threshold are set to 255, and the grayscale values ​​of pixels less than the grayscale threshold are set to 0, and finally the position of the crack pixel is obtained.

[0073] It can be seen that in the embodiment of the present invention, when the welding quality of the steel cage is detected, the welding texture image of the steel cage can be processed by an edge detection algorithm to obtain an edge connected domain of each welding texture, and the edge connected domain contains multiple pixel points; any edge connected domain is used as a target connected domain, the ordinate of the pixel point of the target connected domain is obtained, and the abnormal amount of the target connected domain is determined, and the abnormal amount is positively correlated with the ordinate of the pixel point of the target connected domain; the absolute difference between the abnormal amount of the target connected domain and the adjacent edge connected domains on the left and right sides is obtained respectively, and the morphological difference amount of the target connected domain is determined, and the morphological difference amount is positively correlated with the absolute difference; any pixel point in the target connected domain is used as a target pixel point, the neighborhood pixel point of the target pixel point is obtained, and the sum of the grayscale differences between the target pixel point and each neighborhood pixel point is determined; and the crack degree of the target pixel point is obtained, and the crack degree is positively correlated with the sum of the grayscale differences and the morphological difference amount; the target grayscale value is obtained by multiplying the crack degree of the target pixel point and the grayscale value, and the grayscale value of the target pixel point is adjusted to the target grayscale value, and the quality detection result of the welding texture image is obtained by detection.

[0074] In this way, when detecting the welding quality of the steel cage, the present invention takes into account that there will be a large difference in the morphology between the crack pixel point and other pixel points. Therefore, by analyzing the changes in each pixel point in the target connected domain, the abnormal amount of the target connected domain is obtained, and combined with the changes in the abnormal amount between the target connected domain and the adjacent edge connected domain, the morphological difference amount of the target connected domain is accurately obtained. On this basis, since there will be a difference in the grayscale value between the crack pixel point and other pixel points, the grayscale value of each pixel point is adjusted according to the morphological difference amount of each pixel point in the target connected domain, thereby increasing the gap between the crack pixel point and the normal pixel point, thereby effectively improving the quality detection efficiency of the welding texture image.

[0075] After obtaining the quality inspection result of the welding pattern image based on the above embodiment, the quality inspection result can also be processed:

[0076] For example, in an embodiment of the present invention, after obtaining the quality inspection result of the welding pattern image, it is also possible to: mark the quality inspection result, and then display the welding pattern image and the corresponding quality inspection result to the outside.

[0077] This processing makes it easier for staff to process the welding texture images with cracks in a timely manner.

[0078] An embodiment of the present invention further discloses a steel cage welding quality intelligent detection system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a steel cage welding quality intelligent detection method according to the present invention is implemented.

[0079] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0080] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM, a dynamic random access memory DRAM, a static random access memory SRAM, an enhanced dynamic random access memory EDRAM, a high bandwidth memory HBM, a hybrid memory cube HMC, etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0081] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0082] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent detection method for steel cage welding quality, characterized in that: include: Processing the welding pattern image of the steel cage by an edge detection algorithm to obtain an edge connected domain of each welding pattern, wherein the edge connected domain includes a plurality of pixel points; Taking any edge connected domain as a target connected domain, determining the abnormal amount of the target connected domain includes: obtaining the minimum circumscribed rectangle of the target connected domain, determining the angle between the minimum circumscribed rectangle and the welding direction in the welding pattern image; calculating the abnormal amount of the target connected domain: ； In the formula, represents the abnormal amount of the i-th target connected domain, represents the angle between the minimum circumscribed rectangle of the i-th target connected domain and the welding direction in the welding pattern image, represents the number of pixels in the i-th target connected domain, represents the ordinate of the jth pixel in the i-th target connected domain, represents the ordinate of the j+1th pixel point in the i-th target connected domain, and the abnormality is positively correlated with the ordinate of the pixel point in the target connected domain; respectively obtain the absolute difference between the abnormality of the target connected domain and the adjacent edge connected domains on the left and right sides, and determine the morphological difference of the target connected domain, and the morphological difference is positively correlated with the absolute difference; Taking any pixel point in the target connected domain as the target pixel point, obtaining the neighboring pixel points of the target pixel point, determining the sum of the grayscale differences between the target pixel point and each neighboring pixel point; and obtaining the crack degree of the target pixel point, wherein the crack degree is positively correlated with the sum of the grayscale differences and positively correlated with the morphological difference amount; The crack degree of the target pixel satisfies the relationship: ； In the formula, Indicates the crack degree of the pth target pixel in the i-th target connected domain, represents the morphological difference of the i-th target connected domain, Represents the morphological difference of the similar connected domain of the i-th target connected domain, Represents the number of neighborhood pixels, Represents the gray value of the pth target pixel in the i-th target connected domain, Represents the gray value of the pth similar pixel in the i-th target connected domain, The qth neighboring pixel of the pth target pixel in the i-th target connected domain, Represents the gray value of the qth neighboring pixel of the pth similar pixel in the i-th target connected domain; The target grayscale value is obtained by multiplying the crack degree and the grayscale value of the target pixel point. After the grayscale value of the target pixel point is adjusted to the target grayscale value, the quality detection result of the welding texture image is obtained by clustering algorithm detection.

2. The intelligent detection method for steel cage welding quality according to claim 1 is characterized in that: The morphological difference of the target connected domain satisfies the relationship: ； In the formula, represents the morphological difference of the i-th target connected domain, represents the number of adjacent edge connected domains of the i-th target connected domain, represents the abnormal amount of the i-th target connected domain, represents the abnormal amount of the Lth adjacent edge connected domain on the left side of the i-th target connected domain, Represents the anomaly of the Lth adjacent edge connected domain to the right of the i-th target connected domain.

3. The intelligent detection method for steel cage welding quality according to claim 1 is characterized in that: The step of obtaining the crack degree of the target pixel point includes: The edge connected domain with the smallest difference in morphological difference with the target connected domain in the welding texture image is used as a similar connected domain of the target connected domain; in the similar connected domain, the pixel point with the shortest distance to the target pixel point is used as a similar pixel point to the target pixel point; The sum of the grayscale differences between the target pixel and each of its neighboring pixels is taken as the first feature, and the sum of the grayscale differences between the similar pixel and each of its neighboring pixels is taken as the second feature to determine the degree of crack of the target pixel, and the degree of crack is positively correlated with the deviation between the first feature and the second feature.

4. The intelligent detection method for steel cage welding quality according to claim 1 is characterized in that: The quality detection result of the welding pattern image obtained by clustering algorithm detection includes: After adjusting the target pixel points based on the target grayscale value, clustering processing is performed on all pixel points in the target connected domain using a K-means clustering algorithm to obtain a quality detection result of the welding texture image.

5. A steel cage welding quality intelligent detection method according to claim 1 or 4, characterized in that: After obtaining the quality inspection result of the welding texture image, the method further includes: After marking the quality inspection result, the welding pattern image and the corresponding quality inspection result are displayed externally.

6. An intelligent detection system for steel cage welding quality, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an intelligent detection method for the welding quality of a steel cage according to any one of claims 1 to 5 is implemented.

Images