A crack defect image recognition device and method for weld magnetic particle inspection

By integrating edge computing modules and grayscale cameras in magnetic powder flaw detection equipment, automatic identification of weld crack defects is solved, and the problem of difficult detection of small cracks and frequent misjudgment in the existing technology is solved, and the accuracy of detection is improved Sex and efficiency.

CN118229634BActive Publication Date: 2024-11-15SHUNDE BRANCH GUANGDONG INST OF SPECIAL EQUIP INSPECTION & RES +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410299898.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-03-15
Publication Date
2024-11-15
Estimated Expiration
2044-03-15

AI Technical Summary

Technical Problem

When detecting small cracks on the surface of the workpiece, the existing magnetic powder flaw detection method relies on naked eye detection, which can easily lead to misjudgment and misjudgment. The level of the detector is high, and it is difficult to accurately identify crack defects.

Method used

Design a device that includes an edge computing module and a grayscale camera. Through real-time image acquisition and algorithm processing, it realizes automatic identification of weld crack defects. Use edge computing module to castrate images, threshold extraction and feature extraction to judge the image in the image. Cracked defects.

Benefits of technology

It effectively reduces misjudgment and misjudgment in crack defect image recognition, improves the accuracy and efficiency of detection, and assists detectors in defect identification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118229634B_ABST
    Figure CN118229634B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of special detection, and specifically discloses a crack defect image recognition device and method for magnetic particle flaw detection of welds; the present invention designs a miniaturized device including an edge computing module and a grayscale camera which is integrated into a magnetic particle flaw detection device, and performs real-time image acquisition of the weld flaw detection position through a grayscale camera connected to the edge computing module, and then uses an algorithm integrated in the edge computing device to distinguish crack defects, thereby realizing offline identification of weld magnetic particle flaw detection crack defects, effectively assisting magnetic particle flaw detection personnel in defect identification, and reducing missed identification and misjudgment in image recognition of workpiece crack defects.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of special detection, and in particular relates to a crack defect image recognition device and method for magnetic particle flaw detection of welds. Background Art

[0002] After the ferromagnetic workpiece is magnetized, the magnetic lines of force on the surface and near the surface of the workpiece are locally distorted due to the existence of discontinuities, generating leakage magnetic fields, which adsorb magnetic powder applied to the surface of the workpiece and form visually visible magnetic marks under appropriate lighting, thereby showing the location, size, shape and severity of the discontinuity.

[0003] In the Chinese patent with publication number CN109856234B, a magnetic particle flaw detection device is mentioned, including a workbench, a plurality of support columns are uniformly welded at the lower end of the workbench, baffles are parallelly welded on both sides of the upper end of the workbench, two sets of connecting mechanisms are arranged at the lower ends of both sides of the workbench, and a collection box is connected through the two sets of connecting mechanisms, a handle is also installed in the middle of one side of the collection box, a sliding mechanism is arranged at the upper end of the workbench, and a straight plate is slidably arranged through the sliding mechanism, a square hole is opened in the middle of the upper end of the straight plate, infusion tubes running through the workbench are opened on both sides of the inside of the workbench, and the infusion tubes are located at the upper end of the collection box, a support rod is vertically fixed on one side of the upper end of the workbench, and the support rod is located between the two baffles, a support plate is horizontally installed on the upper end of the support rod, and an actuator is also arranged on the upper end of the support plate. The present invention has a stable structure, is easy to operate, and has a good clamping effect on the workpiece.

[0004] During the application process, the applicant discovered that the above application had defects during use. The magnetic particle flaw detection method disclosed in the application magnetizes the workpiece, adjusts the distance between the light source and the magnetized workpiece, and then detects crack damage on the surface of the workpiece by naked eye. However, there are fine cracks on the surface of some workpieces that are not easy to find with the naked eye. It requires a high level of detection from the inspectors, which can easily lead to missed and misjudgment of defects. Summary of the invention

[0005] The object of the present invention is to provide a crack defect image recognition device and method for weld magnetic particle inspection to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A crack defect image recognition device for weld magnetic particle inspection, comprising:

[0008] A device housing, wherein a mounting hole is provided on the device housing and the device housing is fixedly mounted on the magnetic particle flaw detection device through the mounting hole;

[0009] An edge computing module, wherein the edge computing module is fixedly installed inside the device housing;

[0010] A grayscale camera is fixedly installed inside the device housing, and the grayscale camera is fixedly connected to the computing module via a thermally conductive silicone pad.

[0011] Preferably, the edge computing module includes a PCB board and a minimum system circuit and an interface circuit integrated on the PCB board, and the grayscale camera is connected to the interface circuit via a cable communication.

[0012] A crack defect image recognition method for weld magnetic particle inspection specifically comprises the following steps:

[0013] S1, after the device is started, the entire algorithm process begins to run;

[0014] S2, using the grayscale camera to collect image information of the flaw detection position in real time and transmit it to the edge computing module;

[0015] S3, the edge computing module parses the image information according to the established camera data transmission protocol and generates a two-dimensional array containing the original image information, recorded as Image_Data0[N0][M0];

[0016] S4. The edge computing module performs regional castration on the image according to the predetermined castration ratio. The castration coefficients are recorded as K1 and K2, where K1 is the column castration coefficient and K2 is the row castration coefficient. The two castration coefficients are determined by the magnetization mode, magnetization intensity and camera height of the deployed magnetic particle flaw detection equipment. After the regional castration, the two-dimensional array containing valid image information is recorded as Image_Data1[N1][M1], where N1 is the number of column pixels after the regional castration and M1 is the number of row pixels after the regional castration.

[0017] S5, performing threshold extraction through the two-dimensional array Image_Data1[N1][M1] containing valid image information, and obtaining the grayscale threshold of the current image, recorded as Threshold;

[0018] S6. By taking the row and column as the calculation direction, respectively, feature extraction is performed with the help of grayscale threshold, the array position corresponding to the feature point is found, and all the array positions are counted into two arrays containing feature positions, Position_0[N2] and Position_1[M2], where N2 is the total number of feature points in the column direction, and M2 is the total number of feature points in the row direction;

[0019] S7, determining whether there is a crack defect by judging the row and column data in the two arrays Position_0[N2] and Position_1[M2] containing the characteristic positions;

[0020] S8. Report the judgment result through the serial port or WI FI, and repeat the algorithm process.

[0021] Preferably, in S3, N0 is the number of pixel points in the camera column, which is 128, and M0 is the number of pixel points in the camera row, which is 128.

[0022] Preferably, in S4, according to the information of the deployed magnetic particle flaw detection equipment, the castration coefficient takes the average value of 0.5, so N1 = K1 × N0 = 0.5 × 128 = 64, M1 = K2 × M0 = 0.5 × 128 = 64, so the two-dimensional array containing valid image information obtained after the defined area is castrated is recorded as Image_Data1[N1][M1], and there is (wherein, the value of i ranges from 0 to N1-1, and the value of j ranges from 0 to M1-1), completing the transfer of image information.

[0023] Preferably, in S5, the grayscale threshold is calculated based on the image information after the region is cut off.

[0024]

[0025] Preferably, in S6, feature extraction is completed through the grayscale threshold, the size of the pixel value and the grayscale threshold is determined, and the vertical coordinates of the pixels less than the grayscale threshold are stored in the Position_0[N2] array, and the horizontal coordinates are stored in the Position_1[M2] array.

[0026] Preferably, the judgment is made based on the continuity of the position of the feature points in the row and column information array, the judgment threshold is set to 3, the difference between the position data between the rows and columns is calculated, and the final defect judgment result is recorded as End, |(Position_0[i]-Position_0[i-1])|≤3(i ranges from 1 to N2)|

[0027] When (Position_1[i]-Position_1[i-1])|≤3(i ranges from 1 to M2) is true, End takes the value of 1, and |(Position_0[i]-Position_0[i-1])|≤3(i ranges from 1 to N2) and |

[0028] When any one of (Position_1[i]-Position_1[i-1])|≤3 (i ranges from 1 to M2) is not true or both of them are not true, End takes the value of 0.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] The present invention designs a miniaturized device including an edge computing module and a grayscale camera which is integrated into a magnetic particle flaw detection device. The grayscale camera connected to the edge computing module is used to collect real-time images of the weld flaw detection position, and then the crack defects are identified with the help of the algorithm integrated in the edge computing device. This enables the identification of weld magnetic particle flaw detection crack defects in an offline form, effectively assisting magnetic particle flaw detection personnel in defect identification, and reducing missed identification and misjudgment in image recognition of workpiece crack defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a structural schematic diagram of a crack defect image recognition device of the present invention;

[0032] Figure 2 The schematic diagram of the circuit board structure of the edge computing module of the present invention is shown in FIG. Figure 1 ;

[0033] Figure 3 The schematic diagram of the circuit board structure of the edge computing module of the present invention is shown in FIG. Figure 2 ;

[0034] Figure 4 is the minimum system circuit diagram of the present invention;

[0035] Figure 5 is an interface circuit diagram of the present invention;

[0036] Figure 6 It is a structural block diagram of the crack defect image recognition method of the present invention;

[0037] In the figure: 1. Device housing; 2. Edge computing module; 3. Grayscale camera. DETAILED DESCRIPTION

[0038] 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.

[0039] Example:

[0040] See also Figure 1 - Figure 6 As shown, a crack defect image recognition device for magnetic particle inspection of welds comprises:

[0041] A device housing 1, wherein a mounting hole is provided on the device housing 1, and the device housing 1 is fixedly mounted on the magnetic particle flaw detection device through the mounting hole;

[0042] An edge computing module 2, wherein the edge computing module 2 is fixedly installed inside the device housing 1;

[0043] Grayscale camera 3, the grayscale camera 3 is fixedly installed inside the device housing 1, and the grayscale camera 3 is fixedly connected to the calculation module 2 through a thermally conductive silicone pad.

[0044] The edge computing module 2 includes a PCB board and a minimum system circuit and an interface circuit integrated on the PCB board, and the grayscale camera 3 is connected to the interface circuit via a cable communication.

[0045] A crack defect image recognition method for weld magnetic particle inspection specifically comprises the following steps:

[0046] S1, after the device is started, the entire algorithm process begins to run;

[0047] S2, using the grayscale camera 3 to collect image information of the flaw detection position in real time, and transmit it to the edge computing module 2;

[0048] S3, edge computing module 2 parses the image information according to the established camera data transmission protocol and generates a two-dimensional array containing the original image information, recorded as Image_Data0[N0][M0], where N0 is the number of camera column pixels, which is 128, and M0 is the number of camera row pixels, which is 128;

[0049] S4. The edge computing module 2 performs regional castration on the image according to the established castration ratio. The castration coefficients are denoted as K1 and K2, where K1 is the column castration coefficient and K2 is the row castration coefficient. The two castration coefficients are determined by the magnetization mode, magnetization intensity and camera height of the deployed magnetic particle flaw detection equipment. After the regional castration, the two-dimensional array containing valid image information is obtained and denoted as Image_Data1[N1][M1]. Where N1 is the number of column pixels after the regional castration, and M1 is the number of row pixels after the regional castration. According to the information of the deployed magnetic particle flaw detection equipment, the castration coefficient takes an average value of 0.5, so N1=K1×N0=0.5×128=64, M1=K2×M0=0.5×128=64, so the two-dimensional array containing valid image information after the regional castration is defined as Image_Data1[N1][M1], and there is (wherein the value of i ranges from 0 to N1-1, and the value of j ranges from 0 to M1-1), completing the transfer of image information;

[0050] S5. Threshold extraction is performed through the two-dimensional array Image_Data1[N1][M1] containing valid image information to obtain the grayscale threshold of the current image, recorded as Threshold, and the grayscale threshold is calculated based on the image information after the region is cut off.

[0051] S6. Perform feature extraction with the help of grayscale thresholds using rows and columns as calculation directions, find the array positions corresponding to the feature points, and count all array positions into two arrays containing feature positions, Position_0[N2] and Position_1[M2], where N2 is the total number of feature points in the column direction and M2 is the total number of feature points in the row direction. Complete feature extraction using grayscale thresholds, determine the size of the pixel value and the grayscale threshold, and store the ordinates of the pixels that are less than the grayscale threshold into the Position_0[N2] array, and the abscissas into the Position_1[M2] array.

[0052] S7. Determine whether there is a crack defect by judging the row and column data in the two arrays Position_0[N2] and Position_1[M2] containing the characteristic positions. Judge according to the continuity of the characteristic point positions in the row and column information arrays. Set the judgment threshold to 3, calculate the difference between the position data between the rows and columns, and obtain the final defect judgment result, which is recorded as End. When both |(Position_0[i]-Position_0[i-1])|≤3 (i from 1 to N2) and |(Position_1[i]-Position_1[i-1])|≤3 (i from 1 to M2) are true, End is 1. When either |(Position_0[i]-Position_0[i-1])|≤3 (i from 1 to N2) or |(Position_1[i]-Position_1[i1])|≤3 (i from 1 to M2) is false or both are false, End is 0.

[0053] S8. Report the judgment result through the serial port or WIFI, and repeat the algorithm process.

[0054] The standard parts used in the present invention can all be purchased from the market, and the special-shaped parts can be customized according to the description and the drawings. The specific connection methods of each part adopt conventional means such as mature bolts, rivets, welding, etc. in the prior art. The machinery, parts and equipment all adopt conventional models in the prior art, and the circuit connection adopts the conventional connection method in the prior art, which will not be described in detail here. The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

[0055] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0056] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may mean that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, a first feature being "above", "above" or "above" a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. A first feature being "below", "below" or "below" a second feature may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is lower in level than the second feature.

[0057] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0058] In the drawings of the embodiments disclosed in the present invention, only the structures related to the embodiments disclosed in the present invention are involved, and other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of the present invention can be combined with each other.

Claims

1. A crack defect image recognition method for weld magnetic particle inspection, using a crack defect image recognition device for weld magnetic particle inspection, the crack defect image recognition device is composed of a shell (1), an edge computing module (2) and a grayscale camera (3), the grayscale camera (3) is fixedly installed inside the device shell (1), and the grayscale camera (3) and the edge computing module (2) are fixedly connected through a thermal conductive silicone pad, the edge computing module (2) includes a PCB board and a minimum system circuit and an interface circuit integrated on the PCB board, the grayscale camera (3) and the interface circuit are connected through a cable communication, characterized in that The specific steps include: S1, after the device is started, the entire algorithm process begins to run; S2, using a grayscale camera (3) to collect image information of the flaw detection position in real time, and transmit it to the edge computing module (2); S3, the edge computing module (2) parses the image information according to the established camera data transmission protocol and generates a two-dimensional array containing the original image information, recorded as Image_Data0[N0][M0]; S4. The edge computing module (2) performs regional castration on the image according to a predetermined castration ratio. The castration coefficients are denoted as K1 and K2, where K1 is the column castration coefficient and K2 is the row castration coefficient. The two castration coefficients are determined by the magnetization mode, magnetization intensity and camera height of the deployed magnetic particle flaw detection equipment. After the regional castration, a two-dimensional array containing valid image information is obtained and denoted as Image_Data1[N1][M1], where N1 is the number of column pixels after the regional castration and M1 is the number of row pixels after the regional castration. S5, performing threshold extraction through the two-dimensional array Image_Data1[N1][M1] containing valid image information, and obtaining the grayscale threshold of the current image, recorded as Threshold; S6, extracting features by using grayscale thresholds in the row and column directions, finding the array positions corresponding to the feature points, and counting all array positions into two arrays containing feature positions, Position_0[N2] and Position_1[M2], where N2 is the total number of feature points in the column direction, and M2 is the total number of feature points in the row direction; S7, determining whether there is a crack defect by judging the row and column data in the two arrays Position_0[N2] and Position_1[M2] containing the characteristic positions; S8. Report the judgment result through the serial port or WIFI, and repeat the algorithm process.

2. The method for crack defect image recognition for weld magnetic particle testing according to claim 1, characterized in that: In S3, N0 is the number of pixel points in the camera column, which is 128, and M0 is the number of pixel points in the camera row, which is 128.

3. The method for crack defect image recognition for weld magnetic particle testing according to claim 2, characterized in that: In S4, according to the information of the deployed magnetic particle flaw detection equipment, the castration coefficient takes the average value of 0.5, so N1 = K1 × N0 = 0.5 × 128 = 64, M1 = K2 × M0 = 0.5 × 128 = 64, so after the defined area is castrated, the two-dimensional array containing valid image information is recorded as Image_Data1[N1][M1], and there is j], where the value of i ranges from 0 to N1-1, and the value of j ranges from 0 to M1-1, completing the transfer of image information.

4. The method for crack defect image recognition for weld magnetic particle testing according to claim 3 is characterized in that: In S5, the grayscale threshold is calculated based on the image information after the region is cut off.

5. The crack defect image recognition method for weld magnetic particle inspection according to claim 4 is characterized in that: In S6, feature extraction is completed through the grayscale threshold, the size of the pixel value and the grayscale threshold is determined, and the vertical coordinates of the pixels less than the grayscale threshold are stored in the Position_0[N2] array, and the horizontal coordinates are stored in the Position_1[M2] array.

6. The method for crack defect image recognition for weld magnetic particle testing according to claim 5, characterized in that: In S7, judgment is made based on the continuity of the position of the feature points in the row and column information array, the judgment threshold is set to 3, the difference in the position data between the rows and columns is calculated, and the final defect judgment result is recorded as End. When |(Position_0[i]-Position_0[i-1])|≤3, i ranges from 1 to N2, and |(Position_1[i]-Position_1[i-1])|≤3, i ranges from 1 to M2, both conditions are met, End is 1, |(Position_0[i]-Position_0[i-1])|≤3, i ranges from 1 to N2, and |(Position_1[i]-Position_1[i-1])|≤3, i ranges from 1 to M2, and when any one of the conditions is not met or both are not met, End is 0.