Defect detection device and system for ocean engineering steel plate

Abstract

The invention discloses a defect detection device and system for an ocean engineering steel plate, and relates to the technical field of material detection. The image processing and control module comprises an image quality verification unit, an image quality verification unit and a control unit; the image quality verification unit performs quality analysis on the surface image data of the steel plate to obtain an image evaluation value; the image quality verification unit sets an image evaluation threshold according to the image evaluation value, and when the image evaluation threshold is lower than the threshold, a recollection signaling is generated. Through the image quality verification and verification unit, the steel plate surface image quality can be accurately analyzed, substandard images can be automatically recollected, the defect detection module judges defects according to multiple indexes such as the color shadow value and the three-dimensional deviation value, the visual detector can adjust the angle to collect the images, and the detection accuracy is improved. An ultrasonic sensor is combined to construct a steel plate three-dimensional model for auxiliary detection, multiple means ensure that the detection result is accurate and reliable, omission defects are avoided, and the quality of the steel plate for ocean engineering is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of material detection, and in particular to a defect detection device and system for marine engineering steel plates. Background Art

[0002] Marine engineering projects such as offshore dam projects, submarine tunnel projects, submarine pipelines, etc., have extremely strict requirements on the quality and performance of the steel plates used because of the complex and harsh marine environment in which they are located, the strong corrosiveness of seawater, and the impact and vibration caused by waves and currents. Any minor defects may cause serious safety problems in long-term use;

[0003] Traditional manual visual inspection methods have significant defects. They only perform a single image acquisition operation and have no solution to the situation of poor image quality caused by poor lighting, stains on the steel plate surface or equipment imaging problems. Low-quality images will inevitably make the defect features unclear, greatly increasing the probability of misjudgment and missed judgment, and seriously threatening the reliability of the inspection results;

[0004] At the same time, the experience of inspectors varies, and fatigue caused by long-term work will distract their attention. Coupled with the inevitability of subjective judgment, the inspection efficiency is extremely low. Therefore, it is necessary to propose a defect detection device and system for marine engineering steel plates. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a defect detection device and system for marine engineering steel plates.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a defect detection device for marine engineering steel plates, comprising a positioning plate and three support columns fixedly connected at equal distances to the lower end of the positioning plate, a conveying motor is installed on one side of the positioning plate, a driving shaft of the conveying motor passes through the positioning plate and is clamped on one end of a roller between the positioning plates, the rollers are horizontally and equidistantly installed between the positioning plates, and a conveyor belt is sleeved and installed on the outer side of the roller, and limiting grooves are equidistantly provided on the conveyor belt, a gantry is installed on the upper middle end of the conveyor belt, a lifting and lowering adjustment mechanism is installed on the gantry, and a horizontal moving mechanism is installed on the gantry;

[0007] The horizontal moving mechanism includes a bolt rod horizontally installed on the inner side of the T-shaped moving groove, the bolt rod is installed on the moving seat through a threaded connection, and one end of the bolt rod passes through the lifting plate and is fixedly connected to one end of the driven wheel on the inner side of the fixed block, the driven wheel and the driving wheel are meshed with each other, and a driving shaft of a driving motor is clamped in the middle of the driving wheel, and the driving motor is installed on one side of the fixed block; a rotating disk is rotatably installed on one side of the moving seat, and a visual detector is fixedly connected to one side of the rotating disk.

[0008] Preferably, both ends of the gantry are fixedly connected to the middle of the positioning plates on both sides, and the upper end of the gantry is provided with positioning grooves at both ends of the inner side.

[0009] Preferably, the lifting and adjusting mechanism includes a through hole opened in the middle of the upper end of the gantry, and a telescopic cylinder is installed in the middle of the upper end of the gantry, the telescopic end of the telescopic cylinder passes through the through hole and is fixedly connected to the middle of the upper end of the lifting plate, and both ends of the lifting plate are vertically slid and clamped in the positioning groove.

[0010] Preferably, T-shaped movable grooves are horizontally opened inward on both sides of one end of the lifting plate, and movable grooves are opened in the vertical direction of the gantries at both ends of the lifting plate.

[0011] A defect detection system for marine engineering steel plates, comprising a data acquisition module, an image processing and control module;

[0012] A data acquisition module is used to collect surface image data of the steel plate through a visual detector and perform preprocessing operations, and send the preprocessed surface image data to the image processing and control module;

[0013] The image processing and control module includes a monitoring setting unit, an image quality checking unit, an image quality verification unit and a control unit; the monitoring setting unit is used to mark the time area between the start and end of the visual detector scanning according to the preset scanning path as a monitoring time zone;

[0014] An image quality verification unit is used to perform quality analysis on the surface image data of the steel plate to obtain a quality assessment result; wherein the quality assessment result includes the grayscale impact value, contrast, gradient amplitude, signal-to-noise ratio difference, resolution difference, quality assessment value, image assessment value, and the quality average value and quality fluctuation value of the surface image;

[0015] An image quality verification unit is used to receive the quality evaluation result of the steel plate surface image data, set an image evaluation threshold, and if the image evaluation value is less than the image evaluation threshold, it means that the image quality of the surface image does not meet the standard, record the position of the surface image corresponding to the acquisition time in the preset scanning path as a re-acquisition position, and generate a re-acquisition signal; record all re-acquisition positions in the scanning path as re-acquisition information;

[0016] The control unit is used to control the drive motor to start when receiving the re-capture signal, so that the visual detector moves horizontally along a preset scanning path, and re-captures the image of the steel plate surface when it is in the re-capture position in the scanning path.

[0017] Preferably, the present invention further comprises a defect detection module and a defect determination module;

[0018] The defect detection module is used to receive the surface image data of the steel plate to perform defect detection analysis and obtain defect analysis results; wherein the defect analysis results include color shadow values, three-dimensional deviation values ​​and defect assessment values;

[0019] The defect determination module is used to receive the defect analysis result and compare the defect analysis result with the preset threshold value. If the color and shadow value is greater than the preset threshold value, it indicates that there are abnormalities in the color and light and shadow of the steel plate surface, and a color abnormality signal is generated; if the three-dimensional deviation value is greater than the preset threshold value, it indicates that the surface shape of the steel plate has a large deviation from the standard shape, and a shape abnormality signal is generated; if the defect assessment value is greater than the preset threshold value, it indicates that there are defects in the steel plate, and a steel plate defect signal is generated;

[0020] Color abnormal signaling, shape abnormal signaling and steel plate abnormal signaling are marked as an abnormal signaling group.

[0021] Preferably, the surface image data of the steel plate is received for defect detection analysis, and the specific analysis method is as follows:

[0022] Acquire surface image data of the steel plate, including the surface image at any acquisition time in the monitoring time zone; identify the edge contour of the steel plate in the surface image by using an edge detection operator; and mark the area within the edge contour as the steel plate area;

[0023] The surface image in the steel plate area is enlarged to obtain a pixel grid image of the steel plate area, the color value of the pixel grid image of the steel plate area is identified, the standard color value of the steel plate is set, and the color value of any pixel grid in the pixel grid image of the steel plate area is subtracted from the standard color value of the steel plate to obtain a color difference value; a number of color intervals are set, and a color weight is assigned to each color interval; the color value of the pixel grid in the pixel grid image of the steel plate area is matched with the value range of the color interval, the pixel grids in the same color interval are marked as similar pixels, and the area of ​​the similar pixels is calculated to obtain the color surface value in the same area; all the color surface values ​​in the same area are weightedly calculated with the color weights of the corresponding color intervals to obtain the color shadow value;

[0024] The visual detector is also provided with an ultrasonic sensor; when the visual detector moves along a preset scanning path, the ultrasonic sensor transmits a signal to the surface of the steel plate in the form of a high-frequency sound wave pulse, and receives an echo reflected from the surface of the steel plate, and by measuring the time difference between the emission and reception of the sound wave, combined with the propagation speed of the sound wave in the air, the formula: spacing = propagation speed × time difference / 2 is used to obtain the spacing between any point on the surface of the steel plate;

[0025] The distance between the visual sensor and all points on the steel plate surface is fused by using the distance surface reconstruction algorithm based on point cloud to obtain the three-dimensional model of the steel plate surface.

[0026] A standard 3D model library for steel plates is set, and a standard 3D model of the current steel plate is extracted from the standard 3D model library; the 3D model of the steel plate surface is compared with its standard 3D model, and difference measurement indicators between the 3D model of the steel plate surface and its standard 3D model are calculated, including average distance deviation and root mean square error;

[0027] The average distance deviation and the root mean square error are weighted to obtain the three-dimensional deviation value;

[0028] The three-dimensional deviation value and the color shadow value are weighted to obtain the defect evaluation value of the steel plate;

[0029] The color shadow value, three-dimensional deviation value and defect assessment value are marked as defect analysis results.

[0030] As a preferred method, the surface image data of the steel plate is subjected to quality analysis, and the specific analysis process is as follows:

[0031] Acquire surface image data of the steel plate, including the surface image at any acquisition time within the monitoring time zone;

[0032] The surface image data is magnified to obtain a pixel grid image, and the grayscale value of any pixel grid in the pixel grid image is identified; a grayscale range group is set, including several grayscale ranges, and a grayscale weight is assigned to each grayscale range; the grayscale values ​​of all pixel grids in the pixel grid image are matched with the value range of the grayscale range group, and the pixel grids in the same grayscale range are marked as pixels with the same grayscale; the area of ​​pixels with the same grayscale is calculated to obtain the same grayscale face value; the grayscale weight corresponding to the same grayscale face value and its grayscale range is weighted to obtain the grayscale influence value;

[0033] The grayscale values ​​of all pixels in the pixel grid image are averaged to obtain the grayscale average value; the maximum grayscale value and the minimum grayscale value in the pixel grid are identified, and the difference between the maximum grayscale value and the minimum grayscale value is divided by the grayscale average value to obtain the image contrast;

[0034] Use edge detection operators to extract image edges and calculate the gradient amplitude of image edges;

[0035] Obtaining the signal-to-noise ratio and actual resolution of the surface image; setting the expected signal-to-noise ratio and expected resolution; subtracting the expected signal-to-noise ratio from the signal-to-noise ratio of the surface image to obtain the signal-to-noise ratio difference, and subtracting the expected resolution from the actual resolution to obtain the resolution difference;

[0036] The grayscale impact value, contrast, gradient amplitude, signal-to-noise ratio difference, and resolution difference of the surface image are weighted to obtain a quality assessment value of the surface image;

[0037] The quality assessment value of the surface image at any acquisition time in the monitoring time zone is calculated by mean and variance to obtain the quality average value and quality fluctuation value; the quality assessment value of the surface image is weighted with the quality average value and quality fluctuation value of the surface image in the monitoring time zone to obtain the image assessment value of the surface image.

[0038] Preferably, the data acquisition module uses an adaptive median filtering algorithm to remove noise from the surface image data of the steel plate during preprocessing operations, wherein the window size of the adaptive median filtering algorithm is dynamically adjusted according to the local noise characteristics of the image, and the adjustment range is 3x3 pixels to 9x9 pixels.

[0039] Preferably, after generating the abnormal signaling, the defect judgment module sends the abnormal signaling group to the remote monitoring terminal through the wireless communication module. After receiving the abnormal signaling, the remote monitoring terminal prompts on the display screen with warning lights of different colors according to the type of abnormal signaling, among which the color abnormal signaling corresponds to the yellow warning light, the shape abnormal signaling corresponds to the blue warning light, and the steel plate defect signaling corresponds to the red warning light.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. The present invention is equipped with a conveying motor and a conveyor belt, which can automatically transport steel plates for inspection, saving manpower and time. At the same time, the lifting and adjusting mechanism and the horizontal moving mechanism can flexibly adjust the position of the visual detector to achieve all-round rapid scanning. Compared with traditional manual visual inspection, the inspection speed is greatly improved, which is suitable for large-scale steel plate inspection needs and effectively improves the inspection efficiency of marine engineering steel plates.

[0042] 2. The present invention can accurately analyze the surface image quality of steel plates through the image quality verification and validation unit, automatically re-collect images that do not meet the standards, and the defect detection module judges defects based on multiple indicators such as color shadow value and three-dimensional deviation value. The visual detector can adjust the angle to collect images, and also combines with ultrasonic sensors to build a three-dimensional model of the steel plate for auxiliary detection. Multiple means ensure that the detection results are accurate and reliable, avoid missing defects, and ensure the quality of steel plates used in marine engineering.

[0043] 3. The defect detection module in the present invention can comprehensively consider indicators such as color shadow value and three-dimensional deviation value to perform multi-dimensional defect detection on the surface of the steel plate and accurately find out the potential defect locations. On this basis, the defect judgment module makes a scientific judgment on the detected abnormal situation according to the set standards and algorithms to clarify whether it is a real defect and the severity of the defect, thereby achieving precise control of the quality of marine engineering steel plates and ensuring product quality and safety in use. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0045] Figure 1 This is a schematic diagram of the overall structure of a defect detection device for marine engineering steel plates proposed by the present invention;

[0046] Figure 2 This is a schematic diagram of the overall three-dimensional structure of the other side proposed by the present invention;

[0047] Figure 3 This is the overall three-dimensional structure proposed by the present invention when viewed from above;

[0048] Figure 4 A side cross-sectional structural schematic diagram of the present invention;

[0049] Figure 5 It is a schematic diagram of the partial overall three-dimensional structure proposed by the present invention;

[0050] Figure 6 The present invention proposes Figure 5 A schematic diagram of the enlarged structure in the middle;

[0051] Figure 7 This is a principle block diagram of a defect detection system for marine engineering steel plates proposed by the present invention.

[0052] Serial numbers in the figure: 1. Support column; 2. Positioning plate; 3. Conveyor belt; 4. Limiting groove; 5. Gantry; 6. Positioning groove; 7. Telescopic cylinder; 8. Lifting plate; 9. Driving motor; 10. Fixing block; 11. Conveying motor; 12. Bolt rod; 13. Driving wheel; 14. Driven wheel; 15. Visual detector; 16. Moving seat; 17. Electric rotating disk. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0054] Example: See Figure 1-Figure 6The defect detection device for marine engineering steel plate in the present invention comprises a positioning plate 2 and three supporting columns 1 fixedly connected to the lower end of the positioning plate 2 at equal distances, a conveying motor 11 is installed on one side of the positioning plate 2, a driving shaft of the conveying motor 11 passes through the positioning plate 2 and is clamped on one end of a roller between the positioning plates 2, the rollers are horizontally and equidistantly installed between the positioning plates 2, and a conveyor belt 3 is sleeved and installed on the outer side of the roller, and a limiting groove 4 is equidistantly provided on the conveyor belt 3, a gantry 5 is installed on the middle upper end of the conveyor belt 3, a lifting and lowering adjustment mechanism is installed on the gantry 5, and a horizontal moving mechanism is installed on the gantry 5, the conveying motor 11 and the conveyor The conveyor belt 3 is used for transporting and inspecting the steel plates; both ends of the gantry 5 are fixedly connected to the middle of the positioning plates 2 on both sides, and the upper end of the gantry 5 is provided with positioning grooves 6 at both inner ends, and the gantry 5 and the positioning grooves 6 are used to improve the stability of the reinforcement mechanism; the lifting and adjusting mechanism includes a through hole opened in the middle of the upper end of the gantry 5, and a telescopic cylinder 7 is installed in the middle of the upper end of the gantry 5, and the telescopic end of the telescopic cylinder 7 passes through the through hole and is fixedly connected to the middle of the upper end of the lifting plate 8, and both ends of the lifting plate 8 are vertically slidably clamped in the positioning grooves 6, and the lifting plate 8 is controlled to move by the telescopic cylinder 7, so that the lifting plate 8 is lifted up and down.

[0055] In the present invention, both sides of one end of the lifting plate 8 are horizontally inwardly provided with T-shaped moving grooves, and the gantry frames 5 at both ends of the lifting plate 8 are vertically provided with moving grooves, and the lifting and horizontal movement are facilitated by the T-shaped moving grooves and the moving grooves; the horizontal movement mechanism includes a bolt rod 12 horizontally installed on the inner side of the T-shaped moving groove, the bolt rod 12 is installed on the moving seat 16 through a threaded connection, and one end of the bolt rod 12 passes through the lifting plate 8 and is fixed to one end of the driven wheel 14 inside the fixed block 10, the driven wheel 14 is meshed with the driving wheel 13, and the driving shaft of the driving motor 9 is clamped in the middle of the driving wheel 13. On one side of the fixed block 10 where the driving motor 9 is installed, the driving motor 9 is used to move the moving seat 16 through the driving wheel 13 and the driven wheel 14; a rotating disk 17 is rotatably installed on one side of the moving seat 16, and a visual detector 15 is fixedly connected to one side of the rotating disk 17. While the driving motor 9 drives the moving seat 16 to move, the visual detector 15 continuously collects image data of the surface of the steel plate. At the same time, the inclination angle of the visual detector 15 can be adjusted by the rotating disk 17, thereby collecting surface image data of steel plates at different angles, thereby further ensuring the accuracy of steel plate surface defect detection.

[0056] Working principle: When the present invention is used, the detection device is first supported and fixed by the support column 1 and the positioning plate 2, and then the steel plate to be monitored is placed on the conveyor belt 3, and the conveyor belt 3 is driven to operate by starting the conveying motor 11, so that the steel plate to be detected is moved, and the limit groove 4 ensures the stability of the steel plate when moving. The lifting plate 8 is vertically moved in the positioning groove 6 and the moving groove by controlling the telescopic cylinder 7. When the lifting plate 8 reaches the specified height, the driving motor 9 is started, and the output shaft of the driving motor 9 drives the moving seat 16 on the bolt rod 12 to move horizontally in the T-shaped moving groove through the driving wheel 13 and the driven wheel 14, so that the visual detector 15 performs a comprehensive scanning and detection of the steel plate surface in the horizontal direction according to the preset scanning path; during the entire detection process, the visual detector 15 continuously collects image data of the steel plate surface; at the same time, the inclination angle of the visual detector 15 can be adjusted by the rotating disk 17, so as to collect surface image data of steel plates at different angles, and further ensure the accuracy of steel plate surface defect detection.

[0057] A defect detection system for marine engineering steel plates, comprising a data acquisition module, an image processing and control module;

[0058] A data acquisition module, used to collect surface image data of the steel plate through the visual detector 15 and perform preprocessing operations, and send the preprocessed surface image data to the image processing and control module;

[0059] The image processing and control module includes a monitoring setting unit, an image quality checking unit, an image quality verification unit and a control unit; the monitoring setting unit is used to mark the time area between the start and end of the scanning of the visual detector 15 according to the preset scanning path as a monitoring time zone;

[0060] An image quality verification unit is used to perform quality analysis on the surface image data of the steel plate to obtain a quality assessment result; wherein the quality assessment result includes the grayscale impact value, contrast, gradient amplitude, signal-to-noise ratio difference, resolution difference, quality assessment value, image assessment value, and the quality average value and quality fluctuation value of the surface image;

[0061] An image quality verification unit is used to receive the quality evaluation result of the steel plate surface image data, set an image evaluation threshold, and if the image evaluation value is less than the image evaluation threshold, it means that the image quality of the surface image does not meet the standard, record the position of the surface image corresponding to the acquisition time in the preset scanning path as a re-acquisition position, and generate a re-acquisition signal; record all re-acquisition positions in the scanning path as re-acquisition information;

[0062] The control unit is used to receive and control the drive motor 9 to start when the re-capture signal is generated, so that the visual detector 15 moves horizontally according to the preset scanning path, and re-captures the image of the steel plate surface when it is in the re-capture position in the scanning path.

[0063] In this application, the present invention also includes a defect detection module and a defect determination module;

[0064] The defect detection module is used to receive the surface image data of the steel plate to perform defect detection analysis and obtain defect analysis results; wherein the defect analysis results include color shadow values, three-dimensional deviation values ​​and defect assessment values;

[0065] The defect determination module is used to receive the defect analysis result and compare the defect analysis result with the preset threshold value. If the color and shadow value is greater than the preset threshold value, it indicates that there are abnormalities in the color and light and shadow of the steel plate surface, and a color abnormality signal is generated; if the three-dimensional deviation value is greater than the preset threshold value, it indicates that the surface shape of the steel plate has a large deviation from the standard shape, and a shape abnormality signal is generated; if the defect assessment value is greater than the preset threshold value, it indicates that there are defects in the steel plate, and a steel plate defect signal is generated;

[0066] Color abnormal signaling, shape abnormal signaling and steel plate abnormal signaling are marked as an abnormal signaling group.

[0067] In this application, the surface image data of the steel plate is received for defect detection and analysis. The specific analysis method is as follows:

[0068] Acquire surface image data of the steel plate, including the surface image at any acquisition time in the monitoring time zone; identify the edge contour of the steel plate in the surface image by using an edge detection operator; and mark the area within the edge contour as the steel plate area;

[0069] The surface image in the steel plate area is enlarged to obtain a pixel grid image of the steel plate area, the color value of the pixel grid image of the steel plate area is identified, the standard color value of the steel plate is set, and the color value of any pixel grid in the pixel grid image of the steel plate area is subtracted from the standard color value of the steel plate to obtain a color difference value; a number of color intervals are set, and a color weight is assigned to each color interval; the color value of the pixel grid in the pixel grid image of the steel plate area is matched with the value range of the color interval, the pixel grids in the same color interval are marked as similar pixels, and the area of ​​the similar pixels is calculated to obtain the color surface value in the same area; all the color surface values ​​in the same area are weightedly calculated with the color weights of the corresponding color intervals to obtain the color shadow value;

[0070] The visual detector 15 is also provided with an ultrasonic sensor; when the visual detector 15 moves along a preset scanning path, the ultrasonic sensor transmits a signal to the surface of the steel plate in the form of a high-frequency sound wave pulse, and receives an echo reflected from the surface of the steel plate. By measuring the time difference between the emission and reception of the sound wave, combined with the propagation speed of the sound wave in the air, the distance between the visual sensor 15 and any point on the surface of the steel plate is obtained using the formula: distance = propagation speed × time difference / 2;

[0071] A point cloud-based spacing surface reconstruction algorithm is used to fuse the spacing between the visual sensor 15 and all points on the steel plate surface to obtain a three-dimensional model of the steel plate surface;

[0072] A standard 3D model library for steel plates is set, and a standard 3D model of the current steel plate is extracted from the standard 3D model library; the 3D model of the steel plate surface is compared with its standard 3D model, and difference measurement indicators between the 3D model of the steel plate surface and its standard 3D model are calculated, including average distance deviation and root mean square error;

[0073] The average distance deviation and the root mean square error are weighted to obtain the three-dimensional deviation value;

[0074] The three-dimensional deviation value and the color shadow value are weighted to obtain the defect evaluation value of the steel plate;

[0075] The color shadow value, three-dimensional deviation value and defect assessment value are marked as defect analysis results.

[0076] In this application, the quality analysis of the surface image data of the steel plate is performed, and the specific analysis process is as follows:

[0077] Acquire surface image data of the steel plate, including the surface image at any acquisition time within the monitoring time zone;

[0078] The surface image data is magnified to obtain a pixel grid image, and the grayscale value of any pixel grid in the pixel grid image is identified; a grayscale range group is set, including several grayscale ranges, and a grayscale weight is assigned to each grayscale range; the grayscale values ​​of all pixel grids in the pixel grid image are matched with the value range of the grayscale range group, and the pixel grids in the same grayscale range are marked as pixels with the same grayscale; the area of ​​pixels with the same grayscale is calculated to obtain the same grayscale face value; the grayscale weight corresponding to the same grayscale face value and its grayscale range is weighted to obtain the grayscale influence value;

[0079] The grayscale values ​​of all pixels in the pixel grid image are averaged to obtain the grayscale average value; the maximum grayscale value and the minimum grayscale value in the pixel grid are identified, and the difference between the maximum grayscale value and the minimum grayscale value is divided by the grayscale average value to obtain the image contrast;

[0080] Use edge detection operators to extract image edges and calculate the gradient amplitude of image edges;

[0081] Obtaining the signal-to-noise ratio and actual resolution of the surface image; setting the expected signal-to-noise ratio and expected resolution; subtracting the expected signal-to-noise ratio from the signal-to-noise ratio of the surface image to obtain the signal-to-noise ratio difference, and subtracting the expected resolution from the actual resolution to obtain the resolution difference;

[0082] The grayscale impact value, contrast, gradient amplitude, signal-to-noise ratio difference, and resolution difference of the surface image are weighted to obtain a quality assessment value of the surface image;

[0083] The quality assessment value of the surface image at any acquisition time in the monitoring time zone is calculated by mean and variance to obtain the quality average value and quality fluctuation value; the quality assessment value of the surface image is weighted with the quality average value and quality fluctuation value of the surface image in the monitoring time zone to obtain the image assessment value of the surface image.

[0084] In the present application, when performing preprocessing operations, the data acquisition module uses an adaptive median filtering algorithm to remove noise from the surface image data of the steel plate, wherein the window size of the adaptive median filtering algorithm is dynamically adjusted according to the local noise characteristics of the image, and the adjustment range is 3x3 pixels to 9x9 pixels.

[0085] In the present application, after generating an abnormal signal, the defect judgment module sends the abnormal signaling group to the remote monitoring terminal through the wireless communication module. After receiving the abnormal signal, the remote monitoring terminal prompts on the display screen with warning lights of different colors according to the type of abnormal signal, among which the color abnormal signal corresponds to the yellow warning light, the shape abnormal signal corresponds to the blue warning light, and the steel plate defect signal corresponds to the red warning light.

[0086] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A defect detection device for marine engineering steel plates, comprising a positioning plate (2) and three support columns (1) fixedly connected to the lower end of the positioning plate (2) at equal distances, characterized in that: A conveying motor (11) is installed on one side of the positioning plate (2), and a driving shaft of the conveying motor (11) passes through the positioning plate (2) and is clamped on one end of a roller between the positioning plates (2). The roller is installed horizontally and equidistantly between the positioning plates (2), and a conveyor belt (3) is sleeved and installed on the outer side of the roller. Limiting grooves (4) are arranged at equal intervals on the conveyor belt (3). A gantry (5) is installed at the upper middle end of the conveyor belt (3), and a lifting and lowering adjustment mechanism is installed on the gantry (5), and a horizontal moving mechanism is installed on the gantry (5); The horizontal moving mechanism comprises a bolt rod (12) horizontally mounted on the inner side of a T-shaped moving groove, the bolt rod (12) being mounted on a moving seat (16) through a threaded connection, and one end of the bolt rod (12) passing through a lifting plate (8) being fixedly connected to one end of a driven wheel (14) on the inner side of a fixed block (10), the driven wheel (14) being meshed with a driving wheel (13), a driving shaft of a driving motor (9) being clamped in the middle of the driving wheel (13), and one side of the fixed block (10) on which the driving motor (9) is mounted; a rotating disk (17) is rotatably mounted on one side of the moving seat (16), and a visual detector (15) is fixedly connected to one side of the rotating disk (17).

2. A defect detection device for marine engineering steel plates according to claim 1, characterized in that: Both ends of the gantry (5) are fixedly connected to the middle of the positioning plates (2) on both sides, and the upper end of the gantry (5) is provided with positioning grooves (6) at both ends of the inner side.

3. A defect detection device for marine engineering steel plates according to claim 1, characterized in that: The lifting and lowering adjustment mechanism comprises a through hole formed in the middle of the upper end of the gantry (5), and a telescopic cylinder (7) is installed in the middle of the upper end of the gantry (5). The telescopic end of the telescopic cylinder (7) passes through the through hole and is fixedly connected to the middle of the upper end of the lifting plate (8). Both ends of the lifting plate (8) are vertically slidably clamped in the positioning groove (6).

4. A defect detection device for marine engineering steel plates according to claim 3, characterized in that: Both sides of one end of the lifting plate (8) are provided with T-shaped moving grooves inwardly and horizontally, and the gantry frames (5) at both ends of the lifting plate (8) are provided with moving grooves in the vertical direction.

5. A defect detection system for marine engineering steel plates, using a defect detection device for marine engineering steel plates according to any one of claims 1 to 4, characterized in that: It includes data acquisition module, image processing and control module; A data acquisition module, used for acquiring surface image data of the steel plate through a visual detector (15) and performing a preprocessing operation, and sending the preprocessed surface image data to an image processing and control module; The image processing and control module comprises a monitoring setting unit, an image quality checking unit, an image quality verification unit and a control unit; the monitoring setting unit is used to mark the time zone between the start and end of scanning of the visual detector (15) according to a preset scanning path as a monitoring time zone; An image quality verification unit is used to perform quality analysis on the surface image data of the steel plate to obtain a quality assessment result; wherein the quality assessment result includes the grayscale impact value, contrast, gradient amplitude, signal-to-noise ratio difference, resolution difference, quality assessment value, image assessment value, and the quality average value and quality fluctuation value of the surface image; An image quality verification unit is used to receive the quality evaluation result of the steel plate surface image data, set an image evaluation threshold, and if the image evaluation value is less than the image evaluation threshold, it means that the image quality of the surface image does not meet the standard, record the position of the surface image corresponding to the acquisition time in the preset scanning path as a re-acquisition position, and generate a re-acquisition signal; record all re-acquisition positions in the scanning path as re-acquisition information; The control unit is used to control the drive motor (9) to start when receiving the re-collection signal, so that the visual detector (15) moves horizontally according to a preset scanning path, and re-collects images of the steel plate surface when it is at a re-collection position in the scanning path.

6. A defect detection system for marine engineering steel plates according to claim 5, characterized in that: It also includes a defect detection module and a defect determination module; A defect detection module is used to receive surface image data of the steel plate to perform defect detection analysis and obtain defect analysis results; The defect analysis results include color shadow value, three-dimensional deviation value and defect assessment value; The defect determination module is used to receive the defect analysis result and compare the defect analysis result with the preset threshold value. If the color and shadow value is greater than the preset threshold value, it indicates that there are abnormalities in the color and light and shadow of the steel plate surface, and a color abnormality signal is generated; if the three-dimensional deviation value is greater than the preset threshold value, it indicates that the surface shape of the steel plate has a large deviation from the standard shape, and a shape abnormality signal is generated; if the defect assessment value is greater than the preset threshold value, it indicates that there are defects in the steel plate, and a steel plate defect signal is generated; Color abnormal signaling, shape abnormal signaling and steel plate abnormal signaling are marked as an abnormal signaling group.

7. A defect detection system for marine engineering steel plates according to claim 6, characterized in that: Receive the surface image data of the steel plate for defect detection and analysis. The specific analysis method is as follows: Acquire surface image data of the steel plate, including the surface image at any acquisition time in the monitoring time zone; identify the edge contour of the steel plate in the surface image by using an edge detection operator; and mark the area within the edge contour as the steel plate area; The surface image in the steel plate area is enlarged to obtain a pixel grid image of the steel plate area, the color value of the pixel grid image of the steel plate area is identified, the standard color value of the steel plate is set, and the color value of any pixel grid in the pixel grid image of the steel plate area is subtracted from the standard color value of the steel plate to obtain a color difference value; a number of color intervals are set, and a color weight is assigned to each color interval; Match the color value of the pixel grid in the pixel grid image of the steel plate area with the value range of the color interval, mark the pixel grids in the same color interval as similar pixels, calculate the area of ​​similar pixels to obtain the color surface value of the same area; perform weighted calculation on all the color surface values ​​of the same area and the color weights of the corresponding color interval to obtain the color shadow value; The visual detector (15) is also provided with an ultrasonic sensor; when the visual detector 15 moves along a preset scanning path, the ultrasonic sensor transmits a signal to the surface of the steel plate in the form of a high-frequency sound wave pulse, and receives an echo reflected from the surface of the steel plate, and by measuring the time difference between the emission and reception of the sound wave, combined with the propagation speed of the sound wave in the air, the formula: spacing = propagation speed × time difference / 2 is used to obtain the spacing between the visual sensor (15) and any point on the surface of the steel plate; A point cloud-based spacing surface reconstruction algorithm is used to fuse the spacing between the visual sensor (15) and all points on the steel plate surface to obtain a three-dimensional model of the steel plate surface; A standard 3D model library for steel plates is set, and a standard 3D model of the current steel plate is extracted from the standard 3D model library; the 3D model of the steel plate surface is compared with its standard 3D model, and difference measurement indicators between the 3D model of the steel plate surface and its standard 3D model are calculated, including average distance deviation and root mean square error; The average distance deviation and the root mean square error are weighted to obtain the three-dimensional deviation value; The three-dimensional deviation value and the color shadow value are weighted to obtain the defect evaluation value of the steel plate; The color shadow value, three-dimensional deviation value and defect assessment value are marked as defect analysis results.

8. A defect detection system for marine engineering steel plates according to claim 5, characterized in that: The quality analysis of the surface image data of the steel plate is carried out, and the specific analysis process is as follows: Acquire surface image data of the steel plate, including the surface image at any acquisition time within the monitoring time zone; The surface image data is magnified to obtain a pixel grid image, and the grayscale value of any pixel grid in the pixel grid image is identified; a grayscale range group is set, including a plurality of grayscale ranges, and a grayscale weight is assigned to each grayscale range; Match the grayscale values ​​of all pixels in the pixel grid image with the value range of the grayscale range group, and mark the pixels in the same grayscale range as pixels with the same grayscale; Calculate the area of ​​pixels with the same gray level to get the face value with the same gray level; The grayscale weight corresponding to the grayscale face value and its grayscale range is weighted to obtain the grayscale influence value; The grayscale values ​​of all pixels in the pixel grid image are averaged to obtain the grayscale average value; the maximum grayscale value and the minimum grayscale value in the pixel grid are identified, and the difference between the maximum grayscale value and the minimum grayscale value is divided by the grayscale average value to obtain the image contrast; Use edge detection operators to extract image edges and calculate the gradient amplitude of image edges; Obtaining the signal-to-noise ratio and actual resolution of the surface image; setting the expected signal-to-noise ratio and expected resolution; subtracting the expected signal-to-noise ratio from the signal-to-noise ratio of the surface image to obtain the signal-to-noise ratio difference, and subtracting the expected resolution from the actual resolution to obtain the resolution difference; The grayscale impact value, contrast, gradient amplitude, signal-to-noise ratio difference, and resolution difference of the surface image are weighted to obtain a quality assessment value of the surface image; The quality assessment value of the surface image at any acquisition time in the monitoring time zone is calculated by mean and variance to obtain the quality average value and quality fluctuation value; the quality assessment value of the surface image is weighted with the quality average value and quality fluctuation value of the surface image in the monitoring time zone to obtain the image assessment value of the surface image.

9. A defect detection system for marine engineering steel plates according to claim 5, characterized in that: When performing preprocessing operations, the data acquisition module uses an adaptive median filtering algorithm to remove noise from the surface image data of the steel plate, wherein the window size of the adaptive median filtering algorithm is dynamically adjusted according to the local noise characteristics of the image, and the adjustment range is 3x3 pixels to 9x9 pixels.

10. A defect detection system for marine engineering steel plates according to claim 6, characterized in that: After generating the abnormal signaling, the defect judgment module sends the abnormal signaling group to the remote monitoring terminal through the wireless communication module. After receiving the abnormal signaling, the remote monitoring terminal prompts on the display screen with warning lights of different colors according to the type of abnormal signaling, among which the color abnormal signaling corresponds to the yellow warning light, the shape abnormal signaling corresponds to the blue warning light, and the steel plate defect signaling corresponds to the red warning light.

Images