A composite material defect statistics method based on ultrasonic C-scan images

Abstract

A composite material defect statistics method based on ultrasonic C-scan images relates to the technical field of ultrasonic non-destructive testing. It includes an ultrasonic C-scan non-destructive testing system, which is composed of an ultrasonic flaw detector, an ultrasonic probe, an automatic scanning device, a probe position calibration device, a workpiece to be inspected, a scanning imaging industrial control computer, an optical projection device, and scanning imaging software. The workpiece to be inspected is placed horizontally; the ultrasonic transmitting probe, the probe position calibration device, and the optical projection device are located above the workpiece to be inspected; the ultrasonic receiving probe is located below the workpiece to be inspected; the automatic scanning device clamps the ultrasonic probe to complete the scanning inspection of the workpiece; and the scanning imaging software generates a C-scan image. In addition, it includes a C-scan image processing software to automatically count and output the defect information of the C-scan image, generate a defect schematic diagram, and realize efficient and accurate statistics of the internal defect information of the material. Combined with the optical projection device, it assists in marking product defects.

Description

Technical Field

[0001] The invention relates to a composite material defect statistics method based on ultrasonic C-scan images, and belongs to the technical field of ultrasonic non-destructive testing. Background Art

[0002] Currently, although ultrasonic nondestructive testing of composite materials and other products used in the aerospace industry has achieved automated scanning processes, some test result evaluation tasks, such as defect information statistics and defect mapping, are still performed manually, resulting in low efficiency and accuracy, which is a weak link restricting ultrasonic C-scan testing technology. This traditional evaluation method has the following limitations:

[0003] First, efficiency is low. If the inspected product is large and has numerous defects, the manual work of marking, recording, calculating, entering, and reporting the defect locations is extremely time-consuming and can take up more than 50% of the total inspection time, becoming a bottleneck restricting inspection efficiency. Furthermore, errors and omissions are inevitable in this repetitive and extensive work.

[0004] Secondly, multiple steps require two people to operate simultaneously, such as one person measuring with a ruler and one recording data, which results in high labor costs. Furthermore, defect mapping results are significantly influenced by the subjective influence of the operator, and different personnel have different evaluation criteria.

[0005] Third, the defects of manual drawing are low precision, no electronic version, and inconvenient traceability.

[0006] Therefore, how to improve the efficiency, accuracy and automation level of ultrasonic testing defect statistics is an urgent problem to be solved in the ultrasonic C-scan testing widely used in composite materials. Summary of the Invention

[0007] The technical problem solved by the present invention is: to overcome the shortcomings of the existing technology and provide a composite material defect statistics method based on ultrasonic C-scan images, so as to solve the problems of large workload and low precision in ultrasonic C-scan defect statistics of metal and composite materials, and to more accurately quantify defect information and improve the efficiency and reliability of ultrasonic detection result processing.

[0008] The technical solution of the present invention is:

[0009] A composite material defect statistics method based on ultrasonic C-scan images comprises the following steps:

[0010] S201, using an ultrasonic C-scan nondestructive testing system to obtain an ultrasonic C-scan image of the workpiece to be inspected;

[0011] S202, determining a grayscale / hue threshold for judging defects using a defect comparison sample, dividing defect areas and non-defect areas in the ultrasonic C-scan image, and obtaining a black and white image;

[0012] S203, using an image processing algorithm to reduce defect noise and weak connection phenomena in the black and white image;

[0013] S204, identifying defect areas in the black and white image, creating structured two-dimensional data, and saving the defect information in the structured two-dimensional data;

[0014] S205, setting a defect threshold, eliminating defect areas smaller than a specified area in the black and white image according to the defect threshold, and updating the black and white image and structured two-dimensional data;

[0015] S206, extracting the defect outline of the defect area in the black and white image, drawing a defect schematic diagram, marking the defect with a serial number, adding a hatching line inside the defect, and obtaining a defect schematic diagram;

[0016] S207, correcting the position of the defective area in the structured two-dimensional data according to the placement of the workpiece;

[0017] S208, saving the defect schematic diagram, obtaining a defect information table based on the structured two-dimensional data and saving it;

[0018] S209: Draw the defect outline on the inspected workpiece with the help of the defect schematic diagram and the optical projection device.

[0019] The ultrasonic C-scan image is a color or grayscale ultrasonic C-scan image of the inspected workpiece obtained by using an ultrasonic non-destructive testing system and a penetrating automatic ultrasonic C-scan non-destructive testing technology;

[0020] Ultrasonic nondestructive testing system includes ultrasonic flaw detector, ultrasonic transmitting probe, ultrasonic receiving probe, automatic scanning device, probe position calibration device, inspected workpiece, scanning imaging industrial computer, optical projection device, scanning imaging software, and image processing software;

[0021] The ultrasonic flaw detector is connected to an ultrasonic transmitting probe and an ultrasonic receiving probe, and stimulates the ultrasonic transmitting probe to generate ultrasonic waves; the workpiece to be inspected is placed horizontally; the ultrasonic transmitting probe and the probe position calibration device are located above the workpiece to be inspected; the ultrasonic receiving probe is located below the workpiece to be inspected; the automatic scanning device drives the ultrasonic transmitting probe and the ultrasonic receiving probe to move and complete the scanning and detection of the workpiece; the scanning imaging industrial computer uses scanning imaging software to process the received ultrasonic flaw detector signal and the position data of the automatic scanning device to obtain an ultrasonic C-scan image; the scanning imaging industrial computer uses image processing software to process the ultrasonic C-scan image to obtain a defect schematic diagram; the optical projection device projects the defect schematic diagram onto the workpiece to assist in defect drawing.

[0022] The probe position calibration device projects two line light sources onto the surface of the workpiece to be inspected, and the intersection of the line light sources is used to mark the center position of the ultrasonic transmitting probe.

[0023] The step S202 includes:

[0024] S2021. Prepare a comparison sample with the same material and thickness as the workpiece to be inspected and containing artificial defects, perform an ultrasonic C-scan, and obtain a grayscale value or RGB value threshold representing the defect from the image;

[0025] S2022. Using the same detection parameters, complete the ultrasonic C-scan of the inspected workpiece to obtain an ultrasonic C-scan image, determine the grayscale value or RGB value in the ultrasonic C-scan image, set the pixels below the threshold to black, representing the defect area, and set the remaining pixels to white, representing the non-defect area.

[0026] In step S203, the image processing algorithm includes:

[0027] S2031, scanning the ultrasonic C-scan image line by line, eliminating pixel blocks with a width smaller than a specified value, where the width is determined based on the Gaussian noise of the detection device;

[0028] S2032. Reduce image noise using the corrosion and dilation algorithm.

[0029] The step S204 includes: scanning the black and white image obtained in step S203 line by line, where the defective area is a black connected domain, identifying the black connected domain, and storing the coordinates of the black pixels in each black connected domain in structured two-dimensional data; the defect information includes a defect array, and the defect array includes the sequence number, position, and area information of the defective area.

[0030] The step S205 includes: determining the number of coordinate points contained in each array corresponding to the defect area in the structured two-dimensional data according to the set defect threshold, and comparing it with the defect threshold. If the number of coordinate points is less than the defect threshold, deleting the array from the structured two-dimensional data to eliminate the defect area less than the defect threshold. Finally, redrawing the black and white binary image with the specified small defects filtered out based on the processed array information.

[0031] In step S206: the defect contour is extracted by drawing the defect contour through neighborhood calculation or using the Canny contour extraction algorithm; a profile line is drawn within the defect contour, and the defect area is labeled with numbers 1, 2, 3..., based on the pixel coordinates (x, y) of the structured two-dimensional data, a +45° profile line is drawn, the profile line spacing can be set, and the defect number is marked at an appropriate position inside the defect to obtain a defect schematic diagram.

[0032] The S207 includes: measuring the scanning starting point, the scanning first line end point, and the position deviation between the ultrasonic transmitting probe axis and the two corners of the inspected workpiece with the help of a probe position calibration device, and correcting the position information of the defect area in the structured two-dimensional data according to the geometric relationship.

[0033] The step S209 includes: using an optical projection device to project a defect diagram onto the surface of the inspected workpiece, adjusting the position and size of the defect diagram so that the projection of the defect diagram coincides with the inspected workpiece; and drawing all defects on the inspected workpiece with a metal pencil according to the defect outlines displayed by the projection.

[0034] In summary, this application has at least the following beneficial technical effects:

[0035] (1) The ultrasonic C-scan image processing method proposed in the present invention directly obtains defect information and defect schematic diagrams through ultrasonic C-scan images, and the statistics are free of omissions and errors. The detection and evaluation accuracy is high, and an electronic version of the evaluation results is available for easy traceability.

[0036] (2) The ultrasonic C-scan image processing method proposed in the present invention can produce test results in seconds regardless of the size of the product being tested or how complex the test results are, thus saving a lot of test result evaluation time and greatly improving test efficiency.

[0037] (3) The ultrasonic C-scan image processing method proposed in the present invention can be completed by one person, saving labor costs and eliminating the process that originally required two people to operate simultaneously.

[0038] (4) The ultrasonic C-scan image processing method proposed in the present invention has good consistency in detection and evaluation results, and avoids the defect mapping results being affected by the subjective influence of the operator.

[0039] (5) The ultrasonic C-scan image processing method of the present invention can be mainly used for square flat parts, and can also be extended to process C-scan images of grooved parts, cylindrical parts, etc., to achieve automatic, rapid and highly sensitive statistics of defects.

[0040] (6) The ultrasonic C-scan image processing method of the present invention is low-cost and easy to implement. It directly utilizes a conventional ultrasonic C-scan detection system without the need to modify the equipment or add accessories. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a schematic diagram of the composition of the ultrasonic C-scanning system involved in the present invention;

[0042] Figure 2 Schematic diagram of the two-dimensional array d for recording defect numbers according to the present invention;

[0043] Figure 3Schematic diagram of a dictionary-type structured two-dimensional data structure for recording defect information according to the present invention;

[0044] Figure 4 This is a schematic diagram of defect position correction according to the present invention;

[0045] Figure 5 This is a schematic diagram of the defect of the present invention;

[0046] Figure 6 This is the workflow of the present invention.

[0047] Explanation of the accompanying symbols: 1. Ultrasonic flaw detector; 2. Ultrasonic transmitting probe; 3. Ultrasonic receiving probe; 4. Automatic scanning device; 5. Probe position calibration device; 6. Inspected workpiece; 7. Scanning imaging industrial computer; 8. Optical projection device. DETAILED DESCRIPTION

[0048] The present application is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0049] like Figure 1 As shown, the ultrasonic C-scanning system used in the present invention includes: an ultrasonic flaw detector 1, an ultrasonic transmitting probe 2, an ultrasonic receiving probe 3, an automatic scanning device 4, a probe position calibration device 5, a workpiece to be inspected 6, a scanning imaging industrial computer 7, and an optical projection device 8.

[0050] Ultrasonic flaw detector 1: Generates electrical oscillations and applies them to the ultrasonic transmitting probe 2, stimulating the ultrasonic transmitting probe 2 to generate ultrasonic waves. After the ultrasonic waves propagate inside the workpiece 6, the ultrasonic receiving probe 3 receives the penetrated or reflected ultrasonic waves and transmits them to the ultrasonic flaw detector 1.

[0051] Automatic scanning device 4: driven by a motor, it can carry the ultrasonic transmitting probe 2 and the ultrasonic receiving probe 3 to perform a smooth plane automatic scanning movement to complete the scanning of the workpiece 6;

[0052] Probe position calibration device 5: uses two laser line light sources to project onto the product to form a cross cursor to mark the center position of the ultrasonic transmitting probe 2;

[0053] Workpiece 6: Place the square workpiece 6 horizontally, ensuring that one side of the workpiece 6 is parallel to the x-direction and the other side is parallel to the y-direction.

[0054] Scanning and imaging industrial computer 7: equipped with scanning and imaging software and image processing software; collects the amplitude A of the ultrasonic signal received by the probe from the ultrasonic flaw detector 1, reads the probe position coordinates (x, y) from the automatic scanning device 4, and uses the scanning and imaging software to draw an ultrasonic C-scan image in real time based on the received coordinates and amplitude. After the scan is completed, the image is saved as an image file; the scanning and imaging industrial computer 7 uses the image processing software to process the ultrasonic C-scan image to obtain a defect schematic diagram;

[0055] Optical projection device 8: installed on the fixed bracket of the automatic scanning device 4, located directly above the inspected workpiece 6, connected to the scanning imaging industrial computer 7 through an image signal cable, and can be used as the second screen of the industrial computer to project the defect schematic diagram onto the inspected workpiece 6.

[0056] The ultrasonic C-scan system described above can produce an ultrasonic C-scan image of the inspected workpiece 6. Defects such as delamination and low density within the material will be reflected in the C-scan image. To ensure scanning imaging accuracy, appropriate scanning line spacing, scanning speed, and sampling accuracy must be employed.

[0057] A composite material defect statistics method based on ultrasonic C-scan images, such as Figure 2-Figure 6 As shown, the following steps are included:

[0058] 1. Prepare a comparison sample with the same material and thickness as the workpiece to be inspected and containing artificial defects, complete an ultrasonic C-scan, and obtain the threshold representing the defect based on the grayscale and RGB values ​​of the artificial defects in the ultrasonic C-scan image.

[0059] 2. Using the same inspection parameters, complete an ultrasonic C-scan of the workpiece to obtain an ultrasonic C-scan image. Determine the grayscale or RGB values ​​in the ultrasonic C-scan image and set pixels below the threshold to grayscale 0, or black (rbg = (0, 0, 0). Set the remaining pixels to grayscale 255, or white (rbg = (255, 255, 255). For example, a grayscale value of 255 represents complete white, and 0 represents complete black. If the grayscale threshold for defects, as verified in step 1, is 20, then pixels in the image with grayscale ≤ 20 are set to grayscale 0, representing defects, and the remaining pixels are set to 255, representing non-defects. This results in a black-and-white binary image (i.e., a black-and-white image) that reflects defect information.

[0060] 3. Reduce defect noise and weak connection phenomena in black and white images caused by system sampling loss and electrical noise crosstalk, and reduce defect noise or edge burrs caused by workpiece surface conditions or internal pores. The following steps are included:

[0061] 1) Scan the black and white image line by line, eliminate pixel blocks with a width less than the specified value, and process the image noise caused by the detection system.

[0062] A weak connection refers to a defect 1-2 pixels wide between two defective areas. This may be caused by a momentary loss of sampling in the flaw detector. Such linear defects generally do not exist in the material and should be eliminated, otherwise the area of ​​the individual defects will exceed the standard. Noise is generally caused by poor system grounding, resulting in Gaussian noise caused by electrical noise crosstalk in the signal during high-speed scanning. This can cause a linear point of good quality 1-2 pixels wide within the defect, which is not a true reflection of the internal defect. In both cases, the detection system causes image noise. By determining the pixel mutation situation line by line and point by point along the scan, repair is performed.

[0063] Taking the example of repairing a 2-pixel-wide sample loss noise point (black pixel), the color values ​​of the current point and the next point are obtained simultaneously. If the current point is white and the next point is black, then continue to judge the two points one by one. If white appears, then all these black pixels are set to white to eliminate this noise point, and continue to judge downward. For the electrical noise crosstalk noise point inside the defect, the elimination method is the same as above, judging and eliminating the white noise pixel in the black defect pixel;

[0064] 2) Reduce image noise through corrosion algorithm. Step 1) mainly removes Gaussian noise generated by flaw detectors, but the effect of removing noise from defects is not ideal, so corrosion algorithm is used to reduce it. Corrosion is an elementary morphological operation on grayscale images. It can remove parts smaller than the specified structural element, separate two objects with small connections, and remove defect burrs. Set A is corroded by set B, which is represented as It is defined as: Where A is the input image and B is the structuring element. This method uses a vertical rectangular structuring element with a vertical length of 10 and a horizontal width of 1, which is close to the image characteristics. It can eliminate weak connections and irregular burrs between defects while preserving the effective defect details as much as possible.

[0065] 4. Identify defective areas in the black and white image. Scan the black and white image processed in step 3 line by line, identify black connected domains, and save the coordinates of black pixels in each connected domain in a structured two-dimensional data dictionary. <int,List <point>>, where the int integer represents the defect number, List <point>All pixels of the defect are recorded in a list. The Point object contains the pixel position and color attributes (the defect is black). The number of pixels with a certain defect number can be counted to obtain the defect area.

[0066] Specifically, according to the size of the processed image [w,h] (unit: pixel), create a two-dimensional array d of the same size [w,h] to record the connected domain number of each pixel, such as Figure 2 As shown. Create an image information extraction function, the function input is the original image, and the function return value is a custom dictionary type structure data Dictionary <int,List <point>>, records the position and color of defective pixels, such as Figure 3 As shown. In the image information extraction function, determine whether the pixel color is black point by point. If the current point is black, then continue to determine the connected domain serial number value of the surrounding 8 pixels and record it in the integer enumeration type variable Labels. If Labels is empty (there are no black pixels around or no judgment has been made), then add a serial number value for the current pixel and set the value of d[x,y] to this value. If there is a record in Labels, then the serial number values ​​of the pixel and the surrounding pixels are all assigned to the minimum value minLabel in the serial numbers of the surrounding pixels, and the values ​​of the two-dimensional array d[x,y] corresponding to these pixels are simultaneously modified to minLabel. After completing the traversal of the entire image, organize it into a structured two-dimensional data Dictionary according to the value of each element in the two-dimensional array d. <int,List <point>>, record all defect information.

[0067] 5. Eliminate defects smaller than the specified area in the black and white image according to the set defect threshold. This includes the following steps:

[0068] 1) Set the threshold value of the defect area to be ignored, such as ignoring defects with an equivalent diameter of 20 mm, and the correspondence between image pixels and sample size in mm is 1:1, then ThresholdValue = π × 10 2 =314;

[0069] 2) Determine the two-dimensional data dictionary in step 4 one by one <int,List <point>>Calculate the number of pixels in the list List corresponding to each defect in the structured two-dimensional data, obtain the area of ​​the connected domain of each defect, compare it with the area threshold, delete the array corresponding to the defect with an area smaller than the area threshold in the structured two-dimensional data, and delete the defect in the schematic diagram at the same time.

[0070] Specifically, first create a new white image of the same size as the original image, and create a new copy of the dictionary data object in step 4. <int,List <point>>, judge the situation of each defect in turn, for example, the first defect Dictionary<1,List <point>> If the number of pixel points in the dictionary is less than the threshold, all pixel data under key 1 is deleted from the dictionary data object copy. Otherwise, the pixel under key 1 is retained and the pixel color at that location is changed to black in the newly created white image based on the position Position.X and Position.Y of each pixel. All defects are judged in sequence, and all defects smaller than the specified area are eliminated and updated to the black and white image and structured two-dimensional data.

[0071] 6. Extract the defect outline from the black and white image and draw a defect diagram. This includes the following steps:

[0072] 1) Draw the defect outline through Laplacian operator neighborhood calculation. The Laplacian operator is an edge detection operator defined based on the second-order partial derivative of the image f(x,y) in the x and y directions. It is defined as: This method is used to extract the actual contour of defects in C-scan images without performing additional filtering or other processing. The specific operations are as follows:

[0073] Create a 3×3 8-neighborhood Laplacian operator matrix template:

[0074]

[0075] Use this matrix to convolve with the image f(x,y) to get the output image I(x,y).

[0076] 2) Alternatively, use the Canny contour extraction algorithm to draw the defect contour. This method extracts a smoother contour edge, closer to the effect of manual drawing, and closer to the actual defect. The specific operation is:

[0077] Using a two-dimensional Gaussian function The first-order derivative of smoothes the image. Its gradient vector is Bundle The two filter templates are decomposed into two one-dimensional row and column filters, and their expressions are Convolve these two one-dimensional filters with the image f(x,y) to obtain the output image I(x,y), and In the formula, k is a constant, and σ is the Gaussian filter distribution parameter, which determines the width of the Gaussian filter and controls the degree of smoothing. A filter with a small σ yields high positioning accuracy but a low signal-to-noise ratio; a large σ has the opposite effect. After comparison, k = 3 achieved the best results. To smooth the edge contours of the drawn defect diagram, a blur smoothing filter can be used to pre-process the image before Canny extraction (a filter size of 3×3 is recommended).

[0078] 7. Draw cross-hatching lines for the defect areas in the defect diagram and label the defects with numbers 1, 2, 3, etc., all based on the saved structured 2D data dictionary. <int,List <point>>Implementation, where the cross-section line drawing is achieved through the position coordinate division algorithm, and the spacing is adjustable. By selecting a picture or automatically adding the workpiece outline to the schematic, and adding x and y axis rulers to the schematic. The specific operations are:

[0079] 1) Set the interval of the hatch line (unit: pixel);

[0080] 2) Use the remainder division algorithm to determine the x and y coordinates of each pixel P in the defect cluster to determine whether to draw a +45° section line at that pixel location. The function of an evenly spaced +45° straight line is y = x + n × interval, where n is an integer. Therefore, in structured two-dimensional data, if the coordinates of a point (Position.y - Position.x) % interval = 0, the section line passes through that point, and the pixel is changed from white to a light gray color (rbg = (100, 100, 100) (to avoid interfering with the black defect outline), thus achieving section line drawing.

[0081] 3) Draw a defect number at an appropriate location within the defect based on the defect size. For defects with an area smaller than the specified value, count the total number of pixels and add a white text box with black font at approximately 1 / 3 of the pixel, drawing the defect number. For defects with an area larger than the specified value, increase the font size of the number and draw the number in multiple locations to enhance recognition.

[0082] 4) Add workpiece outline to the schematic diagram. For regular rectangular plates, a rectangular outer frame is automatically added; for irregularly shaped workpieces or workpieces with internal structural features, the workpiece outline can be drawn by selecting the workpiece schematic file and overlaying it on the current schematic diagram.

[0083] 5) Add x- and y-axis rulers to the diagram. Expand the scanned image by 20 pixels below and to the right. Draw an x-axis ruler 5 to 10 pixels below the scanned area. Draw 6-pixel vertical lines at 10-pixel intervals in the x-direction as minor divisions, and 8-pixel vertical lines at 100-pixel intervals as major divisions. Similarly, draw a y-axis ruler to the right of the image.

[0084] 8. Correct the defect location in the structured two-dimensional data.

[0085] The following steps are included: if the boundary of the inspected workpiece is not completely consistent with the scanning origin and direction of the detection system when the inspected workpiece is placed, the defect position information is corrected by comparing the actual position of the two specified points on the inspected workpiece with the position deviation of the C-scan image, such as Figure 4 The specific operations are as follows:

[0086] 1) With the help of the cross cursor projected by the probe position calibration device and the steel ruler, measure the position deviation △X1 and △Y1 between the ultrasonic transmitting probe at the scanning starting point (i.e. the first point in the first row of the C-scan image) and the upper left corner of the inspected workpiece;

[0087] 2) Move the ultrasonic transmitting probe to the end point of the first scan line, and use the cross cursor projected by the probe position calibration device and the steel ruler to measure the position deviation △X2 and △Y2 between the probe and the upper right corner of the product;

[0088] 3) According to △X1, △Y1, △Y2 and the width W of the C-scan image, the deflection angle θ can be obtained:

[0089] θ=arctan((△Y2-△Y1) / (W-△X2-△X1))

[0090] 4) If the position coordinates of the defect area in the structured two-dimensional data established based on the black and white image are (x, y), and the actual defect position coordinates on the inspected workpiece are (x', y'), according to the translation and rotation relationship of the coordinate system, the relationship between the two is:

[0091] x=x'cosθ+y'sinθ+△X1

[0092] y=y'cosθ-x'sinθ+△Y1

[0093] 5) Calculate the corrected actual defect position coordinates (x', y').

[0094] 9. Save the defect diagram and the defect information table obtained from the structured two-dimensional data. Save the defect serial number, location, area and other information in the defect information table file in the specified format; save the defect diagram after the section line is drawn, the serial number is marked, and the product outline is added as an image file. Specifically:

[0095] 1) The distance between the defect and the left edge of the product is Dictionary <int,List <point>>The minimum x value of the coordinates of each defect in Point corresponds to;

[0096] 2) The distance between the defect and the left edge of the product is Dictionary <int,List <point>>The minimum y value of the coordinate of each defect in Point corresponds to;

[0097] 3) Defect area is Dictionary <int,List <point>>The number of points corresponding to each defect;

[0098] 4) Adjust the defect location and area values ​​according to the dimension (mm or cm);

[0099] 4) Save the defect diagram in jpg or bmp format, such as Figure 5 shown.

[0100] 10. Draw the defect outline on the workpiece to be inspected. This includes the following steps:

[0101] 1) An optical projection device is placed above the scanning system to project the defect diagram obtained in step 9 onto the surface of the inspected workpiece. The position and size of the defect diagram are adjusted so that the projection of the defect diagram coincides with the inspected workpiece.

[0102] 2) Use a metal pencil to draw all defects on the inspected workpiece according to the defect outline projected by the defect diagram.

[0103] A composite material defect statistics method based on ultrasonic C-scan images is implemented based on the ultrasonic automatic detection system, ultrasonic C-scan images, and image processing methods. Based on the above detection system, the implementation steps are as follows:

[0104] like Figure 6 As shown, step 1: turn on the ultrasonic flaw detector and industrial computer, and connect the probe and defect location device;

[0105] Step 2: Adjust the detection sensitivity and threshold position of the ultrasonic flaw detector;

[0106] Step 3: Set the scanning range, line spacing, and speed, start ultrasonic C-scan testing, and complete the inspection of a comparison sample with the same material and thickness as the workpiece being inspected and containing artificial defects. Obtain the grayscale value or RGB value threshold representing the defect from the image;

[0107] Step 4: Using the same detection parameters, complete the ultrasonic C-scan of the inspected workpiece and save the ultrasonic C-scan image;

[0108] Step 5: Preprocess the ultrasonic C-scan image of the workpiece according to the grayscale value or RGB value threshold obtained in step 3 to obtain a black and white binary image (i.e., a black and white image), with black representing defects;

[0109] Step 6: If there are many noise points or weak connections in the black and white image, eliminate Gaussian noise pixels line by line and combine it with the corrosion and expansion algorithm to reduce the noise points and weak connections caused by material factors such as the surface condition of the inspected workpiece to avoid affecting subsequent processing;

[0110] Step 7: Identify black connected domains in the black and white binary image and record the defect information in the image using structured two-dimensional data;

[0111] Step 8: Set an area threshold and delete the connected domain information smaller than the threshold in the black and white binary image and structured two-dimensional data;

[0112] Step 9: Use the Laplacian neighborhood calculation method or the Canny image contour extraction algorithm to draw the defect contour;

[0113] Step 10: Draw a +45° section line and defect number within the defect outline, and add the outline of the inspected workpiece by automatically adding a border or overlaying the image to obtain a defect schematic diagram;

[0114] Step 11: Using a defect marking device to measure the position deviation of the inspected workpiece, and correcting the defect position in the structured two-dimensional data;

[0115] Step 12: Save the defect information in the structured two-dimensional data to a defect information table file;

[0116] Step 13: Save the defect diagram as a jpg or bmp format image file;

[0117] Step 14: Project the defect diagram onto the inspected workpiece through an optical projection device, and use a metal pencil to draw all defects on the inspected workpiece.

[0118] The contents not described in detail in the specification of the present invention belong to the common knowledge of professionals in this field.

[0119] Although the present invention is disclosed above in terms of preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the scope defined by the claims of the present invention.< / point> < / point> < / point> < / point> < / point> < / point> < / point> < / point> < / point> < / point> < / point>

Claims

1. A composite material defect statistics method based on ultrasonic C-scan images, characterized by: The following steps are involved: S201, using an ultrasonic C-scan nondestructive testing system to obtain an ultrasonic C-scan image of the workpiece (6); S202, determining a grayscale / hue threshold for judging defects using a defect comparison sample, dividing defect areas and non-defect areas in the ultrasonic C-scan image, and obtaining a black and white image; S203, using an image processing algorithm to reduce defect noise and weak connections in the black and white image; weak connections refer to defects with a width of 1 to 2 pixels between two defective areas; S204, identifying defect areas in the black and white image, creating structured two-dimensional data, and saving the defect information in the structured two-dimensional data; S205, setting a defect threshold, eliminating defect areas smaller than a specified area in the black and white image according to the defect threshold, and updating the black and white image and structured two-dimensional data; S206, extracting the defect outline of the defect area in the black and white image, drawing a defect schematic diagram, marking the defect with a serial number, adding a hatching line inside the defect, and obtaining a defect schematic diagram; S207, correcting the position of the defective area in the structured two-dimensional data according to the placement of the workpiece; S208, saving the defect schematic diagram, obtaining a defect information table based on the structured two-dimensional data and saving it; S209, drawing a defect outline on the inspected workpiece (6) using a defect schematic diagram and an optical projection device (8); Wherein, the step S202 includes: S2021. Prepare a comparison sample with the same material and thickness as the workpiece (6) and containing artificial defects, perform ultrasonic C scanning, and obtain a grayscale value or RGB value threshold representing the defect from the image; S2022, using the same detection parameters, complete the ultrasonic C-scan of the inspected workpiece (6), obtain the ultrasonic C-scan image, determine the grayscale value or RGB value in the ultrasonic C-scan image, set the pixels below the threshold to black, representing the defect area, and set the remaining pixels to white, representing the non-defect area; In step S203, the image processing algorithm includes: S2031. Scan the ultrasonic C-scan image line by line, and eliminate pixel blocks with a width smaller than a specified value, where the width is determined based on the Gaussian noise of the detection device; S2032, reducing image noise through corrosion and expansion algorithm; The step S204 includes: The black and white image obtained by the line-by-line scanning step S203 has defective areas as black connected domains. The black connected domains are identified, and the coordinates of the black pixels in each black connected domain are stored in structured two-dimensional data. The defect information includes a defect array, which includes the sequence number, location, and area information of the defective area.

2. The composite material defect statistics method based on ultrasonic C-scan images according to claim 1, characterized in that: The ultrasonic C-scan image is a color or grayscale ultrasonic C-scan image of the inspected workpiece (6) obtained by using an ultrasonic non-destructive testing system and a penetrating automatic ultrasonic C-scan non-destructive testing technology; The ultrasonic nondestructive testing system comprises an ultrasonic flaw detector (1), an ultrasonic transmitting probe (2), an ultrasonic receiving probe (3), an automatic scanning device (4), a probe position calibration device (5), a workpiece to be inspected (6), a scanning imaging industrial computer (7), an optical projection device (8), scanning imaging software, and image processing software; The ultrasonic flaw detector (1) is connected to an ultrasonic transmitting probe (2) and an ultrasonic receiving probe (3), and excites the ultrasonic transmitting probe (2) to generate ultrasonic waves; the workpiece (6) to be inspected is placed horizontally; the ultrasonic transmitting probe (2) and the probe position calibration device (5) are located above the workpiece (6) to be inspected; and the ultrasonic receiving probe (3) is located below the workpiece (6) to be inspected. The automatic scanning device (4) drives the ultrasonic transmitting probe (2) and the ultrasonic receiving probe (3) to move and complete the scanning detection of the workpiece; the scanning imaging industrial control computer (7) uses scanning imaging software to process the received signal of the ultrasonic flaw detector (1) and the position data of the automatic scanning device (4) to obtain an ultrasonic C-scan image; the scanning imaging industrial control computer (7) uses image processing software to process the ultrasonic C-scan image to obtain a defect schematic diagram; the optical projection device (8) projects the defect schematic diagram onto the inspected workpiece (6) to assist in defect drawing.

3. The composite material defect statistics method based on ultrasonic C-scan images according to claim 2, characterized in that: The probe position calibration device (5) projects two line light sources onto the surface of the workpiece (6) to be inspected, and the intersection of the line light sources is used to mark the center position of the ultrasonic transmitting probe (2).

4. The composite material defect statistics method based on ultrasonic C-scan images according to claim 1, characterized in that: Step S205 includes determining the number of coordinate points contained in each array corresponding to the defective area in the structured two-dimensional data according to a set defect threshold, comparing the number of coordinate points with the defect threshold, and deleting the array from the structured two-dimensional data if the number of coordinate points is less than the defect threshold to eliminate the defective area less than the defect threshold. Finally, redrawing the black and white binary image based on the processed array information, after filtering out the specified small defects.

5. The composite material defect statistics method based on ultrasonic C-scan images according to claim 1, characterized in that: In step S206: Defect contour extraction: Defect contour is drawn by neighborhood calculation or by using Canny contour extraction algorithm; Draw a section line within the defect outline and label the defect area with numbers 1, 2, 3, etc. Based on the pixel coordinates (x, y) of the structured 2D data, draw a +45° section line with adjustable spacing. The defect number is marked at an appropriate location within the defect to obtain a defect schematic diagram.

6. The composite material defect statistics method based on ultrasonic C-scan images according to claim 1, characterized in that: The S207 includes: The probe position calibration device (5) is used to measure the scanning starting point, the scanning first line end point, and the position deviations between the axis of the ultrasonic transmitting probe (2) and the two corners of the inspected workpiece (6), and the position information of the defect area in the structured two-dimensional data is corrected according to the geometric relationship.

7. The composite material defect statistics method based on ultrasonic C-scan images according to claim 1, characterized in that: The step S209 includes: using an optical projection device (8) to project a defect diagram onto the surface of the inspected workpiece (6), and adjusting the position and size of the defect diagram so that the projection of the defect diagram coincides with the inspected workpiece (6); All defects are drawn on the inspected workpiece (6) using a metal pencil according to the defect outlines displayed by the projection.

Images