CN112037126B - An Image Synthesis Method Based on Laser Scanning Method to Detect Surface Crack Defects - Google Patents

Abstract

The invention relates to an image synthesis method for detecting a structural member with a complex surface appearance by adopting a laser scanning thermal imaging nondestructive testing technology. In the invention, the surface of the structural member is uniformly thermally excited by adopting line laser scanning, and the surface change temperature of the structural member in the laser scanning process is recorded in real time by a thermal imager. Aiming at the phenomena of gradual change of a laser curve, step sudden change and over-temperature points outside the laser curve in an image caused by the change of the surface topography of an object, the three phenomena are distinguished by a distance-setting squat and peak value judging method, and a segment fitting method is adopted to obtain the correct laser curve position. And extracting and interpolating according to the required cooling time after excitation to obtain the gray value of the curve with the fixed interval with the laser curve, and then splicing the gray value into a new image in sequence. The invention not only can compensate the excitation time delay difference of each position of the object surface in a single frame image collected by the thermal imager, but also realizes the correct synthesis of the image under the condition that the object surface has complex morphology.

Description

An Image Synthesis Method Based on Laser Scanning Method to Detect Surface Crack Defects

technical field

The invention relates to an image processing method in the detection process of laser scanning thermal imaging, which is used for the synthesis of detection images of crack defects on the surface of workpieces with complex shapes, and belongs to the technical field of infrared nondestructive detection.

Background technique

Taking line laser scanning detection as an example, the usual detection process is that the line laser scans from one end of the surface of the object to the other. During the scanning process, the thermal imager continuously collects the surface temperature of the object. After the scanning is completed, the thermal imager collects the corresponding sequence of images. As shown in Figure 2, if the scanning speed of the laser is constant, the collected images can be browsed frame by frame, and the laser line can be seen to gradually move from one end of the object surface to the other at fixed intervals, and the interval Δx of each spatial movement is the thermal image The product of the time Δt for acquiring a frame of image by the instrument and the laser scanning speed Δv. Randomly select a frame of images from the sequence images, set the position of the laser line as x, then the surface area of the object corresponding to the position x in the image is in the excitation of the laser line, assuming the current moment is t ₀ . Then, it can be known from the above analysis that the laser line moves a distance of Δx in the next image frame, that is, the position of the laser line at time t ₀ +Δt should be x+Δx. This means that the surface area of the object corresponding to the x+Δx position at time t ₀ +Δt is in the laser excitation, and the area corresponding to the x position at this moment should be the cooling time of Δt after the laser excitation. Therefore, in a single image, different positions on the surface of the object have different laser scanning delay times, and it cannot simultaneously represent the state of each position on the surface at the same time delay during laser scanning. If you want to have the same laser scanning delay time at each position of the object surface in a single frame image, as in the case of using the whole frame excitation, it is necessary to reorganize the pixels in each frame of the sequence image according to the delay difference of laser scanning to achieve the post-recombination Every pixel in the image has the same laser scanning delay. The usual practice is to extract pixel point columns with the same delay in each frame of image frame by frame, and then recombine them into a new image in order. After recombination, each point in the image has the same laser scanning delay.

If the surface of the object is flat and the laser is parallel to the thermal imager, the laser line in the image collected by the thermal imager is a vertical straight line (assuming that the laser line is perpendicular to the thermal imager). However, in the actual detection process, the surface of the object usually has a complex shape, and the images collected by the thermal imager may show gradual changes in the form of laser curve inclination and bending, sudden changes in the form of step jumps, and excessive changes in the surface of objects outside the laser curve. temperature phenomenon. Among them, the change of step jump form and the phenomenon of over-temperature point both appear as position mutations when traversing to find the maximum gray value. Figure 3 is a schematic diagram of the change of the laser curve under different shapes on the surface of the object, where (a) represents the field of view of the laser vertical thermal imager, (b) represents the inclination and bending of the laser curve (c) represents the step jump of the laser curve (d) represents the object There are hot spots on the surface.

As a focal plane detector, a thermal imager is structured as a two-dimensional array composed of a large number of detection units, and its output is a planar image in units of pixels, so the minimum resolution of a thermal imager is the size of the area corresponding to a single pixel. When the above-mentioned changes occur in the laser curve, it will appear across two pixels in the image collected by the thermal imager, and the image extraction can only be based on the pixel as the smallest unit, so it is impossible to extract the middle position of the two pixels. grayscale value. At this time, the gray value of the extracted pixels may not be the real maximum value, and horizontal stripes will also appear in the synthesized image.

As shown in Figure 4, for the vertical and gradual changes of the laser line, fitting and interpolation methods are usually used to obtain the actual laser curve position and its gray value. The specific steps are as follows:

1. Take out a single frame image in the sequence image, and traverse the image line by line to extract the maximum gray value position point of the row, and after the traversal is completed, the maximum gray value position point column corresponding to the number of image columns can be taken out. Using it as a parameter for fitting, the obtained curve can be expressed as the actual laser curve.

2. Extract the position point column of the distance x from the laser line position in the image, and use the interpolation algorithm to calculate the actual gray value of the curve position, which can represent the gray value at the time t=x/Δv after laser excitation.

3. Extract the pixel point columns at a distance of x from the laser line frame by frame in the same steps, and arrange the pixel point columns sequentially to form an image, which can represent the temperature distribution image with the same time delay t at each position on the surface of the object.

When there is a step mutation on the surface of the object, it is difficult for the above method to accurately fit the position of the laser peak curve. Due to the smoothing effect of the fitting, there will be an error between the fitting curve and the actual laser position at the jump of the laser curve. One feasible method is to split the abrupt laser curve segment from the main curve segment, and each segment performs fitting independently. This method not only avoids the fitting error at the transition point, but also realizes the simultaneous display of both ends of the sudden change of the object surface.

When there is an over-temperature point on the surface of the object, the existence of the over-temperature point will cause confusion in judging the position of the maximum value. For example, a small piece of black paint is sprayed on a smooth iron block. Since the absorption rate of the black paint is greater than that of the iron block, in the few frames of images where the laser has just scanned the black paint area, the gray value of the black paint area will be greater than the gray value of the laser curve, resulting in an overtemperature point. Due to the existence of over-temperature points, there will be two gray value peaks in a single row of pixels, one of which is the laser curve position and the other is the over-temperature point position, and the gray value of the over-temperature point position is greater than the gray value of the laser curve position. Then the maximum value obtained after traversal is the gray value of the over-temperature point, not the gray value of the laser curve. A feasible method is to use a small-range fitting method to perform line-by-line pixel fitting in the horizontal direction, compare it with the peak position of the previous line and extract the peak point with a small distance, which is the position of the laser curve.

In actual detection, due to the uncertainty of the surface topography of the object, one or more of the above phenomena will appear in the images collected by the thermal imager at the same time, which leads to the diversity of fitting error forms. However, any one of the above fitting methods cannot be fully applied to different situations to fit the correct laser curve, so that the laser scanning infrared detection method is greatly limited by the surface topography of the detected object.

Contents of the invention

In view of the shortcomings of the above-mentioned image synthesis technology, the present invention provides an image synthesis technology for laser scanning infrared imaging detection of objects with complex surface topography. The invention can actively distinguish the gradual change of the laser curve, the jump change and the phenomenon of over-temperature point, separate the complicated situations, and simplify the complex situations. Different fitting methods are used for different phenomena to fit the correct laser curve position.

By setting the distance criterion Δp, the gradual change and sudden change of the position point of the maximum gray value in the image are distinguished, and it is judged whether there is a step jump or an over-temperature point in the laser curve in the image. The size of the distance criterion Δp changes according to the change of the surface topography of the object, and the more serious the surface of the object is inclined or curved, the size of the criterion should also increase.

Traverse each row of pixel points from top to bottom in the image to find the maximum gray value position point, and compare the traversed position point with the previous row position point from the second row. If it is greater than the distance criterion Δp, it is determined that there is a sudden change in the position of the maximum gray value, and the position of the sudden change pixel row is recorded. Fit the above mutation pixel row, if there is only one peak point in the fitting curve, it means that the mutation is caused by the adjustment jump of the laser curve; if there are two or more peak points in the fitting curve, it means that the position exists overheated. According to the above method, the gradual change of the laser curve in the image, the sudden change and the over-temperature point can be distinguished. Use different fitting methods for different phenomena

In actual detection, by combining the above methods, the laser curve fitting error caused by the change of the surface topography of the object can be effectively compensated, so that the temperature distribution images with the same time delay at each position on the surface of the object can be correctly spliced.

Description of drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the accompanying drawings used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, so It should be regarded as a limitation on the scope, and those skilled in the art can also obtain other related drawings based on these drawings without creative work.

Fig. 1 is a structural schematic diagram of a laser scanning thermal imaging detection system;

Fig. 2 is a curve diagram of the surface temperature distribution of the laser scanning excitation object;

Fig. 3 is a schematic diagram of a laser scanning peak curve under a complex shape;

Fig. 4 is a schematic diagram of the execution steps of image reconstruction in the case of laser line vertical and gradient changes;

Fig. 5 is a schematic diagram of the execution steps of the image reconstruction method described in this patent.

Detailed ways

The present invention will be further described below in conjunction with specific drawings and embodiments.

FIG. 1 shows a schematic diagram of a laser scanning thermal imaging detection system, including a host computer 1, a thermal imager 2, a light source 3, a vibrating mirror 4, a platform 5 and an object 6 to be detected. The upper computer 1 has the following functions: control the switch of the light source 3 and output light intensity; control the vibrating mirror 4 to realize laser scanning at a constant speed; control the thermal imager 3 and receive image data, and can perform post-image processing at the same time. The image synthesis technology described in the present invention is the application in the image post-processing process of the upper computer. The third part of the light source provides a line laser light source for the detection system. It is assumed that the length of the line laser is greater than the field of view of the thermal imager, and its width is extremely thin and does not exceed the field of view of a single pixel of the thermal imager. The function of the thermal imager 2 is to record the changing temperature of the surface of the object during the laser excitation process. The resolution of the thermal imager is X×Y, the acquisition frame frequency is f, and the time Δt=1/f for acquiring a single frame image. Assuming that the duration of a single detection is T, the thermal imager can collect N=T/Δt images after detection. The host computer controls the variable speed rotation of the vibrating mirror 4 to realize the uniform scanning of the line laser on the sample surface. Assuming the scanning speed is v, the laser moves as X _l within T time, and moves Δx distance per unit time Δt. After a single scan excitation, the thermal imager acquires a sequence of images, and image synthesis is performed on the sequence of images. Figure 5 shows a flow chart of image synthesis, and its specific steps are as follows:

1. Select an image in the sequence image as the start of image synthesis. Under the premise of ensuring that the synthesized image can contain the area of interest on the surface of the object, any frame image in the sequence image can be selected as the starting point. The same is true for the end. Appropriately reducing the number of composite frames can reduce the amount of image composite calculations and reduce the burden on the host computer. It is assumed that the first frame image is used as the starting point of this synthesis, and the last frame image is used as the ending point, and the image synthesis is performed within the range of the entire sequence of images.

2. Select a rectangular frame of appropriate size to frame the area of interest on the surface of the object, and traverse the frame line by line to find the maximum gray value position point. Similarly, range selection can reduce the amount of traversal data and shorten the time for image synthesis. Assuming that the selected range is the entire picture, a column of position points equal to the number of rows Y of the thermal imager can be obtained after the traversal.

3. Set the distance criterion Δp. Calculate the distance difference between two adjacent position points in the sequence of maximum gray value position points sequentially. If the distance difference is greater than the distance criterion Δp, it means that there is a sudden change in position at the position point. Record and set this point as a ₀ , followed by a ₁ , a ₂ . . . According to the difference in the flatness of the surface topography of the object, the size of the distance criterion Δp should be adjusted appropriately. If the flatness of the surface of the object is very poor, such as the surface is seriously inclined or the curvature is large, the value of Δp should be increased to prevent misjudgment. Conversely, the value of Δp should be reduced to improve accuracy. Usually the value of Δp should not be less than 2. During the laser scanning process, when the laser line crosses the field of view of two adjacent pixels of the thermal imager, the maximum position may fluctuate between the two pixels. In order to avoid misjudgment, Δp should be greater than the position difference between two adjacent pixels, that is, 1 . This step is used to distinguish between gradual changes in position points and abrupt changes.

4. For the position mutation points a ₀ , a ₁ , a ₂ ... recorded in the previous step, take out the gray value set of the row pixel corresponding to the position point for fitting, and calculate the peak points of the fitted curve number. If the number of peak points is 1, it means that the peak point is the position of the laser line, and there is no over-temperature point. If the number of peak points is greater than 1, it indicates that there may be over-temperature points in this line besides the laser line. This step can perform preliminary screening on the step jump and over-temperature point phenomenon of the laser curve.

5. Traversing the sequence of images frame by frame, repeating steps 2 to 4 above until the end image. Finally, N columns of maximum gray value position point columns are extracted.

6. For the mutation point of the maximum gray value position whose number of peak points after fitting in step 4 is greater than 1, take the images within the range of multiple frames before and after it, and observe the movement of the peak point position. If the position where there is a peak point does not move in several frames of images before and after, it indicates that this position is an over-temperature point. From the above analysis, it can be seen that the position of the laser curve in the sequence image moves frame by frame at a fixed interval Δx, while the position of the over-temperature point does not change. Combining steps 4 and 6, we can distinguish the step jump of the laser curve from the over-temperature point.

7. Through the above steps, the gradual change of the laser curve, the sudden change of the laser curve and the phenomenon of over-temperature point can be separated in sequence. At this time, corresponding methods can be used to compensate the error for different phenomena. When the laser curve gradually changes, use the direct fitting method; when the laser curve jumps, use the segmented fitting method to divide the abrupt laser curve segment from the main curve segment, and each segment performs fitting separately; when there is an overtemperature When the point is selected, the peak value of the over-temperature point is discarded and the peak point close to the peak point position of the previous line is selected as the position of the laser curve to eliminate the influence of the over-temperature point, and then segmented fitting is performed.

Through the above steps, the present invention can separate the gradual change of the laser curve, the sudden change of the laser curve and the phenomenon of over-temperature point in the image, and then compensate the errors caused by each respectively, reducing the complexity of image synthesis in the case of complex surface topography of objects.

Claims (4)

1. An image processing method in a laser scanning thermal imaging detection process, used for image synthesis when detecting a structure with complex surface topography, is characterized in that it comprises the following steps: S1: Set the distance criterion Δp, adjust and set the size of the distance criterion Δp according to the unevenness of the surface topography of the object, and then execute S2; S2: Select a single frame image, traverse line by line and obtain the position point corresponding to the maximum gray value in a single row of pixel point columns, obtain a column of data point columns containing the position points corresponding to the maximum gray value of all rows after traversing, and then execute S3; S3: Calculate the difference between the adjacent points of the data point column obtained in step S2, that is, the distance difference of the maximum position point, and compare the distance difference with the distance criterion Δp set in S1, if the distance difference is greater than the distance The criterion Δp indicates that there is a position mutation at the position point; if there is a position mutation, execute S4, and if not, execute S7; S4: Calculate and obtain all mutation position points, and respectively fit the gray value values of the row pixels corresponding to each position point, calculate the number of peak points of each fitted curve, if the number of peak points is 1, it indicates that the If the peak point is the position of the laser line and there is no over-temperature point, then execute S6; if the number of peak points is greater than 1, it indicates that there may be over-temperature points in this line besides the laser line, then execute S5; S5: Extract the mutation position point with the number of peak points greater than 1, calculate and compare the maximum gray value position point in several frames of images before and after this position, if the position of the peak point remains unchanged, it means that there is an overtemperature point at this position, then From the fitted peak points, select the peak position point close to the maximum gray value position point in the previous line as the laser curve position point, and then execute S6; S6: Divide the whole column of data into multiple segments with the position mutation point whose number of peak points is 1 as the interval, and execute S7 for each segment; S7: Fit the maximum gray value position points obtained above to obtain the actual maximum gray value position curve, and then perform S8; S8: Set the required laser excitation delay time t, traverse the curve of the selected distance fitting laser curve x, where the distance x is the product of the laser scanning speed v and the required delay time t, and then execute S9; S9: Obtain the actual gray value of each point on the delay curve through traversal interpolation, and then execute S10; S10: From the start frame to the end frame, traverse the sequence of images and repeatedly execute the above steps S2-S8, and combine the finally obtained multi-column gray value series into a new image. 2. The image processing method according to claim 1, characterized in that a line laser is used as the laser scanning light source, wherein the length of the laser line should be greater than the width of the area of interest on the surface of the detected object, and the laser scanning speed should remain constant during actual detection , to ensure that the excitation time of each position on the surface of the object within the scanning range is consistent. 3. The image processing method according to claim 1, characterized in that the selection of the start frame and the end frame in the sequence image is arbitrarily selected under the premise that the combined image can completely contain the area of interest on the surface of the object, and an appropriate frame should be selected simultaneously. The sequence image traverses the size of the frame interval to ensure that the size of the object in the composite image is the same as the original image. 4. The image processing method according to claim 1, characterized in that the value of Δp in the step S1 changes with the flatness of the surface of the object, specifically the more uneven the surface, the greater the value of Δp.

Images