Carbon fiber plate defect detection method based on line laser scanning

Abstract

The invention relates to the technical field of carbon fiber plate defect detection, in particular to a carbon fiber plate defect detection method based on line laser scanning, which is realized by adopting the following steps of: 1, scanning the surface of a detected carbon fiber plate by using line laser; in the scanning process, the thermal infrared imager is used for detecting the temperature field change of the surface of the detected carbon fiber plate in real time, and a time-domain temperature image sequence is generated in real time according to the temperature field change; 2, converting the time domain temperature image sequence into a quasi-static time domain temperature image sequence; 3, converting the quasi-static time domain temperature image sequence into a frequency domain frequency amplitude image sequence; 4, converting the frequency domain frequency amplitude image sequence into a reconstructed frequency domain frequency amplitude image sequence; and 5, predicting depth data of the defect by the support vector regression model according to the two variables. The problems that an existing carbon fiber plate defect detection method is low in detection efficiency and inaccurate in detection result are solved, and the method is suitable for carbon fiber plate defect detection.

Description

Technical Field

[0001] The invention relates to the technical field of carbon fiber plate defect detection, and in particular to a carbon fiber plate defect detection method based on line laser scanning. Background Art

[0002] During the processing and use of carbon fiber boards, defects such as holes and delamination are prone to occur, which seriously affect their service life and performance. Therefore, in order to ensure the service life and performance of carbon fiber boards, it is necessary to perform defect detection on carbon fiber boards. At present, there are mainly two methods for detecting defects in carbon fiber boards: the first is a fixed detection method, which cannot quickly detect defects over a large area, so there is a problem of low detection efficiency. The second is a mobile detection method. On the one hand, this method cannot characterize and correct blur and distortion defects, and on the other hand, it cannot accurately predict the depth of defects, so there is a problem of inaccurate detection results. Based on this, it is necessary to invent a carbon fiber board defect detection method based on line laser scanning to solve the problems of low detection efficiency and inaccurate detection results of existing carbon fiber board defect detection methods. Summary of the invention

[0003] In order to solve the problems of low detection efficiency and inaccurate detection results of existing carbon fiber plate defect detection methods, the present invention provides a carbon fiber plate defect detection method based on line laser scanning.

[0004] The present invention is achieved by adopting the following technical solutions:

[0005] A method for detecting defects of carbon fiber plates based on line laser scanning is implemented by the following steps:

[0006] Step 1: Use a line laser to scan the surface of the carbon fiber plate to be tested; during the scanning process, use an infrared thermal imager to detect the temperature field changes on the surface of the carbon fiber plate to be tested in real time, and generate a time domain temperature image sequence in real time based on this;

[0007] Step 2: Determine the coordinates of the defect in each frame of the time-domain temperature image sequence, and crop each frame accordingly so that the defect is located at the center of the image in each frame, thereby converting the time-domain temperature image sequence into a quasi-static time-domain temperature image sequence;

[0008] Step 3: using discrete Fourier transform to convert the quasi-static time-domain temperature image sequence from the time domain to the frequency domain, thereby converting the quasi-static time-domain temperature image sequence into a frequency-domain frequency amplitude image sequence;

[0009] Step 4: Use the bicubic interpolation algorithm to perform super-resolution reconstruction on each frame of the frequency domain frequency amplitude image sequence, thereby converting the frequency domain frequency amplitude image sequence into a reconstructed frequency domain frequency amplitude image sequence, and then use the principal component analysis algorithm to denoise the reconstructed frequency domain frequency amplitude image sequence;

[0010] Step 5: The temperature features of the time domain temperature image sequence and the frequency amplitude features of the reconstructed frequency domain frequency amplitude image sequence are input as two variables into the support vector regression model of the support vector machine algorithm. The support vector regression model predicts the depth data of the defect based on the two variables, thereby realizing defect detection.

[0011] In step 1, the specific steps of using a line laser to scan the surface of the carbon fiber plate to be tested are as follows:

[0012] Firstly, a three-dimensional mobile platform is selected and the carbon fiber plate to be tested is fixed on the three-dimensional mobile platform;

[0013] Then, a line laser source is constructed; the line laser source comprises a semiconductor laser and a lens combination; the output end of the semiconductor laser faces the incident end of the lens combination; the output end of the lens combination faces the surface of the carbon fiber plate to be tested;

[0014] Then, the semiconductor laser is started, and the semiconductor laser emits a point laser, which is converted into a line laser by a lens combination and then irradiated onto the surface of the carbon fiber plate to be tested;

[0015] Then, the three-dimensional moving platform is started, and the three-dimensional moving platform drives the carbon fiber plate to be tested to move, and the line laser scans the surface of the carbon fiber plate to be tested.

[0016] In step 2, the specific steps of determining the coordinates of the defect in each frame of the time-domain temperature image sequence are as follows:

[0017] First, determine the coordinates (x0, y) of the defect in the first frame of the time-domain temperature image sequence;

[0018] Then, the coordinates (x n ,y); where v represents the scanning speed; f represents the frame rate of the infrared thermal imager.

[0019] In step five, the temperature feature of the time domain temperature image sequence refers to the temperature peak value of each frame of the time domain temperature image sequence; the frequency amplitude feature of the reconstructed frequency domain frequency amplitude image sequence refers to the frequency amplitude peak value of each frame of the reconstructed frequency domain frequency amplitude image sequence.

[0020] Compared with the existing carbon fiber plate defect detection method, the carbon fiber plate defect detection method based on line laser scanning described in the present invention has the following advantages: First, compared with the fixed detection method, the present invention realizes large-area rapid detection of defects by scanning the surface of the carbon fiber plate to be tested, thereby making the detection efficiency higher. Second, compared with the mobile detection method, the present invention realizes the characterization and correction of blur and distortion defects through super-resolution reconstruction on the one hand, and realizes accurate prediction of defect depth through dual variables on the other hand, thereby making the detection result more accurate.

[0021] The present invention effectively solves the problems of low detection efficiency and inaccurate detection results of existing carbon fiber plate defect detection methods, and is suitable for carbon fiber plate defect detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the process of the present invention.

[0023] Figure 2 It is a schematic diagram of the principles of steps 2 to 4 in the present invention. DETAILED DESCRIPTION

[0024] A method for detecting defects of carbon fiber plates based on line laser scanning is implemented by the following steps:

[0025] Step 1: Use a line laser to scan the surface of the carbon fiber plate to be tested; during the scanning process, use an infrared thermal imager to detect the temperature field changes on the surface of the carbon fiber plate to be tested in real time, and generate a time domain temperature image sequence in real time based on this;

[0026] Step 2: Determine the coordinates of the defect in each frame of the time-domain temperature image sequence, and crop each frame accordingly so that the defect is located at the center of the image in each frame, thereby converting the time-domain temperature image sequence into a quasi-static time-domain temperature image sequence;

[0027] Step 3: using discrete Fourier transform to convert the quasi-static time-domain temperature image sequence from the time domain to the frequency domain, thereby converting the quasi-static time-domain temperature image sequence into a frequency-domain frequency amplitude image sequence;

[0028] Step 4: Use the bicubic interpolation algorithm to perform super-resolution reconstruction on each frame of the frequency domain frequency amplitude image sequence, thereby converting the frequency domain frequency amplitude image sequence into a reconstructed frequency domain frequency amplitude image sequence, and then use the principal component analysis algorithm to denoise the reconstructed frequency domain frequency amplitude image sequence;

[0029] Step 5: The temperature features of the time domain temperature image sequence and the frequency amplitude features of the reconstructed frequency domain frequency amplitude image sequence are input as two variables into the support vector regression model of the support vector machine algorithm. The support vector regression model predicts the depth data of the defect based on the two variables, thereby realizing defect detection.

[0030] In step 1, the specific steps of using a line laser to scan the surface of the carbon fiber plate to be tested are as follows:

[0031] Firstly, a three-dimensional mobile platform is selected and the carbon fiber plate to be tested is fixed on the three-dimensional mobile platform;

[0032] Then, a line laser source is constructed; the line laser source comprises a semiconductor laser and a lens combination; the output end of the semiconductor laser faces the incident end of the lens combination; the output end of the lens combination faces the surface of the carbon fiber plate to be tested;

[0033] Then, the semiconductor laser is started, and the semiconductor laser emits a point laser, which is converted into a line laser by a lens combination and then irradiated onto the surface of the carbon fiber plate to be tested;

[0034] Then, the three-dimensional moving platform is started, and the three-dimensional moving platform drives the carbon fiber plate to be tested to move, and the line laser scans the surface of the carbon fiber plate to be tested.

[0035] In step 2, the specific steps of determining the coordinates of the defect in each frame of the time-domain temperature image sequence are as follows:

[0036] First, determine the coordinates (x0, y) of the defect in the first frame of the time-domain temperature image sequence;

[0037] Then, the coordinates (x n ,y); where v represents the scanning speed; f represents the frame rate of the infrared thermal imager.

[0038] In step five, the temperature feature of the time domain temperature image sequence refers to the temperature peak value of each frame of the time domain temperature image sequence; the frequency amplitude feature of the reconstructed frequency domain frequency amplitude image sequence refers to the frequency amplitude peak value of each frame of the reconstructed frequency domain frequency amplitude image sequence.

[0039] In a specific implementation, the lens combination is a combination of a plano-convex lens and a plano-convex cylindrical lens.

[0040] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that these are only examples, and the protection scope of the present invention is defined by the appended claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, but these changes and modifications all fall within the protection scope of the present invention.

Claims

1. A method for detecting defects in carbon fiber plates based on line laser scanning, characterized in that: This method is implemented by the following steps: Step 1: Use a line laser to scan the surface of the carbon fiber plate to be tested; during the scanning process, use an infrared thermal imager to detect the temperature field changes on the surface of the carbon fiber plate to be tested in real time, and generate a time domain temperature image sequence in real time based on this; Step 2: Determine the coordinates of the defect in each frame of the time-domain temperature image sequence, and crop each frame accordingly so that the defect is located at the center of the image in each frame, thereby converting the time-domain temperature image sequence into a quasi-static time-domain temperature image sequence; Step 3: using discrete Fourier transform to convert the quasi-static time-domain temperature image sequence from the time domain to the frequency domain, thereby converting the quasi-static time-domain temperature image sequence into a frequency-domain frequency amplitude image sequence; Step 4: Use the bicubic interpolation algorithm to perform super-resolution reconstruction on each frame of the frequency domain frequency amplitude image sequence, thereby converting the frequency domain frequency amplitude image sequence into a reconstructed frequency domain frequency amplitude image sequence, and then use the principal component analysis algorithm to denoise the reconstructed frequency domain frequency amplitude image sequence; Step 5: The temperature features of the time domain temperature image sequence and the frequency amplitude features of the reconstructed frequency domain frequency amplitude image sequence are input as two variables into the support vector regression model of the support vector machine algorithm. The support vector regression model predicts the depth data of the defect based on the two variables, thereby realizing defect detection.

2. The method for detecting defects in a carbon fiber plate based on line laser scanning according to claim 1, characterized in that: In step 1, the specific steps of using a line laser to scan the surface of the carbon fiber plate to be tested are as follows: Firstly, a three-dimensional mobile platform is selected and the carbon fiber plate to be tested is fixed on the three-dimensional mobile platform; Then, a line laser source is constructed; the line laser source comprises a semiconductor laser and a lens combination; the output end of the semiconductor laser faces the incident end of the lens combination; the output end of the lens combination faces the surface of the carbon fiber plate to be tested; Then, the semiconductor laser is started, and the semiconductor laser emits a point laser, which is converted into a line laser by a lens combination and then irradiated onto the surface of the carbon fiber plate to be tested; Then, the three-dimensional moving platform is started, and the three-dimensional moving platform drives the carbon fiber plate to be tested to move, and the line laser scans the surface of the carbon fiber plate to be tested.

3. The method for detecting defects in a carbon fiber plate based on line laser scanning according to claim 1, characterized in that: In step 2, the specific steps of determining the coordinates of the defect in each frame of the time-domain temperature image sequence are as follows: First, determine the coordinates (x0, y) of the defect in the first frame of the time-domain temperature image sequence; Then, the coordinates (x n ,y); where v represents the scanning speed; f represents the frame rate of the infrared thermal imager.

4. The method for detecting defects of carbon fiber plates based on line laser scanning according to claim 1, characterized in that: In step five, the temperature feature of the time domain temperature image sequence refers to the temperature peak value of each frame of the time domain temperature image sequence; the frequency amplitude feature of the reconstructed frequency domain frequency amplitude image sequence refers to the frequency amplitude peak value of each frame of the reconstructed frequency domain frequency amplitude image sequence.

Images