Infrared phase-locked nondestructive testing method based on adaptive normalization improvement technology

Abstract

The invention discloses an infrared phase-locked nondestructive testing method based on an adaptive normalization improvement technology, belongs to the technical field of infrared phase-locked nondestructive testing methods, and invents a problem point filtering algorithm which can find out and filter problem pixel points in an original infrared video. And a problem pixel point is prevented from causing noise interference on a phase result image obtained by detection, and an adaptive normalization algorithm is invented. The self-adaptive normalization algorithm can effectively improve the contrast ratio of a phase result image on the premise of keeping the original phase data structure distribution characteristics, so that the visual perception is improved, the effect is remarkably improved compared with the traditional two normalization algorithms, the defect depth prediction method based on the phase contrast ratio is invented, and the defect depth prediction efficiency is improved. And the capability of predicting the depth of the defect inside or on the back of the detected plate is realized.

Description

Technical Field

[0001] The invention relates to an infrared phase-locked nondestructive testing method, in particular to an infrared phase-locked nondestructive testing method based on an adaptive normalization improvement technology, and belongs to the technical field of infrared phase-locked nondestructive testing methods. Background Art

[0002] The traditional infrared phase-locked thermal wave nondestructive testing algorithm has the following four problems:

[0003] 1. Lack of pre-processing of problematic points in infrared video: In infrared phase-locked thermal wave nondestructive testing technology, a heat source with sinusoidal changes is used to thermally excite the inspected plate, and the heating process of the plate is recorded by an infrared thermal imager to obtain the original infrared video. In real scenes, due to the software settings of the infrared thermal imager, damage to the photosensitive element in the infrared thermal imager, and certain external interference in the experimental environment, the temperature change values ​​of some pixels in the original infrared video we obtained do not conform to the law of sinusoidal rise. If these problematic pixels are not filtered out, it will cause great noise interference to the phase result diagram calculated by the traditional algorithm;

[0004] 2. The algorithm used to normalize the original data into the phase result map has limitations: the traditional normalization algorithms include direct normalization algorithm and mean standard deviation normalization algorithm. The direct normalization algorithm can completely restore the structural distribution law of the original phase data, but the phase result map calculated by this algorithm lacks the specificity of defects. The contrast between defects and background in the phase result map is low, and the defects are not obvious to the naked eye.

[0005] The standard deviation average algorithm normalizes the original data based on the probability distribution characteristics of the data, which can effectively improve the contrast between defects and background in the phase result image, improve the naked eye perception of defects, and make it easier to observe the existence of defects. However, the algorithm has the defect of poor scene adaptability. This is because there is a fixed control parameter k in the algorithm. The imaging quality of the phase result image obtained by using different k values ​​for infrared videos shot in different scenes is also different. Therefore, the algorithm has poor adaptability. In different experiments, the experimenter needs to repeatedly adjust the k value to improve the image quality of the phase result image.

[0006] 3. Only the phase result map of the infrared video is output, but the buried depth of each defect cannot be obtained: the traditional infrared phase-locked thermal wave nondestructive testing algorithm can only output the phase result map of the plate, and the inspectors cannot obtain the depth distribution of each defect. Therefore, it is necessary to study the method of detecting the depth of defects;

[0007] 4. It cannot automatically identify the type, location and diameter of defects, and requires manual analysis by professionals: the traditional infrared phase-locked thermal wave nondestructive testing algorithm can only output the phase result map of the plate, and the inspectors need to manually measure and analyze the defects;

[0008] Therefore, an infrared phase-locked nondestructive testing method based on adaptive normalization improved technology is designed to solve the above problems. Summary of the invention

[0009] The main purpose of the invention is to provide an infrared phase-locked nondestructive testing method based on an adaptive normalization improved technology.

[0010] The purpose of the present invention can be achieved by adopting the following technical solutions:

[0011] The infrared phase-locked nondestructive testing method based on the adaptive normalization improved technology includes the following steps:

[0012] Step 1: Calculate the SSIM structural similarity index of the phase result graph to be evaluated based on the standard phase result graph;

[0013]

[0014] in:

[0015] x and y are the pixel values ​​of the two images;

[0016] μ x is the mean value of image x;

[0017] μ y is the mean value of image y;

[0018] is the x-variance of the image;

[0019] is the y variance of the image;

[0020] C1 and C2 are two constants used to stabilize the denominator.

[0021] Step 2: According to the standard phase result image, calculate the high-frequency similarity index of the GMSD image of the phase result image to be evaluated;

[0022] Gradient amplitude similarity formula:

[0023]

[0024] Where: G x and G y are the gradient magnitudes of images x and y respectively;

[0025] c is a constant used to stabilize the calculation.

[0026]

[0027] in:

[0028] M is the total number of pixels in the image;

[0029] GMS i is the gradient magnitude similarity of the i-th pixel;

[0030] μ GMS All GMS i The average value of .

[0031] Step 3: Calculate the RMS contrast difference score of the phase result image to be evaluated;

[0032] Step 4: Perform a joint weighted score on the SSIM index and the GMSD index, and discard the phase result graph to be evaluated whose joint weighted score is less than the ideal structural similarity score;

[0033] Step 5: Select the phase result image with the highest RMS contrast score from the phase result images to be evaluated that have not been discarded as the final output phase result image of the algorithm.

[0034] Preferably, in step 1, SSIM is used to measure the similarity between two images, and the formula is as follows:

[0035]

[0036] in:

[0037] x and y are the pixel values ​​of the two images;

[0038] μ x is the mean value of image x;

[0039] μ y is the mean value of image y;

[0040] is the x-variance of the image;

[0041] is the y variance of the image;

[0042] C1 and C2 are two constants used to stabilize the denominator.

[0043] Preferably, in step 4, GMSD is used to evaluate the similarity of image quality, and the formula is as follows:

[0044] Gradient amplitude similarity formula:

[0045]

[0046] Where: G x and Gy are the gradient magnitudes of images x and y respectively;

[0047] c is a constant used to stabilize the calculation.

[0048] Preferably, the definition of GMSD;

[0049]

[0050] in:

[0051] M is the total number of pixels in the image;

[0052] GMS i is the gradient magnitude similarity of the i-th pixel;

[0053] μ GMS All GMS i The average value of .

[0054] Preferably, in step 3, RMS is used to measure the contrast of the image, and the formula is as follows:

[0055]

[0056] in:

[0057] N is the total number of pixels in the image;

[0058] I i is the gray value of the pixel;

[0059] μ is the average gray value of the image.

[0060] Preferably, application of infrared lock-in nondestructive testing method based on adaptive normalization improvement technology;

[0061] When it is necessary to detect the depth of defects in a certain type of plate, the defect depth can be predicted by phase contrast. The specific method flow is as follows:

[0062] S11: Prepare a calibration plate with defects of different depths and types, fix the excitation frequency and excitation time, perform infrared thermal wave phase-locked excitation on the calibration plate, and record the infrared video of the excitation process;

[0063] S12: Analyze and obtain phase result image data of the calibration plate according to the infrared video;

[0064] S13: Calculate the phase contrast of defects at different depths;

[0065] S14: Fitting the mapping formula between defect depth and phase contrast of different defect types;

[0066] S15: Repeat the above S11-S13 for the plate to be inspected, calculate the phase contrast of the defect to be inspected, substitute the phase contrast into the mapping formula obtained in S14, and obtain the predicted value of the defect depth.

[0067] Preferably, the formula for the phase contrast in S13 is as follows:

[0068]

[0069] in:

[0070] P background is the average value of the background original phase in the phase result image;

[0071] P defect is the average value of the original phase of the defect in the phase result diagram;

[0072] K is a constant fitting factor.

[0073] Preferably, application of infrared lock-in nondestructive testing method based on adaptive normalization improvement technology;

[0074] Specific method flow:

[0075] S21: Customized plates with different types of internal defects, including slit-type, circular, polygonal, and anisotropic defects. Different light source excitation frequencies and light source powers were set to perform sinusoidal heating on the plates, thereby recording the original infrared videos under various testing conditions.

[0076] S22: Perform algorithm processing on the original infrared video to obtain a phase result map, rotate, scale, shear, process noise, and perform affine transformation on the phase result map to obtain a large amount of original data, manually calibrate the original data, and divide the calibrated data into a training set and a test set of the model;

[0077] S23: Select a suitable neural network model to train the training set, so as to obtain a final neural network model, and use the model to identify the defect type, position and diameter of the phase result image of the plate to be tested.

[0078] Beneficial technical effects of the present invention:

[0079] The infrared phase-locked nondestructive detection method based on the adaptive normalization improvement technology provided by the present invention (1) invents a problem point filtering algorithm, which can find and filter out the problem pixels in the original infrared video to prevent the problem pixels from causing noise interference to the phase result image obtained by the detection.

[0080] (2) Invented an adaptive normalization algorithm. The use of the adaptive normalization algorithm can effectively improve the contrast of the phase result image while maintaining the structural distribution characteristics of the original phase data, thereby improving the visual experience of the naked eye. Compared with the two traditional normalization algorithms, the effect is significantly improved.

[0081] (3) A method for predicting defect depth based on phase contrast was invented, which enables the prediction of defect depth inside or on the back of the inspected plate.

[0082] (4) The use of neural networks enables automatic identification of the type, location and diameter of defects in the phase result image, and the ability to automatically generate defect detection reports. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 The present invention is a flowchart of a preferred embodiment of an infrared phase-locked nondestructive testing method based on the adaptive normalization improvement technology according to the present invention. DETAILED DESCRIPTION

[0084] In order to make the technical solution of the present invention more clear and specific to those skilled in the art, the present invention is further described in detail below in conjunction with embodiments and drawings, but the implementation manner of the present invention is not limited thereto.

[0085] In infrared phase-locked thermal wave nondestructive testing, after using sinusoidal thermal excitation to heat the plate, the normal original infrared video obtained should conform to the law that the amplitude of all pixels presents a sinusoidal rise, and the overall amplitude average of the time domain waveform of all pixels should be within a reasonable fluctuation range. Therefore, based on these two characteristics, we wrote a problem pixel search algorithm, which can quickly find the pixel with problems. After finding the problem pixel, replace the value of the problem pixel with the average value of the pixels in the whole image, so as to achieve the goal of filtering the problem pixel.

[0086] The principle of the problem pixel search algorithm is:

[0087] Step 1: First, calculate the time domain waveform of the average value of all pixels, use the time domain waveform as the standard reference waveform, and calculate the standard deviation, average value, and data range of the standard reference waveform.

[0088] Step 2: Then count the standard deviation, average value, and data range of the time domain waveform of the pixel point to be analyzed.

[0089] Step 3: Finally, the three major parameters of the pixel to be analyzed are analyzed and compared with the three major parameters of the standard reference waveform to determine whether the pixel to be analyzed is a problem pixel.

[0090] The standard deviation formula of the amplitude of the time domain waveform of a single pixel in infrared video is:

[0091]

[0092] Where σ is the standard deviation, μ is the mean amplitude, N is the total number of recorded points, and Ii is the amplitude of the i-th point.

[0093] The average value formula of the amplitude of the time domain waveform of a single pixel in the infrared video is:

[0094]

[0095] Among them, μ is the average amplitude, N is the total number of recorded points, and Ii is the amplitude of the i-th point. The data range formula of the amplitude of the time domain waveform of a single pixel point in the infrared video is:

[0096] R=Imax-Imin

[0097] Imax is the maximum amplitude in the time domain waveform of the pixel point, and Imin is the minimum amplitude in the time domain waveform of the pixel point.

[0098] like Figure 1 As shown, the infrared phase-locked nondestructive testing method based on the adaptive normalization improvement technology provided in this embodiment includes the following steps:

[0099] Step 1: Calculate the SSIM structural similarity index of the phase result graph to be evaluated based on the standard phase result graph;

[0100]

[0101] in:

[0102] x and y are the pixel values ​​of the two images;

[0103] μ x is the mean value of image x;

[0104] μ y is the mean value of image y;

[0105] is the x-variance of the image;

[0106] is the y variance of the image;

[0107] C1 and c2 are two constants used to stabilize the denominator.

[0108] Step 2: According to the standard phase result image, calculate the high-frequency similarity index of the GMSD image of the phase result image to be evaluated;

[0109] Gradient amplitude similarity formula:

[0110]

[0111] Where: G x and G y are the gradient magnitudes of images x and y respectively;

[0112] c is a constant used to stabilize the calculation.

[0113] In this embodiment, the definition of GMSD is

[0114]

[0115] in:

[0116] M is the total number of pixels in the image;

[0117] GMS i is the gradient magnitude similarity of the i-th pixel;

[0118] μ GMS All GMS i The average value of .

[0119] Step 3: Calculate the RMS contrast difference score of the phase result image to be evaluated;

[0120] Step 4: Perform a joint weighted score on the SSIM index and the GMSD index, and discard the phase result graph to be evaluated whose joint weighted score is less than the ideal structural similarity score;

[0121] Step 5: Select the phase result image with the highest RMS contrast score from the phase result images to be evaluated that have not been discarded as the final output phase result image of the algorithm.

[0122] (1) A problem point filtering algorithm was invented, which can find and filter out problem pixels in the original infrared video to prevent the problem pixels from causing noise interference to the detected phase result image.

[0123] (2) Invented an adaptive normalization algorithm. The use of the adaptive normalization algorithm can effectively improve the contrast of the phase result image while maintaining the structural distribution characteristics of the original phase data, thereby improving the visual experience of the naked eye. Compared with the two traditional normalization algorithms, the effect is significantly improved.

[0124] (3) A method for predicting defect depth based on phase contrast was invented, which enables the prediction of defect depth inside or on the back of the inspected plate.

[0125] (4) The use of neural networks enables automatic identification of the type, location and diameter of defects in the phase result image, and the ability to automatically generate defect detection reports.

[0126] In this embodiment, in step 1, SSIM is used to measure the similarity between two images, and the formula is as follows:

[0127]

[0128] in:

[0129] x and y are the pixel values ​​of the two images;

[0130] μ x is the mean value of image x;

[0131] μ y is the mean value of image y;

[0132] is the x-variance of the image;

[0133] is the y variance of the image;

[0134] C1 and C2 are two constants used to stabilize the denominator.

[0135] In this embodiment, in step 4, GMSD is used to evaluate the similarity of image quality, and the formula is as follows:

[0136] Gradient amplitude similarity formula:

[0137]

[0138] Where: G x and G y are the gradient magnitudes of images x and y respectively;

[0139] c is a constant used to stabilize the calculation.

[0140] In this embodiment, the definition of GMSD is

[0141]

[0142] in:

[0143] M is the total number of pixels in the image;

[0144] GMS i is the gradient magnitude similarity of the i-th pixel;

[0145] μ GMS All GMS i The average value of .

[0146] In this embodiment, in step 3, RMS is used to measure the contrast of the image, and the formula is as follows:

[0147]

[0148] in:

[0149] N is the total number of pixels in the image;

[0150] I i is the gray value of the pixel;

[0151] μ is the average gray value of the image.

[0152] In this embodiment, the infrared phase lock-in nondestructive testing method based on the adaptive normalization improvement technology is applied;

[0153] When it is necessary to detect the depth of defects in a certain type of plate, the defect depth can be predicted by phase contrast. The specific method flow is as follows:

[0154] S11: Prepare a calibration plate with defects of different depths and types, fix the excitation frequency and excitation time, perform infrared thermal wave phase-locked excitation on the calibration plate, and record the infrared video of the excitation process;

[0155] S12: Analyze and obtain phase result image data of the calibration plate according to the infrared video;

[0156] S13: Calculate the phase contrast of defects at different depths;

[0157] S14: Fitting the mapping formula between defect depth and phase contrast of different defect types;

[0158] S15: Repeat the above S11-S13 for the plate to be inspected, calculate the phase contrast of the defect to be inspected, substitute the phase contrast into the mapping formula obtained in S14, and obtain the predicted value of the defect depth.

[0159] In this embodiment, the formula of phase contrast in S13 is as follows:

[0160]

[0161] in:

[0162] P background is the average value of the background original phase in the phase result image;

[0163] P defect is the average value of the original phase of the defect in the phase result diagram;

[0164] K is a constant fitting factor.

[0165] In this embodiment, the infrared phase lock-in nondestructive testing method based on the adaptive normalization improvement technology is applied;

[0166] Specific method flow:

[0167] S21: Customized plates with different types of internal defects, including slit-type, circular, polygonal, and anisotropic defects. Different light source excitation frequencies and light source powers were set to perform sinusoidal heating on the plates, thereby recording the original infrared videos under various testing conditions.

[0168] S22: Perform algorithm processing on the original infrared video to obtain a phase result map, rotate, scale, shear, process noise, and perform affine transformation on the phase result map to obtain a large amount of original data, manually calibrate the original data, and divide the calibrated data into a training set and a test set of the model;

[0169] S23: Select a suitable neural network model to train the training set, so as to obtain a final neural network model, and use the model to identify the defect type, position and diameter of the phase result image of the plate to be tested.

[0170] The above are only further embodiments 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 solutions and concepts of the present invention within the scope disclosed by the present invention, which belong to the protection scope of the present invention.

Claims

1. Infrared phase lock-in nondestructive testing method based on adaptive normalization improvement technology, characterized by: The steps include: Step 1: Calculate the SSIM structural similarity index of the phase result graph to be evaluated based on the standard phase result graph; in: x and y are the pixel values ​​of the two images; μ x is the mean value of image x; μ y is the mean value of image y; is the x-variance of the image; is the y variance of the image; C1 and C2 are two constants used to stabilize the denominator. Step 2: According to the standard phase result image, calculate the high-frequency similarity index of the GMSD image of the phase result image to be evaluated; Gradient amplitude similarity formula: Where: G x and G y are the gradient magnitudes of images x and y respectively; c is a constant used to stabilize the calculation. in: M is the total number of pixels in the image; GMS i is the gradient magnitude similarity of the i-th pixel; μ GMS All GMS i The average value of Step 3: Calculate the RMS contrast difference score of the phase result image to be evaluated; in: N is the total number of pixels in the image; I i is the gray value of the pixel; Step 4: Perform a joint weighted score on the SSIM index and the GMSD index, and discard the phase result graph to be evaluated whose joint weighted score is less than the ideal structural similarity score; Step 5: Select the phase result image with the highest RMS contrast score from the phase result images to be evaluated that have not been discarded as the final output phase result image of the algorithm.

2. The infrared phase-locked nondestructive testing method based on the adaptive normalization improvement technology according to claim 1 is characterized in that: In step 1, SSIM is used to measure the similarity between two images. The formula is as follows: in: x and y are the pixel values ​​of the two images; μ x is the mean value of image x; μ y is the mean value of image y; It is a graphic x variance; is the y variance of the image; C1 and C2 are two constants used to stabilize the denominator.

3. The infrared phase lock-in nondestructive testing method based on the adaptive normalization improvement technology according to claim 1 is characterized in that: In step 4, GMSD is used to evaluate the similarity of image quality, and the formula is as follows: Gradient amplitude similarity formula: Where: G x and G y are the gradient magnitudes of images x and y respectively; c is a constant used to stabilize the calculation.

4. The infrared phase lock-in nondestructive testing method based on the adaptive normalization improvement technology according to claim 3 is characterized in that: Definition of GMSD in: M is the total number of pixels in the image; GMS i is the gradient magnitude similarity of the i-th pixel; μ GMS All GMS i The average value of .

5. The infrared phase lock-in nondestructive testing method based on the adaptive normalization improvement technology according to claim 4 is characterized in that: In step 3, RMS is used to measure the contrast of the image. The formula is as follows: in: N is the total number of pixels in the image; I i is the gray value of the pixel; μ is the average gray value of the image.

6. The infrared phase-locked nondestructive testing method based on the adaptive normalization improvement technology according to claim 5 is characterized in that: Application of infrared lock-in nondestructive testing method based on adaptive normalization improvement technology; When it is necessary to detect the depth of defects in a certain type of plate, the defect depth can be predicted by phase contrast. The specific method flow is as follows: S11: Prepare a calibration plate with defects of different depths and types, fix the excitation frequency and excitation time, perform infrared thermal wave phase-locked excitation on the calibration plate, and record the infrared video of the excitation process; S12: Analyze and obtain phase result image data of the calibration plate according to the infrared video; S13: Calculate the phase contrast of defects at different depths; S14: Fitting the mapping formula between defect depth and phase contrast of different defect types; S15: Repeat the above S11-S13 for the plate to be inspected, calculate the phase contrast of the defect to be inspected, substitute the phase contrast into the mapping formula obtained in S14, and obtain the predicted value of the defect depth.

7. The infrared phase lock-in nondestructive testing method based on the adaptive normalization improvement technology according to claim 6 is characterized in that: The formula for phase contrast in S13 is as follows: in: P background is the average value of the background original phase in the phase result image; P defect is the average value of the original phase of the defect in the phase result diagram; K is a constant fitting factor.

8. The infrared phase lock-in nondestructive testing method based on the adaptive normalization improvement technology according to claim 5 is characterized in that: Application of infrared lock-in nondestructive testing method based on adaptive normalization improvement technology; Specific method flow: S21: Customized plates with different types of internal defects, including slit-type, circular, polygonal, and anisotropic defects. Different light source excitation frequencies and light source powers were set to perform sinusoidal heating on the plates, thereby recording the original infrared videos under various testing conditions. S22: Perform algorithm processing on the original infrared video to obtain a phase result map, rotate, scale, shear, process noise, and perform affine transformation on the phase result map to obtain a large amount of original data, manually calibrate the original data, and divide the calibrated data into a training set and a test set of the model; S23: Select a suitable neural network model to train the training set, so as to obtain a final neural network model, and use the model to identify the defect type, position and diameter of the phase result image of the plate to be tested.

Images