A method and system for generating high-voltage thyristor images for quality inspection

Abstract

The present invention belongs to the technical field of quality inspection, and specifically relates to a high-voltage thyristor image generation method and system for quality inspection. Through multi-directional dynamic lighting, real-time brightness control and polarization filtering, structured light projection and synchronous recording, virtual lighting scene simulation and optimization and other measures, all-round uniform lighting is achieved, shadows and overexposure are reduced; image clarity and feature point recognition are improved; data integrity and traceability are ensured; and the optimal image sample is finally selected, which significantly improves image quality and reliability and consistency of detection, effectively solves the problems of uneven lighting and poor image quality in the prior art, and provides strong support for automated detection.

Description

Technical Field

[0001] The invention belongs to the technical field of quality inspection, and in particular relates to a method and system for generating a high-voltage thyristor image for quality inspection. Background Art

[0002] In modern industrial production, quality inspection of high-voltage thyristors (GTOs) is an important part of ensuring their performance and reliability. Traditional quality inspection methods usually rely on manual visual inspection or imaging using simple optical equipment, which is not only inefficient but also easily affected by human factors, resulting in inconsistent inspection results. With the advancement of automation technology and machine vision systems, more and more companies are beginning to adopt image processing-based methods for GTO quality inspection.

[0003] Existing image generation and processing technologies mainly include the following aspects:

[0004] Fixed light source illumination: Many systems use a fixed light source configuration to illuminate the object under test. Although this provides stable lighting conditions, it is often difficult to obtain ideal image quality for high-voltage thyristors with complex geometries and surface characteristics, especially in areas with high reflectivity, which are prone to shadows and overexposure.

[0005] Projecting structured light from a single angle: Some systems use structured light projection technology to enhance feature point recognition, but usually only project patterns from a fixed angle, which limits the ability to capture multi-view features of the object and affects the comprehensiveness and accuracy of detection.

[0006] Static image capture: Image capture in existing technologies is mostly a one-time operation, lacking dynamic adjustment and recording of light source configuration parameters. It is unable to optimize image quality according to specific needs and is not conducive to subsequent image analysis and comparison.

[0007] The main problem with the existing technology is that it is difficult to achieve uniform and high-quality image generation, especially when facing high-voltage thyristors with complex surface characteristics. Problems such as uneven lighting, shadows and overexposure seriously affect the image quality and detection accuracy. Summary of the invention

[0008] The purpose of the present invention is to provide a high-voltage thyristor image generation method and system for quality inspection, which can significantly improve the efficiency and accuracy of high-voltage thyristor quality inspection and solve the problems of uneven illumination, poor image quality and the like existing in the prior art.

[0009] To achieve the above object, the present invention provides a method for generating a high-voltage thyristor image for quality inspection, comprising the following steps:

[0010] Determine the shape parameters and surface characteristics of the high-voltage thyristor to be tested, select an illumination mode according to the information of the shape parameters and surface characteristics, and construct a framework including multiple light source positions based on the selected illumination mode;

[0011] According to the set framework, adjust the angle and distance of each light source, monitor and adjust the output intensity of each light source in real time through the dynamic brightness control system, and apply a polarization filter in front of the imaging device to improve the clarity;

[0012] Combined with the improved clarity, structured light projection is performed to project a texture pattern onto the object being measured, and the generated texture pattern information is used to capture the image, while simultaneously recording all current light source configuration parameters;

[0013] After completing the image capture, various virtual lighting scenarios are simulated according to the saved light source configuration parameters, and the optimal image samples are selected based on the different lighting effects obtained.

[0014] Preferably, the determining of the shape parameters and surface characteristics of the high-voltage thyristor to be tested, and selecting the lighting mode according to the information of the shape parameters and surface characteristics, comprises:

[0015] Measure the size and surface reflection characteristics of the high-voltage thyristor to be tested, and select the minimum light intensity according to the size and surface reflection characteristics;

[0016] Based on the determined minimum light intensity, select the type of light source that will provide uniform illumination without exceeding the maximum permissible heat;

[0017] Design a multi-angle illumination scheme so that the angle of each light source can directly illuminate the thyristor surface without being blocked by other structures and reduce the generation of shadows;

[0018] Build support structures to accommodate the positions and angles of all light sources.

[0019] Preferably, the step of constructing a framework including multiple light source positions based on the selected lighting mode includes:

[0020] Determine the number of light sources and set them according to the size of the high-voltage thyristor to be tested and the required illumination uniformity;

[0021] Plan the position coordinates of each light source to ensure that the position coordinates cover all surface areas of the thyristor and meet multi-directional illumination requirements;

[0022] Adjust the relative distance between each light source to ensure uniform distribution of light intensity without overlapping overexposure, and build a support structure to adapt to the position coordinates and relative distances of all light sources.

[0023] Preferably, adjusting the angle and distance of each light source according to the set framework includes:

[0024] For each light source, measure the distance from the center of the light source to any point on the surface of the high-voltage thyristor to ensure that the optimal irradiation angle of each light source enables the light to directly irradiate the surface of the thyristor without being blocked;

[0025] Determine the optimal illumination angle for each light source to avoid shadows or overexposure caused by the intersection of light beams between adjacent light sources;

[0026] Adjust the direction of the light source and fine-tune the distance between each light source and the thyristor so that the light intensity is evenly distributed on the surface of the thyristor.

[0027] Preferably, the real-time monitoring and adjustment of the output intensity of each light source by a dynamic brightness control system includes:

[0028] After the light source angle and distance are adjusted, light intensity sensors are installed at multiple predetermined points on the surface of the thyristor to collect initial light intensity data;

[0029] When the target light intensity is set, the dynamic brightness control system starts to monitor the output of each light source in real time and adjusts the light source power through the feedback loop so that the actual light intensity is close to the target value;

[0030] The dynamic brightness control system periodically checks the deviation between the actual light intensity and the target light intensity at all predetermined points. If the deviation exceeds the allowable range, the fine-tuning process is triggered and the light source power is adjusted again until the conditions are met.

[0031] Preferably, applying a polarization filter before an imaging device to improve clarity includes:

[0032] After completing the dynamic brightness control, the polarization filter is used to reduce the impact of reflected light from non-metallic surfaces;

[0033] Install the selected polarization filter in front of the imaging device lens and adjust the rotation angle of the filter so that the polarization direction of the light passing through the filter is perpendicular to the polarization direction of the reflected light;

[0034] Measure the actual light intensity after passing through the polarization filter and compare it with the original light intensity without the filter to evaluate the impact of transmittance changes on image quality;

[0035] If the transmittance change exceeds the preset threshold, fine-tune the angle of the polarization filter or replace it with a filter with different characteristic parameters until the image clarity and contrast meet the detection requirements.

[0036] Preferably, performing structured light projection in combination with the improved clarity to project a texture pattern onto the object under test comprises:

[0037] After the clarity provided is improved, a structured light pattern is designed, which contains stripes or dots of different frequencies and phases, and the pattern is projected through a projection device, and the projection angle is adjusted so that the light covers the surface of the object being measured;

[0038] Implementing structured light projection, while recording projection parameters corresponding to each pattern, including but not limited to projection angle, brightness and contrast;

[0039] The image sequence under the projected pattern is collected, and the enhanced feature point recognition gain is calculated by comparing the original image and the image with the structured light pattern.

[0040] Preferably, the method of using the generated texture pattern information to capture an image and simultaneously recording all current light source configuration parameters includes:

[0041] After the texture information is generated, the image capture device is started, and the capture resolution and frame rate are set to ensure that the image with the structured light pattern can be recorded;

[0042] Synchronously record the configuration parameters of each light source, including light source power, angle, distance, and angle of the polarization filter, to form a light source configuration data set;

[0043] Each time an image is captured, a corresponding metadata file is generated, which contains all light source configuration information and the timestamp of image capture;

[0044] The captured image and its corresponding metadata file are stored in the specified data storage unit.

[0045] Preferably, after the image capture is completed, a plurality of virtual lighting scenarios are simulated according to the saved light source configuration parameters, based on the different lighting effects obtained, including:

[0046] After the image is acquired and the light source configuration parameters are saved, the metadata file and the corresponding image data in the storage unit are read for simulation;

[0047] Set up virtual lighting scenarios, each of which contains different light source power, angle, distance, and polarization filter angle to generate a variety of lighting conditions;

[0048] The ray tracing method is applied to simulate the lighting effect, and the light intensity of each point on the thyristor surface in each scenario is calculated. The simulated light intensity is applied to the original image to generate a new image sequence with different lighting scenarios. At the same time, the brightness difference between each new image and the original image is recorded.

[0049] On the other hand, the present invention provides a high-voltage thyristor image generation system for quality inspection, comprising:

[0050] The lighting mode selection and framework construction module is used to determine the shape parameters and surface characteristics of the high-voltage thyristor to be tested, select the lighting mode according to the information of the shape parameters and surface characteristics, and construct a framework containing multiple light source positions based on the selected lighting mode;

[0051] Light source configuration and dynamic brightness control module, used to adjust the angle and distance of each light source according to the set framework, monitor and adjust the output intensity of each light source in real time through the dynamic brightness control system, and apply polarization filter in front of the imaging device to improve the clarity;

[0052] The structured light projection and image capture module is used to perform structured light projection with improved clarity, project texture patterns onto the object under test, use the generated texture pattern information to capture images, and simultaneously record all current light source configuration parameters;

[0053] The virtual lighting scenario simulation and image evaluation module is used to simulate a variety of virtual lighting scenarios according to the saved light source configuration parameters after completing image capture, and select the optimal image sample based on the different lighting effects obtained.

[0054] Technical effects and advantages of the present invention: Compared with the prior art, the method and system for generating a high-voltage thyristor image for quality inspection proposed by the present invention have the following advantages:

[0055] The present invention achieves all-round uniform lighting, reduces shadows and overexposure, improves image clarity and feature point recognition, ensures data integrity and traceability, and ultimately selects the best image samples, significantly improves image quality and detection reliability and consistency, effectively solves the problems of uneven lighting and poor image quality in the prior art, and provides strong support for automated detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a flow chart of the method for generating a high-voltage thyristor image for quality inspection of the present invention;

[0057] Figure 2 The block diagram of the high-voltage thyristor image generation system for quality inspection of the present invention. DETAILED DESCRIPTION

[0058] The following will be combined with the accompanying drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0059] The present invention provides a method for generating a high-voltage thyristor image for quality inspection, which realizes all-round uniform illumination, reduces shadows and overexposure, and improves image clarity and feature point recognition through multi-directional dynamic illumination, real-time brightness control and polarization filtering, structured light projection and synchronous recording, and virtual illumination scene simulation and optimization. The specifics are as follows:

[0060] like Figure 1 As shown, the high-voltage thyristor image generation method for quality inspection in this embodiment includes the following steps:

[0061] a. Determine the shape parameters and surface characteristics of the high-voltage thyristor to be tested, and select the lighting mode according to the information of the shape parameters and surface characteristics; further include the following sub-processes:

[0062] a1. Measure the size L and surface reflectivity Rf of the high-voltage thyristor to be tested, and calculate the required minimum illumination intensity I_min by the formula I_min=C*Rf / L^2, where C is a constant related to the sensitivity of the device; by measuring the specific size and surface reflectivity characteristics of the thyristor, the required minimum illumination intensity can be determined more accurately. This helps avoid unnecessary high brightness settings, reduces heat generation, and ensures sufficient illumination to capture clear images.

[0063] Assuming that the length of a high-voltage thyristor is L=50mm, the surface reflectivity is Rf=0.7, and the constant C=100lux·mm^2 / (W·sr) is known, the minimum light intensity I_min is calculated as:

[0064] I_min=100*0.7 / (50)^2;

[0065] I_min=0.028lux.

[0066] a2. Based on the minimum light intensity I_min determined in step a1, select a light source type that can provide uniform illumination and does not exceed the maximum allowable heat Q_max, ensuring that Q<=Q_max, where Q is the heat generated by the light source; based on the calculated minimum light intensity, select a light source type that can provide sufficient illumination without exceeding the safe heat limit to ensure that the object being tested will not be damaged or the image quality will not be affected due to overheating during the detection process.

[0067] If the maximum allowable heat Q_max = 100W, and the selected light source generates Q = 80W of heat per unit time, then the light source is suitable because it meets the condition Q <= Q_max.

[0068] a3. According to the minimum light intensity I_min and heat limit Q_max determined in step a2, design a multi-angle illumination scheme so that the angle θ of each light source satisfies the formula tan(θ)=H / (0.5*L), where H is the distance from the light source to the center of the thyristor; design a multi-angle illumination scheme so that the angle θ of each light source meets specific conditions, thereby achieving uniform illumination of the thyristor surface, reducing shadows and overexposure, and improving image quality.

[0069] Assuming that the distance from the light source to the center of the thyristor is H = 100 mm and the length of the thyristor is L = 50 mm, the light source angle θ is calculated as:

[0070] tan(θ)=100 / (0.5*50);

[0071] tan(θ)=4;

[0072] θ = arctan(4);

[0073] θ≈75.96°.

[0074] a4. Based on the light source selected in step a3, a support structure S is constructed. The structure needs to adapt to the position and angle θ of all light sources while ensuring structural stability and flexibility. The constructed support structure not only needs to adapt to the position and angle of all light sources, but also needs to ensure structural stability and flexibility during use, so that it can be adjusted as needed to ensure long-term reliability.

[0075] b. Based on the selected lighting mode, construct a framework including multiple light source positions; further including the following sub-processes:

[0076] b1. Determine the number of light sources N, and calculate it according to the size L of the high-voltage thyristor to be tested and the required illumination uniformity U by the formula N=A*L^2 / U, where A is a coefficient related to the characteristics of the device; by calculating the number of light sources, ensure that each light source can effectively cover the surface area of ​​the high-voltage thyristor to be tested and meet the illumination uniformity requirements. This helps to avoid the problem of partial overbrightness or overdarkness and improve image quality.

[0077] Assuming that the length of a high-voltage thyristor is L=50mm, the required illumination uniformity is U=0.05lux, and the coefficient A=0.02 is known, the number of light sources N is calculated as:

[0078] N=0.02*(50)^2 / 0.05;

[0079] N=0.02*2500 / 0.05;

[0080] N=1000.

[0081] b2. Based on the number of light sources N determined in step b1, plan the position coordinates (Px, Py, Pz) of each light source to ensure that these positions can cover all surface areas of the thyristor and meet the multi-directional illumination requirements; Plan the positions of the light sources to ensure that they can cover all surface areas of the thyristor and provide illumination from multiple angles. This reduces shadows and reflections and enhances the recognition of feature points in the image.

[0082] According to the number of light sources N=1000 determined in step b1, the position coordinates (Px, Py, Pz) of each light source are planned. For example, the light sources are distributed in the three-dimensional space around the thyristor to ensure that the position of each light source can effectively illuminate different parts of the thyristor. The specific coordinates are designed in detail according to the geometric shape of the thyristor, such as evenly distributing the light sources around the cylindrical thyristor.

[0083] b3. According to the position coordinates (Px, Py, Pz) set in step b2, adjust the relative distance Dij between each light source, using the formula Dij=sqrt((Px_i-Px_j)^2+(Py_i-Py_j)^2+(Pz_i-Pz_j)^2) to ensure that the light intensity is evenly distributed and there is no overlapping overexposure; by adjusting the relative distance between the light sources, ensure that the light intensity is evenly distributed on the entire measured surface to avoid overlapping overexposure. This helps to improve the contrast and clarity of the image.

[0084] Assuming that the position coordinates of light source 1 are (Px_1, Py_1, Pz_1) = (100, 100, 100) mm, and the position coordinates of light source 2 are (Px_2, Py_2, Pz_2) = (150, 150, 150) mm, the distance Dij between the light sources is calculated as:

[0085] Dij=sqrt((100-150)^2+(100-150)^2+(100-150)^2);

[0086] Dij=sqrt((-50)^2+(-50)^2+(-50)^2);

[0087] Dij = sqrt(7500);

[0088] Dij≈86.6mm.

[0089] b4. After completing the adjustment of the distance between light sources in step b3, construct a support structure S, which needs to adapt to the position coordinates (Px, Py, Pz) and relative distance Dij of all light sources while ensuring structural stability and flexibility. The constructed support structure not only needs to adapt to the position and relative distance of all light sources, but also needs to ensure the stability and flexibility of the structure during use, so that it can be adjusted as needed to ensure reliability for long-term use.

[0090] c. According to the set framework, adjust the angle and distance of each light source; further including the following sub-processes:

[0091] c1. For each light source i, measure the distance Dij from the center of the light source to any point j on the surface of the high-voltage thyristor, using the formula Dij=sqrt((Xi-Xj)^2+(Yi-Yj)^2+(Zi-Zj)^2), where (Xi,Yi,Zi) are the position coordinates of the light source i, and (Xj,Yj,Zj) are the position coordinates of the point j on the thyristor surface; by calculating the distance from each light source to multiple points on the thyristor surface, ensure that the illumination angle and distance setting of each light source can cover the entire surface to avoid missing or repeated lighting.

[0092] Assuming that the position coordinates of light source i are (Xi, Yi, Zi) = (100, 100, 100) mm, and the position coordinates of point j on the thyristor surface are (Xj, Yj, Zj) = (150, 150, 150) mm, the distance Dij is calculated as:

[0093] Dij=sqrt((100-150)^2+(100-150)^2+(100-150)^2);

[0094] Dij=sqrt((-50)^2+(-50)^2+(-50)^2);

[0095] Dij = sqrt(7500);

[0096] Dij≈86.6mm.

[0097] c2. Based on each distance Dij calculated in step c1, determine the optimal illumination angle θi of each light source i so that the light can directly illuminate the surface of the thyristor without being blocked by other structures. The formula is used for calculation, tan(θi)=Hl / Dij, where Hl is the vertical height from the light source to the center of the thyristor. By calculating the optimal illumination angle, ensure that the light directly illuminates the surface of the thyristor without being blocked by other structures, reduce shadows and reflections, and improve image quality.

[0098] Assuming that the vertical height Hl from the light source to the center of the thyristor is 100 mm, and according to the distance Dij calculated in step c1 = 86.6 mm, the optimal irradiation angle θi is calculated as:

[0099] tan(θi)=100 / 86.6;

[0100] θi=arctan(100 / 86.6);

[0101] θi≈49.39°.

[0102] c3. According to the angle θi set in step c2, adjust the direction of light source i, and ensure that the relative angle difference Δθ between all light sources satisfies Δθ<=θ_max, so as to avoid shadows or overexposure caused by the intersection of light beams between adjacent light sources; by adjusting the direction of the light source, ensure that the relative angle difference between all light sources is within a reasonable range, avoid shadows or overexposure caused by the intersection of light beams between adjacent light sources, and ensure uniform light distribution.

[0103] Assuming that the maximum allowable angle difference θ_max = 10°, and the angles of two adjacent light sources are θ1 = 49.39° and θ2 = 55°, their relative angle difference Δθ is calculated as:

[0104] Δθ=|θ2-θ1|;

[0105] Δθ=|55-49.39|;

[0106] Δθ≈5.61°;

[0107] Since 5.61°<10°, the condition is met.

[0108] c4. After completing the light source angle adjustment in step c3, fine-tune the distance Di between each light source and the thyristor to make the light intensity I evenly distributed on the surface of the thyristor. Use the formula I=P / (4*π*Di^2), where P is the light source power, to ensure that the illumination has both comprehensive coverage and no excessive concentration. By fine-tuning the distance between the light source and the thyristor, ensure that the light intensity is evenly distributed on the entire thyristor surface, which has both comprehensive coverage and avoids excessive concentration, thereby improving image quality and detection accuracy.

[0109] Assuming the light source power P = 100W, the initial distance Di = 100mm, the light intensity I is calculated as:

[0110] I = 100 / (4 * π * (100) ^ 2);

[0111] I≈0.000796lux;

[0112] If a higher light intensity is required, reduce Di appropriately. For example, if Di is reduced to 80mm, the new light intensity I' is calculated as:

[0113] I'=100 / (4*π*(80)^2);

[0114] I'≈0.001244lux.

[0115] Through the above steps, the present invention can accurately adjust the angle and distance of each light source to ensure that the light directly irradiates the surface of the thyristor, avoid beam crossing and shadow problems, and achieve uniform distribution of light intensity. This not only improves the quality and clarity of the image, but also enhances the recognition of feature points, thereby significantly improving the quality inspection effect of high-voltage thyristors.

[0116] d. Monitor and adjust the output intensity of each light source in real time through a dynamic brightness control system; further including the following sub-processes:

[0117] d1. After completing the adjustment of the light source angle and distance in step c, install light intensity sensors at multiple predetermined points Pi on the surface of the thyristor to collect initial light intensity data Si; by installing light intensity sensors at multiple predetermined points on the surface of the thyristor, the initial light intensity data can be accurately collected. This provides a reliable basis for the subsequent adjustment of the light source output intensity.

[0118] Assume that 5 predetermined points P1, P2, P3, P4, and P5 are selected on the surface of the thyristor and corresponding light intensity sensors are installed. After measurement, the initial light intensity of each point is S1=100lux, S2=105lux, S3=98lux, S4=102lux, and S5=101lux.

[0119] d2. Based on the data Si obtained in step d1, set the target light intensity Ti to ensure that the light intensity of each predetermined point Pi meets the detection requirements, using the formula Ti=Si+ΔS, where ΔS is the light intensity increment determined according to the detection requirements; set the target light intensity of each predetermined point according to the initial light intensity data and the detection requirements. Ensure that the light intensity meets the detection requirements, improve image quality and detection accuracy.

[0120] Assuming that the required light intensity increment ΔS=10lux, the target light intensities T1, T2, T3, T4, and T5 are calculated as follows:

[0121] T1=S1+ΔS=100+10=110lux;

[0122] T2=S2+ΔS=105+10=115lux;

[0123] T3=S3+ΔS=98+10=108lux;

[0124] T4=S4+ΔS=102+10=112lux;

[0125] T5=S5+ΔS=101+10=111lux.

[0126] d3. According to the target illumination intensity Ti set in step d2, the dynamic brightness control system starts to monitor the output of each light source in real time, and adjusts the light source power Pl through the feedback loop so that the actual illumination intensity Si is close to the target value Ti, using the formula Pl=K*(Ti-Si), where K is the proportional coefficient to ensure the sensitivity of the adjustment; through the dynamic brightness control system, the output of each light source is monitored in real time, and the light source power is adjusted according to the feedback loop so that the actual illumination intensity is close to the target value. This ensures the stability and consistency of the illumination intensity.

[0127] Assuming the proportionality coefficient K = 0.5, and the initial light intensity S1 = 100 lux and the target light intensity T1 = 110 lux, the light source power Pl is calculated as:

[0128] Pl = 0.5 * (110-100);

[0129] Pl = 0.5 * 10;

[0130] Pl=5W.

[0131] Similarly, for other predetermined points, adjustments are made according to their respective actual light intensities and target light intensities.

[0132] d4. Based on step d3, the system periodically checks the deviation εi between the actual illumination intensity Si and the target illumination intensity Ti of all predetermined points Pi. If |εi|>ε_max (ε_max is the maximum allowable deviation), the fine-tuning process is triggered, and the light source power Pl is adjusted again until the condition |εi|<=ε_max is met. The system periodically checks the deviation between the actual illumination intensity and the target illumination intensity of all predetermined points. If the deviation exceeds the allowable range, the fine-tuning process is triggered to ensure that the illumination intensity is always in the best state.

[0133] Assume that the maximum allowable deviation ε_max = 5 lux, and in a certain inspection, it is found that the actual light intensity S1 = 107 lux at a predetermined point, while the target light intensity T1 = 110 lux, then the deviation ε1 is calculated as:

[0134] ε1=|107-110|;

[0135] ε1=3lux;

[0136] Since 3<5, the current deviation is within the allowable range. If S1=102lux at a certain moment, the deviation ε1 becomes:

[0137] ε1=|102-110|;

[0138] ε1=8lux;

[0139] Since 8>5, the fine-tuning process is triggered, and the light source power Pl is adjusted again until the deviation ε1 satisfies the condition |ε1|<=5.

[0140] Through the above steps, the present invention realizes the precise control of the light source output intensity, ensures the stability and consistency of the light intensity, and thus improves the image quality and the accuracy of the detection. The dynamic brightness control system can respond to changes in real time and automatically adjust the light source power, ensuring that the lighting conditions during the detection process always remain in the best state.

[0141] e. Applying a polarization filter before the imaging device to improve the clarity; further comprising the following sub-processes:

[0142] e1. After completing the dynamic brightness control in step d, select a suitable polarization filter Fp, which can effectively reduce the impact of light reflected from the surface of non-metallic materials, while ensuring sufficient light transmittance Tf to maintain image brightness; By selecting a suitable polarization filter, the impact of light reflected from the surface of non-metallic materials is effectively reduced, while ensuring sufficient light transmittance to maintain image brightness. This helps to improve image clarity and contrast.

[0143] Assuming that the surface reflectivity of the high-voltage thyristor to be tested is Rf=0.7 and that it is necessary to ensure at least 80% of the light transmittance Tf=0.8, a polarization filter Fp is selected that meets these conditions.

[0144] e2. Install the polarization filter Fp selected in step e1 in front of the imaging device lens, and adjust the rotation angle αf of the filter so that the polarization direction of the light passing through the filter is perpendicular to the polarization direction of the reflected light. Use the formula cos²(αf)=Rf / Tf, where Rf is the reflectivity of the non-metallic material. By adjusting the rotation angle of the polarization filter, the polarization direction of the light passing through the filter is perpendicular to the polarization direction of the reflected light, thereby minimizing the impact of reflected light on imaging.

[0145] Assuming that the reflectivity of the non-metallic material is Rf=0.7 and the light transmittance is Tf=0.8, the rotation angle αf is calculated as:

[0146] cos²(αf)=0.7 / 0.8;

[0147] cos²(αf)=0.875;

[0148] cos(αf)=sqrt(0.875);

[0149] αf = arccos(sqrt(0.875));

[0150] αf≈29.74°.

[0151] e3. According to the rotation angle αf set in step e2, measure the actual light intensity Iaf after passing through the polarization filter, and compare it with the original light intensity Io without the filter, calculate the transmittance change ΔTf, and use the formula ΔTf=(Iaf-Io) / Io; by measuring the actual light intensity after passing through the polarization filter and comparing it with the original light intensity, calculate the transmittance change and evaluate the effect of the filter.

[0152] Assuming that the original light intensity Io=100lux without the filter, and the actual light intensity Iaf=85lux after passing through the polarization filter, the transmittance change ΔTf is calculated as:

[0153] ΔTf=(85-100) / 100;

[0154] ΔTf=-0.15 or -15%.

[0155] e4. Based on step e3, evaluate the impact of the transmittance change ΔTf on the image quality. If ΔTf exceeds the preset threshold ΔTf_max, fine-tune the angle αf of the polarization filter or replace the filter Fp with different characteristic parameters until ΔTf<=ΔTf_max to ensure that the image clarity and contrast meet the detection requirements. Evaluate the impact of the transmittance change on the image quality. If it exceeds the preset threshold, fine-tune or replace the filter to ensure that the image clarity and contrast meet the detection requirements.

[0156] Assume that the maximum allowable transmittance change threshold ΔTf_max = -0.1 (i.e. -10%), and the calculated transmittance change ΔTf = -0.15 (i.e. -15%). Since |-0.15|>|-0.1|, it is necessary to trigger the fine-tuning process to adjust the angle αf of the polarization filter or replace the filter Fp with different characteristic parameters until the condition |ΔTf|<=ΔTf_max is met.

[0157] Through the above steps, the present invention can effectively apply polarization filters to improve image clarity and contrast. Selecting a suitable polarization filter, adjusting its rotation angle, measuring the transmittance change and making necessary fine adjustments ensures that the image quality is always in the best state, thereby significantly improving the quality inspection effect of high-voltage thyristors.

[0158] f. Performing structured light projection in combination with the improved clarity to project a texture pattern onto the object under test further includes the following sub-processes:

[0159] f1. After completing the clarity improvement provided in step e, design a series of structured light patterns Pn, which contain stripes or dot matrices of different frequencies and phases to adapt to the surface features of the high-voltage thyristor; by designing structured light patterns containing stripes or dot matrices of different frequencies and phases, the geometric features of the surface of the high-voltage thyristor are better adapted to enhance the recognition of feature points in the image.

[0160] Assume that structured light patterns P1, P2, and P3 need to be designed for a specific model of high-voltage thyristor. These patterns will include stripes or dot arrays of high, medium, and low frequencies to cover different surface details. For example, P1 is a high-frequency stripe used to capture subtle features; P2 is a medium-frequency stripe used for general feature recognition; and P3 is a low-frequency stripe used for large-scale surface contours.

[0161] f2. Based on the structured light pattern Pn designed in step f1, select an appropriate projection device to ensure that the pattern can be accurately projected, and adjust the projection angle βp so that the light can evenly cover the surface of the object to be measured, using the formula sin(βp)=Hp / Dp, where Hp is the vertical height from the projection device to the center of the thyristor, and Dp is the horizontal distance between the projection device and the thyristor; select an appropriate projection device and adjust it according to the calculated projection angle to ensure that the light can evenly cover the surface of the object to be measured and avoid the problem of local overbrightness or overdarkness.

[0162] Assuming that the vertical height from the projection device to the center of the thyristor Hp = 100 mm, and the horizontal distance between the projection device and the thyristor Dp = 200 mm, the projection angle βp is calculated as:

[0163] sin(βp)=100 / 200;

[0164] βp=arcsin(0.5);

[0165] βp≈30°.

[0166] f3. Implement structured light projection according to the projection angle βp set in step f2, and record the projection parameter group Gnp corresponding to each pattern Pn, including but not limited to the projection angle βp, brightness Bnp and contrast Cnp, for subsequent analysis; when implementing structured light projection, record the projection parameters corresponding to each pattern in detail, such as projection angle, brightness and contrast, to ensure that subsequent analysis has a basis for optimization and improvement.

[0167] Assuming that for pattern P1, the projection angle βp=30°, the brightness Bnp=80cd / m², and the contrast Cnp=10:1, then record Gnp={βp=30°, Bnp=80cd / m², Cnp=10:1}.

[0168] f4. Based on step f3, collect the image sequence Inp under the projection pattern, and calculate the enhanced feature point recognition gain Gf by comparing the original image Io and the image Inp with the structured light pattern. Use the formula Gf=(F(Inp)-F(Io)) / F(Io), where F represents the feature point extraction process, to ensure that the prominence of the feature points in the image is significantly improved. By comparing the original image and the image with the structured light pattern, calculate the feature point recognition gain, ensure that the prominence of the feature points in the image is significantly improved, thereby improving the detection accuracy.

[0169] Assuming that the number of feature points extracted from the original image Io is F(Io)=500, and the number of feature points extracted from the image Inp with the structured light pattern is F(Inp)=750, the feature point recognition gain Gf is calculated as:

[0170] Gf=(750-500) / 500;

[0171] Gf=250 / 500;

[0172] Gf=0.5 or 50%.

[0173] Through the above steps, the present invention realizes structured light projection combined with clarity improvement, and enhances the recognition of feature points in the image. Design a suitable structured light pattern, select an appropriate projection device and adjust the projection angle, accurately record the projection parameters of each pattern, and finally collect the image sequence under the projected pattern and calculate the feature point recognition gain, which ensures a significant improvement in image quality and detection accuracy.

[0174] g. Using the generated texture pattern information to capture the image and simultaneously record all current light source configuration parameters; further including the following sub-processes:

[0175] g1. After completing the rich texture information generated in step f, start the image capture device, set the capture resolution Rc and frame rate Fp, and ensure that the image with the structured light pattern can be recorded with high quality; by setting the appropriate capture resolution and frame rate, ensure that the image with the structured light pattern can be obtained with high quality.

[0176] Assuming that you need to capture a high-resolution image sequence, choose a capture resolution of Rc=4K (3840x2160 pixels) and a frame rate of Fp=30fps (30 frames per second).

[0177] Rc=4K;

[0178] Fp=30fps.

[0179] g2. Based on the capture parameters set in step g1, synchronously record the configuration parameters of each light source, including the light source power Pl, angle θl, distance Dl, and the angle αf of the polarization filter, to form a light source configuration data set D_set={Pl,θl,Dl,αf}; synchronously record the configuration parameters of each light source, including power, angle, distance, and the angle of the polarization filter, to help reproduce or adjust the lighting conditions in post-processing and improve image consistency.

[0180] For a specific light source, its configuration is light source power Pl = 500W, light source angle θl = 45°, light source distance Dl = 2m, polarization filter angle αf = 90°. Then the light source configuration data set D_set can be expressed as:

[0181] D_set={Pl=500W, θl=45°, Dl=2m, αf=90°}.

[0182] g3. According to the light source configuration data set D_set recorded in step g2, a corresponding metadata file Mg is generated each time an image is captured. The file contains all light source configuration information and the timestamp Tt of the image capture. The formula Mg=(D_set, Tt) is used to ensure a one-to-one correspondence between the image and the light source configuration. A corresponding metadata file is generated each time an image is captured. The file contains all light source configuration information and a timestamp, ensuring a one-to-one correspondence between the image and its generation conditions, which is convenient for tracking and verification.

[0183] If an image is captured at time Tt=2024-12-24T14:30:00Z and the light source configuration is as described above, the metadata file Mg is:

[0184] Mg=({Pl=500W,θl=45°,Dl=2m,αf=90°},2024-12-24T14:30:00Z).

[0185] g4. Based on step g3, the captured image and its corresponding metadata file Mg are stored in the designated data storage unit Sg to ensure the integrity and traceability of the image and light source configuration information, and an index Ig is established to facilitate subsequent retrieval and analysis. The image and its corresponding metadata file are stored in the designated data storage unit, and an index is established to ensure the integrity and traceability of the data, and to facilitate subsequent retrieval and analysis.

[0186] Assume that there is a data storage unit named Sg for storing images and metadata files, and an index Ig is established for fast search.

[0187] The captured image and its metadata file Mg are stored in Sg.

[0188] An index Ig is a database table or file system directory structure that allows fast retrieval based on image attributes such as timestamps, light source configurations, etc.

[0189] Example: Store the image Image_20241224_143000.png and its corresponding metadata file Metadata_20241224_143000.txt in Sg, and add a record pointing to these two files in the index Ig. The record contains information such as the image path, timestamp, and light source configuration summary.

[0190] h. After completing the image capture, simulate multiple virtual lighting scenarios according to the saved light source configuration parameters, and select the best image sample based on the different lighting effects obtained; further including the following sub-processes:

[0191] h1. After completing the image acquisition and light source configuration parameter saving in step g, read the metadata file Mg and the corresponding image data Id in the storage unit Sg, ready for simulation; by reading the saved metadata file and the corresponding image data, provide accurate basic information for the subsequent simulation of virtual lighting scenes.

[0192] Assume that the metadata file is read from the data storage unit Sg:

[0193] Mg=({Pl=500W,θl=45°,Dl=2m,αf=90°},2024-12-24T14:30:00Z) and image file Id=Image_20241224_143000.png. These data will be used for the following virtual lighting scenario simulation.

[0194] h2. Based on the data read in step h1, a series of virtual lighting scenarios Vsc are set, each scenario contains different light source power Plv, angle θlv, distance Dlv and polarization filter angle αfv to generate various lighting conditions; a variety of different virtual lighting scenarios are set, including changes in light source power, angle, distance and polarization filter angle to generate diverse lighting conditions and simulate lighting effects in different environments.

[0195] Assume that three virtual lighting scenarios Vsc1, Vsc2, and Vsc3 are defined, and their parameters are:

[0196] Vsc1:Plv1=600W, θlv1=50°, Dlv1=2.2m, αfv1=80°;

[0197] Vsc2: Plv2=700W, θlv2=40°, Dlv2=1.8m, αfv2=100°;

[0198] Vsc3: Plv3=400W, θlv3=55°, Dlv3=2.5m, αfv3=85°.

[0199] h3. According to the virtual lighting scenario Vsc defined in step h2, use the ray tracing method to simulate the lighting effect, calculate the illumination intensity Isurface at each point on the thyristor surface in each scenario, and use the formula Isurface=Σ(Plv*cos²(θlv) / Dlv^2), where Σ represents the sum of all light sources; use the ray tracing method to simulate the illumination intensity at each point on the thyristor surface in each virtual lighting scenario to ensure the authenticity and accuracy of the simulation results.

[0200] Assume that there are two light sources in the Vsc1 scenario, the parameters of the first light source are Plv1=600W,θlv1=50°,Dlv1=2.2m, and the parameters of the second light source are Plv2=500W,θlv2=45°,Dlv2=2m. Then the light intensity Isurface at a point on the thyristor surface is calculated as:

[0201] Isurface=(600*cos²(50°) / 2.2²)+(500*cos²(45°) / 2²);

[0202] Isurface≈(600*0.643 / 4.84)+(500*0.5 / 4);

[0203] Isurface≈(78.66)+(62.5);

[0204] Isurface≈141.16lux.

[0205] h4. Based on step h3, apply the simulated illumination intensity Isurface to the original image Id to generate a new image sequence Iseq with different illumination scenarios, and record the brightness difference ΔBh between each new image and the original image, using the formula ΔBh=|Iseq-Id|, to ensure that detailed comparative data is provided for subsequent evaluation. Apply the simulated illumination intensity to the original image to generate a new image sequence with different illumination scenarios, and record the brightness difference between each new image and the original image, to provide detailed comparative data for subsequent evaluation.

[0206] Assuming that the average brightness of the original image Id is 100 lux, and the average brightness of the new image Iseq1 generated under the Vsc1 scenario is 120 lux, the brightness difference ΔBh1 is calculated as:

[0207] ΔBh1=|120-100|;

[0208] ΔBh1=20lux.

[0209] Similarly, for other virtual lighting scenarios Vsc2 and Vsc3, ΔBh2 and ΔBh3 are calculated respectively, and the optimal image samples are selected based on these differences to ensure the best image quality.

[0210] Through step h, we have achieved the simulation of multiple virtual lighting scenarios based on the saved light source configuration parameters, and accurately calculated the lighting intensity through the ray tracing method to generate new image sequences with different lighting scenarios. This not only provides a variety of lighting effect simulations, but also provides detailed comparison data for subsequent evaluation through the comparison of brightness differences, so that the best image samples can be selected, which significantly improves the quality inspection effect of high-voltage thyristors.

[0211] On the other hand, the present invention provides a high-voltage thyristor image generation system for quality inspection, such as Figure 2 As shown, including:

[0212] The lighting mode selection and framework construction module is used to determine the shape parameters and surface characteristics of the high-voltage thyristor to be tested, select the lighting mode according to the information of the shape parameters and surface characteristics, and construct a framework containing multiple light source positions based on the selected lighting mode;

[0213] Light source configuration and dynamic brightness control module, used to adjust the angle and distance of each light source according to the set framework, monitor and adjust the output intensity of each light source in real time through the dynamic brightness control system, and apply polarization filter in front of the imaging device to improve the clarity;

[0214] The structured light projection and image capture module is used to perform structured light projection with improved clarity, project texture patterns onto the object under test, use the generated texture pattern information to capture images, and simultaneously record all current light source configuration parameters;

[0215] The virtual lighting scenario simulation and image evaluation module is used to simulate a variety of virtual lighting scenarios according to the saved light source configuration parameters after completing image capture, and select the optimal image sample based on the different lighting effects obtained.

[0216] In addition, the above-mentioned lighting mode selection and framework construction module, light source configuration and dynamic brightness control module, structured light projection and image capture module, and virtual lighting scene simulation and image evaluation module are also used to implement the other steps of the above-mentioned high-voltage thyristor image generation method for quality inspection when executed, which are not described one by one here.

[0217] In summary, the present invention achieves all-round uniform lighting, reduces shadows and overexposure, improves image clarity and feature point recognition, ensures data integrity and traceability, and ultimately selects the best image samples, significantly improving image quality and the reliability and consistency of detection, through measures such as multi-directional dynamic lighting, real-time brightness control and polarization filtering, structured light projection and synchronous recording, and virtual lighting scene simulation and optimization.

[0218] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for generating a high-voltage thyristor image for quality inspection, characterized in that: The following steps are involved: Determine the shape parameters and surface characteristics of the high-voltage thyristor to be tested, select an illumination mode according to the information of the shape parameters and surface characteristics, and construct a framework including multiple light source positions based on the selected illumination mode; According to the set framework, adjust the angle and distance of each light source, monitor and adjust the output intensity of each light source in real time through the dynamic brightness control system, and apply a polarization filter in front of the imaging device to improve the clarity; Combined with the improved clarity, structured light projection is performed to project a texture pattern onto the object being measured, and the generated texture pattern information is used to capture the image, while simultaneously recording all current light source configuration parameters; After completing the image capture, various virtual lighting scenarios are simulated according to the saved light source configuration parameters, and the optimal image samples are selected based on the different lighting effects obtained.

2. A method for generating a high-voltage thyristor image for quality inspection according to claim 1, characterized in that: The step of determining the shape parameters and surface characteristics of the high-voltage thyristor to be tested, and selecting the lighting mode according to the information of the shape parameters and surface characteristics, comprises: Measure the size and surface reflection characteristics of the high-voltage thyristor to be tested, and select the minimum light intensity according to the size and surface reflection characteristics; Based on the determined minimum light intensity, select the type of light source that will provide uniform illumination without exceeding the maximum permissible heat; Design a multi-angle illumination scheme so that the angle of each light source can directly illuminate the thyristor surface without being blocked and reduce the generation of shadows; Build support structures to accommodate the positions and angles of all light sources.

3. A method for generating a high-voltage thyristor image for quality inspection according to claim 2, characterized in that: Based on the selected lighting mode, a framework including multiple light source positions is constructed, including: Determine the number of light sources and set them according to the size of the high-voltage thyristor to be tested and the required illumination uniformity; Plan the position coordinates of each light source to ensure that the position coordinates cover all surface areas of the thyristor and meet multi-directional illumination requirements; Adjust the relative distance between each light source to ensure uniform distribution of light intensity without overlapping overexposure, and build a support structure to adapt to the position coordinates and relative distances of all light sources.

4. The method for generating a high-voltage thyristor image for quality inspection according to claim 3, characterized in that: The adjusting the angle and distance of each light source according to the set framework includes: For each light source, measure the distance from the center of the light source to any point on the surface of the high-voltage thyristor to ensure that the optimal irradiation angle of each light source enables the light to directly irradiate the surface of the thyristor without being blocked; Determine the optimal illumination angle for each light source to avoid shadows or overexposure caused by the intersection of light beams between adjacent light sources; Adjust the direction of the light source and fine-tune the distance between each light source and the thyristor so that the light intensity is evenly distributed on the surface of the thyristor.

5. A method for generating a high-voltage thyristor image for quality inspection according to claim 4, characterized in that: The dynamic brightness control system monitors and adjusts the output intensity of each light source in real time, including: After the light source angle and distance are adjusted, light intensity sensors are installed at multiple predetermined points on the surface of the thyristor to collect initial light intensity data; When the target light intensity is set, the dynamic brightness control system starts to monitor the output of each light source in real time and adjusts the light source power through the feedback loop so that the actual light intensity is close to the target value; The dynamic brightness control system periodically checks the deviation between the actual light intensity and the target light intensity at all predetermined points. If the deviation exceeds the allowable range, the fine-tuning process is triggered and the light source power is adjusted again until the conditions are met.

6. The method for generating a high-voltage thyristor image for quality inspection according to claim 5, characterized in that: The method of applying a polarization filter before an imaging device to improve the clarity includes: After completing the dynamic brightness control, the polarization filter is used to reduce the impact of reflected light from non-metallic surfaces; Install the selected polarization filter in front of the imaging device lens and adjust the rotation angle of the filter so that the polarization direction of the light passing through the filter is perpendicular to the polarization direction of the reflected light; Measure the actual light intensity after passing through the polarization filter and compare it with the original light intensity without the filter to evaluate the impact of transmittance changes on image quality; If the transmittance change exceeds the preset threshold, fine-tune the angle of the polarization filter or replace it with a filter with different characteristic parameters until the image clarity and contrast meet the detection requirements.

7. A method for generating a high-voltage thyristor image for quality inspection according to claim 6, characterized in that: The method of performing structured light projection in combination with the improved clarity to project a texture pattern onto the object under test includes: After the clarity provided is improved, a structured light pattern is designed, which contains stripes or dots of different frequencies and phases, and the pattern is projected through a projection device, and the projection angle is adjusted so that the light covers the surface of the object being measured; Implementing structured light projection, while recording projection parameters corresponding to each pattern, including but not limited to projection angle, brightness and contrast; The image sequence under the projected pattern is collected, and the enhanced feature point recognition gain is calculated by comparing the original image and the image with the structured light pattern.

8. The method for generating a high-voltage thyristor image for quality inspection according to claim 7, characterized in that: The generated texture pattern information is used to capture the image and simultaneously record all current light source configuration parameters, including: After the texture information is generated, the image capture device is started, and the capture resolution and frame rate are set to ensure that the image with the structured light pattern can be recorded; Synchronously record the configuration parameters of each light source, including light source power, angle, distance, and angle of the polarization filter, to form a light source configuration data set; Each time an image is captured, a corresponding metadata file is generated, which contains all light source configuration information and the timestamp of image capture; The captured image and its corresponding metadata file are stored in the specified data storage unit.

9. A method for generating a high-voltage thyristor image for quality inspection according to claim 8, characterized in that: After the image capture is completed, a variety of virtual lighting scenarios are simulated according to the saved light source configuration parameters, based on the different lighting effects obtained, including: After the image is acquired and the light source configuration parameters are saved, the metadata file and the corresponding image data in the storage unit are read for simulation; Set up virtual lighting scenarios, each of which contains different light source power, angle, distance, and polarization filter angle to generate a variety of lighting conditions; The ray tracing method is applied to simulate the lighting effect, and the light intensity of each point on the thyristor surface in each scenario is calculated. The simulated light intensity is applied to the original image to generate a new image sequence with different lighting scenarios. At the same time, the brightness difference between each new image and the original image is recorded.

10. A high-voltage thyristor image generation system for quality inspection for executing the method according to any one of claims 1 to 9, characterized in that: include: The lighting mode selection and framework construction module is used to determine the shape parameters and surface characteristics of the high-voltage thyristor to be tested, select the lighting mode according to the information of the shape parameters and surface characteristics, and construct a framework containing multiple light source positions based on the selected lighting mode; Light source configuration and dynamic brightness control module, used to adjust the angle and distance of each light source according to the set framework, monitor and adjust the output intensity of each light source in real time through the dynamic brightness control system, and apply polarization filter in front of the imaging device to improve the clarity; The structured light projection and image capture module is used to perform structured light projection with improved clarity, project texture patterns onto the object under test, use the generated texture pattern information to capture images, and simultaneously record all current light source configuration parameters; The virtual lighting scenario simulation and image evaluation module is used to simulate a variety of virtual lighting scenarios according to the saved light source configuration parameters after completing image capture, and select the optimal image sample based on the different lighting effects obtained.

Images