LED light source defect detection method and system

Abstract

The invention discloses an LED light source defect detection method and system, and relates to the technical field of LED light source defect detection. According to the LED light source defect detection method, light intensity time sequence image data and temperature time sequence image data of an LED light source when a plurality of test currents are introduced are obtained, a defect index set of the LED light source is analyzed, a comprehensive defect evaluation index is analyzed based on the defect index set of the LED light source, judgment and analysis are carried out on the comprehensive defect evaluation index and a preset defect evaluation interval, and a defect evaluation result is obtained. According to the method, the light intensity time sequence image data and the temperature time sequence image data under different current conditions are collected at the same time, multi-dimensional feature analysis is combined, the defects of the LED light source can be comprehensively evaluated, performance degradation of the LED in different working environments can be effectively revealed, especially the influence of current fluctuation and temperature change on the light intensity and stability, and the LED light source detection accuracy is improved. By analyzing the defect index set of the LED light source, not only can a single defect be captured, but also the comprehensive defect of the light source can be accurately identified.

Description

Technical Field

[0001] The present invention relates to the technical field of LED light source defect detection, and specifically to a method and system for detecting LED light source defects. Background Art

[0002] LED (Light Emitting Diode) light sources have been widely used in the fields of lighting, display, communication, etc. due to their advantages such as high efficiency, energy saving, and long lifespan. However, with the expansion of the application range of LED light sources and the continuous improvement of performance requirements, their long-term stability and reliability have become an important research direction in the development of LED technology.

[0003] During the operation of an LED light source, affected by environmental factors such as current and temperature, its performance may decline or malfunctions may occur. Defects in LED light sources, such as uneven light intensity, brightness attenuation, overheating, etc., seriously affect their luminous efficiency, lifespan, and user experience. Therefore, timely and effectively detecting the defects of LED light sources is of great significance for extending their lifespan, improving work efficiency, and reducing maintenance costs.

[0004] Existing technologies, such as the LED light source defect detection method and system disclosed in the patent application with the publication number CN117437170A, include: turning on the drive power supply of the LED light source to be tested and obtaining the illumination image of the LED light source; obtaining the spot contour of the bright area in the illumination image; calculating the area of the spot contour to obtain the spot area; comparing the spot area with the standard area and detecting the current LED light source according to the comparison result. Through the above solution, it can completely replace manual detection of LED light sources, effectively saving labor costs, and has the advantages of accurate detection results and fast detection speed.

[0005] Based on the above solution, it is found that the limitations of existing technologies at least include the following problems. First, existing LED light source defect detection methods are often limited to the analysis of traditional single data sources, such as only using static light intensity images or temperature data for evaluation. This makes it difficult for existing technologies to comprehensively evaluate the comprehensive defect state of LED light sources under dynamically changing current and temperature conditions. Since existing technologies fail to fully utilize the timing data under different current tests, in environments with various current fluctuations and temperature changes, they cannot effectively identify potential complex defects of LED light sources. For example, existing technologies are difficult to accurately capture the mutual relationship between current fluctuations and light intensity differences, and it is also difficult to effectively analyze the specific impact of temperature changes and current interaction effects on the performance of LED light sources, resulting in relatively low accuracy of defect identification and difficulty in comprehensively improving the stability evaluation of LED light sources. Summary of the Invention

[0006] Aiming at the deficiencies of the prior art, the present invention provides a method and system for detecting defects in LED light sources, which solves the problem in the prior art that it is difficult to comprehensively consider the interaction of multi-dimensional factors such as current fluctuations, temperature changes, and light intensity differences, and thus it is difficult to accurately identify defects.

[0007] To achieve the above objectives, the present invention is realized through the following technical solutions: A method for detecting defects in an LED light source, comprising the following steps: obtaining the light intensity time-series image data and temperature time-series image data of the LED light source when passing through several test currents, and respectively performing preprocessing; respectively performing feature analysis on the preprocessed light intensity time-series image data and temperature time-series image data of the LED light source when passing through several test currents to obtain a defect index set of the LED light source, including a light intensity difference defect index, a current fluctuation defect index, and a local thermal effect defect index; comprehensively analyzing the defect index set of the LED light source to obtain a comprehensive defect evaluation index of the LED light source, and its calculation formula is as follows: Wherein, ZqX is the comprehensive defect evaluation index of the LED light source, GqQ is the light intensity difference defect index of the LED light source, λ1 is the light intensity influence coefficient stored in the database, DqQ is the current fluctuation defect index of the LED light source, λ2 is the current influence coefficient stored in the database, RqQ is the local thermal effect defect index of the LED light source, μ1 is the thermal effect adjustment coefficient stored in the database, λ3 is the thermal effect influence coefficient stored in the database, μ2 is the light intensity-current interaction adjustment coefficient stored in the database, and λ4 is the light intensity-current interaction influence coefficient stored in the database; judging and analyzing the comprehensive defect evaluation index of the LED light source with a preset defect evaluation interval. If the comprehensive defect evaluation index of the LED light source is within the preset defect evaluation interval, it is regarded that the LED light source has defects.

[0008] Further, the light intensity time-series image data includes light intensity images at several time points, and each light intensity image includes pixel values of several light source points. The specific steps for obtaining the light intensity difference defect index of the LED light source are as follows: comprehensively analyzing the light intensity images at several time points of the LED light source when passing through several test currents to obtain a light intensity difference feature set of the LED light source, including the average pixel values, pixel time variance values, and time pixel means at several time points of several light source points when passing through several test currents; respectively combining the pixel values of several light source points in the light intensity images at several time points of the LED light source when passing through several test currents with the light intensity difference feature set of the LED light source for comprehensive analysis to obtain the light intensity difference defect index of the LED light source.

[0009] Further, the specific steps to obtain the light intensity difference feature set of the LED light source are as follows: Read the pixel values of several light source points in the light intensity images of the LED light source at several time points when several test currents are applied, and perform time point analysis respectively to obtain the average pixel values of several light source points of the LED light source when several test currents are applied; Read the pixel values of several light source points in the light intensity images of the LED light source at several time points when several test currents are applied, and perform variance analysis in combination with the average pixel values of several light source points of the LED light source when several test currents are applied to obtain the pixel time variance values of several light source points of the LED light source when several test currents are applied; Read the pixel values of several light source points in the light intensity images of the LED light source at several time points when several test currents are applied, and perform light source point pixel analysis respectively to obtain the time pixel means of several time points of the LED light source when several test currents are applied.

[0010] Further, the specific formula for calculating the light intensity difference defect index of the LED light source is as follows: where GqQ is the light intensity difference defect index of the LED light source, XsZ i(u+1)r is the pixel value of the r-th light source point in the light intensity image at the (u + 1)-th time point when the i-th test current is applied to the LED light source, XsZ iur is the pixel value of the r-th light source point in the light intensity image at the u-th time point when the i-th test current is applied to the LED light source, PXS ir is the average pixel value of the r-th light source point when the i-th test current is applied to the LED light source, XsF ir is the pixel time variance value of the r-th light source point when the i-th test current is applied to the LED light source, δ1 is the time variance influence coefficient stored in the database, SxS iu is the time pixel mean at the u-th time point when the i-th test current is applied to the LED light source, δ2 is the time pixel influence coefficient stored in the database, i = 1, 2, 3, …, i0, i0 is the number of types of test currents, u = 1, 2, 3, …, u0, u0 is the number of time points, r = 1, 2, 3, …, r0, r0 is the number of light source points.

[0011] Further, the specific steps to obtain the current fluctuation defect index of the LED light source are as follows: Obtain the test current values at several time points for each type of test current, and perform comprehensive analysis to obtain several groups of test current difference values for each type of test current; Combine several groups of test current difference values for each type of test current with the pixel values of several light source points in the light intensity images of the LED light source at several time points when several test currents are applied, and perform comprehensive analysis to obtain the current fluctuation defect index of the LED light source.

[0012] Furthermore, the specific formula for calculating the current fluctuation defect index of the LED light source is as follows: where DqQ is the current fluctuation defect index of the LED light source, and XsZ i(u+1)r is the pixel value of the r-th light source point in the light intensity image at the (u + 1)-th time point when the i-th test current is applied to the LED light source, and XsZ iur is the pixel value of the r-th light source point in the light intensity image at the u-th time point when the i-th test current is applied to the LED light source, ε is the pixel adjustment factor stored in the database, and CfD ia is the a-th test current difference value of the i-th test current, χ is the current adjustment factor stored in the database, β is the differential current adjustment coefficient stored in the database, i = 1, 2, 3, …, i0, where i0 is the number of types of test currents, u = 1, 2, 3, …, u0, where u0 is the number of time points, r = 1, 2, 3, …, r0, where r0 is the number of light source points, and a = 1, 2, 3, …, a0, where a0 is the number of groups of test current difference values.

[0013] Furthermore, the temperature time-series image data includes temperature images at a number of time points, and each temperature image includes temperature values of a number of light source points. The specific steps for obtaining the local thermal effect defect index of the LED light source are as follows: perform feature analysis on the temperature time-series image data of the LED light source when a number of test currents are applied respectively to obtain a temperature difference feature set of the LED light source, including the time-temperature mean value and the temperature-time gradient value at a number of time points when a number of test currents are applied; comprehensively analyze the temperature values of a number of light source points in the temperature images at a number of time points when a number of test currents are applied to the LED light source respectively in combination with the temperature difference feature set of the LED light source to obtain the local thermal effect defect index of the LED light source.

[0014] Further, the specific steps to obtain the temperature difference feature set of the LED light source are as follows: Read the temperature values of several light source points in the temperature images of the LED light source at several time points when several test currents are applied, and perform light source point temperature analysis respectively to obtain the time-temperature means of the LED light source at several time points when several test currents are applied; For each light source point in the temperature images of the LED light source at several time points when several test currents are applied, identify several adjacent light source points within the set neighborhood range; Comprehensively analyze the temperature values of each light source point in the temperature images of the LED light source at several time points when several test currents are applied with the temperature values of several adjacent light source points within the set neighborhood range of the corresponding light source point to obtain the temperature-time gradient values of the LED light source at several time points when several test currents are applied; Comprehensively analyze the local temperature-time gradient values of each light source point in the temperature images of the LED light source at several time points when several test currents are applied to obtain the temperature-time gradient values of the LED light source at several time points when several test currents are applied.

[0015] Further, the specific formula for calculating the local thermal effect defect index of the LED light source is as follows: where \(RqQ\) is the local thermal effect defect index of the LED light source, \(WdZ\) iur is the temperature value of the \(r\)-th light source point in the temperature image of the LED light source at the \(u\)-th time point when the \(i\)-th test current is applied, \(SwZ\) iu is the time-temperature mean of the LED light source at the \(u\)-th time point when the \(i\)-th test current is applied, \(\theta\) is the temperature adjustment factor stored in the database, \(SwT\) iu is the temperature-time gradient value of the LED light source at the \(u\)-th time point when the \(i\)-th test current is applied, \(\omega\) is the temperature gradient adjustment coefficient stored in the database, \(\xi\) is the temperature gradient influence coefficient stored in the database, \(i = 1, 2, 3, \ldots, i0\), \(i0\) is the number of test current types, \(u = 1, 2, 3, \ldots, u0\), \(u0\) is the number of time points, \(r = 1, 2, 3, \ldots, r0\), \(r0\) is the number of light source points.

[0016] An LED light source defect detection system, comprising: an image data acquisition unit for acquiring the light intensity time-series image data and temperature time-series image data of the LED light source under several test currents and respectively performing preprocessing; an image feature analysis unit for respectively performing feature analysis on the preprocessed light intensity time-series image data and temperature time-series image data of the LED light source under several test currents to obtain a defect index set of the LED light source, including a light intensity difference defect index, a current fluctuation defect index, and a local thermal effect defect index; a comprehensive defect evaluation unit for comprehensively analyzing the defect index set of the LED light source to obtain a comprehensive defect evaluation index of the LED light source; and a defect judgment unit for performing judgment analysis on the comprehensive defect evaluation index of the LED light source and a preset defect evaluation interval. If the comprehensive defect evaluation index of the LED light source is within the preset defect evaluation interval, it is considered that the LED light source has a defect.

[0017] The present invention has the following beneficial effects:

[0018] (1). This LED light source defect detection method can comprehensively evaluate the defects of the LED light source by simultaneously collecting the light intensity time-series images and temperature time-series image data under different current conditions and combining multi-dimensional feature analysis. This method can effectively reveal the performance degradation of the LED under different working environments, especially the influence of current fluctuation and temperature change on light intensity and stability. However, the prior art often relies only on a single data source, such as a static light intensity image or a temperature image, which makes it difficult to comprehensively evaluate the defects of the LED light source under dynamic current and temperature change conditions. By comprehensively calculating the light intensity difference defect index, the current fluctuation defect index, and the local thermal effect defect index, this method can not only capture single defects but also accurately identify the comprehensive defects of the light source, improving the accuracy and comprehensiveness of detection. This is of great significance for improving the stability and lifespan of the LED light source in practical applications.

[0019] (2). This LED light source defect detection method can accurately analyze the mutual influence of current and temperature by combining multi-dimensional data of current fluctuation, light intensity difference, and temperature change, so as to dynamically evaluate the performance of the LED light source. The prior art usually ignores the interaction effect between current fluctuation and temperature change and only relies on static light intensity or temperature data to evaluate the state of the light source, which leads to the risk of difficultly accurately identifying potential defects in the actual application of the LED light source. By introducing adjustment coefficients stored in the database, such as a current adjustment factor and a temperature gradient adjustment factor, this method can perform real-time evaluation on the LED light source under different current and temperature conditions and accurately identify complex defect patterns, enabling the LED light source to maintain good adaptability in various working environments, ensuring its long-term stability and high efficiency, and effectively avoiding the problem that it is difficult to reliably detect defects by traditional methods in extreme environments.

[0020] (3) The LED light source defect detection method can calculate the comprehensive defect evaluation index of the LED light source through real-time analysis of the light intensity time-series image and the temperature time-series image, and compare it with the preset defect evaluation interval, so as to judge in real time whether there are defects in the LED light source. This technology based on time-series data and dynamic evaluation provides an efficient real-time monitoring and early warning mechanism, ensuring that feedback and intervention can be obtained in a timely manner when performance problems occur in the LED light source. However, the existing technologies usually rely on regular inspections or manual maintenance, lacking comprehensive monitoring of the real-time state of the LED light source. By combining real-time data analysis with automated judgment, this method significantly improves the real-time performance and efficiency of detection, can quickly take measures when any potential faults occur, reduces the risks of misjudgment and missed judgment, reduces the maintenance cost of the equipment, and extends the service life of the LED light source.

[0021] (4) The LED light source defect detection system realizes comprehensive monitoring and intelligent evaluation of the LED light source through the organic combination of an image data acquisition unit, an image feature analysis unit, a comprehensive defect evaluation unit, and a defect judgment unit. The system can automatically acquire the light intensity time-series image data and the temperature time-series image data under different current conditions, and reduce human intervention through automatic preprocessing and feature extraction, improving the automation degree of the detection process. However, the existing technologies often rely on manual operations or traditional manual inspections, resulting in a cumbersome detection process and being easily affected by human factors. Through automated image processing and feature analysis, this system can not only quickly process a large amount of data, but also dynamically adjust the evaluation parameters according to real-time data, making the defect detection of the LED light source more accurate and efficient. By fully automating the defect evaluation process, this system can significantly reduce human errors, improve the detection speed, and reduce the labor cost, while enhancing the intelligent level of detection, providing more reliable support for real-time monitoring and maintenance in large-scale LED light source applications.

[0022] Of course, it is not necessary for any product implementing the present invention to simultaneously achieve all the above-mentioned advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a flowchart of a method for detecting defects of an LED light source according to the present invention.

[0024] Figure 2 It is a schematic diagram of the light intensity image and the temperature image when a test current is applied to the LED light source in a method for detecting defects of an LED light source according to the present invention.

[0025] Figure 3 It is a flowchart of the specific steps for obtaining the light intensity difference feature set of the LED light source in a method for detecting defects of an LED light source according to the present invention.

[0026] Figure 4 This is a block diagram of a defect detection system for an LED light source according to the present invention. Specific embodiments

[0027] Please refer to Figure 1 , an embodiment of the present invention provides a technical solution: a method for detecting defects of an LED light source, including the following steps: obtaining light intensity time-series image data and temperature time-series image data of the LED light source under several test currents, and respectively performing preprocessing; respectively performing feature analysis on the preprocessed light intensity time-series image data and temperature time-series image data of the LED light source under several test currents to obtain a defect index set of the LED light source, including a light intensity difference defect index, a current fluctuation defect index, and a local thermal effect defect index; comprehensively analyzing the defect index set of the LED light source to obtain a comprehensive defect evaluation index of the LED light source; judging and analyzing the comprehensive defect evaluation index of the LED light source with a preset defect evaluation interval. If the comprehensive defect evaluation index of the LED light source is within the preset defect evaluation interval, it is considered that the LED light source has a defect, and light intensity images and temperature image examples at a certain time point when the test currents (0.7A, 1.0A, 1.2A) are respectively applied to the LED light source are provided, such as Figure 2 shown.

[0028] The specific formula for calculating the comprehensive defect evaluation index of the LED light source is as follows: Among them, ZqX is the comprehensive defect evaluation index of the LED light source, GqQ is the light intensity difference defect index of the LED light source, λ1 is the light intensity influence coefficient stored in the database, DqQ is the current fluctuation defect index of the LED light source, λ2 is the current influence coefficient stored in the database, RqQ is the local thermal effect defect index of the LED light source, μ1 is the thermal effect adjustment coefficient stored in the database, λ3 is the thermal effect influence coefficient stored in the database, μ2 is the light intensity-current interaction adjustment coefficient stored in the database, and λ4 is the light intensity-current interaction influence coefficient stored in the database.

[0029] It should be explained that the specific steps for obtaining the light intensity influence coefficient λ1, current influence coefficient λ2, thermal effect influence coefficient λ3, and light intensity-current interaction influence coefficient λ4 stored in the database are as follows: First, in the experimental stage, the light intensity time-series images of the LED light source under various different current intensities are recorded and collected, while simultaneously monitoring the current fluctuations and temperature changes. For the light intensity, temperature, and current data under each different current condition, mathematical models between current fluctuations and light intensity differences, and temperature changes are established using regression analysis methods (such as multiple linear regression, decision tree, support vector machine, etc.). Through these models, we can extract the mutual relationships between current, temperature, and light intensity. These relationships are stored in the database in the form of coefficients, forming the light intensity influence coefficient, current influence coefficient, and thermal effect influence coefficient. Specifically, the light intensity influence coefficient reflects the influence of different current changes on the LED light intensity; the current influence coefficient reflects the influence of current fluctuations on the LED stability; the thermal effect influence coefficient measures the influence of insufficient heat dissipation or local overheating on the LED performance through temperature data; the light intensity-current interaction influence coefficient evaluates the coupling effect of current fluctuations on light intensity changes. These coefficients are stored and updated in real time in the database according to different current, temperature, and light intensity conditions.

[0030] The specific steps for obtaining the thermal effect adjustment coefficient μ1 and light intensity-current interaction adjustment coefficient μ2 stored in the database are as follows: They are obtained through systematic experiments and modeling of the thermal effect and the interaction effect of current and light intensity of the LED light source under different working environment conditions. During the experiment, the LED light source operates under different current and temperature environments, and temperature data and light intensity data at multiple time points are collected. By statistically analyzing and processing these data using optimization algorithms (such as genetic algorithm, particle swarm optimization algorithm, or neural network model), the adjustment effect of temperature fluctuations on the LED performance, that is, the thermal effect adjustment coefficient, can be obtained. In addition, for the interaction effect of current and light intensity, a similar method is used, combined with the light intensity data under different current conditions, to establish a mathematical model for the interaction adjustment between current and light intensity. These adjustment coefficients help the system accurately reflect the complex relationship between current fluctuations and light intensity. The acquisition of these coefficients is stored in the database through continuous optimization and adjustment, so as to adjust the adjustment parameters in real time during the actual operation of the LED to ensure its stability and efficiency. Through this method, the database not only stores static influence coefficients but also records the dynamic adjustment parameters during the operation of the system in real time, ensuring more accurate performance evaluation and adjustment.

[0031] The specific implementation example of calculating the comprehensive defect assessment index of the LED light source is as follows. The following parameters are available:

[0032] The light intensity difference defect index of the LED light source is approximately: 0.763.

[0033] The light intensity influence coefficient stored in the database is approximately: 1.452.

[0034] The current fluctuation defect index of the LED light source is approximately: 0.540.

[0035] The current influence coefficient stored in the database is approximately: 1.621.

[0036] The local thermal effect defect index of the LED light source is approximately: 0.815.

[0037] The thermal effect adjustment coefficient stored in the database is approximately: 1.134.

[0038] The thermal effect influence coefficient stored in the database is approximately: 0.942.

[0039] The light intensity - current interaction adjustment coefficient stored in the database is approximately: 1.365.

[0040] The light intensity - current interaction influence coefficient stored in the database is approximately: 0.890.

[0041] Substitute the above - mentioned data into the specific formula for calculating the comprehensive defect evaluation index of the LED light source respectively, and we get:

[0042] The comprehensive defect evaluation index of the LED light source=(1.452×0.763)+(1.621×0.540)+(1 + 0.942×((0.815)^1.134))+(1 + 0.890×((0.763×0.540)^1.365))≈4.995.

[0043] Specifically, the light intensity time - series image data includes light intensity images at several time points, and each light intensity image includes pixel values of several light source points, that is, the light intensity (brightness) at the light source point positions. The light source point positions and quantities in each light intensity image are the same and correspond one - to - one to each light source point in the LED light source, which is used to reflect the brightness intensity distribution of the LED light source points. The specific steps to obtain the light intensity difference defect index of the LED light source are as follows: comprehensively analyze the light intensity images at several time points of the LED light source under several test currents respectively to obtain the light intensity difference feature set of the LED light source, including the average pixel values of several light source points, pixel - time variance values, and time - pixel mean values at several time points under several test currents; comprehensively analyze the pixel values of several light source points in the light intensity images at several time points of the LED light source under several test currents respectively in combination with the light intensity difference feature set of the LED light source to obtain the light intensity difference defect index of the LED light source.

[0044] The specific formula for calculating the light intensity difference defect index of the LED light source is as follows: Among them, GqQ is the light intensity difference defect index of the LED light source, XsZ i(u+1)r is the pixel value of the r-th light source point in the light intensity image at the (u + 1)-th time point when the i-th test current is applied to the LED light source, XsZ iur is the pixel value of the r-th light source point in the light intensity image at the u-th time point when the i-th test current is applied to the LED light source, PxS ir is the average pixel value of the r-th light source point when the i-th test current is applied to the LED light source, XsF ir is the pixel time variance value of the t-th light source point when the i-th test current is applied to the LED light source, δ1 is the time variance influence coefficient stored in the database, SxS in is the time pixel mean value of the y-th time point when the i-th test current is applied to the LED light source, δ2 is the time pixel influence coefficient stored in the database, i = 1, 2, 3, …, i0, i0 is the number of types of test currents, u = 1, 2, 3, …, u0, u0 is the number of time points, r = 1, 2, 3, …, r0, r0 is the number of light source points.

[0045] It should be explained that the specific steps for obtaining the time variance influence coefficient δ1 and the time pixel influence coefficient δ2 stored in the database are as follows: Record the light intensity image data of the LED light source at multiple different time points (such as t1, t2, t3, …), and then use analysis of variance to calculate the time variance of the light intensity change in the image data, that is, the degree of dispersion of the light intensity change between different time points, as the time variance influence coefficient. At the same time, through pixel point analysis of the light intensity image at each time point, calculate the performance of each pixel point at different times, and further obtain the time pixel influence coefficient. These two coefficients are obtained through regression analysis or data fitting methods and stored in the database for subsequent use in calculating the light source difference and evaluating the light intensity change.

[0046] In this implementation scheme, by defining and calculating the light intensity difference defect index of the LED light source in detail, and through the time variance influence coefficient and the time pixel influence coefficient, the accurate evaluation of the light source performance is realized. By performing variance analysis and pixel point analysis on the light intensity image data at multiple different time points, not only can the temporal changes of the light source be effectively captured, but also the light intensity fluctuations and non-uniformities of the LED light source under different current conditions can be revealed, so as to identify potential defects. This method can combine time series data with pixel-level data to provide a more comprehensive and accurate defect assessment than traditional static detection methods. First, through the comprehensive analysis of the light intensity time series image data of the LED light source under different current conditions, combined with the influence of time variance and time pixels, the stability and health status of the LED light source under long-term and various current conditions can be reflected in detail, which is crucial for revealing the problems of brightness unevenness or decline caused by current fluctuations and temperature changes. Second, by storing and using the time variance influence coefficient and the time pixel influence coefficient in the database, it provides an important reference for subsequent light intensity difference defect detection, and makes the detection method more automated and intelligent, reducing the deviation and error of manual operation. This method not only improves the accuracy of defect detection of the LED light source, but also comprehensively considers the influence of current and temperature changes on light intensity through the analysis of time series data, effectively improving the reliability and applicability of the detection results. By comprehensively using these coefficients, the performance of the LED light source can be dynamically monitored, and a scientific basis can be provided for subsequent maintenance and performance optimization, enhancing the stability and lifespan of the LED light source in practical applications.

[0047] Specifically, as Figure 3 shown, the specific steps to obtain the light intensity difference feature set of the LED light source are as follows: Read the pixel values of several light source points in the light intensity images at several time points when the LED light source is applied with several test currents, and perform time point analysis respectively (that is, perform mean analysis on several time points) to obtain the average pixel values of several light source points of the LED light source when applied with several test currents; Read the pixel values of several light source points in the light intensity images at several time points when the LED light source is applied with several test currents, and perform variance analysis in combination with the average pixel values of several light source points of the LED light source when applied with several test currents to obtain the pixel time variance values of several light source points of the LED light source when applied with several test currents; Read the pixel values of several light source points in the light intensity images at several time points when the LED light source is applied with several test currents, and perform light source point pixel analysis respectively (that is, perform pixel mean analysis on several light source points) to obtain the time pixel means (the pixel means of several light source points in the light intensity image) of several time points of the LED light source when applied with several test currents.

[0048] In this implementation, through a multi-dimensional data analysis method, it is detailed how to calculate the light intensity difference feature set of the LED light source, including the average pixel value, pixel time variance value, and time pixel mean, thereby providing key data support for accurately evaluating the defects of the LED light source. First, by performing a mean analysis on the light intensity images at each time point, the light intensity stability of the light source points under different current conditions can be understood; second, through variance analysis, the volatility between the light source points at multiple time points can be captured, further revealing whether there are non-uniformity or decay problems in the LED light source; finally, pixel mean analysis provides a data basis for detecting the brightness change trend of each light source point. These steps can help the system more accurately identify the potential defects of the LED light source and improve the accuracy and comprehensiveness of defect detection.

[0049] Specifically, the specific steps to obtain the current fluctuation defect index of the LED light source are as follows: Obtain the test current values at several time points for each test current and perform comprehensive analysis (i.e., the difference in test currents between adjacent time points) to obtain several groups of test current difference values for each test current; Combine the several groups of test current difference values for each test current with the pixel values of several light source points in the light intensity images at several time points when several test currents are applied to the LED light source for comprehensive analysis to obtain the current fluctuation defect index of the LED light source.

[0050] The specific formula for calculating the current fluctuation defect index of the LED light source is as follows: where DqQ is the current fluctuation defect index of the LED light source, XsZ i(u+1)r is the pixel value of the r-th light source point in the light intensity image at the (u + 1)-th time point when the i-th test current is applied to the LED light source, XsZ iur is the pixel value of the r-th light source point in the light intensity image at the u-th time point when the i-th test current is applied to the LED light source, ε is the pixel adjustment factor stored in the database, used to prevent the denominator from being 0, CfD ia is the a-th group of test current difference values of the i-th test current, χ is the current adjustment factor stored in the database, used to prevent the denominator from being 0, β is the differential current adjustment coefficient stored in the database, i = 1, 2, 3,..., i0, i0 is the number of test current types, u = 1, 2, 3,..., u0, u0 is the number of time points, r = 1, 2, 3,..., r0, r0 is the number of light source points, a = 1, 2, 3,..., a0, a0 is the number of groups of test current difference values.

[0051] It should be noted that the specific steps for obtaining the differential current adjustment coefficient β stored in the database are as follows: First, at different current intensities, record the current fluctuation data of the LED light source at multiple time points, that is, the change amount of the current at each moment. Through statistical analysis of these data, calculate the differential value of the current fluctuation, and evaluate the impact of the current fluctuation on the LED performance. Then, use regression analysis or optimization algorithms (such as the least squares method, gradient descent, etc.) to fit the relationship between the current fluctuation and the LED performance indicators (such as brightness change, temperature fluctuation, etc.), so as to obtain the differential current adjustment coefficient. These coefficients reflect the adjustment effect of the current fluctuation on the LED stability and efficiency, and are stored in the database for subsequent calculation and real-time adjustment.

[0052] In this implementation scheme, by combining the calculation of the test current differential value and the light intensity image data, it is possible to effectively identify the impact of current fluctuations on the performance of the LED light source. Especially at different time points and under current changes, whether the light intensity fluctuates. The advantage of this method is that through the differential current adjustment coefficient, it can reflect the specific impact of current fluctuations on the stability and efficiency of the LED light source, further improving the accuracy of LED light source defect detection. The prior art often ignores the details of current fluctuations, resulting in an incomplete evaluation of the stability and efficacy of the LED light source. The current adjustment coefficient obtained through regression analysis or optimization algorithms can combine current fluctuations with factors such as the light intensity change and temperature fluctuation of the LED, ensuring that the performance of the LED light source can be accurately evaluated in an environment with current changes. In addition, using the adjustment coefficient in the database for real-time adjustment not only improves the automation level of the system, but also provides a more scientific basis for subsequent maintenance and adjustment. Generally speaking, this method effectively improves the dynamic stability detection of the LED light source, providing strong technical support for the reliability and efficiency optimization of LED products in practical applications.

[0053] Specifically, the temperature time-series image data includes temperature images at several time points, and each temperature image includes the temperature values of several light source points, that is, the surface temperature at the light source point position. The position and quantity of the light source points in each temperature image are the same, and they all correspond one-to-one to each light source point in the LED light source, which is used to reflect the temperature distribution of the LED light source points. The specific steps for obtaining the local thermal effect defect index of the LED light source are as follows: Perform feature analysis on the temperature time-series image data of the LED light source under several test currents respectively to obtain the temperature difference feature set of the LED light source, including the time-temperature mean value and temperature-time gradient value at several time points under several test currents; Combine the temperature values of several light source points in the temperature images of the LED light source at several time points under several test currents with the temperature difference feature set of the LED light source respectively for comprehensive analysis to obtain the local thermal effect defect index of the LED light source.

[0054] The specific formula for calculating the local thermal effect defect index of the LED light source is as follows: Among them, RqQ is the local thermal effect defect index of the LED light source, WdZ iur is the temperature value of the r-th light source point in the temperature image of the u-th time point when the LED light source is applied with the i-th test current, SwZ iu is the time-temperature average value of the u-th time point when the LED light source is applied with the i-th test current, θ is the temperature adjustment factor stored in the database, which is used to prevent the denominator from being 0, SwT iu is the temperature-time gradient value of the u-th time point when the LED light source is applied with the i-th test current, ω is the temperature gradient adjustment coefficient stored in the database, ξ is the temperature gradient influence coefficient stored in the database, i = 1, 2, 3,..., i0, where i0 is the number of types of test currents, u = 1, 2, 3,..., u0, where u0 is the number of time points, r = 1, 2, 3,..., r0, where r0 is the number of light source points.

[0055] It should be explained that the specific steps for obtaining the temperature gradient adjustment coefficient ω and the temperature gradient influence coefficient ξ stored in the database are as follows: First, in the experiment, record the temperature changes of the LED light source under different current conditions, especially the temperature gradients at different time points. The temperature image data at each time point will be used to calculate the temperature gradient, that is, the temperature change rate of each pixel point. Then, through regression analysis or machine learning methods (such as multiple regression or support vector machines), analyze the relationship between the temperature gradient and the LED performance (such as light intensity, stability, etc.) to obtain the temperature gradient adjustment coefficient, which represents the adjustment effect of the temperature gradient on the LED performance. At the same time, the temperature gradient influence coefficient reflects the influence of the temperature gradient under different current and temperature conditions, and is obtained through data fitting or optimization algorithms. Finally, these coefficients are stored in the database and used as key parameters in subsequent defect evaluations to ensure the stability and efficiency of the LED under different working conditions.

[0056] In this implementation, by describing in detail how to calculate the local thermal effect defect index of the LED light source using temperature time-series image data, a comprehensive analysis framework is provided that can accurately evaluate the thermal effect of the LED light source under different current and temperature conditions. First, the temperature time-series image data can effectively capture the local temperature fluctuations and thermal non-uniformity of the LED light source by analyzing the temperature values of each light source point one by one and combining the time-temperature mean and temperature gradient, thus providing important data support for defect detection. Through regression analysis or machine learning methods, the calculated temperature gradient adjustment coefficient and temperature gradient influence coefficient can accurately reflect the influence of different temperature gradients on the performance of the LED light source, providing an accurate basis for real-time evaluation. The prior art usually ignores the complex influence of the temperature gradient, while this method makes up for this deficiency through a detailed analysis of the temperature gradient, providing an in-depth understanding of the thermal effect of the LED light source. In addition, the temperature gradient adjustment coefficient and temperature gradient influence coefficient stored in the database enable the dynamic adjustment and real-time feedback of the system. In practical applications, these coefficients can dynamically adjust the performance evaluation of the LED light source according to different current and temperature changes, ensuring the stability and reliability of the LED light source under various working conditions.

[0057] Specifically, the specific steps to obtain the temperature difference feature set of the LED light source are as follows: Read the temperature values of several light source points in the temperature images of the LED light source at several time points when several test currents are applied, and perform light source point temperature analysis respectively (that is, perform temperature mean analysis on several light source points) to obtain the time-temperature means (the temperature means of several light source points in the temperature image) of the LED light source at several time points when several test currents are applied; For each light source point in the temperature images of the LED light source at several time points when several test currents are applied, identify several adjacent light source points within the set neighborhood range respectively; Perform comprehensive analysis on the temperature value of each light source point in the temperature images of the LED light source at several time points when several test currents are applied with the temperature values of several adjacent light source points within the set neighborhood range of the corresponding light source point to obtain the temperature-time gradient values of the LED light source at several time points when several test currents are applied. Specifically, perform absolute difference analysis on the temperature value of the light source point and the temperature values of several adjacent light source points within the set neighborhood range to obtain several temperature absolute differences. Among them, the absolute difference is the absolute value of the difference, and its quantity corresponds to the number of adjacent light source points. Then, calculate the mean value of several temperature absolute differences to obtain the local temperature-time gradient value of the light source point. Then, perform mean analysis on the local temperature-time gradient values of several light source points to obtain the temperature-time gradient value; Perform comprehensive analysis (that is, mean analysis) on the local temperature-time gradient values of each light source point in the temperature images of the LED light source at several time points when several test currents are applied to obtain the temperature-time gradient values of the LED light source at several time points when several test currents are applied.

[0058] In this implementation scheme, through detailed analysis steps, using temperature image data and local temperature-time gradient values, an accurate thermal effect evaluation method is provided for LED light source defect detection. First, through the temperature mean analysis of each light source point, the temperature stability of the LED light source under different current conditions can be identified; Second, through the analysis of the temperature value differences of adjacent light source points within the neighborhood range, the influence of local temperature fluctuations and temperature gradients on the performance of the LED light source is further revealed. Finally, this analysis method based on time-temperature means and local temperature-time gradient values can effectively capture the possible local overheating problems of the LED light source and provide a reliable basis for subsequent defect evaluation. Furthermore, it can comprehensively and accurately analyze the thermal effects of the LED light source under different current and time conditions, providing a more refined evaluation result than traditional methods. This method significantly improves the accuracy of thermal effect detection, makes the defect identification of the LED light source more comprehensive, and can play a positive role in the performance optimization and stability guarantee of the LED light source in practical applications.

[0059] Please refer to Figure 4, an embodiment of the present invention provides a technical solution: an LED light source defect detection system, including: an image data acquisition unit, configured to acquire the light intensity time-series image data and temperature time-series image data of the LED light source under several kinds of test currents, and perform preprocessing respectively; an image feature analysis unit, configured to perform feature analysis on the preprocessed light intensity time-series image data and temperature time-series image data of the LED light source under several kinds of test currents respectively, to obtain a defect index set of the LED light source, including a light intensity difference defect index, a current fluctuation defect index, and a local thermal effect defect index; a comprehensive defect evaluation unit, configured to perform comprehensive analysis on the defect index set of the LED light source to obtain a comprehensive defect evaluation index of the LED light source; a defect judgment unit, configured to perform judgment analysis on the comprehensive defect evaluation index of the LED light source and a preset defect evaluation interval, and if the comprehensive defect evaluation index of the LED light source is within the preset defect evaluation interval, it is regarded that the LED light source has a defect.

[0060] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they know the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present invention.

[0061] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims

1. A method for detecting defects of an LED light source, characterized in that, Including the following steps: Obtain the light intensity time - series image data and temperature time - series image data of the LED light source under several kinds of test currents, and perform pre - processing respectively; Perform feature analysis on the pre - processed light intensity time - series image data and temperature time - series image data of the LED light source under several kinds of test currents respectively, to obtain the defect index set of the LED light source, including the light intensity difference defect index, current fluctuation defect index, and local thermal effect defect index; Perform comprehensive analysis on the defect index set of the LED light source to obtain the comprehensive defect evaluation index of the LED light source. The calculation formula is as follows: Where, ZqX, GqQ, DqQ, and RqQ are the comprehensive defect evaluation index, light intensity difference defect index, current fluctuation defect index, and local thermal effect defect index of the LED light source respectively; λ1, λ2, λ3, and λ4 are the light intensity influence coefficient, current influence coefficient, thermal effect influence coefficient, and light intensity - current interaction influence coefficient stored in the database respectively; μ1 and μ2 are the thermal effect adjustment coefficient and light intensity - current interaction adjustment coefficient stored in the database respectively; Perform judgment analysis on the comprehensive defect evaluation index of the LED light source and the preset defect evaluation interval. If the comprehensive defect evaluation index of the LED light source is within the preset defect evaluation interval, it is considered that the LED light source has defects.

2. The LED light source defect detection method according to claim 1, characterized in that, The light intensity time - series image data includes light intensity images at several time points, and each light intensity image includes pixel values of several light source points. The specific steps to obtain the light intensity difference defect index of the LED light source are as follows: Perform comprehensive analysis on the light intensity images at several time points of the LED light source under several kinds of test currents respectively, to obtain the light intensity difference feature set of the LED light source, including the average pixel values, pixel - time variance values, and time - pixel means at several light source points under several kinds of test currents; Perform comprehensive analysis on the pixel values of several light source points in the light intensity images at several time points of the LED light source under several kinds of test currents respectively, in combination with the light intensity difference feature set of the LED light source, to obtain the light intensity difference defect index of the LED light source.

3. The LED light source defect detection method according to claim 2, characterized in that, The specific steps to obtain the light intensity difference feature set of the LED light source are as follows: Read the pixel values of several light source points in the light intensity images at several time points of the LED light source under several kinds of test currents, and perform time - point analysis respectively, to obtain the average pixel values of several light source points of the LED light source under several kinds of test currents; Read the pixel values of several light source points in the light intensity images at several time points of the LED light source under several kinds of test currents, and perform variance analysis in combination with the average pixel values of several light source points of the LED light source under several kinds of test currents, to obtain the pixel - time variance values of several light source points of the LED light source under several kinds of test currents; Read the pixel values of several light source points in the light intensity images at several time points of the LED light source under several kinds of test currents, and perform light - source - point pixel analysis respectively, to obtain the time - pixel means at several time points of the LED light source under several kinds of test currents.

4. The LED light source defect detection method according to claim 2, wherein The specific formula for calculating the light intensity difference defect index of the LED light source is as follows: Among them, GqQ is the light intensity difference defect index of the LED light source, XsZ i(u+1)r is the pixel value of the r-th light source point in the light intensity image of the LED light source at the (u + 1)-th time point when the i-th test current is applied, XsZ iur is the pixel value of the r-th light source point in the light intensity image of the LED light source at the u-th time point when the i-th test current is applied, PxS ir 、XsF ir are the average pixel value and the pixel time variance value of the r-th light source point of the LED light source when the i-th test current is applied, respectively, SxS iu is the time pixel mean value of the LED light source at the u-th time point when the i-th test current is applied. δ1 and δ2 are the time variance influence coefficient and the time pixel influence coefficient stored in the database, respectively. i = 1, 2, 3, …, i0, where i0 is the number of test current types, u = 1, 2, 3, …, u0, where u0 is the number of time points, and r = 1, 2, 3, …, r0, where r0 is the number of light source points.

5. The LED light source defect detection method according to claim 2, wherein, The specific steps for obtaining the current fluctuation defect index of the LED light source are as follows: Obtain the test current values at several time points for each test current, and conduct comprehensive analysis to obtain several groups of test current difference values for each test current; Conduct comprehensive analysis on several groups of test current difference values for each test current in combination with the pixel values of several light source points in the light intensity images at several time points when the LED light source is applied with several test currents, to obtain the current fluctuation defect index of the LED light source.

6. The LED light source defect detection method according to claim 5, characterized in that, The specific formula for calculating the current fluctuation defect index of the LED light source is as follows: Among them, DqQ is the current fluctuation defect index of the LED light source, XsZ i(u+1)r is the pixel value of the r-th light source point in the light intensity image of the LED light source at the (u + 1)-th time point when the i-th test current is applied, XsZ iur is the pixel value of the r-th light source point in the light intensity image of the LED light source at the u-th time point when the i-th test current is applied, CfD ia is the a-th group of test current difference values of the i-th test current, ε, χ, β are the pixel adjustment factor, current adjustment factor, and differential current adjustment coefficient stored in the database in sequence, i = 1, 2, 3, …, i0, i0 is the number of test current types, u = 1, 2, 3, …, u0, u0 is the number of time points, r = 1, 2, 3, …, r0, r0 is the number of light source points, a = 1, 2, 3, …, a0, a0 is the number of groups of test current difference values.

7. The LED light source defect detection method according to claim 1, wherein The temperature time-series image data includes temperature images at several time points, and each temperature image includes temperature values of several light source points. The specific steps for obtaining the local thermal effect defect index of the LED light source are as follows: Conduct feature analysis on the temperature time-series image data of the LED light source when applied with several test currents respectively, to obtain the temperature difference feature set of the LED light source, including the time-temperature mean value and the temperature-time gradient value at several time points when applied with several test currents; Conduct comprehensive analysis on the temperature values of several light source points in the temperature images at several time points when the LED light source is applied with several test currents respectively in combination with the temperature difference feature set of the LED light source, to obtain the local thermal effect defect index of the LED light source.

8. The LED light source defect detection method according to claim 7, wherein The specific steps for obtaining the temperature difference feature set of the LED light source are as follows: Read the temperature values of several light source points in the temperature images at several time points when the LED light source is applied with several test currents, and conduct light source point temperature analysis respectively to obtain the time-temperature mean value at several time points when the LED light source is applied with several test currents; For each light source point in the temperature images at several time points when the LED light source is applied with several test currents, identify several adjacent light source points within the set neighborhood range respectively; Conduct comprehensive analysis on the temperature value of each light source point in the temperature images at several time points when the LED light source is applied with several test currents respectively with the temperature values of several adjacent light source points within the set neighborhood range of the corresponding light source point, to obtain the temperature-time gradient value at several time points when the LED light source is applied with several test currents; Conduct comprehensive analysis on the local temperature-time gradient values of each light source point in the temperature images at several time points when the LED light source is applied with several test currents, to obtain the temperature-time gradient value at several time points when the LED light source is applied with several test currents.

9. The LED light source defect detection method according to claim 7, wherein The specific formula for calculating the local thermal effect defect index of the LED light source is as follows: Among them, RqQ is the local thermal effect defect index of the LED light source, WdZ iur is the temperature value of the r-th light source point in the temperature image of the LED light source at the u-th time point when the i-th test current is applied, SwZ iu , SwT iu are successively the time-temperature mean value and the temperature-time gradient value of the LED light source at the u-th time point when the i-th test current is applied. θ, ω, and ξ are successively the temperature adjustment factor, the temperature gradient adjustment coefficient, and the temperature gradient influence coefficient stored in the database. i = 1, 2, 3, …, i0, where i0 is the number of types of test currents, u = 1, 2, 3, …, u0, where u0 is the number of time points, and r = 1, 2, 3, …, r0, where r0 is the number of light source points.

10. An LED light source defect detection system, which applies the LED light source defect detection method described in any one of claims 1-9, and is characterized in that, Including: An image data acquisition unit, configured to acquire the light intensity time-series image data and the temperature time-series image data of the LED light source when applied with several test currents, and conduct preprocessing respectively; An image feature analysis unit is used to perform feature analysis on the light intensity time-series image data and temperature time-series image data of the preprocessed LED light source under several kinds of test currents respectively, so as to obtain a defect index set of the LED light source, including a light intensity difference defect index, a current fluctuation defect index, and a local thermal effect defect index; A comprehensive defect evaluation unit is used to comprehensively analyze the defect index set of the LED light source to obtain a comprehensive defect evaluation index of the LED light source; A defect judgment unit is used to perform judgment analysis on the comprehensive defect evaluation index of the LED light source and a preset defect evaluation interval. If the comprehensive defect evaluation index of the LED light source is within the preset defect evaluation interval, it is regarded that the LED light source has a defect.

Images