LED module light color visual detection method and system based on multispectrum

Abstract

The invention discloses an LED module light color visual detection method and system based on multispectrum, and the method comprises the steps: initializing to a preset detection gear, the spectral range of the preset detection gear comprises a first buffer region of a wavelength lower limit neighborhood and a second buffer region of an upper limit neighborhood, and the upper limit of the first buffer region is smaller than the lower limit of the second buffer region; acquiring spectral characteristic parameters of the LED module; if the main wavelength is the main wavelength, the relation between the main wavelength and the upper limit and the lower limit of the current gear is judged, if the condition is met, a low-level gear or a high-level gear is switched to, and a corresponding buffer area threshold value is verified after switching; if the statistical value is the wavelength distribution statistical value, the relationship between the statistical value and a preset threshold value is compared, and if the condition is met, corresponding gear switching is carried out; and finally, switching the detection to the decided target gear, and synchronously adjusting the integral time and the gain parameter. According to the invention, the adaptive capability and stability of the detection process are improved, and misjudgment and measurement fluctuation during multi-model switching are reduced.

Description

Technical Field

[0001] This application relates to the field of light and color detection technology, and in particular to a method and system for visual detection of light and color in LED modules based on multispectral light. Background Technology

[0002] In the rapidly developing automotive electronics manufacturing sector, especially in the high-precision production and quality control of automotive LED lighting modules, enterprises and manufacturers face the challenge of producing a wide variety of products in small batches with high consistency requirements. Automotive lighting modules need to possess high optical consistency, environmental adaptability, and long lifespan characteristics, encompassing various products including exterior lighting and interior functions. The production of these products requires rigorous calibration of light and color parameters, functional testing, and reliability verification, placing higher demands on the accuracy, efficiency, and adaptability of testing equipment. The main problem with existing technologies is the insufficient adaptability and stability of the spectral detection process, leading to misjudgments of light levels and measurement fluctuations when rapidly switching between testing multiple LED module models. Summary of the Invention

[0003] The main technical problem addressed in this application is to provide a multispectral-based method and system for visual inspection of the light and color of LED modules, which improves the adaptability and stability of the inspection process and reduces misjudgment of gear positions and measurement fluctuations when switching between multiple models.

[0004] To address the aforementioned technical problems, this application provides a technical solution: a multispectral-based LED module light and color visual detection method. This method includes: initializing the detection process to a preset detection level, wherein the spectral range of the preset detection level includes a first buffer region in the neighborhood of the lower wavelength limit and a second buffer region in the neighborhood of the upper wavelength limit, the upper wavelength limit of the first buffer region being less than the lower wavelength limit of the second buffer region; acquiring spectral feature parameters of the target LED module, the spectral feature parameters being selected from any of the following: a measured value of the dominant wavelength, or a statistical value of wavelength distribution; and performing a level decision operation based on the spectral feature parameters: when the spectral feature parameter is a measured value of the dominant wavelength, determining its first magnitude relationship with the lower wavelength limit of the preset detection level. The first value relationship with the upper limit of wavelength is considered; if the first value relationship meets the first switching condition, the system switches to a lower detection level and verifies the threshold of the first buffer region after switching; if the second value relationship meets the second switching condition, the system switches to a higher detection level and verifies the threshold of the second buffer region after switching; when the spectral characteristic parameter is a wavelength distribution statistical value, its first numerical relationship with the first preset threshold and its second numerical relationship with the second preset threshold are compared; if the first numerical relationship meets the third switching condition, the system switches to a lower detection level; if the second numerical relationship meets the fourth switching condition, the system switches to a higher detection level; the detection process is switched to the determined target detection level, and the integration time parameter and gain parameter are adjusted synchronously.

[0005] The preset detection level is the narrowband spectral level. The level decision operation based on spectral feature parameters includes: when the measured value of the main wavelength is lower than the lower limit of the narrowband spectral level wavelength, switching to the full spectrum level; when the measured value of the main wavelength is higher than the upper limit of the narrowband spectral level wavelength, switching to the mixed light level; after switching to the full spectrum level, if the proportion of pixels with wavelength distribution statistics higher than the upper limit of the full spectrum level wavelength meets the first back-cut condition, then switching back to the narrowband spectral level; the first buffer area is set as the first offset range of the lower limit of the narrowband spectral level wavelength, and the second buffer area is set as the second offset range of the upper limit of the narrowband spectral level wavelength.

[0006] The narrowband spectral mode also includes a blue light safe mode. The spectral range of the blue light safe mode is included within the range of the narrowband spectral mode. The spectral characteristic parameters perform the mode decision operation as follows: when the measured value of the main wavelength is within the wavelength range of the blue light safe mode, switch to the blue light safe mode; in the blue light safe mode, set the gain parameter to the first gain mode and the integration time parameter to the first integration mode; if the proportion of pixels in the wavelength distribution statistics that exceed the wavelength range of the blue light safe mode meets the exit condition, switch to the narrowband spectral mode and start the spectral calibration process.

[0007] The adjustment of the integration time and gain parameters includes: in the narrowband spectral setting, setting the integration time to the first duration value and the gain parameter to the first gain value; in the full-spectrum setting, setting the integration time to the second duration value and the gain parameter to the second gain value; in the mixed light setting, setting the integration time to the third duration value and the gain parameter to the third gain value; wherein the first gain value is greater than the second gain value, the second gain value is greater than the third gain value, the first duration value is greater than the second duration value, and the second duration value is greater than the third duration value.

[0008] The width of the first buffer region is defined as the difference between the lower limit of the preset detection wavelength and a first boundary value.

[0009] The width of the second buffer region is defined as the difference between the upper limit of the preset detection wavelength and a second boundary value. The setting of the first boundary value and the second boundary value ensures that the upper limit of the wavelength of the first buffer region is less than the lower limit of the wavelength of the second buffer region.

[0010] The verification of the first buffer region threshold includes: if, in a lower detection level, the measured value of the main wavelength is higher than the boundary value of the first buffer region in multiple consecutive samples, then switch back to the preset detection level. The verification of the second buffer region threshold includes: if, in a higher detection level, the percentage of pixels whose wavelength distribution statistics are lower than the boundary value of the second buffer region continuously exceeds a time threshold, then switch back to the preset detection level.

[0011] The first numerical relationship is the comparison between the percentage of pixels in the wavelength distribution statistics that are below the lower limit of the current wavelength range and the first preset threshold; the third switching condition is that the percentage of pixels is greater than the first preset threshold; after switching to a lower detection range, if the percentage of pixels in the wavelength distribution statistics that are above the upper limit of the wavelength range is greater than the second preset threshold, then switch back to the preset detection range.

[0012] To address the aforementioned technical problems, another technical solution adopted in this application is: providing a multispectral LED module light and color vision inspection system. This system includes an initialization module, an acquisition module, a decision module, and an execution module. The initialization module initializes the detection process to a preset detection level, wherein the spectral range of the preset detection level includes a first buffer region in the neighborhood of the lower wavelength limit and a second buffer region in the neighborhood of the upper wavelength limit, with the upper wavelength limit of the first buffer region being less than the lower wavelength limit of the second buffer region. The acquisition module acquires the spectral characteristic parameters of the target LED module, which are selected from any of the following: a measured value of the dominant wavelength or a statistical value of wavelength distribution. The decision module is configured to determine whether the spectral characteristic parameter matches the preset detection level when the spectral characteristic parameter is a measured value of the dominant wavelength. The system considers the following parameters: a first relationship between the lower wavelength limit and a second relationship between the upper wavelength limit and the lower wavelength limit; if the first relationship satisfies the first switching condition, it switches to a lower detection level and verifies the threshold of the first buffer region after switching; if the second relationship satisfies the second switching condition, it switches to a higher detection level and verifies the threshold of the second buffer region after switching; when the spectral characteristic parameter is a wavelength distribution statistical value, it compares its first numerical relationship with a first preset threshold and its second numerical relationship with a second preset threshold; if the first numerical relationship satisfies the third switching condition, it switches to a lower detection level; if the second numerical relationship satisfies the fourth switching condition, it switches to a higher detection level; and an execution module switches the detection process to the determined target detection level and simultaneously adjusts the integration time parameter and gain parameter.

[0013] Unlike existing technologies, the advantages of this application are as follows: By initializing the detection process to a preset detection level and setting buffer regions at both ends of its spectral range, the fault tolerance range of the level coverage is effectively expanded; by acquiring the spectral characteristic parameters of the target LED module, i.e., the measured value of the dominant wavelength or the statistical value of the wavelength distribution, and based on the magnitude or numerical relationship between this parameter and a preset threshold, the level switching judgment is adaptively executed, accurately jumping to the adjacent level when the switching conditions are met, and simultaneously verifying the buffer region threshold; after the switching is completed, the integration time and gain parameters are further adjusted collaboratively to achieve rapid response and system matching in the detection process. Thus, by introducing a strategy combining buffer regions and multi-dimensional spectral criteria, the adaptive recognition capability and level stability of the spectral detection system for multiple varieties and small batches of LED modules are significantly enhanced, reducing measurement fluctuations and misjudgments caused by model switching, improving the accuracy of light and color parameter calibration and detection efficiency, thereby better meeting the quality control requirements for high-consistency production in automotive electronics manufacturing. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating one embodiment of the multispectral LED module light color visual detection method of this application.

[0015] Figure 2 This is a schematic diagram of one embodiment of the LED module light color vision inspection system based on multispectral light. Detailed Implementation

[0016] To make the purpose, technical solution and effects of this application clearer and more explicit, the following describes this application in further detail with reference to the accompanying drawings and embodiments.

[0017] The terms "first" and "second" used herein are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. Furthermore, the various embodiments described herein are not necessarily mutually exclusive, as some embodiments may be combined with one or more other embodiments to form new embodiments.

[0018] In some related technologies, spectral detection equipment suffers from unstable detection results during the rapid switching detection of multiple LED modules. To address this, this application provides the following technical solution.

[0019] See Figure 1 , Figure 1 This is a flowchart illustrating one embodiment of the multispectral LED module light color visual detection method of this application. The method includes: Step S11: Initialize the detection process to the preset detection level, wherein the spectral range of the preset detection level includes a first buffer region in the neighborhood of the lower wavelength limit and a second buffer region in the neighborhood of the upper wavelength limit, and the upper wavelength limit of the first buffer region is less than the lower wavelength limit of the second buffer region.

[0020] Step S12: Obtain the spectral characteristic parameters of the target LED module. The spectral characteristic parameters are selected from any of the following: the measured value of the dominant wavelength or the statistical value of the wavelength distribution.

[0021] Step S13: Perform a gear selection decision based on spectral feature parameters: When the spectral feature parameter is the main wavelength measurement value, determine its first magnitude relationship with the preset detection gear wavelength lower limit and its second magnitude relationship with the wavelength upper limit; if the first magnitude relationship meets the first switching condition, switch to a lower detection gear and verify the first buffer area threshold after switching; if the second magnitude relationship meets the second switching condition, switch to a higher detection gear and verify the second buffer area threshold after switching; when the spectral feature parameter is a wavelength distribution statistical value, compare its first numerical relationship with the first preset threshold and its second numerical relationship with the second preset threshold; if the first numerical relationship meets the third switching condition, switch to a lower detection gear; if the second numerical relationship meets the fourth switching condition, switch to a higher detection gear.

[0022] Step S14: Switch the detection process to the determined target detection level and simultaneously adjust the integration time parameter and gain parameter.

[0023] Specifically, the detection process may include a series of preparatory actions such as starting the spectral imaging equipment, self-testing, parameter loading, and triggering the enable signal. The preset detection level can be the system's default starting working state, and its selection can be determined based on the optimal detection range corresponding to the LED module type that appears most frequently in historical detection data. The lower wavelength limit neighborhood and the upper wavelength limit neighborhood can refer to a nearby wavelength interval surrounding the lower and upper wavelength values ​​of the preset detection level. The first buffer region and the second buffer region can be specific wavelength intervals located below the lower limit and above the upper limit of the spectral range of the preset detection level, respectively. The setting of these two regions can prevent frequent level switching caused by small fluctuations in the detection value near the level boundary. The upper wavelength limit of the first buffer region is smaller than the lower wavelength limit of the second buffer region; this feature ensures that the two buffer regions do not overlap spectrally, thus avoiding confusion in logical judgments. Spectral characteristic parameters can be obtained using a spectrophotometer, a spectral camera, or a multi-channel filter combined with an image sensor. The dominant wavelength measurement value can be a representative value obtained by calculating the centroid of the spectral power distribution or finding its peak point. Wavelength distribution statistics can be the number of pixels falling within a specific wavelength range or the percentage of pixels in the total number of pixels on the acquired spectral image or data. Integration time parameters control the exposure time of the image sensor or spectral detector to the light signal. Gain parameters adjust the amplification factor of the signal amplifier, thus adjusting the sensitivity of the detection system. In some applications, this is used in online inspection lines for automotive taillight LED modules. The preset detection level can be set to a narrowband detection level for common red LEDs. When a production line switches batches and encounters a red module with a dominant wavelength within the first buffer region, the system may still maintain the preset detection level. When switching to an amber turn signal module with a significantly shorter dominant wavelength, if the first switching condition is met, the system automatically switches to a lower detection level with a shorter wavelength range, such as the level for amber light, and adjusts the integration and gain to the optimal settings for that level to ensure measurement accuracy.

[0024] The above scheme effectively expands the tolerance range of the detection range by initializing the detection process to a preset detection level and setting buffer regions at both ends of its spectral range. It acquires the spectral characteristic parameters of the target LED module, i.e., the measured value of the dominant wavelength or the statistical value of the wavelength distribution, and adaptively executes level switching judgment based on the magnitude or numerical relationship between these parameters and preset thresholds. When the switching conditions are met, it accurately jumps to the adjacent level and simultaneously verifies the buffer region threshold. After the switching is completed, it further coordinates and adjusts the integration time and gain parameters to achieve rapid response and system matching in the detection process. Thus, by introducing a strategy combining buffer regions and multi-dimensional spectral criteria, the spectral detection system significantly enhances its adaptive recognition capability and level stability for multiple varieties and small batches of LED modules, reduces measurement fluctuations and misjudgments caused by model switching, and improves the accuracy of light and color parameter calibration and detection efficiency, thereby better meeting the quality control requirements for high-consistency production in automotive electronics manufacturing.

[0025] In some embodiments, the preset detection level is a narrowband spectral level, and the level decision operation based on spectral feature parameters includes: when the measured value of the main wavelength is lower than the lower limit of the wavelength of the narrowband spectral level, switching to the full spectrum level; when the measured value of the main wavelength is higher than the upper limit of the wavelength of the narrowband spectral level, switching to the mixed light level; after switching to the full spectrum level, if the proportion of pixels in the wavelength distribution statistics that are higher than the upper limit of the wavelength of the full spectrum level meets the first back-cutting condition, then switching back to the narrowband spectral level; the first buffer region is set as a first offset range of the lower limit of the wavelength of the narrowband spectral level, and the second buffer region is set as a second offset range of the upper limit of the wavelength of the narrowband spectral level.

[0026] Specifically, the narrowband spectral mode can refer to a working mode of a spectral detection device, where the effective detection wavelength range is typically set narrower, perhaps only covering a band width of tens of nanometers, to accommodate high-precision, high-sensitivity measurement of the color parameters of specific monochromatic LEDs, such as pure red, pure green, and pure blue LEDs. The full-spectrum mode can refer to another working mode, where the effective detection wavelength range is typically significantly wider than the narrowband spectral mode, such as covering the entire visible light band or even wider, used for measuring LED modules with broad-spectrum characteristics or for initial capture when the dominant wavelength of the LED under test exceeds the narrowband range. The mixed-spectrum mode can refer to the working mode set for LED modules with multiple distinct peaks or complex spectral components in their emission spectrum, such as certain white LEDs or multi-chip integrated modules; its detection strategy or algorithm may differ from the mode for processing single-peak spectra. The first switchback condition can refer to a judgment logic used to determine whether to switch back from the full-spectrum mode to the narrowband spectral mode. This condition can be based on statistics. For example, after switching to the full-spectrum mode, if the percentage of pixels with wavelengths higher than the upper limit of the full-spectrum mode, calculated from several consecutive samples, is lower than a preset percentage threshold, such as 5%, then a back-cut action is triggered. The first offset range and the second offset range can be tolerance intervals set for the lower and upper limits of the narrowband spectrum mode's wavelength, respectively. For example, the first offset range can be the interval [λ_lower - 10nm, λ_lower] formed by shifting downwards by 10nm, and the second offset range can be the interval [λ_upper, λ_upper + 10nm] formed by shifting upwards by 10nm. These two intervals constitute a part of the first buffer region and the second buffer region, respectively. The specific value of the offset can be adjusted according to the general fluctuation range of the LED wavelength and the required stabilization sensitivity.

[0027] In some specific embodiments, a specific narrowband spectral range can be set to detect wavelengths in the range of 620nm to 645nm, specifically for detecting a particular type of red LED.

[0028] Switching to full-spectrum mode can be triggered when the measured dominant wavelength is below 620nm, such as when an amber LED with a dominant peak at 610nm is detected. Switching to mixed-light mode can be triggered when the measured dominant wavelength is above 645nm, such as when a deep red LED with a dominant peak at 660nm is detected, or when a white LED with dual peaks at 460nm and 555nm is detected. The first back-cut condition can be specifically set as follows: monitoring begins after switching to full-spectrum mode. If the proportion of pixels with wavelengths above 700nm in the wavelength statistics is less than 2% for three consecutive samples, the first back-cut condition is met, and the system switches the detection mode back to the original narrowband spectrum mode. The first offset range can be specifically set to [610nm, 620nm]. The second offset range can be specifically set to [645nm, 655nm].

[0029] In some applications, this technology is used for online sorting of LED modules for automotive taillights. The production line's default preset detection range is a narrowband spectrum of 620-645nm for standard red LEDs. When a batch of incoming materials contains individual LEDs with shorter wavelengths, such as a main wavelength of 615nm, falling into the first buffer zone [610nm, 620nm], the system can maintain stable measurement at the current range. When the equipment detects a clearly amber-colored faulty or mixed LED with a main wavelength of 595nm, below the narrowband range's lower limit of 620nm, it immediately switches to the full spectrum range for capture and measurement, and marks the abnormal unit. In subsequent inspections, if several consecutive normal red LEDs do not trigger any further abnormalities, the system meets the first cut-off condition and automatically switches back to the narrowband spectrum range to maintain efficient and high-precision detection of mainstream products.

[0030] In some embodiments, the narrowband spectral setting also includes a blue light safe setting, the spectral range of which is included within the narrowband spectral setting range. The spectral characteristic parameters perform the setting decision operation as follows: when the measured value of the dominant wavelength is within the wavelength range of the blue light safe setting, switch to the blue light safe setting; in the blue light safe setting, set the gain parameter to a first gain mode and the integration time parameter to a first integration mode; if the proportion of pixels in the wavelength distribution statistics that exceed the wavelength range of the blue light safe setting meets the exit condition, switch to the narrowband spectral setting and start the spectral calibration process.

[0031] Specifically, the blue light safety setting can be a sub-setting with a more finely divided wavelength range nested within a broader narrow-band spectral setting. Its spectral range can be set according to the wavelength weighting function for photobiological safety in international standards, with particular attention to wavelengths with higher weighting for retinal blue light hazard, such as specific ranges like 435nm-445nm or 460nm-480nm, for high-precision assessment and control of potential blue light hazard risks from LED modules. The first gain mode and first integration mode can refer to a specific set of parameter configurations. Considering the extremely high requirements for measurement accuracy and stability in blue light safety assessment, this mode typically means setting a relatively low gain value to reduce signal noise and may set a relatively long integration time to ensure a sufficiently strong effective signal is obtained even at low gain, thereby guaranteeing the signal-to-noise ratio and accuracy of the data at this critical setting. The exit condition can be a threshold value used to trigger exit from the high-precision blue light safety detection mode. This condition can be based on statistics. For example, during continuous monitoring in the blue light safe setting, if the percentage of pixels with wavelengths exceeding the strictly defined range of the blue light safe setting exceeds a very small preset tolerance threshold, then the spectral purity of the module under test may not meet the prerequisite for accurate safety assessment in this setting. Initiating the spectral calibration process can be an action performed after exiting the blue light safe setting. This process may include recalibrating the device's spectral responsivity using a built-in standard reference light source to ensure the accuracy of subsequent measurements in the narrowband spectral setting.

[0032] In some applications, this technology is used for the safety inspection of LED indicator lights in children's wearable devices. The default production line process initially enters the narrowband spectrum range (430nm-490nm) when inspecting blue indicator lights. When a module's main wavelength measurement is detected at 460nm, falling within the blue light safety range, the system automatically switches to the blue light safety range and uses the first gain mode and first integration mode for high-precision measurement to accurately calculate whether its blue light hazard weighted irradiance is below the safety limit. If the inspection reveals significant stray light of other wavelengths in the spectrum due to manufacturing defects—for example, if the wavelength distribution statistics show that the proportion exceeding the 455-465nm range reaches 1.5%, meeting the exit criteria—the system determines that this module is unsuitable for measurement at this high-precision range. It automatically switches to the narrowband spectrum range and initiates calibration. After confirming the equipment is functioning correctly, other parameters can be tested at a wider range, such as the narrowband spectrum range, and the module is ultimately deemed unqualified.

[0033] In some embodiments, adjusting the integration time parameter and the gain parameter includes: in the narrowband spectral setting, setting the integration time to a first duration value and the gain parameter to a first gain value; in the full-spectrum setting, setting the integration time to a second duration value and the gain parameter to a second gain value; in the mixed light setting, setting the integration time to a third duration value and the gain parameter to a third gain value; wherein the first gain value is greater than the second gain value, the second gain value is greater than the third gain value, the first duration value is greater than the second duration value, and the second duration value is greater than the third duration value.

[0034] Specifically, this parameter adjustment strategy is based on the optical characteristics and signal strength requirements of different detection ranges. Narrowband spectral ranges typically target monochromatic light of specific wavelengths, with relatively low light throughput. Therefore, longer integration times are needed, employing a first duration value to accumulate sufficient photogenerated charge. Higher gain may also be required, using a first gain value to amplify the relatively weak electrical signal, ensuring measurement sensitivity and signal-to-noise ratio. Full-spectrum ranges cover a wider wavelength range, resulting in a larger total light throughput into the detector. Therefore, sufficient signal strength can be obtained without excessively long integration times. Setting a second duration value helps improve detection speed and prevents saturation of strong light signals. Simultaneously, due to the strong signal background, the gain can be reduced, employing a second gain value to minimize noise introduced by the amplification circuit, prioritizing the signal's dynamic range and linearity. Mixed-spectrum ranges may deal with multiple peaks with significant intensity differences, exhibiting the widest signal dynamic range. Setting a third duration value primarily aims to prevent oversaturation of the strongest wavelength band. Meanwhile, a third gain value is used to prioritize preventing clipping distortion of strong signal components during signal amplification, thus preserving complete spectral details for analysis, even if this may sacrifice sensitivity to weak signal components. In some applications, it's used in the spectral performance testing station of an LED chip sorting machine. When detecting a weak, deep red LED chip, the system, in narrowband spectral mode, using a longer integration time of 150ms and a higher gain of 80, successfully captured its weak spectral signal and accurately measured its dominant wavelength as 650nm. When the production line switches to a high-brightness white LED module, whose spectrum covers the entire visible light range, the system automatically switches to full-spectrum mode and adjusts the parameters to a shorter integration time of 50ms and a medium gain of 40, quickly completing the full-spectrum scan without signal saturation. When encountering an RGB three-in-one LED, the light intensity of its red, green, and blue chips differs significantly. The system activates the mixed light mode, using a very short integration time of 20ms and a low gain of 20 to ensure that the strong blue light signal is not overexposed, while still capturing the spectral profiles of the relatively weak red and green light signals, thereby accurately analyzing its light mixing effect.

[0035] In some specific embodiments, the width of the first buffer region is defined as the difference between a preset detection range wavelength lower limit and a first boundary value. The width of the second buffer region is defined as the difference between a preset detection range wavelength upper limit and a second boundary value. The setting of the first and second boundary values ​​ensures that the wavelength upper limit of the first buffer region is less than the wavelength lower limit of the second buffer region.

[0036] Specifically, this definition provides a quantifiable and configurable geometric description of the buffer region. It transforms the buffer region from a vague "neighborhood" concept into a range with clear mathematical boundaries, facilitating precise logical judgments and parameter settings in software algorithms. The width of the first buffer region: its physical meaning is the extent to which the lower limit of the preset detection range wavelength extends downwards. This width value is the lower limit of the preset detection range wavelength – the first boundary value. The magnitude of this width directly determines the system's fault tolerance and response sensitivity to spectral signals below the preset range. A larger width results in a wider fault tolerance range, but the response switching may be slightly sluggish; a smaller width makes the system more sensitive to wavelength down-biasing, but also more susceptible to noise interference and jitter. The width of the second buffer region: its physical meaning is the extent to which the upper limit of the preset detection range wavelength extends upwards. This width value is the second boundary value – the upper limit of the preset detection range wavelength. This width value controls the system's sensitivity to spectral signals above the preset range. The constraint setting the first and second boundary values: This constraint, namely, the upper wavelength limit of the first buffer region < the lower wavelength limit of the second buffer region, is used to ensure that the two buffer regions are completely separated on the spectral axis and do not overlap. The upper limit of the first buffer region is the lower wavelength limit of the preset detection range itself, and the lower limit of the second buffer region is the upper wavelength limit of the preset detection range itself. Therefore, this constraint is equivalent to requiring the lower wavelength limit of the preset detection range to be < the upper wavelength limit of the preset detection range, which is itself a basic premise for defining a detection range. A more relevant constraint might be ensuring that the two buffer regions themselves do not intersect, but this is usually naturally satisfied by their respective characteristics of being located outside the upper and lower limits of the range. The wording here may be intended to emphasize that the settings of the two buffer regions are independent and do not interfere with each other.

[0037] In some embodiments, verifying the first buffer region threshold includes: in a lower detection level, if the measured value of the main wavelength is higher than the boundary value of the first buffer region in multiple consecutive samplings, then switching back to a preset detection level. Verifying the second buffer region threshold includes: in a higher detection level, if the percentage of pixels whose wavelength distribution statistics are lower than the boundary value of the second buffer region continuously exceeds a time threshold, then switching back to the preset detection level.

[0038] Specifically, this verification mechanism, as a debouncing measure, is used to confirm the correctness of the initial switching decision and prevent frequent oscillations between two adjacent levels due to normal minor fluctuations in measured values. For the verification of the first buffer region: the logic is that if the system switches to a lower level due to a lower dominant wavelength value, then subsequent continuous measurements of the dominant wavelength should stabilize within the typical range of that lower level. If they instead repeatedly exceed a boundary, namely the boundary value of the first buffer region, which is usually set near the lower limit of the preset level, this indicates that the spectral characteristics of the target LED may be closer to the preset level. The previous switching may have been a misjudgment, or the characteristics of the target itself may have changed, such as wavelength drift after warm-up, thus triggering a re-switching. For the verification of the second buffer region, switching back to the preset level from a higher level: its logic is symmetrical. After the system switches to a higher level due to higher spectral components, it monitors the percentage of pixels in the signal below a certain boundary, namely the boundary value of the second buffer region, which is usually set near the upper limit of the preset level. If this high percentage persists for a period of time, rather than just in a single sampling, it indicates that the main spectral components of the target LED are still concentrated within the preset range. The previously observed high-wavelength signal may be noise or transient interference. Therefore, switching back to the preset range after the condition is met is more stable. The introduction of a time threshold enhances the anti-interference capability of the judgment.

[0039] In some applications, this technology is used in LED aging test lines. During the initial heating phase, the wavelength of a red LED may briefly drift to 615nm due to temperature effects (triggering a switch to the "amber" setting). After the system switches, during the verification of the first buffer region threshold, as the chip temperature stabilizes, its dominant wavelength quickly stabilizes at 618nm. Several consecutive measurements (618nm, 617nm, 619nm) are all higher than the set boundary value of 615nm. Therefore, the system determines that the initial switch was due to a transient process and switches back to the "red" setting for subsequent long-term stabilization testing, effectively avoiding unnecessary oscillations in the setting during the early stages of aging.

[0040] The above scheme effectively expands the tolerance range of the detection range by initializing the detection process to a preset detection level and setting buffer regions at both ends of its spectral range. It acquires the spectral characteristic parameters of the target LED module, i.e., the measured value of the dominant wavelength or the statistical value of the wavelength distribution, and adaptively executes level switching judgment based on the magnitude or numerical relationship between these parameters and preset thresholds. When the switching conditions are met, it accurately jumps to the adjacent level and simultaneously verifies the buffer region threshold. After the switching is completed, it further coordinates and adjusts the integration time and gain parameters to achieve rapid response and system matching in the detection process. Thus, by introducing a strategy combining buffer regions and multi-dimensional spectral criteria, the spectral detection system significantly enhances its adaptive recognition capability and level stability for multiple varieties and small batches of LED modules, reduces measurement fluctuations and misjudgments caused by model switching, and improves the accuracy of light and color parameter calibration and detection efficiency, thereby better meeting the quality control requirements for high-consistency production in automotive electronics manufacturing.

[0041] Please see Figure 2 , Figure 2 This is a schematic diagram of one embodiment of the LED module light color vision inspection system based on multispectral light.

[0042] The multispectral LED module color vision inspection system 20 includes: an initialization module 21, an acquisition module 22, a decision module 23, and an execution module 24. The initialization module 21 initializes the inspection process to a preset inspection level, where the spectral range of the preset inspection level includes a first buffer region in the neighborhood of the lower wavelength limit and a second buffer region in the neighborhood of the upper wavelength limit, with the upper wavelength limit of the first buffer region being less than the lower wavelength limit of the second buffer region. The acquisition module 22 acquires the spectral characteristic parameters of the target LED module, which are selected from any of the following: a measured value of the dominant wavelength or a statistical value of wavelength distribution. The decision module 23 is configured to, when the spectral characteristic parameter is a measured value of the dominant wavelength, determine whether it is significantly different from the lower wavelength limit of the preset inspection level. The system compares the first value with the wavelength upper limit and the second value with the wavelength upper limit. If the first value meets the first switching condition, it switches to a lower detection level and verifies the threshold of the first buffer region after switching. If the second value meets the second switching condition, it switches to a higher detection level and verifies the threshold of the second buffer region after switching. When the spectral characteristic parameter is a wavelength distribution statistical value, it compares its first numerical relationship with the first preset threshold and its second numerical relationship with the second preset threshold. If the first numerical relationship meets the third switching condition, it switches to a lower detection level. If the second numerical relationship meets the fourth switching condition, it switches to a higher detection level. The execution module 24 is used to switch the detection process to the determined target detection level and simultaneously adjust the integration time parameter and gain parameter.

[0043] The preset detection level is the narrowband spectral level. The decision module 23 is further configured to perform level decision operations based on spectral feature parameters, including: when the measured value of the main wavelength is lower than the lower limit of the wavelength of the narrowband spectral level, switch to the full spectrum level; when the measured value of the main wavelength is higher than the upper limit of the wavelength of the narrowband spectral level, switch to the mixed light level; after switching to the full spectrum level, if the proportion of pixels in the wavelength distribution statistics that are higher than the upper limit of the wavelength of the full spectrum level meets the first back-cutting condition, switch back to the narrowband spectral level; the first buffer area is set as the first offset range of the lower limit of the wavelength of the narrowband spectral level, and the second buffer area is set as the second offset range of the upper limit of the wavelength of the narrowband spectral level.

[0044] The narrowband spectral mode also includes a blue light safe mode. The spectral range of the blue light safe mode is included within the range of the narrowband spectral mode. The spectral characteristic parameters perform mode decision-making operations, including: the decision module 23 is further configured to: switch to the blue light safe mode when the main wavelength measurement value is within the wavelength range of the blue light safe mode; in the blue light safe mode, set the gain parameter to the first gain mode and the integration time parameter to the first integration mode; if the proportion of pixels in the wavelength distribution statistics that exceed the wavelength range of the blue light safe mode meets the exit condition, switch to the narrowband spectral mode and start the spectral calibration process.

[0045] The execution module 24 is configured to adjust the integration time parameter and gain parameter as follows: in the narrowband spectral mode, the integration time is set to a first duration value and the gain parameter is set to a first gain value; in the full-spectrum mode, the integration time is set to a second duration value and the gain parameter is set to a second gain value; in the mixed light mode, the integration time is set to a third duration value and the gain parameter is set to a third gain value; wherein the first gain value is greater than the second gain value, the second gain value is greater than the third gain value, the first duration value is greater than the second duration value, and the second duration value is greater than the third duration value.

[0046] The width of the first buffer region defined in the initialization module 21 is defined as the difference between the lower limit of the preset detection wavelength and a first boundary value.

[0047] The width of the second buffer region defined in the initialization module 21 is defined as the difference between the upper limit of the preset detection wavelength and a second boundary value. The setting of the first boundary value and the second boundary value ensures that the upper limit of the wavelength of the first buffer region is less than the lower limit of the wavelength of the second buffer region.

[0048] The decision module 23 is configured to verify the threshold of the first buffer region, including: in a lower detection level, if the measured value of the main wavelength is higher than the boundary value of the first buffer region in multiple consecutive samplings, then switch back to the preset detection level. Verifying the threshold of the second buffer region includes: in a higher detection level, if the proportion of pixels whose wavelength distribution statistics are lower than the boundary value of the second buffer region continuously exceeds a time threshold, then switch back to the preset detection level.

[0049] The first numerical relationship configured in the decision module 23 is the comparison between the proportion of pixels in the wavelength distribution statistics that are lower than the lower limit of the current wavelength and the first preset threshold; the third switching condition is that the proportion of pixels is greater than the first preset threshold; after switching to a lower detection level, if the proportion of pixels in the wavelength distribution statistics that are higher than the upper limit of the wavelength of that level is detected to be greater than the second preset threshold, then switch back to the preset detection level.

[0050] The above scheme effectively expands the tolerance range of the detection range by initializing the detection process to a preset detection level and setting buffer regions at both ends of its spectral range. It acquires the spectral characteristic parameters of the target LED module, i.e., the measured value of the dominant wavelength or the statistical value of the wavelength distribution, and adaptively executes level switching judgment based on the magnitude or numerical relationship between these parameters and preset thresholds. When the switching conditions are met, it accurately jumps to the adjacent level and simultaneously verifies the buffer region threshold. After the switching is completed, it further coordinates and adjusts the integration time and gain parameters to achieve rapid response and system matching in the detection process. Thus, by introducing a strategy combining buffer regions and multi-dimensional spectral criteria, the spectral detection system significantly enhances its adaptive recognition capability and level stability for multiple varieties and small batches of LED modules, reduces measurement fluctuations and misjudgments caused by model switching, and improves the accuracy of light and color parameter calibration and detection efficiency, thereby better meeting the quality control requirements for high-consistency production in automotive electronics manufacturing.

[0051] The above are merely embodiments of this application and do not limit the scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the protection scope of this application.

Claims

1. A method for visual detection of light and color in LED modules based on multispectral light, characterized in that, The method includes: initializing the detection process to a preset detection level, wherein the spectral range of the preset detection level includes a first buffer region in the lower wavelength limit neighborhood and a second buffer region in the upper wavelength limit neighborhood, and the upper wavelength limit of the first buffer region is less than the lower wavelength limit of the second buffer region. Obtain the spectral characteristic parameters of the target LED module, wherein the spectral characteristic parameters are selected from any of the following: the measured value of the dominant wavelength, or the statistical value of the wavelength distribution; Perform gear selection operation based on the spectral characteristic parameters: When the spectral characteristic parameter is the main wavelength measurement value, determine its first relationship with the lower limit of the preset detection range wavelength and its second relationship with the upper limit of the wavelength; If the first size relationship meets the first switching condition, then switch to the lower detection level, and verify the threshold of the first buffer area after switching. If the second size relationship meets the second switching condition, switch to a higher detection level and verify the threshold of the second buffer area after switching. When the spectral feature parameter is a wavelength distribution statistical value, compare its first numerical relationship with a first preset threshold and its second numerical relationship with a second preset threshold; If the first numerical relationship meets the third switching condition, switch to the lower detection level; if the second numerical relationship meets the fourth switching condition, switch to the higher detection level. Switch the detection process to the determined target detection level and simultaneously adjust the integration time parameter and gain parameter.

2. The method according to claim 1, wherein the preset detection level is a narrowband spectral level, and the level decision operation based on the spectral feature parameters includes: When the measured value of the main wavelength is lower than the lower limit of the narrowband spectrum mode, switch to the full spectrum mode; When the measured value of the dominant wavelength is higher than the upper limit of the narrowband spectrum mode, switch to the mixed light mode; After switching to the full spectrum mode, if the percentage of pixels in the wavelength distribution statistics that are higher than the upper limit of the full spectrum mode wavelength meets the first back-cut condition, then switch back to the narrowband spectrum mode; the first buffer area is set as the first offset range of the lower limit of the narrowband spectrum mode wavelength, and the second buffer area is set as the second offset range of the upper limit of the narrowband spectrum mode wavelength.

3. The method according to claim 2, characterized in that, The narrowband spectral setting also includes a blue light safe setting, the spectral range of which is included within the narrowband spectral setting range. The spectral characteristic parameters perform the setting decision operation as follows: when the main wavelength measurement value is within the blue light safe setting wavelength range, switch to the blue light safe setting; in the blue light safe setting, set the gain parameter to the first gain mode and the integration time parameter to the first integration mode; if the percentage of pixels exceeding the blue light safe setting wavelength range in the wavelength distribution statistics meets the exit condition, switch to the narrowband spectral setting and start the spectral calibration process.

4. The method according to claim 1, characterized in that, The adjustment of the integration time parameter and gain parameter includes: in the narrowband spectral setting, setting the integration time to a first duration value and the gain parameter to a first gain value; in the full-spectrum setting, setting the integration time to a second duration value and the gain parameter to a second gain value; in the mixed light setting, setting the integration time to a third duration value and the gain parameter to a third gain value; wherein, the first gain value is greater than the second gain value, the second gain value is greater than the third gain value, the first duration value is greater than the second duration value, and the second duration value is greater than the third duration value.

5. The method according to claim 1 or 4, characterized in that, The width of the first buffer region is defined as the difference between the lower limit of the preset detection wavelength and a first boundary value.

6. The method according to claim 5, characterized in that, The width of the second buffer region is defined as the difference between the upper limit of the preset detection wavelength and a second boundary value. The setting of the first boundary value and the second boundary value ensures that the upper limit of the wavelength of the first buffer region is less than the lower limit of the wavelength of the second buffer region.

7. The method according to claim 5, characterized in that, The verification of the first buffer region threshold includes: if the measured value of the main wavelength is higher than the boundary value of the first buffer region in multiple consecutive samplings at a lower detection level, then switch back to the preset detection level.

8. The method according to claim 6, characterized in that, The verification of the second buffer area threshold includes: if, at a higher detection level, the percentage of pixels whose wavelength distribution statistics are lower than the boundary value of the second buffer area continues to exceed a time threshold, then switch back to the preset detection level.

9. The method according to claim 1, characterized in that, The first numerical relationship is a comparison between the percentage of pixels in the wavelength distribution statistics that are below the lower limit of the current wavelength range and the first preset threshold; the third switching condition is that the percentage of pixels is greater than the first preset threshold; after switching to a lower detection range, if the percentage of pixels in the wavelength distribution statistics that are above the upper limit of the wavelength range is greater than the second preset threshold, then switch back to the preset detection range.

10. A multispectral LED module light and color vision inspection system, characterized in that, include: An initialization module is used to initialize the detection process to a preset detection level, wherein the spectral range of the preset detection level includes a first buffer region in the lower wavelength limit neighborhood and a second buffer region in the upper wavelength limit neighborhood, and the upper wavelength limit of the first buffer region is less than the lower wavelength limit of the second buffer region. The acquisition module is used to acquire the spectral characteristic parameters of the target LED module. The spectral characteristic parameters are selected from any of the following: the measured value of the dominant wavelength, or the statistical value of the wavelength distribution. The decision module is configured to determine the first size relationship between the spectral feature parameter and the lower limit of the preset detection level wavelength, and the second size relationship between the spectral feature parameter and the upper limit of the wavelength when the spectral feature parameter is the main wavelength measurement value; if the first size relationship meets the first switching condition, the module switches to a lower detection level and verifies the threshold of the first buffer area after switching. If the second size relationship satisfies the second switching condition, switch to a higher detection level and verify the threshold of the second buffer area after switching; when the spectral feature parameter is a wavelength distribution statistical value, compare its first numerical relationship with the first preset threshold and its second numerical relationship with the second preset threshold; if the first numerical relationship satisfies the third switching condition, switch to a lower detection level; if the second numerical relationship satisfies the fourth switching condition, switch to a higher detection level. The execution module is used to switch the detection process to the determined target detection level and simultaneously adjust the integration time parameter and gain parameter.

Images