Uniform light mixing algorithm for omnidirectional light-emitting lamp

Abstract

The invention provides a uniform light mixing algorithm for an omni-directional light-emitting lamp. The uniform light mixing algorithm comprises the steps of obtaining a reflection spectrum of a target object, dynamically calculating a target SPD based on the data, driving a multi-channel LED to synthesize output, and achieving closed-loop control through continuous spectrum monitoring. According to the algorithm, a perception-decision-execution-feedback loop is constructed according to a visual effect of an optimization target after object-light interaction from a light-emitting surface static parameter to match the environment and requirements in real time, and the problems of color development distortion and insufficient comfort of static light mixing in a dynamic scene are solved.

Description

Technical Field

[0001] This application relates to the field of light emission control, and more particularly to a uniform light mixing algorithm for omnidirectional luminaires. Background Technology

[0002] Although existing uniform light mixing technology can achieve uniform and stable color on the light-emitting surface of lamps, it is essentially a static and passive optimization mode that cannot solve the fundamental contradiction that arises when light interacts with the real physical world: that is, the fixed light source spectrum cannot adaptively match the ever-changing characteristics of the illuminated objects such as paintings, fabrics, dynamic ambient light fields, and the user's differentiated visual tasks such as reading and screen work. This directly leads to deep user experience defects such as inaccurate color rendering and insufficient visual comfort in key application scenarios. Summary of the Invention

[0003] The purpose of this application is to provide a uniform light mixing algorithm for omnidirectional luminaires.

[0004] According to one aspect of this application, a uniform light mixing algorithm for omnidirectional luminaires is provided, comprising: acquiring spectral reflectance data of the surface of a target object; dynamically calculating a target spectral power distribution for optimizing the visual experience of the target object based on the spectral reflectance data; driving a multi-channel LED light source to synthesize and output corresponding light according to the target spectral power distribution; and performing closed-loop control on the output of the multi-channel LED light source based on continuous spectral monitoring results of the target object.

[0005] In one specific embodiment, acquiring the spectral reflectance data of the target object surface includes: collecting continuous spectral information of the target surface through a spectral sensor; and acquiring ambient light intensity and color temperature data through an ambient light sensor.

[0006] In one specific embodiment, the method further includes, before the data acquisition step, identifying the user's activity type using an image sensor or a ToF sensor; wherein the dynamic calculation of the target spectral power distribution is also based on the user's activity type.

[0007] In one specific embodiment, the dynamic calculation of the target spectral power distribution specifically includes: if the activity type is identified as art appreciation, a spectral optimization algorithm is executed, which calculates the spectrum that optimizes the color reproduction or artistic expression of the target object by matching the spectral reflectance data with a pre-stored standard color database; if the activity type is identified as reading or screen work, a visual comfort adjustment algorithm is executed, which calculates lighting parameters for neutralizing screen glare or reducing the contrast between ambient and screen brightness based on the ambient light data.

[0008] In one specific embodiment, the spectral optimization algorithm is configured to: solve for a non-standard lamp output spectrum such that, after being reflected by the target object, the spectral power distribution of the spectrum is closest to the reflection spectrum of the object under a standard illuminator.

[0009] In one specific embodiment, the visual comfort adjustment algorithm is configured to: automatically lower the color temperature of the multi-channel LED light source when the ambient light color temperature is detected to be higher than a preset threshold; and / or, automatically reduce the brightness of the multi-channel LED light source to maintain a constant desktop illuminance when the ambient light intensity is detected to be enhanced.

[0010] In one specific embodiment, the multi-channel LED light source includes at least five types of LED chips: red light, green light, blue light, cool white light, and warm white light.

[0011] In one specific embodiment, the multi-channel LED light source further includes cyan LED chips and magenta LED chips to expand its color gamut range and spectral shaping capabilities.

[0012] In one specific embodiment, the closed-loop control based on continuous spectral monitoring results includes: comparing the monitored actual chromaticity coordinates and luminance values ​​with the target values; if the difference between the actual chromaticity coordinates and the target values ​​exceeds a first tolerance, or the difference between the actual luminance values ​​and the target values ​​exceeds a second tolerance, then dynamically adjusting the PWM duty cycle driving the multi-channel LED light source to bring the deviation back within the tolerance.

[0013] According to another aspect of this application, a smart lighting fixture is provided, comprising: a memory, a processor, and a spectral sensor and an ambient light sensor as described in claim 2; the memory is used to store a computer program; the processor is used to, when the computer program is invoked, to control the multi-channel LED light source and execute the method described in any of the preceding claims.

[0014] Therefore, this application shifts the optimization target of the light mixing algorithm from the static parameters of the light-emitting surface of the lamp to the dynamic visual effect after the light interacts with the object, and constructs a closed-loop control loop of "perception-decision-execution-feedback". The system dynamically generates the optimal lighting spectrum based on the real-time acquired spectral characteristics of the target object, and ensures that the output spectrum is accurately matched with the environmental requirements through continuous monitoring and adjustment. Thus, at the principle level, it solves the problems of color distortion and insufficient visual comfort caused by the inability of traditional static light mixing technology to adapt to complex dynamic scenes. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 This is a flowchart of a uniform light mixing algorithm for omnidirectional luminaires. Detailed Implementation

[0017] To facilitate understanding of this application, a more complete description will be provided below with reference to the accompanying drawings. Preferred embodiments of this application are shown in the drawings. However, this application can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of this application.

[0018] It should be noted that when a component is said to be "fixed to" another component, it can be directly attached to the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the specification of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0020] Please refer to Figure 1 One embodiment of this application provides a uniform light mixing algorithm for omnidirectional luminaires, comprising: acquiring spectral reflectance data of the surface of a target object; dynamically calculating a target spectral power distribution for optimizing the visual experience of the target object based on the spectral reflectance data; driving a multi-channel LED light source to synthesize and output corresponding light according to the target spectral power distribution; and performing closed-loop control on the output of the multi-channel LED light source based on the continuous spectral monitoring results of the target object.

[0021] Furthermore, firstly, the system needs to reliably obtain the reflectivity of the target object under the current lighting conditions, which is the starting point for all subsequent calculations. To this end, the luminaire's built-in spectral sensor (typical range 350–780 nm, resolution 5–10 nm) performs integrated sampling of the target surface at a common operating distance of 0.8–1.2 m. When a significant change in the relative attitude between the luminaire and the target is detected (e.g., an angle change exceeding 10° or a distance change exceeding 0.2 m), the sampling period is automatically increased from the usual 100 ms to no more than 10 ms to avoid error accumulation caused by sudden scene changes. To ensure data availability, the system first performs dark / white field and wavelength calibration, then establishes a baseline using a high-reflectivity standard white board, thereby converting the sensor readings into the reflectivity R(λ) of the target object. If high-quality R(λ) cannot be directly obtained due to site conditions, the system can also select an approximate reflectance spectrum from a pre-set standard sample library and gradually correct it using monitoring data during subsequent operation. The purpose of this step is to characterize the object itself with "reflection" as the core, rather than just staying at the chromaticity level, thus leaving more room for optimization in subsequent spectral shaping. After obtaining R(λ), the system enters the calculation stage of the target SPD. Unlike the traditional "recipe table" style open-loop method, this method takes "visual experience" as the optimization goal and concretizes it into two measurable dimensions: one is color reproduction (e.g., ΔE2000 and Ra / R9 should be as optimal as possible), and the other is visual comfort (e.g., Δuv is close to the target, UGR decreases, shortwave energy is controlled, and illuminance is stable near E_set). Based on the channel fundamental spectrum {S_k(λ)} of a multi-channel LED, the system constructs a synthesized light S_out(λ) = ∑_k w_k·S_k(λ) and solves for the weights w to minimize the spectral distance between R(λ)·S_out(λ) and R(λ)·S_ref(λ), while satisfying power and quality constraints (e.g., 0≤w_k≤W_k^max, ∑w_k≤P_max, Δuv≤0.002, Ra≥90, and short-wavelength energy of 400–460 nm not exceeding the threshold). The solution can be obtained using weighted least squares or quadratic programming with box constraints. To balance real-time performance and embedded computing power, the system defaults to using the projected gradient method, with a single iteration taking no more than 2ms and typically converging within 20 iterations. The direct effect of this setup is to transform the "SPD for optimizing visual experience" from an abstract goal into a set of feasible solutions under multi-objective constraints: for example, in the "assessment" scenario of medical rehabilitation, the system will prioritize satisfying ΔE and Δuv; while in the "detailed operation" scenario, it will moderately increase illuminance and suppress glare while ensuring the color rendering index. Once w is obtained, the system immediately maps it to the PWM duty cycle and current amplitude of each channel; this mapping is completed by looking up a table based on the "duty cycle / current - luminous flux / SPD" generated by factory calibration. To avoid flicker interfering with rehabilitation training, the PWM frequency is set to no less than 2 kHz, and the target SVM is no higher than 0.4.Meanwhile, through an omnidirectional light-emitting structure composed of a mixing cavity and a diffuser, the system achieves an illuminance uniformity of no less than 0.7 on a sphere with a radius of 1 m, minimizing "color spots" and "stripes." In terms of channel configuration, the RGB and cool / warm white channels provide the framework energy, while the cyan (C) and magenta (M) channels are used to finely shape the energy distribution in the 480–520 nm and 560–640 nm wavelength bands, which is particularly crucial for the presentation of sensitive objects such as artwork and skin tones. To resist environmental and device drift, the system does not stop at one-time output but continuously monitors and performs closed-loop correction. Specifically, the system calculates the chromaticity error e_chroma=Δuv(actual, target) and the illuminance error e_lux=E_actual in real time. When either error exceeds the tolerance (|Δuv|≤0.002 and |e_lux| / E_set≤10%), the controller applies PID regulation to both errors based on the feedforward w, and uses inverse integral limiting and slope limiting to suppress overshoot and jitter. Considering the luminous efficacy and peak position shift of the LED with temperature, the system also performs feedforward temperature drift compensation on w based on NTC temperature estimation. If a change in posture or activity type is detected (e.g., from "evaluation" to "fine operation"), the sampling and control cycle will be temporarily accelerated until the tolerance requirements are met again. Through this closed-loop strategy, the lamp output can return to the stable range within a typical 1 second, thereby ensuring the consistency of visual cues during rehabilitation training. Combining the above process, rehabilitation institutions or home users can use the system in the following order: ① Install the lamps 0.8–1.2 m above the training table and complete one-click calibration; ② Select the "Rehabilitation—Assessment" or "Rehabilitation—Fine Maneuvering" scenario in the application interface (the system automatically sets the E_set and CCT range); ③ If conditions permit, collect R(λ) data for the training equipment; otherwise, select the standard library approximation; ④ The system calculates and outputs the target SPD, continuously monitors it during training, and quickly corrects it when necessary; ⑤ Commonly used tasks (such as grasping training and contrast sensitivity assessment) can be saved as recipes for later one-click recall. The direct benefit of this method is maintaining consistency between chromaticity and illuminance during actual training, reducing visual fatigue, and improving the accuracy of color / contrast recognition.

[0022] In one specific embodiment, acquiring the spectral reflectance data of the target object surface includes: collecting continuous spectral information of the target surface through a spectral sensor; and acquiring ambient light intensity and color temperature data through an ambient light sensor.

[0023] Furthermore, the spectral sensor acquires continuous spectral information from the target surface, including sampling geometry and anti-glare: the sensor's field of view is limited to 20–60° using a cosine-corrected probe, and the angle between the optical axis and the target surface normal is preferably controlled at 0–30° to reduce specular reflection contamination; if necessary, polarizers are added to suppress highlights in specular material scenes. Measurement range and resolution: operating wavelength 350–780 nm, grating resolution 5–10 nm, A / D quantization 12–16 bit; to cover low-reflectivity dark objects and high-reflectivity white boards, the integration time is adaptively set to 0.5–50 ms, combined with automatic gain control (AGC). Calibration and conversion: first, dark field subtraction and white field normalization are performed, then a reflectivity conversion coefficient is established using a standard white board to obtain R(λ); if the target area is small (e.g., a small rehabilitation training piece with a diameter <10 mm), the measurement domain can be limited by a pinhole aperture and signal-to-noise weighting can be performed at the algorithm end. Continuity Guarantee: During regular monitoring, the sampling period is set to 100 ms; when a pose change event is detected, it is temporarily increased to ≤10 ms to ensure the timeliness of R(λ) updates. Ambient light sensor acquires ambient light intensity and color temperature data, including illuminance measurement: The ambient light sensor has a cosine correction structure to measure illuminance E_env (1–10000 lx); Color temperature estimation: CCT_env (1800–10000 K) is estimated using a dual-channel or three-channel sensor, and nonlinearity errors are corrected using the factory-fitted curve; Anti-false triggering: When instantaneous changes in the screen or moving light source cause a sudden increase in E_env, time window mid-range filtering and threshold confirmation are used to prevent erroneous entry into the high-brightness strategy; Spatial consistency: The sensor and spectrometer are arranged on the same side to avoid obstruction causing reading deviations; when necessary, two ambient light probes are used for least-squares fusion to estimate the actual E_env of the working surface.

[0024] In one specific embodiment, the method further includes, before the data acquisition step, identifying the user's activity type using an image sensor or a ToF sensor; wherein the dynamic calculation of the target spectral power distribution is also based on the user's activity type.

[0025] Furthermore, before collecting reflectance spectrum and ambient light data, the luminaire first performs a "light scan" of the work area using an image sensor (such as 720p, 30 frames / second) or a ToF sensor (approximately 0.2–2.5 meters range). This extracts only some intuitive and task-related signals locally on the device—such as how fast the hand or tool is moving, whether the screen occupies the frame, whether the desktop has fine textures like paper, and whether the distance between the person and the luminaire is stable. The system summarizes these signals into labels indicating "what the user is doing" within a small time window of 2–5 seconds (such as appreciating art, reading / screen work, rehabilitation assessment, fine motor rehabilitation, or rest and relaxation), and sets entry and exit thresholds to prevent back-and-forth shaking. If the user turns off the camera for privacy, only the distance and movement rhythm of the ToF sensor are used for coarse classification, and the entire judgment is completed on the device side without being uploaded. Next, the calculation of the "target spectral power distribution" begins. This can be understood as a "recipe," telling each LED channel (red, green, blue, cool white, warm white, and cyan and magenta if necessary) how bright it should be and how much energy should be released at different wavelengths of visible light. The reason for identifying the activity type first is that different tasks have different priorities regarding "fidelity and comfort": if it's identified as art appreciation, the system prioritizes color fidelity—using channel ratios to bring the target object's reflectance spectrum under local light as close as possible to the reflectance under standard sunlight. A common practice is to fine-tune the cyan / magenta auxiliary channels to correct the pigment-sensitive blue-green and magenta regions, thus making the painting appear "color-accurate." For reading or screen work, the system prioritizes comfort and anti-glare—the overall tone is slightly warmer, with blue light (approximately 400–460). The illumination intensity (nm) converges, ensuring that the light is bright but not glaring when projected onto paper or a keyboard, without excessively interfering with screen brightness, thus reducing eye strain. For rehabilitation assessments, the system sets the illumination to a medium level, emphasizing discernible subtle color differences and contrasts to facilitate standardized scales. For fine motor skills rehabilitation, the illuminance is moderately increased, the color tone is slightly warmer, and more weight is given to "red reproduction" in the formula to ensure clearer skin and blood color details and more stable hand-eye coordination. If the system is judged to be for rest and relaxation, the illuminance and "coolness" are adjusted downwards for a softer overall feel. The entire "identification first, formula later" chain is completed within approximately 120–300 milliseconds, after which the formula is distributed to each channel driver. During operation, if a person or object moves, the system continuously measures the spectrum and brightness while making minor corrections to stabilize the visual perception within the preset "clear / long-lasting visibility" range. In this way, "activity type" is not just a label, but the principle of "why the light is matched in this way" is embedded in the calculation: when the task emphasizes realism, the reflectance spectrum is made close to the standard reference; when the task emphasizes comfort, blue light and glare are controlled and the contrast between the environment and the screen is balanced, so that the "dynamically calculated target formula" truly corresponds causally with the user's usage scenario.

[0026] In one specific embodiment, the dynamic calculation of the target spectral power distribution specifically includes: if the activity type is identified as art appreciation, a spectral optimization algorithm is executed, which calculates the spectrum that optimizes the color reproduction or artistic expression of the target object by matching the spectral reflectance data with a pre-stored standard color database; if the activity type is identified as reading or screen work, a visual comfort adjustment algorithm is executed, which calculates lighting parameters for neutralizing screen glare or reducing the contrast between ambient and screen brightness based on the ambient light data.

[0027] Furthermore, the "dynamic calculation of target spectral power distribution" process operates in two paths based on the identified activity type: When it is determined to be art appreciation, the system first uses the aforementioned reflectance spectral data to calculate "the color that the viewed object should present under standard daylight" (which can be understood as projecting the appearance of each color point on the target surface back to the standard reference light). Then, it searches in the multi-channel spectral library of the lamp to find the closest matching ratio, so that the color reflected from the object after illumination by this light is as small as possible compared with the standard appearance. To avoid "some pigments appearing gray or off-color" due to relying solely on the three color channels, the algorithm will prioritize mobilizing auxiliary channels such as cyan and magenta to supplement the energy of the blue-green and red-magenta bands. It can also slightly enhance specific bands in the "expressiveness priority" sub-mode to improve texture level and saturation without significantly compromising overall fidelity. When it is determined to be reading or screen work, the system no longer pursues strict consistency with standard daylight. Instead, it works backwards from "comfort and non-glare" to derive lighting parameters: it first reads the intensity and color cast of the ambient light, estimates the brightness ratio between the current screen and the surroundings, and then stabilizes the tabletop illuminance within a moderate range that is easy to read for extended periods (such as around 500 lux). At the same time, it slightly shifts the light hue towards a gentler direction and reduces the short-wavelength blue light component of 400 to 460 nanometers to neutralize screen glare and bring the brightness contrast between the screen and the environment back to a more acceptable range (such as about 1:3 to 3:1), avoiding situations where "the screen is too bright and the surroundings are too dark" or "the surroundings are too cold, causing glare." The outputs of both paths are ultimately converted into the driving intensity formula of each LED channel (red, green, blue, cool white, warm white, and cyan and magenta if necessary). The former solves the problem of "seeing clearly," while the latter solves the problem of "seeing for a long time and being comfortable," thus concretely implementing the abstract goal of "visual experience" into an executable spectrum and brightness ratio.

[0028] In one specific embodiment, the spectral optimization algorithm is configured to: solve for a non-standard lamp output spectrum such that, after being reflected by the target object, the spectral power distribution of the spectrum is closest to the reflection spectrum of the object under a standard illuminator.

[0029] Furthermore, spectral optimization is not simply about "aligning the lamp's chromaticity coordinates with sunlight." Instead, it uses the actual light reflected by an object after it is illuminated as a comparison to infer the light distribution: the system first obtains the reflectivity of the target object (which can be understood as "how much light this object can reflect at each wavelength"), then calculates "the reflected light that the object should present under standard sunlight" using a selected standard illuminator (such as standard sunlight). Subsequently, it treats each LED channel of the lamp as a set of adjustable "spectral bases" and searches for a set of non-negative channel ratios constrained by power and smoothness so that the overall shape of the reflected light after the light emitted by the lamp illuminates the object is closest to "standard sunlight conditions." The significance of doing this is to bypass the common error of "only looking at the chromaticity point" and directly reduce the problem of heterochromatic color caused by pigments, materials, and fine textures, thereby allowing the visual experience to truly return to "what it should look like under sunlight." To avoid ambient light interference, the system first estimates and subtracts the contribution of the environment to each wavelength, only calculating the portion that the luminaire needs to compensate for. The calculated ratio is then mapped to the driving intensity of each channel, and a small-step closed loop of rapid retesting and fine-tuning brings the color difference and power within tolerance. Taking oil paintings containing strong cyan pigments or rehabilitation training scenarios focusing on skin as examples, the algorithm appropriately boosts the blue-green or magenta related wavelengths and suppresses irrelevant energy, restoring the painting / skin to its proper gradation and saturation under sunlight, while avoiding sacrificing realistic texture for the sake of "looking white." This implementation transforms the principle in the claims of "solving the output spectrum of non-standard luminaires and making the spectrum after reflection from objects approximate the results of standard luminaires" into an executable engineering process: modeling centered on reflection, solving with constrained channel ratios, and concluding with environmental subtraction and small-step closed-loop convergence, with a clear purpose, clear cause and effect, and quantifiable results.

[0030] In one specific embodiment, the visual comfort adjustment algorithm is configured to: automatically lower the color temperature of the multi-channel LED light source when the ambient light color temperature is detected to be higher than a preset threshold; and / or, automatically reduce the brightness of the multi-channel LED light source to maintain a constant desktop illuminance when the ambient light intensity is detected to be enhanced.

[0031] Furthermore, the visual comfort adjustment algorithm operates on the principle of "reducing glare, stabilizing brightness, and minimizing stimulation": The system continuously reads the color temperature and illuminance of the ambient light sensor. Once it determines that the overall environment is too cool or too blue (for example, exceeding the set "cool threshold" for an extended period—which can be understood as the cooler side of typical sunlight), the light distribution of the lamps will "take a small step towards a gentler direction." This is not done by simply turning it yellow, but by moderately reducing the energy on the blue / cyan side and increasing the warm white component, so that the color tone of the screen and its surroundings falls back to a softer range. At the same time, when it detects that the surroundings are brighter (such as increased sunlight near a window or the overhead light being turned on), the system will proportionally reduce its own brightness to maintain the desktop at a stable and easy-to-view level (which can be understood as the medium brightness commonly used for office reading). In this way, the brightness ratio between the screen and the environment will not be unbalanced, neither "the screen is bright and the surroundings are dark" nor "the surroundings are bright and the screen is glaring," fundamentally reducing glare. The entire adjustment process is slow and consistent: the algorithm sets entry / exit thresholds and an upper limit for the rate of change to avoid glare from sudden fluctuations or abrupt changes; and it combines this with closed-loop fine-tuning using spectral monitoring within the lamp to ensure that each correction to hue and brightness is truly reflected on the desktop rather than remaining at the set value. For example, when reading by the window at midday, the environment is cool and bright, so the lamp will become slightly softer and automatically dim, keeping the desktop at a familiar reading brightness; as evening light weakens, the lamp smoothly increases its brightness and pulls the hue back to neutral to avoid straining the eyes while reading. Through this self-adjusting mechanism of "being gentler when the environment is cool and yielding when the environment is brighter," users can have a more stable visual background when reading or working on the screen, glare is neutralized, and the desktop brightness does not fluctuate wildly, making it more comfortable for prolonged use.

[0032] In one specific embodiment, the multi-channel LED light source includes at least five types of LED chips: red light, green light, blue light, cool white light, and warm white light.

[0033] Furthermore, the multi-channel LED light source adopts a "five-channel skeleton" configuration, namely, red, green, and blue narrowband chips combined with cool white and warm white broadband chips. Each channel is driven by an independent constant current and controlled by a high-resolution dimming unit (e.g., 12-16 bit duty cycle control, with a dimming frequency higher than the common flicker perception threshold to ensure comfort). The design intention is to use two types of white light to provide "high throughput and broadband base" to provide continuous and balanced visible light components with high efficiency; at the same time, RGB three colors are used as "fine-tuning tools" to supplement and trim key wavelengths, correcting problems such as color flattening, pale skin tones, or grayish pigments that are common when relying solely on white light. Specifically, warm white is responsible for providing energy on the red-yellow side, suitable for reading and prolonged close-range eye use; cool white supplements the blue-green side, which helps improve clarity and detail; red, green, and blue are used in scenarios with high requirements for color accuracy, such as art or medical rehabilitation training, to finely adjust the colors that actually reach the eyes after reflection according to the ratio given by the algorithm, making them closer to what they should look like under standard reference. To achieve smooth, omnidirectional light emission without "color spots," the five channels are first fully superimposed within the mixing cavity and then homogenized by a diffuser, typically achieving high illuminance uniformity on a one-meter spherical surface. To ensure long-term stability and consistency, each channel is calibrated at the factory using a "current / duty cycle - luminous flux and spectrum" table and its temperature coefficient is recorded. During operation, minor compensation is made based on temperature sensors. In practice, this "five-channel framework" can cover the core requirements from "seeing clearly" (reducing the risk of heterometachromatic color matching through RGB color correction) to "seeing for longer periods" (suppressing glare and stimulation through white light underlay), while reserving interfaces for optional cyan and magenta expansion channels. Even without expansion, this configuration is sufficient to ensure the stable implementation of the uniform light mixing algorithm in most home, office, and rehabilitation training scenarios.

[0034] In one specific embodiment, the multi-channel LED light source further includes cyan LED chips and magenta LED chips to expand its color gamut range and spectral shaping capabilities.

[0035] Furthermore, based on the aforementioned "five-channel framework," two additional channels, cyan and magenta, are added to fill the two common "spectral valleys" in RGB + cool / warm white: the blue-green transition zone and the red-blue mixing zone. The cyan channel (peak position approximately 500±10nm) can directly fill the energy gap between blue and green, preventing materials containing cyan-green pigments, aquamarine fabrics, and jade-like materials from appearing grayish under light. The magenta channel enhances the red-magenta side in a "red + blue synergy" manner, providing an independent and adjustable boost to areas that are extremely sensitive to skin rosiness, blood-red details, flowers, and the magenta layer in oil paintings, avoiding the overall imbalance caused by simply boosting red or blue. The benefit on the algorithm side is that when the system subtracts the "expected appearance under standard daylight" from the "reflected appearance under this lamp," the error peak often falls precisely in these two bands. With the addition of C / M, the optimizer can smooth out the error "point-to-point" with less power increment, reducing color difference and minimizing the impact on other channels, thus refining and stabilizing the ability to "see clearly." At the same time, in comfort-priority scenarios such as reading or screen work, the control strategy sets an upper limit on short-wavelength stimulation, and C / M only intervenes lightly when it is necessary to correct local color cast, ensuring a soft and non-glaring overall visual experience. In terms of engineering implementation, the cyan and magenta chips are each equipped with independent constant current and high-resolution dimming channels. A calibration table of "current / duty time - luminous flux and emitted light spectrum" is established at the factory and the temperature coefficient is recorded. During operation, they participate in temperature drift compensation and closed-loop fine-tuning together with the original five channels. To ensure that there are no color spots in omnidirectional light emission, the two channels are mixed with the other channels in the mixing cavity and then homogenized by a diffuser. For example, in museum mode, the cyan channel can accurately restore the brightness and saturation of cyan pigments, while the magenta channel is used to restore the layers of magenta and purple-red fabrics. In medical rehabilitation scenarios, the magenta channel makes it easier to distinguish between hand skin and slightly congested areas. Combined with the cyan channel for color calibration of small colored instruments, it can improve the stability of target object recognition and hand-eye coordination while maintaining comfort. Even if one of the extended channels fails, the system can revert to the five-channel framework and continue to work with minimal performance loss, thus achieving a better overall balance between feasibility, reliability, and visual quality.

[0036] In one specific embodiment, the closed-loop control based on continuous spectral monitoring results includes: comparing the monitored actual chromaticity coordinates and luminance values ​​with the target values; if the difference between the actual chromaticity coordinates and the target values ​​exceeds a first tolerance, or the difference between the actual luminance values ​​and the target values ​​exceeds a second tolerance, then dynamically adjusting the PWM duty cycle driving the multi-channel LED light source to bring the deviation back within the tolerance.

[0037] Furthermore, the closed-loop control operates continuously based on the principle of "accurate measurement first—fine-tuning then—until it returns to within the range": the spectral / environmental sensor inside the lamp reads the current light output at a fixed period (usually about 100 milliseconds, automatically speeding up to about 10 milliseconds when a person or object approaches), and processes it to obtain two most intuitive evaluation quantities—"color point" (which can be understood as the current position on the chromaticity diagram) and "desktop brightness," which are compared with the target values ​​set by the system in this scenario; when the color deviation exceeds the first tolerance (which can be understood as "a slight color shift that can just be perceived by the naked eye," which is often set as a very small threshold in engineering), the system will take action. If the color deviation is less than a few thousandths (e.g., a deviation of no more than a few thousandths on the chromaticity map) or the brightness error exceeds the second tolerance (e.g., ±10% relative to the target brightness), the controller begins to make very small corrections to the PWM duty cycle of each channel: if the main issue is color deviation, it prioritizes adjustments to the channels that "determine the color" (e.g., red, green, blue, and optional cyan and magenta), and tries to maintain the ratio of the two types of white light channels unchanged to prevent brightness from fluctuating wildly; if the main issue is brightness deviation, it first fine-tunes the white light channels as a whole to bring the "brightness" back to the target, and then uses the three color channels for minor color correction to avoid new brightness fluctuations caused by color correction. To prevent "back and forth" adjustments, the controller sets a slight priority level for color and brightness (e.g., first bringing the brightness closer to the target, then tightening the color to a smaller deviation range), with an upper limit on the rate of change and an inverse integral limit for each step, ensuring that the change is smooth and without noticeable "jumps"; at the same time, it combines temperature sensors to perform feedforward compensation for LED thermal drift, reducing unnecessary back and forth adjustments. In actual execution, the system uses a small lookup table of "channel-color / brightness response" to determine whether to increase or decrease the brightness of each channel each time, and by how much (this can be understood as "the sensitivity of color and brightness to each channel"). Therefore, only a small portion of the duty cycle needs to be changed each time, and the deviation will be brought back to within the tolerance within tens to hundreds of milliseconds. The entire dimming is completed under high-frequency PWM (e.g., above 2 kHz), which is neither flickering nor dazzling. If the environment changes suddenly in a short period of time (e.g., the sunlight outside the window suddenly increases), the algorithm will temporarily increase the sampling and control frequency to speed up convergence. If the sensor data is noisy, a small window median / mean filter will be applied before deciding whether to take action. In this way, the system can stably keep the "detected actual color points and brightness" within two preset "lines": one for color deviation and one for brightness. If the line is crossed, it will be gently pulled back; if it is not crossed, it will not be disturbed, which is both stable and natural.

[0038] According to another aspect of this application, a smart lighting fixture is provided, comprising: a memory, a processor, and a spectral sensor and an ambient light sensor as described in claim 2; the memory is used to store a computer program; the processor is used to, when the computer program is invoked, to control the multi-channel LED light source and execute the method described in any of the preceding claims.

[0039] Furthermore, the intelligent lighting fixture comprises a housing, an omnidirectional multi-channel LED light-emitting component, a driver and power supply module, a processor and a memory, and a spectral sensor and an ambient light sensor for acquiring on-site information. The memory pre-stores a computer program for executing the method of this application, calibration data for each LED channel (current / duty cycle and luminous flux, channel light emission spectrum, temperature coefficient), and a standard color / reference spectrum library. The processor connects to the two types of sensors via a bus to acquire the "reflection characteristics of the target object" and the "intensity and color cast of the on-site light." Then, according to the program logic, it converts these data into a "light distribution prescription," that is, determines how much each LED channel should output and in which wavelengths it should be enhanced or converged. Subsequently, the processor sends instructions to the driver module to modulate the current and duty cycle of the red, green, blue, cool white, and warm white (and optional cyan and magenta) channels according to the prescription. First, it achieves open-loop synthetic light emission, and then continuously performs small closed-loop corrections on the reflected color points and the brightness of the desktop, so that the deviation automatically returns to the set tolerance. To ensure feasibility and long-term stable operation, the luminaire employs a mixing cavity and a diffuser in its structure, ensuring that light from different channels is thoroughly "mixed" before entering the space, thus avoiding color spots under omnidirectional illumination. For reliability, the processor performs feedforward compensation for each channel based on temperature sensor readings to reduce thermal drift effects, and briefly accelerates sampling and adjustment to quickly stabilize when a sudden environmental change (such as a sudden increase in sunlight) is detected. In terms of user experience, users only need to press a button or select a scenario such as "Art Appreciation," "Reading / Screen Work," or "Rehabilitation Assessment / Fine Operations" through an application. The processor then calls upon the corresponding target and weights, automatically completing reflection data acquisition, target light distribution calculation, and closed-loop tracking using any of the methods described in this application. If an extended channel becomes unavailable, the system can revert to the "five-channel skeleton" to continue working, without affecting the basic visual experience or interrupting the task. Through the cooperation of the aforementioned hardware and software, when the processor calls the computer program stored in the memory, it can control the multi-channel LED light source to execute the entire process of any of the aforementioned methods—an integrated closed loop of "collection, calculation, combination, and adjustment"—thereby implementing the goal of "seeing clearly, seeing for a long time, and seeing stably" in a concrete device in an engineering manner.

[0040] Therefore, this application shifts the optimization target of the light mixing algorithm from the static parameters of the light-emitting surface of the lamp to the dynamic visual effect after the light interacts with the object, and constructs a closed-loop control loop of "perception-decision-execution-feedback". The system dynamically generates the optimal lighting spectrum based on the real-time acquired spectral characteristics of the target object, and ensures that the output spectrum is accurately matched with the environmental requirements through continuous monitoring and adjustment. Thus, at the principle level, it solves the problems of color distortion and insufficient visual comfort caused by the inability of traditional static light mixing technology to adapt to complex dynamic scenes.

[0041] The embodiments described above are merely examples of several implementations of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the scope of protection of this application.

Claims

1. A uniform light mixing algorithm for an omnidirectional light emitting luminaire, characterized in that, The method comprises: acquiring spectral reflectance data of a target object surface; based on the spectral reflectance data, dynamically calculating a target spectral power distribution for optimizing visual experience of the target object; driving a multi-channel LED light source to synthesize and output corresponding light rays according to the target spectral power distribution; based on continuous spectral monitoring results of the target object, performing closed-loop control on the output of the multi-channel LED light source.

2. The uniform light mixing algorithm for an omnidirectional light emitting luminaire of claim 1, wherein, The acquiring of the spectral reflectance data of the target object surface comprises: acquiring continuous spectral information of the target surface by a spectral sensor; and acquiring intensity and color temperature data of ambient light by an ambient light sensor.

3. The uniform light mixing algorithm for an omnidirectional light emitting luminaire of claim 2, wherein, Before the data acquisition step, the method further comprises: identifying an activity type of a user by an image sensor or a ToF sensor; wherein the dynamic calculation of the target spectral power distribution is further based on the activity type of the user.

4. The uniform light mixing algorithm for an omnidirectional light emitting luminaire of claim 3, wherein, The dynamic calculation of the target spectral power distribution specifically comprises: if the activity type is identified as art appreciation, performing a spectral optimization algorithm, which calculates a spectrum that makes the color restoration degree or artistic expressiveness of the target object optimal by matching the spectral reflectance data with a pre-stored standard color database; if the activity type is identified as reading or screen work, performing a visual comfort adjustment algorithm, which calculates illumination parameters for neutralizing screen glare or reducing the contrast between ambient light and screen brightness based on the ambient light data.

5. The uniform light mixing algorithm for an omnidirectional light emitting luminaire of claim 4, wherein, The spectral optimization algorithm is configured to solve a non-standard lamp output spectrum, so that after the spectrum is reflected by the target object, its spectral power distribution is closest to the reflection spectrum of the object under standard illumination.

6. The uniform light mixing algorithm for an omnidirectional light emitting luminaire of claim 4, wherein, The visual comfort adjustment algorithm is configured to automatically lower the color temperature of the multi-channel LED light source when the color temperature of the ambient light is detected to be higher than a preset threshold; and / or automatically reduce the brightness of the multi-channel LED light source to maintain constant desktop illuminance when the intensity of the ambient light is detected to be enhanced.

7. A uniform light mixing algorithm for an omnidirectional light emitting luminaire according to any one of claims 1-6, characterized in that, The multi-channel LED light source comprises at least five types of LED chips, i.e. red, green, blue, cool white and warm white.

8. The uniform light mixing algorithm for an omnidirectional light emitting luminaire of claim 7, wherein, The multi-channel LED light source further comprises cyan and magenta LED chips to expand its color gamut range and spectral shaping capability.

9. The uniform light mixing algorithm for an omnidirectional light emitting luminaire of claim 7, wherein, The closed-loop control based on the continuous spectral monitoring results comprises: comparing the monitored actual chromaticity coordinates and brightness values with target values; if the difference between the actual chromaticity coordinates and the target values exceeds a first tolerance, or the difference between the actual brightness values and the target values exceeds a second tolerance, dynamically adjusting the PWM duty cycle for driving the multi-channel LED light source to make the deviation return to within the tolerance.

10. A smart light fixture, characterized in that The method comprises: a memory, a processor, and the spectral sensor and the ambient light sensor as claimed in claim 2; the memory is used to store a computer program; the processor is used to execute the method as claimed in any one of claims 1-9 by controlling the multi-channel LED light source when the computer program is called.

Images