LED intelligent adjusting method and system based on full spectrum analysis

Abstract

The invention discloses an LED intelligent adjustment method and system based on full spectrum analysis, and the method comprises the steps: carrying out the spectrum analysis of a plurality of LED lights, extracting a peak wavelength and a full width at half maximum parameter, carrying out the spectrum fitting based on a Gaussian function, and screening out a qualified LED group; the method comprises the following steps: carrying out actual illumination application by utilizing an LED group, obtaining an original image and an illumination feedback image, dividing regions according to image sizes, comparing image attribute differences of different regions, screening out a comparison image region, carrying out color feature extraction on the basis of the comparison image region, forming a first color feature set and a second color feature set, and fusing the feature sets by adopting a DCA feature fusion algorithm to obtain a first color feature set and a second color feature set. Forming first and second fusion features; and calculating the difference between the first fusion feature and the second fusion feature through a standard Euclidean distance method, and evaluating and adjusting the lamp effect of the LED group according to the difference. According to the invention, accurate model selection and application of full-spectrum-based LED light can be realized, accurate evaluation of the full-spectrum LED illumination effect is realized, and illumination requirements under multiple scenes are met.

Description

Technical Field

[0001] The present invention relates to the field of LED spectrum analysis, and more specifically, to an LED intelligent adjustment method and system based on full spectrum analysis. Background Art

[0002] With the rapid development of LED lighting technology, LED light sources have been widely used in the lighting field due to their high efficiency, energy saving, and environmental protection. However, traditional LED lighting systems often use fixed spectrum output, which is difficult to adapt to complex and changing lighting needs and full-spectrum lighting needs. In addition, in the existing technology, the selection and lighting effect evaluation of LEDs based on the full spectrum are poor, often relying on manual judgment, with poor practicality, and difficult to adapt to the application of lighting groups in multiple scenarios.

[0003] Therefore, there is an urgent need for an efficient method of LED intelligent adjustment, analysis and evaluation based on full spectrum analysis. Summary of the invention

[0004] The present invention overcomes the defects of the prior art and proposes an LED intelligent adjustment method and system based on full spectrum analysis.

[0005] The first aspect of the present invention provides an LED intelligent adjustment method based on full spectrum analysis, comprising:

[0006] Perform spectral power distribution tests on a plurality of preset single-primary-color LED lights to obtain corresponding measured spectral power distribution curves;

[0007] The peak wavelength and half-width are extracted by measuring the spectral power distribution curve. The peak wavelength and half-width are used as parameters, and the spectrum is fitted by Gaussian function to obtain the fitting spectral power distribution curve corresponding to each single-primary color LED light.

[0008] Combine multiple single-primary-color LED lights to form N LED groups. For each LED group, perform spectral fitting on multiple fitted spectral power distribution curves in the group to obtain a synthetic spectrum. Perform spectral matching analysis on the synthetic spectrum based on the spectral mismatch error of the least squares method, and screen out qualified LED groups, marking them as preferred LED groups.

[0009] By optimizing the LED group to perform actual LED lighting application, an original image and a lighting feedback image are obtained, and the original image is divided into multiple image regions based on the size. The image region is used as an analysis unit, and a difference analysis of preset image attributes is performed in the original image and the lighting feedback image. Through the preset attribute range, multiple comparison image regions are screened out;

[0010] Based on multiple contrast image regions, color features are extracted from the illumination feedback image and the preset standard illumination image respectively to obtain a first color feature set and a second color feature set. Based on the DCA feature fusion algorithm, features in the first color feature set are fused to form a first fused feature, and features in the second color feature set are fused to form a second fused feature.

[0011] The standard Euclidean distance method is introduced to vectorize the first fusion feature and the second fusion feature and calculate the difference between the data. The lighting effect of the preferred LED group is evaluated and adjusted based on the difference.

[0012] In this scheme, the peak wavelength and half-width are extracted by measuring the spectral power distribution curve, and the peak wavelength and half-width are used as parameters to perform spectrum fitting through Gaussian function to obtain the fitting spectral power distribution curve corresponding to each single-primary color LED light, specifically:

[0013] For a measured spectral power distribution curve, the peak wavelength and half-height width are extracted;

[0014] The peak wavelength and half-height width are used as parameters to fit the full spectrum through a preset modified Gaussian function;

[0015] The measured spectral power distribution curves corresponding to the single-primary-color LED lights are all fitted to obtain the fitted spectral power distribution curve corresponding to each single-primary-color LED light;

[0016] The Gaussian function is as follows:

[0017]

[0018] Where S(λ) is the relative spectral power distribution of the LED, λ 0 is the peak wavelength, and Δλ is the full width at half maximum.

[0019] In this solution, multiple single-primary color LED lights are combined to form N LED groups. For each LED group, multiple fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum. The synthetic spectrum is subjected to spectral matching analysis based on the spectral mismatch error of the least squares method, and qualified LED groups are screened out and marked as preferred LED groups, specifically:

[0020] Among multiple single-primary-color LED lights, analyze the types and quantities of the current primary colors, select one of each primary color to combine to obtain a light group simulating a full spectrum, which is recorded as N LED groups;

[0021] For each LED group, multiple fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum;

[0022] The spectrum mismatch error based on the least square method is used to perform spectrum matching analysis on the synthetic spectrum and obtain the total spectrum mismatch evaluation index;

[0023] By presetting the evaluation index range, qualified LED groups are screened out and marked as preferred LED groups;

[0024] The fitting expression of the synthetic spectrum is as follows:

[0025]

[0026] Among them, S m (λ) is the spectral power distribution after fitting, S i is the spectral power distribution of light i, D i is the duty cycle of light i, λ is the wavelength of the independent variable;

[0027] The spectrum mismatch error based on the least squares method and the corresponding total spectrum mismatch evaluation index are specifically calculated as follows:

[0028]

[0029] Where ε is the total spectral mismatch evaluation index, S T (λ) is the target spectral power distribution; S i (λ) is the spectral power distribution of the i-th light, n is the total number of lights; D i is the duty cycle of light i.

[0030] In this solution, the actual LED lighting application is performed by the preferred LED group to obtain the original image and the lighting feedback image, which are divided into multiple image areas based on the size of the original image. The image area is used as the analysis unit, and the difference analysis of the preset image attributes is performed in the original image and the lighting feedback image. Through the preset attribute range, multiple comparison image areas are screened out, specifically:

[0031] Select a preferred LED group for analysis;

[0032] In an actual lighting scene, an original image of a target scene is obtained by a preset camera device, actual LED lighting application is performed by using a preferred LED group, and a lighting feedback image of the target scene is obtained;

[0033] Grid division is performed based on the size of the original image to form multiple image regions;

[0034] Taking an image area as the analysis unit, based on the three attribute dimensions of brightness, contrast, and color saturation, image difference analysis is performed between the original image and the illumination feedback image. If the differences in the three attribute dimensions all exceed the preset attribute range, the image area is marked as a comparison image area;

[0035] All image regions are analyzed and screened to obtain multiple comparison image regions.

[0036] In this solution, based on multiple contrast image areas, color features are extracted from the illumination feedback image and the preset standard illumination image respectively to obtain a first color feature set and a second color feature set. Based on the DCA feature fusion algorithm, the features in the first color feature set are fused to form a first fused feature, and the features in the second color feature set are fused to form a second fused feature, specifically:

[0037] Using a plurality of contrast image regions as feature extraction regions, extracting color features from the illumination feedback image, and integrating the feature data to form a first color feature set;

[0038] Using a plurality of contrasting image regions as feature extraction regions, extracting color features from a preset standard illumination image, and integrating the feature data to form a second color feature set;

[0039] In the process of color feature extraction, feature extraction is performed by calculating the color histogram of the corresponding image area;

[0040] In the first color feature set and the second color feature set, a DCA feature fusion algorithm is used to generate a category activation map for each feature data in the color feature set and feature fusion is performed based on preset weights to obtain a first fused feature and a second fused feature, respectively.

[0041] In this solution, the standard Euclidean distance method is introduced to vectorize the first fusion feature and the second fusion feature and calculate the difference between the data. The lighting effect of the preferred LED group is evaluated and adjusted based on the difference. Specifically:

[0042] Performing feature vectorization on the first fused feature and the second fused feature to obtain a first feature vector and a second feature vector;

[0043] The distance between the first eigenvector and the second eigenvector is calculated by the standard Euclidean distance method, and the difference between the data is evaluated based on the distance value;

[0044] Through difference analysis, the lighting effect of the preferred LED group is evaluated;

[0045] Perform lighting effect evaluation on all preferred LED groups and generate lighting effect evaluation results;

[0046] Based on the lighting effect evaluation results, different preferred LED groups are put into practical application and the lighting parameters of the LED groups are adjusted.

[0047] The second aspect of the present invention further provides an LED intelligent adjustment system based on full spectrum analysis, the system comprising: a memory, a processor, the memory comprising an LED intelligent adjustment program based on full spectrum analysis, the LED intelligent adjustment program based on full spectrum analysis being executed by the processor to implement the following steps:

[0048] Perform spectral power distribution tests on a plurality of preset single-primary-color LED lights to obtain corresponding measured spectral power distribution curves;

[0049] The peak wavelength and half-width are extracted by measuring the spectral power distribution curve. The peak wavelength and half-width are used as parameters, and the spectrum is fitted by Gaussian function to obtain the fitting spectral power distribution curve corresponding to each single-primary color LED light.

[0050] Combine multiple single-primary-color LED lights to form N LED groups. For each LED group, perform spectral fitting on multiple fitted spectral power distribution curves in the group to obtain a synthetic spectrum. Perform spectral matching analysis on the synthetic spectrum based on the spectral mismatch error of the least squares method, and screen out qualified LED groups, marking them as preferred LED groups.

[0051] By optimizing the LED group to perform actual LED lighting application, an original image and a lighting feedback image are obtained, and the original image is divided into multiple image regions based on the size. The image region is used as an analysis unit, and a difference analysis of preset image attributes is performed in the original image and the lighting feedback image. Through the preset attribute range, multiple comparison image regions are screened out;

[0052] Based on multiple contrast image regions, color features are extracted from the illumination feedback image and the preset standard illumination image respectively to obtain a first color feature set and a second color feature set. Based on the DCA feature fusion algorithm, features in the first color feature set are fused to form a first fused feature, and features in the second color feature set are fused to form a second fused feature.

[0053] The standard Euclidean distance method is introduced to vectorize the first fusion feature and the second fusion feature and calculate the difference between the data. The lighting effect of the preferred LED group is evaluated and adjusted based on the difference.

[0054] The third aspect of the present invention also provides a computer-readable storage medium, which includes an LED intelligent adjustment program based on full spectrum analysis. When the LED intelligent adjustment program based on full spectrum analysis is executed by a processor, the steps of the LED intelligent adjustment method based on full spectrum analysis as described in any one of the above items are implemented.

[0055] The present invention discloses an LED intelligent adjustment method and system based on full spectrum analysis. By performing spectrum analysis on multiple LED lights, peak wavelength and half-width parameters are extracted to perform spectrum fitting based on Gaussian function, and qualified LED groups are screened out; the LED groups are used for actual lighting application, the original image and the lighting feedback image are obtained, the regions are divided according to the image size, the image attribute differences of different regions are compared, the contrast image regions are screened out, color features are extracted based on the contrast image regions, and the first and second color feature sets are formed. The feature sets are fused using the DCA feature fusion algorithm to form the first and second fused features; the difference between the first and second fused features is calculated by the standard Euclidean distance method, and the lighting effect of the LED group is evaluated and adjusted accordingly. Through the present invention, accurate selection and application of LED lights based on the full spectrum can be achieved, accurate evaluation of the full spectrum LED lighting effect can be achieved, and lighting requirements in multiple scenarios can be met. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A flow chart of an LED intelligent adjustment method based on full spectrum analysis of the present invention is shown;

[0057] Figure 2 A block diagram of an LED intelligent adjustment system based on full spectrum analysis of the present invention is shown. DETAILED DESCRIPTION

[0058] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0059] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0060] Figure 1 The flowchart of the LED intelligent adjustment method based on full spectrum analysis of the present invention is shown.

[0061] like Figure 1 As shown, the first aspect of the present invention provides an LED intelligent adjustment method based on full spectrum analysis, comprising:

[0062] S102, performing a spectral power distribution test on a plurality of preset single-primary-color LED lights to obtain a corresponding measured spectral power distribution curve;

[0063] S104, extracting the peak wavelength and half-width by measuring the spectral power distribution curve, using the peak wavelength and half-width as parameters, performing spectral fitting through a Gaussian function, and obtaining a fitting spectral power distribution curve corresponding to each single-primary color LED light;

[0064] S106, combining multiple single-primary-color LED lights to form N LED groups, performing spectral fitting on multiple fitted spectral power distribution curves in each LED group to obtain a synthetic spectrum, performing spectral matching analysis on the synthetic spectrum based on the spectral mismatch error of the least square method, and screening out qualified LED groups, marking them as preferred LED groups;

[0065] S108, performing actual LED lighting application by optimizing the LED group, obtaining an original image and a lighting feedback image, dividing the original image into multiple image regions based on the size of the original image, taking the image region as an analysis unit, performing a difference analysis of preset image attributes in the original image and the lighting feedback image, and screening out multiple comparison image regions according to the preset attribute range;

[0066] S110, based on multiple contrast image regions, respectively extracting color features from the illumination feedback image and the preset standard illumination image to obtain a first color feature set and a second color feature set, and based on a DCA feature fusion algorithm, fusing features in the first color feature set to form a first fused feature, and fusing features in the second color feature set to form a second fused feature;

[0067] S112, introduce the standard Euclidean distance method, vectorize the first fusion feature and the second fusion feature and calculate the difference between the data, and evaluate and adjust the lighting effect of the preferred LED group based on the difference.

[0068] According to an embodiment of the present invention, the peak wavelength and half-width are extracted by measuring the spectral power distribution curve, and the peak wavelength and half-width are used as parameters to perform spectral fitting through a Gaussian function to obtain a fitting spectral power distribution curve corresponding to each single-primary color LED light, specifically:

[0069] For a measured spectral power distribution curve, the peak wavelength and half-height width are extracted;

[0070] The peak wavelength and half-height width are used as parameters to fit the full spectrum through a preset modified Gaussian function;

[0071] The measured spectral power distribution curves corresponding to the single-primary-color LED lights are all fitted to obtain the fitted spectral power distribution curve corresponding to each single-primary-color LED light;

[0072] The Gaussian function is as follows:

[0073]

[0074] Where S(λ) is the relative spectral power distribution of the LED, λ 0 is the peak wavelength, and Δλ is the full width at half maximum.

[0075] It should be noted that in the spectral power distribution curve, the highest point of the curve, that is, the position where the radiation intensity of the light source is the largest, corresponds to the existence of a wavelength, which is called the peak wavelength. The energy level position of the luminous center is generally the position of the peak wavelength, and different peak wavelengths represent the color of light emitted by different LEDs. The half-height full width is also called the half-height width, which is specifically the difference between the two wavelengths at the half-peak value of a peak on the relative spectral power distribution curve.

[0076] According to an embodiment of the present invention, a plurality of single-primary-color LED lights are combined to form N LED groups. For each LED group, a plurality of fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum. A spectral matching analysis is performed on the synthetic spectrum based on the spectral mismatch error of the least squares method, and a qualified LED group is screened out and marked as a preferred LED group, specifically:

[0077] Among multiple single-primary-color LED lights, analyze the types and quantities of the current primary colors, select one of each primary color to combine to obtain a light group simulating a full spectrum, which is recorded as N LED groups;

[0078] For each LED group, multiple fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum;

[0079] The spectrum mismatch error based on the least square method is used to perform spectrum matching analysis on the synthetic spectrum and obtain the total spectrum mismatch evaluation index;

[0080] By presetting the evaluation index range, qualified LED groups are screened out and marked as preferred LED groups;

[0081] The fitting expression of the synthetic spectrum is as follows:

[0082]

[0083] Among them, S m (λ) is the spectral power distribution after fitting, S i is the spectral power distribution of light i, D i is the duty cycle of light i, λ is the wavelength of the independent variable;

[0084] The spectrum mismatch error based on the least squares method and the corresponding total spectrum mismatch evaluation index are specifically calculated as follows:

[0085]

[0086] Where ε is the total spectral mismatch evaluation index, S T (λ) is the target spectral power distribution; S i (λ) is the spectral power distribution of the i-th light, n is the total number of lights; D i is the duty cycle of light i.

[0087] It should be noted that the multiple single-primary color LED lights generally include three-primary color (RGB) LED lights, and the number of each primary color is not fixed. Therefore, multiple LED light groups can be formed by any combination, and each group is measured to select the group that meets the requirements for lighting application. The synthetic spectrum is a mixed spectrum, which is represented by the corresponding spectral power distribution and is obtained by fitting using the above method. In the spectral mismatch error calculation formula, its specific role is to evaluate the similarity between the synthetic spectrum and the target spectrum (i.e., the full spectrum). Among them, the smaller the ε value, the higher the similarity of the two spectra and the lower the spectral mismatch. S i is the spectral power distribution of light i, and according to the embodiment of the present invention, the spectral power distribution of light i can be calculated based on a spectral power distribution curve formed by Gaussian function fitting. Preferably, the LED group includes a plurality of groups.

[0088] According to an embodiment of the present invention, the actual LED lighting application is performed by the preferred LED group to obtain the original image and the lighting feedback image, and the original image is divided into multiple image areas based on the size. The image area is used as the analysis unit, and the difference analysis of preset image attributes is performed in the original image and the lighting feedback image. Through the preset attribute range, multiple comparison image areas are screened out, specifically:

[0089] Select a preferred LED group for analysis;

[0090] In an actual lighting scene, an original image of a target scene is obtained by a preset camera device, actual LED lighting application is performed by using a preferred LED group, and a lighting feedback image of the target scene is obtained;

[0091] Grid division is performed based on the size of the original image to form multiple image regions;

[0092] Taking an image area as the analysis unit, based on the three attribute dimensions of brightness, contrast, and color saturation, image difference analysis is performed between the original image and the illumination feedback image. If the differences in the three attribute dimensions all exceed the preset attribute range, the image area is marked as a comparison image area;

[0093] All image regions are analyzed and screened to obtain multiple comparison image regions.

[0094] It should be noted that the original image and the illumination feedback image are images of the same scene (location) and have the same image size. The preset attribute range includes a brightness difference threshold, a contrast difference threshold, and a color saturation difference threshold.

[0095] According to an embodiment of the present invention, based on multiple contrast image areas, color features are extracted from the illumination feedback image and the preset standard illumination image respectively to obtain a first color feature set and a second color feature set, and based on the DCA feature fusion algorithm, the features in the first color feature set are fused to form a first fused feature, and the features in the second color feature set are fused to form a second fused feature, specifically:

[0096] Using a plurality of contrast image regions as feature extraction regions, extracting color features from the illumination feedback image, and integrating the feature data to form a first color feature set;

[0097] Using a plurality of contrasting image regions as feature extraction regions, extracting color features from a preset standard illumination image, and integrating the feature data to form a second color feature set;

[0098] In the process of color feature extraction, feature extraction is performed by calculating the color histogram of the corresponding image area;

[0099] In the first color feature set and the second color feature set, a DCA feature fusion algorithm is used to generate a category activation map for each feature data in the color feature set and feature fusion is performed based on preset weights to obtain a first fused feature and a second fused feature, respectively.

[0100] It should be noted that the color feature set includes feature data for each contrasting area. The first fusion feature is obtained based on the analysis of the first color feature set, and the second fusion feature is obtained based on the analysis of the second color feature set, and both are analyzed based on the same DCA feature fusion algorithm. The preset standard illumination image is an idealized effect image based on full-spectrum lighting in the corresponding scene, which can be obtained by user setting or idealized model image simulation for feature comparison. The color histogram includes the frequency values ​​of different colors appearing in the pixel area, thereby forming color feature data.

[0101] According to an embodiment of the present invention, the standard Euclidean distance method is introduced to vectorize the first fusion feature and the second fusion feature and calculate the difference between the data, and the lighting effect of the preferred LED group is evaluated and adjusted based on the difference, specifically:

[0102] Performing feature vectorization on the first fused feature and the second fused feature to obtain a first feature vector and a second feature vector;

[0103] The distance between the first eigenvector and the second eigenvector is calculated by the standard Euclidean distance method, and the difference between the data is evaluated based on the distance value;

[0104] Through difference analysis, the lighting effect of the preferred LED group is evaluated;

[0105] Perform lighting effect evaluation on all preferred LED groups and generate lighting effect evaluation results;

[0106] Based on the lighting effect evaluation results, different preferred LED groups are put into practical application and the lighting parameters of the LED groups are adjusted.

[0107] It should be noted that DCA feature fusion is a feature fusion algorithm that uses a deep learning network. It generates corresponding category activation maps by analyzing feature data for weighted fusion, effectively highlighting the key information of the features, reducing data complexity, and improving recognition and analysis efficiency. The smaller the distance value, the greater the similarity and the smaller the difference. The distance value can effectively evaluate the difference in data values, and evaluate the lighting effect level of the light group based on the distance value. The higher the level, the better the effect and the smaller the distance value.

[0108] It is worth mentioning that in the prior art, most of the traditional full-spectrum lighting effect evaluation is based on human eye observation or simple numerical analysis based on brightness and contrast, and the evaluation effect is poor. The lighting evaluation effect in multiple scenes and complex lighting environments is quite different, and it is not standardized and practical. Based on this, the present invention compares the actual image with the standard lighting image. In the process, the original image and the light feedback image are regionalized and compared, and the areas with high distinction are screened out. These areas are used for subsequent feature fusion analysis based on DCA. By comparing the feature fusion, the accuracy and analysis efficiency of the light effect analysis are greatly improved. The fusion process can improve the system's understanding and judgment of key lighting features. At the same time, the full-spectrum effect can be accurately evaluated in the form of color features, and the evaluation accuracy of the full-spectrum LED lighting effect can be improved. Fast and accurate full-spectrum effect evaluation is achieved in multiple scenes and complex lighting environments. At the same time, through the evaluation results, the lighting parameters in the LED group can be dynamically adjusted to achieve better lighting effects and meet the lighting needs in multiple scenes.

[0109] In the present invention, LED groups are screened by fitting calculations of spectral curves and analyzed in theoretical form. Then, lighting effect evaluation is performed for practical applications by integrating features and analyzed with actual application images to achieve screening of optimal lighting groups and adjustment of poor lighting groups. The present invention achieves precise matching and selection of full-spectrum requirements for preset LED lights through dual analysis of theory and practice, realizes precise evaluation of lighting effects, reduces human intervention, and improves the objectivity of lighting effect analysis.

[0110] According to an embodiment of the present invention, it also includes:

[0111] Acquire multiple original images under multiple scenes;

[0112] Taking an original image as the analysis unit, color features of the original image are extracted based on multiple contrast image regions to obtain multiple environmental color features;

[0113] Characteristic vectorization is performed on a plurality of environmental color features to obtain an environmental color feature vector;

[0114] A DBSCAN clustering model was constructed, with the environmental color feature vector as the clustering sample data and the original image as the clustering unit. During the clustering process, the similarity between the data was analyzed by the standard Euclidean distance between the environmental color feature vectors, and multiple clustering groups were finally generated.

[0115] Each cluster group includes one or more original images, and multiple scenes are divided according to the scenes corresponding to the original images in the cluster group to obtain multiple scene groups;

[0116] Select a scene group, select a scene in the scene group to perform full-spectrum LED light matching, and obtain a corresponding optimal LED group;

[0117] The environmental color feature vectors corresponding to all original images in a scene group are integrated to form an environmental feature set;

[0118] Establish a mapping relationship between the environmental feature set and the optimal LED group and generate an environmental feature and light group association record;

[0119] Perform corresponding correlation analysis on multiple scenes to form multiple environmental characteristics and light group correlation records, and import all environmental characteristics and light group correlation records into the system cache database;

[0120] In actual lighting applications, it is determined whether a new scene is generated. If so, a corresponding environmental color feature vector is obtained based on the new scene and marked as a real-time feature vector;

[0121] By associating environmental features with light groups, the real-time feature vector is evaluated for similarity with the environmental feature set in each data record. Based on the similarity, the highly similar data record results are screened out and the corresponding optimal LED group is obtained.

[0122] The optimal LED group obtained by screening is applied to the new scene.

[0123] It should be noted that each of the contrast image areas corresponds to an environmental color feature. Each scene corresponds to an original image. It is worth mentioning that in different scenes, there are certain differences and similarities in environmental color features. The present invention clusters the environmental color features in different scenes to achieve clustering and grouping analysis of similar scenes, and selects a group of optimal LED groups for application based on the clustering group. In actual lighting applications, by importing associated records, the real-time environmental features in the new scene are compared with the associated records, and the LED group suitable for the current scene can be quickly screened out, thereby improving the adaptability and practicality of multi-scene and real-time lighting applications, reducing repetitive scene analysis and LED matching analysis processes, and improving the efficiency of actual lighting applications.

[0124] Figure 2 A block diagram of an LED intelligent adjustment system based on full spectrum analysis of the present invention is shown.

[0125] The second aspect of the present invention further provides an LED intelligent adjustment system 2 based on full spectrum analysis, the system comprising: a memory 21, a processor 22, the memory comprising an LED intelligent adjustment program based on full spectrum analysis, the LED intelligent adjustment program based on full spectrum analysis being executed by the processor to implement the following steps:

[0126] Perform spectral power distribution tests on a plurality of preset single-primary-color LED lights to obtain corresponding measured spectral power distribution curves;

[0127] The peak wavelength and half-width are extracted by measuring the spectral power distribution curve. The peak wavelength and half-width are used as parameters, and the spectrum is fitted by Gaussian function to obtain the fitting spectral power distribution curve corresponding to each single-primary color LED light.

[0128] Combine multiple single-primary-color LED lights to form N LED groups. For each LED group, perform spectral fitting on multiple fitted spectral power distribution curves in the group to obtain a synthetic spectrum. Perform spectral matching analysis on the synthetic spectrum based on the spectral mismatch error of the least squares method, and screen out qualified LED groups, marking them as preferred LED groups.

[0129] By optimizing the LED group to perform actual LED lighting application, an original image and a lighting feedback image are obtained, and the original image is divided into multiple image regions based on the size. The image region is used as an analysis unit, and a difference analysis of preset image attributes is performed in the original image and the lighting feedback image. Through the preset attribute range, multiple comparison image regions are screened out;

[0130] Based on multiple contrast image regions, color features are extracted from the illumination feedback image and the preset standard illumination image respectively to obtain a first color feature set and a second color feature set. Based on the DCA feature fusion algorithm, features in the first color feature set are fused to form a first fused feature, and features in the second color feature set are fused to form a second fused feature.

[0131] The standard Euclidean distance method is introduced to vectorize the first fusion feature and the second fusion feature and calculate the difference between the data. The lighting effect of the preferred LED group is evaluated and adjusted based on the difference.

[0132] According to an embodiment of the present invention, the peak wavelength and half-width are extracted by measuring the spectral power distribution curve, and the peak wavelength and half-width are used as parameters to perform spectral fitting through a Gaussian function to obtain a fitting spectral power distribution curve corresponding to each single-primary color LED light, specifically:

[0133] For a measured spectral power distribution curve, the peak wavelength and half-height width are extracted;

[0134] The peak wavelength and half-height width are used as parameters to fit the full spectrum through a preset modified Gaussian function;

[0135] The measured spectral power distribution curves corresponding to the single-primary-color LED lights are all fitted to obtain the fitted spectral power distribution curve corresponding to each single-primary-color LED light;

[0136] The Gaussian function is as follows:

[0137]

[0138] Where S(λ) is the relative spectral power distribution of the LED, λ 0 is the peak wavelength, and Δλ is the full width at half maximum.

[0139] It should be noted that in the spectral power distribution curve, the highest point of the curve, that is, the position where the radiation intensity of the light source is the largest, corresponds to the existence of a wavelength, which is called the peak wavelength. The energy level position of the luminous center is generally the position of the peak wavelength, and different peak wavelengths represent the color of light emitted by different LEDs. The half-height full width is also called the half-height width, which is specifically the difference between the two wavelengths at the half-peak value of a peak on the relative spectral power distribution curve.

[0140] According to an embodiment of the present invention, a plurality of single-primary-color LED lights are combined to form N LED groups. For each LED group, a plurality of fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum. A spectral matching analysis is performed on the synthetic spectrum based on the spectral mismatch error of the least squares method, and a qualified LED group is screened out and marked as a preferred LED group, specifically:

[0141] Among multiple single-primary-color LED lights, analyze the types and quantities of the current primary colors, select one of each primary color to combine to obtain a light group simulating a full spectrum, which is recorded as N LED groups;

[0142] For each LED group, multiple fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum;

[0143] The spectrum mismatch error based on the least square method is used to perform spectrum matching analysis on the synthetic spectrum and obtain the total spectrum mismatch evaluation index;

[0144] By presetting the evaluation index range, qualified LED groups are screened out and marked as preferred LED groups;

[0145] The fitting expression of the synthetic spectrum is as follows:

[0146]

[0147] Among them, S m (λ) is the spectral power distribution after fitting, S i is the spectral power distribution of light i, D i is the duty cycle of light i, λ is the wavelength of the independent variable;

[0148] The spectrum mismatch error based on the least squares method and the corresponding total spectrum mismatch evaluation index are specifically calculated as follows:

[0149]

[0150] Where ε is the total spectral mismatch evaluation index, S T (λ) is the target spectral power distribution; S i (λ) is the spectral power distribution of the i-th light, n is the total number of lights; D i is the duty cycle of light i.

[0151] It should be noted that the multiple single-primary color LED lights generally include three-primary color (RGB) LED lights, and the number of each primary color is not fixed. Therefore, multiple LED light groups can be formed by any combination, and each group is measured to select the group that meets the requirements for lighting application. The synthetic spectrum is a mixed spectrum, which is represented by the corresponding spectral power distribution and is obtained by fitting using the above method. In the spectral mismatch error calculation formula, its specific role is to evaluate the similarity between the synthetic spectrum and the target spectrum (i.e., the full spectrum). Among them, the smaller the ε value, the higher the similarity of the two spectra and the lower the spectral mismatch. S iis the spectral power distribution of light i, and according to the embodiment of the present invention, the spectral power distribution of light i can be calculated based on a spectral power distribution curve formed by Gaussian function fitting. Preferably, the LED group includes a plurality of groups.

[0152] According to an embodiment of the present invention, the actual LED lighting application is performed by the preferred LED group to obtain the original image and the lighting feedback image, and the original image is divided into multiple image areas based on the size. The image area is used as the analysis unit, and the difference analysis of preset image attributes is performed in the original image and the lighting feedback image. Through the preset attribute range, multiple comparison image areas are screened out, specifically:

[0153] Select a preferred LED group for analysis;

[0154] In an actual lighting scene, an original image of a target scene is obtained by a preset camera device, actual LED lighting application is performed by using a preferred LED group, and a lighting feedback image of the target scene is obtained;

[0155] Grid division is performed based on the size of the original image to form multiple image regions;

[0156] Taking an image area as the analysis unit, based on the three attribute dimensions of brightness, contrast, and color saturation, image difference analysis is performed between the original image and the illumination feedback image. If the differences in the three attribute dimensions all exceed the preset attribute range, the image area is marked as a comparison image area;

[0157] All image regions are analyzed and screened to obtain multiple comparison image regions.

[0158] It should be noted that the original image and the illumination feedback image are images of the same scene (location) and have the same image size. The preset attribute range includes a brightness difference threshold, a contrast difference threshold, and a color saturation difference threshold.

[0159] According to an embodiment of the present invention, based on multiple contrast image areas, color features are extracted from the illumination feedback image and the preset standard illumination image respectively to obtain a first color feature set and a second color feature set, and based on the DCA feature fusion algorithm, the features in the first color feature set are fused to form a first fused feature, and the features in the second color feature set are fused to form a second fused feature, specifically:

[0160] Using a plurality of contrast image regions as feature extraction regions, extracting color features from the illumination feedback image, and integrating the feature data to form a first color feature set;

[0161] Using a plurality of contrasting image regions as feature extraction regions, extracting color features from a preset standard illumination image, and integrating the feature data to form a second color feature set;

[0162] In the process of color feature extraction, feature extraction is performed by calculating the color histogram of the corresponding image area;

[0163] In the first color feature set and the second color feature set, a DCA feature fusion algorithm is used to generate a category activation map for each feature data in the color feature set and feature fusion is performed based on preset weights to obtain a first fused feature and a second fused feature, respectively.

[0164] It should be noted that the color feature set includes feature data for each contrasting area. The first fusion feature is obtained based on the analysis of the first color feature set, and the second fusion feature is obtained based on the analysis of the second color feature set, and both are analyzed based on the same DCA feature fusion algorithm. The preset standard illumination image is an idealized effect image based on full-spectrum lighting in the corresponding scene, which can be obtained by user setting or idealized model image simulation for feature comparison. The color histogram includes the frequency values ​​of different colors appearing in the pixel area, thereby forming color feature data.

[0165] According to an embodiment of the present invention, the standard Euclidean distance method is introduced to vectorize the first fusion feature and the second fusion feature and calculate the difference between the data, and the lighting effect of the preferred LED group is evaluated and adjusted based on the difference, specifically:

[0166] Performing feature vectorization on the first fused feature and the second fused feature to obtain a first feature vector and a second feature vector;

[0167] The distance between the first eigenvector and the second eigenvector is calculated by the standard Euclidean distance method, and the difference between the data is evaluated based on the distance value;

[0168] Through difference analysis, the lighting effect of the preferred LED group is evaluated;

[0169] Perform lighting effect evaluation on all preferred LED groups and generate lighting effect evaluation results;

[0170] Based on the lighting effect evaluation results, different preferred LED groups are put into practical application and the lighting parameters of the LED groups are adjusted.

[0171] It should be noted that DCA feature fusion is a feature fusion algorithm that uses a deep learning network. It generates corresponding category activation maps by analyzing feature data for weighted fusion, effectively highlighting the key information of the features, reducing data complexity, and improving recognition and analysis efficiency. The smaller the distance value, the greater the similarity and the smaller the difference. The distance value can effectively evaluate the difference in data values, and evaluate the lighting effect level of the light group based on the distance value. The higher the level, the better the effect and the smaller the distance value.

[0172] It is worth mentioning that in the prior art, most of the traditional full-spectrum lighting effect evaluation is based on human eye observation or simple numerical analysis based on brightness and contrast, and the evaluation effect is poor. The lighting evaluation effect in multiple scenes and complex lighting environments is quite different, and it is not standardized and practical. Based on this, the present invention compares the actual image with the standard lighting image. In the process, the original image and the light feedback image are regionalized and compared, and the areas with high distinction are screened out. These areas are used for subsequent feature fusion analysis based on DCA. By comparing the feature fusion, the accuracy and analysis efficiency of the light effect analysis are greatly improved. The fusion process can improve the system's understanding and judgment of key lighting features. At the same time, the full-spectrum effect can be accurately evaluated in the form of color features, and the evaluation accuracy of the full-spectrum LED lighting effect can be improved. Fast and accurate full-spectrum effect evaluation is achieved in multiple scenes and complex lighting environments. At the same time, through the evaluation results, the lighting parameters in the LED group can be dynamically adjusted to achieve better lighting effects and meet the lighting needs in multiple scenes.

[0173] In the present invention, LED groups are screened by fitting calculations of spectral curves and analyzed in theoretical form. Then, lighting effect evaluation is performed for practical applications by integrating features and analyzed with actual application images to achieve screening of optimal lighting groups and adjustment of poor lighting groups. The present invention achieves precise matching and selection of full-spectrum requirements for preset LED lights through dual analysis of theory and practice, realizes precise evaluation of lighting effects, reduces human intervention, and improves the objectivity of lighting effect analysis.

[0174] The third aspect of the present invention also provides a computer-readable storage medium, which includes an LED intelligent adjustment program based on full spectrum analysis. When the LED intelligent adjustment program based on full spectrum analysis is executed by a processor, the steps of the LED intelligent adjustment method based on full spectrum analysis as described in any one of the above items are implemented.

[0175] The present invention discloses an LED intelligent adjustment method and system based on full spectrum analysis. By performing spectrum analysis on multiple LED lights, peak wavelength and half-width parameters are extracted to perform spectrum fitting based on Gaussian function, and qualified LED groups are screened out; the LED groups are used for actual lighting application, the original image and the lighting feedback image are obtained, the regions are divided according to the image size, the image attribute differences of different regions are compared, the contrast image regions are screened out, color features are extracted based on the contrast image regions, and the first and second color feature sets are formed. The feature sets are fused using the DCA feature fusion algorithm to form the first and second fused features; the difference between the first and second fused features is calculated by the standard Euclidean distance method, and the lighting effect of the LED group is evaluated and adjusted accordingly. Through the present invention, accurate selection and application of LED lights based on the full spectrum can be achieved, accurate evaluation of the full spectrum LED lighting effect can be achieved, and lighting requirements in multiple scenarios can be met.

[0176] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0177] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0178] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0179] Those skilled in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), disks or optical disks, and other media that can store program codes.

[0180] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0181] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An LED intelligent adjustment method based on full spectrum analysis, characterized in that: include: Perform spectral power distribution tests on a plurality of preset single-primary-color LED lights to obtain corresponding measured spectral power distribution curves; The peak wavelength and half-width are extracted by measuring the spectral power distribution curve. The peak wavelength and half-width are used as parameters, and the spectrum is fitted by Gaussian function to obtain the fitting spectral power distribution curve corresponding to each single-primary color LED light. Combine multiple single-primary-color LED lights to form N LED groups. For each LED group, perform spectral fitting on multiple fitted spectral power distribution curves in the group to obtain a synthetic spectrum. Perform spectral matching analysis on the synthetic spectrum based on the spectral mismatch error of the least squares method, and screen out qualified LED groups, marking them as preferred LED groups. By optimizing the LED group to perform actual LED lighting application, an original image and a lighting feedback image are obtained, and the original image is divided into multiple image regions based on the size. The image region is used as an analysis unit, and a difference analysis of preset image attributes is performed in the original image and the lighting feedback image. Through the preset attribute range, multiple comparison image regions are screened out; Based on multiple contrast image regions, color features are extracted from the illumination feedback image and the preset standard illumination image respectively to obtain a first color feature set and a second color feature set. Based on the DCA feature fusion algorithm, features in the first color feature set are fused to form a first fused feature, and features in the second color feature set are fused to form a second fused feature. The standard Euclidean distance method is introduced to vectorize the first fusion feature and the second fusion feature and calculate the difference between the data. The lighting effect of the preferred LED group is evaluated and adjusted based on the difference.

2. The LED intelligent adjustment method based on full spectrum analysis according to claim 1 is characterized in that: The peak wavelength and half-width are extracted by measuring the spectral power distribution curve, and the peak wavelength and half-width are used as parameters to perform spectrum fitting through a Gaussian function to obtain the fitting spectral power distribution curve corresponding to each single-primary color LED light, specifically: For a measured spectral power distribution curve, the peak wavelength and half-height width are extracted; The peak wavelength and half-height width are used as parameters to fit the full spectrum through a preset modified Gaussian function; The measured spectral power distribution curves corresponding to the single-primary-color LED lights are all fitted to obtain the fitted spectral power distribution curve corresponding to each single-primary-color LED light; The Gaussian function is as follows: Where S(λ) is the relative spectral power distribution of the LED, λ0 is the peak wavelength, and Δλ is the full width at half maximum.

3. The LED intelligent adjustment method based on full spectrum analysis according to claim 2 is characterized in that: The plurality of single-primary-color LED lights are combined to form N LED groups. For each LED group, a plurality of fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum. The synthetic spectrum is subjected to spectral matching analysis based on the spectral mismatch error of the least square method, and a qualified LED group is screened out and marked as a preferred LED group, specifically: Among multiple single-primary-color LED lights, analyze the types and quantities of the current primary colors, select one of each primary color and combine them to obtain a light group simulating a full spectrum, which is recorded as N LED groups; For each LED group, multiple fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum; The spectrum mismatch error based on the least square method is used to perform spectrum matching analysis on the synthetic spectrum and obtain the total spectrum mismatch evaluation index; By presetting the evaluation index range, qualified LED groups are screened out and marked as preferred LED groups; The fitting expression of the synthetic spectrum is as follows: Among them, S m (λ) is the spectral power distribution after fitting, S i is the spectral power distribution of light i, D i is the duty cycle of light i, λ is the wavelength of the independent variable; The spectrum mismatch error based on the least squares method and the corresponding total spectrum mismatch evaluation index are specifically calculated as follows: Among them, ε is the total spectrum mismatch evaluation index, S T (λ) is the target spectral power distribution; S i (λ) is the spectral power distribution of the i-th light, n is the total number of lights; D i is the duty cycle of light i.

4. The LED intelligent adjustment method based on full spectrum analysis according to claim 3 is characterized in that: The actual LED lighting application is performed by the preferred LED group to obtain the original image and the lighting feedback image, and the original image is divided into multiple image areas based on the size. The image area is used as the analysis unit, and the difference analysis of the preset image attributes is performed in the original image and the lighting feedback image. Through the preset attribute range, multiple comparison image areas are screened out, specifically: Select a preferred LED group for analysis; In an actual lighting scene, an original image of a target scene is obtained by a preset camera device, actual LED lighting application is performed by using a preferred LED group, and a lighting feedback image of the target scene is obtained; Grid division is performed based on the size of the original image to form multiple image regions; Taking an image area as the analysis unit, based on the three attribute dimensions of brightness, contrast, and color saturation, image difference analysis is performed between the original image and the illumination feedback image. If the differences in the three attribute dimensions all exceed the preset attribute range, the image area is marked as a comparison image area; All image regions are analyzed and screened to obtain multiple comparison image regions.

5. The LED intelligent adjustment method based on full spectrum analysis according to claim 4 is characterized in that: The method extracts color features from the illumination feedback image and the preset standard illumination image based on multiple contrast image regions, obtains a first color feature set and a second color feature set, and fuses the features in the first color feature set to form a first fused feature, and fuses the features in the second color feature set to form a second fused feature based on a DCA feature fusion algorithm, specifically: Using a plurality of contrast image regions as feature extraction regions, extracting color features from the illumination feedback image, and integrating the feature data to form a first color feature set; Using a plurality of contrasting image regions as feature extraction regions, extracting color features from a preset standard illumination image, and integrating the feature data to form a second color feature set; In the process of color feature extraction, feature extraction is performed by calculating the color histogram of the corresponding image area; In the first color feature set and the second color feature set, a DCA feature fusion algorithm is used to generate a category activation map for each feature data in the color feature set and feature fusion is performed based on preset weights to obtain a first fused feature and a second fused feature, respectively.

6. The LED intelligent adjustment method based on full spectrum analysis according to claim 5, characterized in that: The standard Euclidean distance method is introduced to vectorize the first fusion feature and the second fusion feature and calculate the difference between the data, and the lighting effect of the preferred LED group is evaluated and adjusted based on the difference, specifically: Performing feature vectorization on the first fused feature and the second fused feature to obtain a first feature vector and a second feature vector; The distance between the first eigenvector and the second eigenvector is calculated by the standard Euclidean distance method, and the difference between the data is evaluated based on the distance value; Through difference analysis, the lighting effect of the preferred LED group is evaluated; Perform lighting effect evaluation on all preferred LED groups and generate lighting effect evaluation results; Based on the lighting effect evaluation results, different preferred LED groups are put into practical application and the lighting parameters of the LED groups are adjusted.

7. An LED intelligent adjustment system based on full spectrum analysis, characterized in that: The system includes: a memory and a processor, wherein the memory includes an LED intelligent adjustment program based on full spectrum analysis, and when the LED intelligent adjustment program based on full spectrum analysis is executed by the processor, the following steps are implemented: Perform spectral power distribution tests on a plurality of preset single-primary-color LED lights to obtain corresponding measured spectral power distribution curves; The peak wavelength and half-width are extracted by measuring the spectral power distribution curve. The peak wavelength and half-width are used as parameters, and the spectrum is fitted by Gaussian function to obtain the fitting spectral power distribution curve corresponding to each single-primary color LED light. Combine multiple single-primary-color LED lights to form N LED groups. For each LED group, perform spectral fitting on multiple fitted spectral power distribution curves in the group to obtain a synthetic spectrum. Perform spectral matching analysis on the synthetic spectrum based on the spectral mismatch error of the least squares method, and screen out qualified LED groups, marking them as preferred LED groups. By optimizing the LED group to perform actual LED lighting application, an original image and a lighting feedback image are obtained, and the original image is divided into multiple image regions based on the size. The image region is used as an analysis unit, and a difference analysis of preset image attributes is performed in the original image and the lighting feedback image. Through the preset attribute range, multiple comparison image regions are screened out; Based on multiple contrast image regions, color features are extracted from the illumination feedback image and the preset standard illumination image respectively to obtain a first color feature set and a second color feature set. Based on the DCA feature fusion algorithm, features in the first color feature set are fused to form a first fused feature, and features in the second color feature set are fused to form a second fused feature. The standard Euclidean distance method is introduced to vectorize the first fusion feature and the second fusion feature and calculate the difference between the data. The lighting effect of the preferred LED group is evaluated and adjusted based on the difference.

8. The LED intelligent adjustment system based on full spectrum analysis according to claim 7, characterized in that: The peak wavelength and half-width are extracted by measuring the spectral power distribution curve, and the peak wavelength and half-width are used as parameters to perform spectrum fitting through a Gaussian function to obtain the fitting spectral power distribution curve corresponding to each single-primary color LED light, specifically: For a measured spectral power distribution curve, the peak wavelength and half-height width are extracted; The peak wavelength and half-height width are used as parameters to fit the full spectrum through a preset modified Gaussian function; The measured spectral power distribution curves corresponding to the single-primary-color LED lights are all fitted to obtain the fitted spectral power distribution curve corresponding to each single-primary-color LED light; The Gaussian function is as follows: Where S(λ) is the relative spectral power distribution of the LED, λ0 is the peak wavelength, and Δλ is the full width at half maximum.

9. The LED intelligent adjustment system based on full spectrum analysis according to claim 7, characterized in that: The plurality of single-primary-color LED lights are combined to form N LED groups. For each LED group, a plurality of fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum. The synthetic spectrum is subjected to spectral matching analysis based on the spectral mismatch error of the least square method, and a qualified LED group is screened out and marked as a preferred LED group, specifically: Among multiple single-primary-color LED lights, analyze the types and quantities of the current primary colors, select one of each primary color and combine them to obtain a light group simulating a full spectrum, which is recorded as N LED groups; For each LED group, multiple fitted spectral power distribution curves in the group are spectrally fitted to obtain a synthetic spectrum; The spectrum mismatch error based on the least square method is used to perform spectrum matching analysis on the synthetic spectrum and obtain the total spectrum mismatch evaluation index; By presetting the evaluation index range, qualified LED groups are screened out and marked as preferred LED groups; The fitting expression of the synthetic spectrum is as follows: Among them, S m (λ) is the spectral power distribution after fitting, S i is the spectral power distribution of light i, D i is the duty cycle of light i, λ is the wavelength of the independent variable; The spectrum mismatch error based on the least squares method and the corresponding total spectrum mismatch evaluation index are specifically calculated as follows: ； Among them, ε is the total spectrum mismatch evaluation index, S T (λ) is the target spectral power distribution; S i (λ) is the spectral power distribution of the i-th light, n is the total number of lights; D i is the duty cycle of light i.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes an LED intelligent adjustment program based on full spectrum analysis. When the LED intelligent adjustment program based on full spectrum analysis is executed by a processor, the steps of the LED intelligent adjustment method based on full spectrum analysis as described in any one of claims 1 to 6 are implemented.

Images