Spectrum control method, device and system for LED plant lamps

Abstract

The present invention relates to a spectrum control method, device and system for LED plant lamps, aiming to improve the intelligent adjustment capability of the plant growth environment. First, plant image data is collected through a camera, and then the morphological data of the plant is extracted. These morphological data are compared with the preset model data to determine the growth state and stage of the plant. If the growth state of the plant fails to meet the preset expectations of the current growth stage, the spectrum requirements will be determined according to the specific growth conditions of the plant, and then a spectrum curve will be fitted according to the spectrum requirements. The curve represents the spectrum requirements of different bands, and finally a control instruction is generated to adjust the spectrum output of the LED plant lamp to match the fitted spectrum curve. The shortcomings of the existing LED plant lamps in spectrum adjustment are effectively solved, the adaptability and efficiency of the plant growth environment are enhanced, the dynamic optimization of spectrum settings is achieved, the utilization efficiency of light is improved, and the technical effect of promoting the healthy growth of plants is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a spectrum control method, device and system for an LED plant lamp. Background Art

[0002] With the development of modern agriculture, plant factories and greenhouse planting technologies have gradually become popular. LED plant lights have become an important tool for promoting plant growth due to their high efficiency, energy saving, long life, and adjustable spectrum. However, the spectrum design of traditional LED plant lights is mostly based on the needs of a certain growth stage of a specific plant. The spectrum adjustment process is complicated and usually requires manual setting. It cannot be automatically adjusted according to the growth conditions of the plant, which limits its wide application under diverse plant growth conditions. Due to the large differences in the light quality requirements of different plants at different growth stages, a single or fixed spectrum setting is difficult to meet the photosynthesis needs of a variety of plants, thereby affecting the overall growth efficiency of the plant.

[0003] In view of the light requirements of plants at different growth stages, such as germination, flowering, and fruiting, different combinations of spectra have different stimulating effects on plant physiological processes. Although existing LED plant lights have light sources of multiple bands, most products cannot dynamically adjust the spectrum output in real time according to the specific health status or growth cycle of the plants, resulting in low light utilization efficiency. In addition, the lack of effective control algorithms means that spectrum adjustment can only rely on experience or manual operation, making it difficult to achieve accurate and timely adjustments, especially in large-scale planting environments, and it is difficult to meet the needs of efficient agriculture for automation and intelligence.

[0004] Therefore, the key issue that needs to be solved in current research is how to automatically and accurately adjust the spectrum combination of LED plant lights according to the growth stage, health status and environmental conditions of the plants. This requires not only the system to monitor and analyze the specific needs of the plants, but also the establishment of a set of intelligent control methods to achieve automatic optimization of the spectrum to improve plant growth efficiency and save energy, thereby providing a more scientific and efficient lighting solution for modern agriculture. Summary of the invention

[0005] The main purpose of the present invention is to provide a spectrum control method, device and system for LED plant lights to solve the technical problem that although existing LED plant lights have light sources of multiple bands, most products cannot dynamically adjust the spectrum output in real time according to the specific health status or growth cycle of the plants, resulting in low light utilization efficiency.

[0006] To achieve the above-mentioned purpose, the present invention provides a spectrum control method for an LED plant lamp, comprising the following steps: collecting image data of a plant through a camera, and determining morphological data of the plant based on the image data of the plant; obtaining preset model data corresponding to the plant, and comparing the model data and the morphological data to obtain the growth status and growth stage of the plant; judging whether the growth status of the plant meets the preset expectations of the current growth stage of the plant, and if not, determining the spectrum requirement based on the growth status and growth stage of the plant; fitting a spectrum curve based on the spectrum requirement, and using the spectrum curve to control the LED plant lamp to irradiate the plant.

[0007] The present invention also provides a spectrum control device for an LED plant lamp, comprising: a determination unit, used to collect image data of a plant through a camera, and determine morphological data of the plant according to the image data of the plant; an acquisition unit, used to acquire preset model data corresponding to the plant, and compare the model data and the morphological data to obtain the growth state and growth stage of the plant; a judgment unit, used to judge whether the growth state of the plant meets the preset expectations of the current growth stage of the plant, and if not, determine the spectrum requirement according to the growth condition and growth stage of the plant; an irradiation unit, used to fit a spectrum curve according to the spectrum requirement, and use the spectrum curve to control the LED plant lamp to irradiate the plant.

[0008] The present invention also provides a spectrum control system for an LED plant lamp, comprising: an image recognition unit, comprising a camera, through which image data of the plant is collected, and morphological data of the plant is determined according to the image data of the plant; a core processor, which is in communication connection with the image recognition unit, and is used to receive the morphological data of the plant sent by the image recognition unit, and compare the morphological data according to preset model data corresponding to the plant, to obtain the growth state and growth stage of the plant, and to judge whether the growth state of the plant meets the preset expectation of the current growth stage of the plant; if not, the spectrum requirement is determined according to the growth state and growth stage of the plant, and then based on the growth state and growth stage of the plant, the spectrum requirement is determined. A spectral curve is fitted according to the spectral demand, and a control instruction for controlling the LED plant lamp to irradiate the plant based on the spectral curve is generated; a spectral control unit includes an LED plant lamp and is communicated with the core processor to receive the control instruction sent by the core processor, and controls the LED plant lamp to irradiate the plant based on the spectral curve according to the control instruction; a spectral detection unit includes a spectral detector and is communicated with the core processor to detect the spectral information of the actual irradiation of the plant by the LED plant lamp, and sends the spectral information to the core processor so that the core processor can adjust the control instruction based on the spectral information.

[0009] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the above methods when executing the computer program.

[0010] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned methods are implemented.

[0011] The spectrum control method, device and system of the LED plant lamp provided by the present invention collect the image data of the plant through a camera, and determine the morphological data of the plant based on the image data. This process involves capturing the appearance characteristics of the plant in detail in order to obtain comprehensive information about the health status of the plant. Next, the preset model data related to the plant is obtained, and the model data is compared with the actual morphological data. This comparison process is intended to evaluate the growth state and growth stage of the plant to ensure the accuracy and validity of the data. If it is found that the growth state of the plant fails to meet the preset expectations of the current growth stage, the specific spectrum requirements will be determined according to the growth status and stage of the plant. After determining the spectrum requirements, a spectrum curve is further fitted. This spectrum curve represents the spectrum requirements of different bands, and is intended to adjust the light source output to best support the growth of the plant. Finally, control instructions for controlling the LED plant lamp are generated according to the spectrum curve. These instructions will guide the LED plant lamp to adjust its spectrum output to meet the specific needs of the plant and optimize the growth environment of the plant.

[0012] This ensures that plants can obtain the most suitable lighting conditions at each growth stage, improves the efficiency of light utilization, and helps plants grow healthily. Through this intelligent spectrum control method, the present invention effectively solves the problem of insufficient spectrum adjustment in the prior art, thereby improving the application effect of LED plant lights. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a schematic diagram of the steps of a spectrum control method of an LED plant lamp in one embodiment of the present invention;

[0014] Figure 2 It is a structural block diagram of a spectrum control device for an LED plant lamp in one embodiment of the present invention;

[0015] Figure 3 It is a structural block diagram of a spectrum control system of an LED plant lamp in one embodiment of the present invention;

[0016] Figure 4 is a schematic diagram of a display page of an interactive display unit in an embodiment of the present invention Figure 1 ;

[0017] Figure 5 is a schematic diagram of a display page of an interactive display unit in an embodiment of the present invention Figure 2 ;

[0018] Figure 6 is a schematic diagram of a photosynthetic photon energy data table in one embodiment of the present invention;

[0019] Figure 7 is a principle block diagram of a core processing unit in one embodiment of the present invention;

[0020] Figure 8 is a block diagram of the internal structure of a core processing unit in one embodiment of the present invention;

[0021] Fig. 9 It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0022] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0024] Reference Figure 1 The present application provides a spectrum control method for an LED plant lamp, comprising the following steps:

[0025] S1, collecting image data of the plant through a camera, and determining the morphological data of the plant according to the image data of the plant.

[0026] S2, obtaining preset model data corresponding to the plant, and performing comparison processing on the model data and the morphological data to obtain the growth state and growth stage of the plant.

[0027] S3, determining whether the growth status of the plant meets the preset expectations of the current growth stage of the plant; if not, determining the spectrum requirement according to the growth status and growth stage of the plant.

[0028] S4, fitting a spectrum curve according to the spectrum requirement, and using the spectrum curve to control the LED plant lamp to irradiate the plant.

[0029] That is, step S1 involves using a camera to collect image data of the plant. These image data will be used to determine the morphological data of the plant. Specifically, the camera will capture multiple frames of images of the plant to obtain detailed images of the plant from different angles. These images are processed by Gaussian filtering and median filtering to remove noise in the image and obtain clearer image data. The processed images will be converted to the HSV color space, and the plant part will be highlighted by adjusting the color tone and filtering out irrelevant background. Next, a background model is constructed based on the dynamic background features, and areas similar to the background features are identified and removed to extract the contour image of the plant. Finally, the Canny edge detection algorithm is used to extract the boundaries of the plant and obtain detailed morphological data of the plant. These morphological data include information such as the structural characteristics, geometric data, and growth stage of the plant, which will be used for further analysis.

[0030] In step S2, the growth state and growth stage of the plant are determined by obtaining the preset model data corresponding to the plant and comparing it with the morphological data actually obtained. The preset model data is usually based on the standard morphological data of different plants at various growth stages, which provide reference standards for the normal growth of plants. By comparing the actual collected morphological data with these standards, the current growth state and growth stage of the plant can be identified. For example, if the preset model data of a certain plant indicates that it should have a specific leaf morphology and flower distribution in the flowering stage, and the actual data does not meet this standard, then it can be determined that the growth state of the plant does not meet expectations and further analysis is performed.

[0031] In step S3, the system needs to determine whether the growth status of the plant meets the preset expectations of its current growth stage. If it is found that the actual growth status of the plant does not meet expectations, this indicates that the current lighting conditions may not be suitable for the needs of the plant. Therefore, it is necessary to determine its spectral requirements based on the actual growth conditions and growth stage of the plant. This usually involves calculating the effects of different spectral bands on plant growth. For example, some bands may be more effective for plant photosynthesis. Through these calculations, the spectral combination that is most suitable for the current plant growth can be determined.

[0032] Finally, in step S4, a spectral curve is fitted according to the determined spectral requirements, and the spectral curve is used to control the irradiation of the LED plant lamp to the plants. Specifically, the spectra of different bands are normalized according to the spectral requirement data, and continuously fitted by cubic spline interpolation to obtain the initial spectral curve. Then, a Gaussian filter is used to remove high frequencies from the initial spectral curve to improve the stability and continuity of the spectral curve. The final spectral curve will be used to adjust the spectral output of the LED plant lamp to ensure that the plants obtain the most suitable light for their growth stage, thereby improving the growth efficiency of the plants.

[0033] For example, suppose a plant is in the flowering stage and its morphological data indicates that its flowers are not growing as expected. The system analyzes and finds that the intensity of certain bands in the spectral curve is insufficient, so the system generates a new spectral curve to increase the intensity of these bands to provide a more appropriate spectral output. This adjustment ensures that the plant can get enough light during its critical growth stage and optimizes the plant's growth environment.

[0034] In summary, the entire method realizes the whole process from image data acquisition, morphological data analysis, model comparison, spectral demand determination to spectral curve fitting and control of LED plant lights. This process not only improves the efficiency of plant growth, but also makes spectral control more intelligent and automated, adapting to the diverse needs of plants.

[0035] In one example, collecting image data of a plant through a camera includes: performing multi-frame image capture processing on the plant through a camera to obtain multiple frames of first images, wherein the multiple frames of first images are images obtained by multiple shots of the plant at different angles at the same time; using Gaussian filtering to perform weighted averaging on each pixel in each of the first images and its adjacent pixels in turn to smooth out fine noise points in the first image, and replacing salt and pepper noise points in the first image through median filtering to obtain multiple frames of second images; converting each frame of the second image into HSV space, and filtering out the noise points in the first image by adjusting the hue of the second image. For pixels in the background of the second image that do not match the color of the plant, the plant part in the second image is highlighted to obtain a third image; a background model is constructed through dynamic and static background features in which the plant does not appear in multiple frames of the first image, and each frame of the third image is compared with the background model one by one, and the areas in the third image that are similar to the dynamic and static background features are again identified and removed to obtain a fourth image; Canny edge detection is used to perform boundary segmentation processing on the areas in the fourth image with large changes in grayscale values, and the plant outline in the fourth image is extracted, and the image inside the plant outline is retained to obtain a fifth image of the plant.

[0036] The process of acquiring plant image data through a camera involves multiple complex steps, each of which plays a key role in different stages of image processing. This process not only ensures the quality of the data obtained from the camera, but also optimizes the recognition and analysis of plant morphology. First, the camera will perform multi-frame image capture processing on the plant. This means that at the same moment, the camera takes multiple shots of the plant from different angles to obtain multiple frames of the first image. In this way, the appearance information of the plant can be fully captured from multiple perspectives, improving the understanding and accuracy of the analysis of the overall morphology of the plant.

[0037] Next, the image data needs to be processed for noise to ensure the clarity and quality of the captured image. First, each pixel in each frame of the first image is weighted averaged with its neighboring pixels, that is, Gaussian filtering is used. This filtering method smoothes the small noise in the image by weighted averaging the pixel values, thereby improving the quality of the image. Next, salt and pepper noise is processed by median filtering. Salt and pepper noise is a common type of noise in image processing, which appears as bright or dark spots in the image. Median filtering effectively removes these noise points by replacing the value of the noise pixel with the median of its surrounding pixels, further improving the clarity of the image. The result of this process is multiple frames of second images, in which the noise is significantly reduced and the quality is improved.

[0038] Subsequently, the second image needs further processing to highlight the plant part. To achieve this, the second image is converted into the HSV color space. In the HSV space, by adjusting the hue of the image, the background pixels that do not match the color of the plant can be effectively filtered out, thereby separating the plant part from the background. The purpose of this processing method is to improve the visibility of the plant and make the plant part more prominent in the image. The image obtained in this way is called the third image, which can visually show the area of ​​the plant more clearly.

[0039] After obtaining the third image, the next step is to build a background model and remove the background of the image. A background model can be established through the background features in multiple frames of the first image, that is, the dynamic and static background features where the plant does not appear. This model is used to identify and remove areas in the third image that are similar to the background features, thereby obtaining the fourth image. The key to this processing method is to accurately distinguish the features of the plant from the background, thereby improving the extraction effect of the plant contour. In the fourth image, most of the background interference has been removed, leaving only the image data related to the plant.

[0040] Finally, the Canny edge detection algorithm is used to perform boundary segmentation processing on the areas with large grayscale value changes in the fourth image to extract the outline of the plant. Canny edge detection is a classic edge detection method that can accurately identify the changes in the edges of the image and extract the clear outline of the plant. In this way, the image inside the plant outline is retained, and the fifth image finally obtained presents the detailed outline information and structural data of the plant. This process not only improves the recognition accuracy of the plant outline, but also provides a reliable basis for subsequent morphological data analysis.

[0041] To illustrate with a practical example, suppose we are processing image data of a tomato plant. The camera takes images of the tomato plant from different angles to obtain multiple frames of the first image. After Gaussian filtering and median filtering, the noise in the image is removed to obtain a clearer second image. Then, through HSV space conversion and hue adjustment, the red area of ​​the tomato is highlighted from the background to obtain the third image. By constructing a background model and removing areas similar to the background, the fourth image is obtained, and finally the outline of the tomato is extracted through Canny edge detection to form the fifth image. This series of processing steps effectively improves the quality and accuracy of the image, making the analysis of the morphological data of the plant more accurate.

[0042] The advantage of this processing method is that it can systematically reduce image noise, improve image clarity, and accurately extract the outline of the plant through background removal and edge detection. Gaussian filtering and smoothing improve the visual quality of the image, median filtering removes noise, HSV space conversion makes the plant stand out from the background, and background model construction and edge detection ensure accurate recognition of plant morphology. These steps work together to ensure the high quality and reliability of the final plant image data, providing a solid foundation for subsequent spectral control and plant growth analysis.

[0043] In one example, determining the morphological data of the plant according to the image data of the plant includes: calculating the gradient of each pixel point in the X and Y directions of the fifth image of the plant, and using the calculated gradient data to construct an autocorrelation matrix of each pixel point; calculating the Harris response value of each pixel point according to the autocorrelation matrix of each pixel point, the Harris response value representing the corner point strength of each pixel point in the fifth image; and screening out a plurality of feature points from the fifth image according to a comparison result of the Harris response value and a preset threshold value, wherein the feature points can represent the most important corner points and corner points in the fifth image. The invention relates to a method for obtaining a plant structure information of a plant by using a method of converting pixel coordinates of the feature points into physical coordinates in an image coordinate system of the fifth image, and calculating the geometric data of the plant according to the physical coordinates; determining the position coordinates and structural information of the feature points, and dividing the feature points into a plurality of categories according to the position coordinates and the structural information, wherein the categories include at least leaves, flowers and fruits; converting the pixel coordinates of the feature points into physical coordinates in an image coordinate system of the fifth image, and calculating the geometric data of the plant according to the physical coordinates; determining the distribution information and arrangement information of the feature points of each category in the fifth image, and extracting the structural information of the plant according to the distribution information and arrangement information; and performing data set processing on the geometric data and structural data of the plant to obtain the morphological data of the plant.

[0044] That is, in the process of determining the morphological data of the plant, a series of complex and detailed image processing and analysis steps are involved, which are aimed at extracting the morphological characteristics of the plant from the fifth image. First, the gradients in the X and Y directions of each pixel are extracted from the fifth image. This step obtains the edge strength and direction information of each pixel by calculating the gradient of the image. The calculation of gradient is a basic operation in image processing, which helps to identify areas with significant changes in the image, usually corresponding to edges or corners in the image. After obtaining the gradient information, the next step is to use these gradient data to construct the autocorrelation matrix of each pixel. The autocorrelation matrix is ​​used to describe the local structural features around the pixel, including the strength and direction of edges and corners. By calculating the autocorrelation matrix of each pixel, the structural information of the pixel and its surrounding area can be more accurately analyzed, thus laying the foundation for further feature point extraction.

[0045] Next, the Harris response value of each pixel is calculated based on the autocorrelation matrix. Harris response value is a commonly used corner detection indicator used to measure the significance of corner points in an image. Corner points are usually key feature points in an image. They have large gradient changes in the image and are therefore of great significance in image processing. By calculating the Harris response value, the corner points in the fifth image can be identified and the strength and importance of these corner points can be understood. After obtaining the Harris response values, these response values ​​are compared with a preset threshold. This threshold is used to filter out important feature points in the image. Based on the comparison results, multiple feature points are filtered out from the fifth image, and these feature points can effectively represent the key corner points and structural information in the image. For example, in a plant image, feature points may include the tip of a leaf, the center of a flower, or the edge of a fruit.

[0046] After the feature points are identified, the next step is to classify the location coordinates of these feature points and the surrounding structural information. This process involves classifying the feature points into different categories, such as leaves, flowers, and fruits. By analyzing the location coordinates of the feature points and the structural information around them, different parts of the plant can be identified more accurately. For example, leaf feature points may be concentrated on the periphery of the plant, while flower feature points may be distributed in the central area. Once the feature points are classified, the next step is to convert the pixel coordinates of these feature points into physical coordinates in the image coordinate system. The physical coordinates can provide the location data of the feature points in the actual space, not just the pixel position in the image. Based on these physical coordinates, the geometric data of the plant, such as the size, shape, and distribution of each part, can be calculated.

[0047] By analyzing the distribution and arrangement information of feature points of each category in the image, the structural information of the plant can be extracted. This information helps to understand the growth status and structural characteristics of the plant, such as the density of leaves, the distribution pattern of flowers, and the number and location of fruits. Finally, the calculated geometric data and structural data are processed by data set processing to obtain the complete morphological data of the plant. Data set processing can integrate various types of data to provide a comprehensive description of the plant morphology. For example, by combining the geometric data of leaves, flowers, and fruits, the overall structure and morphological characteristics of the plant can be obtained, thereby providing valuable information for plant growth monitoring and management.

[0048] For example, if we perform morphological analysis on a tomato plant, we first calculate the gradient of each pixel in the image and construct an autocorrelation matrix using these gradient data. Subsequently, the corner features in the tomato plant, such as the edge of the fruit and the tip of the leaf, are identified using the Harris response value. By classifying these feature points, their specific categories (such as leaves, flowers, or fruits) are determined, and the pixel coordinates of these feature points are converted into physical coordinates, and finally the geometric data and structural information of the tomato plant are calculated. Through data set processing, the complete morphological data of the tomato plant can be obtained, including the distribution of leaves, the location of flowers, and the number of fruits, thereby providing accurate data support for plant growth status monitoring and further agricultural management. The advantage of this process is that it can comprehensively process the detailed information in the image and extract accurate plant morphological characteristics, thereby improving the scientificity and effectiveness of plant management.

[0049] It should be noted that the morphological data at least includes: leaf characteristic data, plant size data and flower and fruit characteristic data, the leaf characteristic data at least includes: the number of leaves, leaf size, leaf hole conditions, and yellowing area of ​​leaf edges, the plant size data at least includes: plant height and plant breast diameter, and the flower and fruit characteristic data at least includes: the number of flowers, flower size, number of fruits and fruit diameter.

[0050] In one example, the model data and the morphological data are compared and processed to obtain the growth stage of the plant, including: integrating the morphological data into a data set containing multiple morphological parameters according to a preset arrangement order; obtaining a preset standard morphological interval range set corresponding to different growth stages, and aligning the integrated morphological data set with the standard morphological interval range set; comparing each morphological parameter in the morphological data set with each standard in the standard morphological interval range set one by one, wherein the comparison process includes: determining whether each morphological parameter is within the corresponding standard interval range, if it falls within the standard interval range, it is marked as meeting the standard, if it exceeds the standard interval range, it is recorded as not meeting the standard, and calculating the minimum distance value between the morphological parameter and the corresponding standard interval endpoint value; based on the comparison result, determining whether there is a first growth stage in which all morphological parameters are marked as meeting the standard, if so, determining the first growth stage as the growth stage of the plant; if not, based on the first weight value corresponding to each morphological parameter, weighted calculation is performed on the minimum distance value corresponding to each morphological parameter to obtain a final evaluation value; determining the second growth stage corresponding to the final evaluation value with the smallest value, and determining the second growth stage as the growth stage of the plant.

[0051] That is, when comparing the model data and morphological data to determine the growth stage of the plant, the morphological data first needs to be integrated into a data set containing multiple morphological parameters according to a preset arrangement order. These morphological parameters may include leaf length, width, thickness, number of flowers, fruit size, etc., which can fully describe the morphological characteristics of the plant. The integrated morphological data set provides a systematic data basis for subsequent comparison and analysis. Next, it is necessary to obtain a set of preset standard morphological interval ranges corresponding to different growth stages. These standard morphological interval range sets are defined based on the typical characteristics of the plant at different growth stages. For example, a certain growth stage may define that the leaf length should be within a certain range, the number of flowers should meet a certain standard, etc. The integrated morphological data set is aligned with these standard morphological interval range sets for effective comparison and analysis.

[0052] During the comparison process, it is necessary to compare each morphological parameter in the morphological data set with each standard in the standard morphological interval range set one by one. Specifically, first determine whether each morphological parameter is within the corresponding standard interval. If the value of a morphological parameter is within the standard interval, it is marked as meeting the standard; if it exceeds the standard interval, it is marked as not meeting the standard. In addition, it is necessary to calculate the minimum distance value between the morphological parameter and the corresponding standard interval endpoint value. This calculation helps to quantify the degree of deviation and provide a basis for subsequent evaluation. Based on these comparison results, it is necessary to determine whether there is a first growth stage in which all morphological parameters are marked as meeting the standards. If such a growth stage exists, it is directly determined as the current growth stage of the plant. This stage usually corresponds to the ideal state of the plant, and its various morphological parameters meet the preset standards.

[0053] If there is no growth stage in which all morphological parameters meet the standards, further processing is required. Based on the first weight value corresponding to each morphological parameter, the minimum distance value of each morphological parameter is weightedly calculated. The weight value usually reflects the importance of each morphological parameter in the determination of the growth stage, and the parameter with a larger weight has a greater impact on the final evaluation result. The final evaluation value obtained by weighted calculation will be used to determine the growth stage that best matches the current morphology of the plant. Finally, the second growth stage corresponding to the final evaluation value with the smallest value is determined and determined as the growth stage of the plant. This process takes into account the degree of deviation and importance of each morphological parameter, thereby providing a comprehensive evaluation method to ensure that the determined growth stage reflects the actual growth status of the plant as accurately as possible.

[0054] For example, suppose we determine the growth stage of a tomato plant. First, we integrate the morphological data of tomatoes, such as leaf length, number of flowers, and fruit size. Then, we align them with the preset standard morphological interval range set. For example, for a specific growth stage, the standard range of leaf length is 10-15 cm, the standard range of flower number is 5-10, etc. During the comparison process, if it is found that the leaf length is 12 cm and the number of flowers is 6, and these data are within the standard range, they are marked as meeting the standard. If some parameters exceed the standard range, the minimum distance from the standard endpoint value is calculated, and a weighted calculation is performed based on these distance values ​​to determine the growth stage that best matches the actual morphology of the plant. Finally, based on the weighted evaluation value, the closest growth stage is determined, thereby accurately determining the growth status of the tomato plant. The advantage of this method is that it not only considers the degree of deviation of each morphological parameter, but also integrates their importance in different growth stages, making the determination of the growth stage more comprehensive and accurate.

[0055] In one example, the spectral requirements are determined according to the morphological data and growth stage of the plant, including: retrieving the basic spectral requirements corresponding to the growth stage of the plant from a preset spectral requirement model; determining the spectral weights of each band corresponding to each morphological parameter of the plant, and weightedly accumulating each morphological parameter of the plant and the spectral weights of each band corresponding to each morphological parameter to obtain the spectral correction requirements of each band; performing nonlinear mapping processing on the spectral correction requirements based on each morphological parameter of the plant, refining the influence of the morphological parameters on the spectral correction requirements, and obtaining the final spectral distribution adjustment requirements; performing dynamic feedback adjustment on the basic spectral requirements based on the spectral distribution adjustment requirements to obtain the final spectral requirements, wherein the spectral requirements include at least: spectral energy distribution, spectral wavelength range, spectral intensity curve and photoperiod requirements.

[0056] That is, in the process of determining the spectral requirements of plants, we first need to retrieve the basic spectral requirements corresponding to the current growth stage of the plant from the preset spectral requirement model. These basic spectral requirements are set according to the spectral requirement standards of plants at different growth stages. For example, a certain growth stage may require a spectrum with a specific wavelength range and intensity to support the best growth effect. Next, it is necessary to determine the spectral weights of each band corresponding to each morphological parameter of the plant. For example, morphological parameters such as leaf length, width, number of flowers, and fruit size may have different effects on the spectral requirements of different bands. These effects are represented by assigning spectral weights, which reflect the importance of each morphological parameter in the spectral requirements of a specific band.

[0057] Once the morphological parameters and their corresponding spectral weights are determined, these spectral weights need to be weighted and accumulated. This means that we add up the spectral weights of all morphological parameters in each band to calculate the spectral correction requirement for each band. The goal of this step is to combine the effects of all morphological parameters to derive the correction requirement for each spectral band. Next, we perform a nonlinear mapping of the spectral correction requirements based on the morphological parameters of the plant. This means that we apply some kind of nonlinear function or algorithm to refine the impact of each morphological parameter on the spectral correction requirement. The purpose of this process is to ensure that the spectral correction requirement can accurately reflect the actual impact of the morphological parameters, rather than a simple linear sum, so as to obtain a more accurate final spectral distribution adjustment requirement.

[0058] Finally, based on the obtained spectral distribution adjustment requirements, we make dynamic feedback adjustments to the basic spectral requirements. The dynamic feedback adjustment here refers to correcting the basic spectral requirements according to the aforementioned spectral distribution adjustment requirements to ensure that the spectral energy distribution, spectral wavelength range, spectral intensity curve, and photoperiod requirements can meet the actual needs of the plants.

[0059] For example, suppose that at a certain stage of plant growth, the basic spectral demand model requires the provision of red and blue light. Through morphological parameter analysis, it is found that the leaves of the plant are long and growing vigorously, which means that the demand for red light increases because red light helps promote photosynthesis. In this case, we will increase the spectral weight of red light. At the same time, if it is found that there are more flowers, this may increase the demand for blue light because blue light helps promote the flowering of flowers. In this way, specific spectral correction requirements are determined, and these requirements are adjusted through nonlinear mapping to ensure the optimal spectral distribution. Finally, the basic spectral requirements are dynamically adjusted to achieve the best growth effect of the plant.

[0060] The advantage of this whole process is that it can accurately adjust the spectrum requirements according to the specific plant morphological data, rather than using a unified spectrum setting. This can better meet the special needs of different plants at different growth stages, improve the support effect of light on plant growth, and thus optimize the health and growth rate of plants.

[0061] In one example, a spectral curve is fitted according to the spectral demand, including: normalizing the spectral data in the spectral demand, normalizing the demand parameters of different bands to the range of 0 to 1, and obtaining the normalized spectral demand; continuously fitting each band in the spectral demand using a cubic spline interpolation method to obtain an initial spectral curve, wherein the initial spectral curve is used to show the energy distribution of different bands in the entire spectral range; and removing high frequencies from the initial spectral curve using a Gaussian filter to obtain a final spectral curve, wherein the stability value and continuity of the final spectral curve are higher than those of the initial spectral curve.

[0062] In this example, the process of fitting the spectral curve involves several key steps to ensure that the final spectral curve can accurately reflect the spectral requirements of the plant and has good stability and continuity.

[0063] First, the spectral data in the spectral requirements are normalized. The purpose of this step is to unify the requirement parameters of different bands in the spectral requirements into a standard range, usually from 0 to 1. The benefit of normalization is that it eliminates the absolute magnitude differences between the requirement parameters of different spectral bands, making the data more consistent and convenient for subsequent processing and analysis. For example, if the spectral requirement of a band is 50 units and another band is 200 units, after normalization, these requirements will be converted to a range of 0 to 1, which can be compared and fitted more fairly. Next, the normalized spectral requirements are continuously fitted using cubic spline interpolation. Cubic spline interpolation is a commonly used mathematical method for generating smooth curves between known data points. Specifically, cubic spline interpolation generates a smooth curve by constructing a series of cubic polynomial functions that have continuous derivatives at each data point. This initial spectral curve shows the energy distribution of different bands across the entire spectral range. This process helps us obtain a continuous spectral curve that accurately reflects the demand for each band and transitions naturally between all bands.

[0064] In order to further improve the stability and continuity of the spectral curve, we apply a Gaussian filter to the initial spectral curve to remove high frequencies. The Gaussian filter is a smoothing filter that reduces noise and high-frequency fluctuations in the signal by weighted averaging of adjacent data points. Specifically, the Gaussian filter reduces small fluctuations in the spectral curve, making the curve smoother and avoiding drastic changes. The stability and continuity of the spectral curve after Gaussian filtering are significantly improved because high-frequency noise and unnecessary fluctuations have been effectively removed. The final spectral curve is the required spectral demand curve, which shows a stable and smooth energy distribution over the entire spectral range. This final spectral curve not only accurately reflects the spectral requirements of the plant, but also has good continuity, which means that in practical applications, the light source can be adjusted more smoothly to provide lighting that meets the requirements.

[0065] For example, suppose we need to provide a plant with a spectrum of a specific wavelength range, including red and blue light. Through normalization, we adjust these required parameters to the range of 0 to 1, and then use cubic spline interpolation to generate an initial spectral curve. This initial spectral curve may have some fluctuations, such as large changes in demand in certain bands. Through Gaussian filtering, we can smooth these fluctuations so that the final spectral curve presents a more stable energy distribution. This final spectral curve can be used to control the light source to ensure that the plant obtains uniform and required light in the actual growth environment, thereby optimizing its growth effect.

[0066] like Figure 2As shown, the embodiment of the present application also provides a spectrum control device for an LED plant lamp, comprising:

[0067] A determination unit 1 is used to collect image data of the plant through a camera, and determine the morphological data of the plant according to the image data of the plant;

[0068] An acquisition unit 2 is used to acquire preset model data corresponding to the plant, and compare the model data with the morphological data to obtain the growth state and growth stage of the plant;

[0069] A judging unit 3 is used to judge whether the growth state of the plant meets the preset expectation of the current growth stage of the plant, and if not, determine the spectrum requirement according to the growth state and growth stage of the plant;

[0070] The irradiation unit 4 is used to fit a spectrum curve according to the spectrum requirement, and use the spectrum curve to control the LED plant lamp to irradiate the plants.

[0071] In this embodiment, for the specific implementation of each unit in the above device embodiment, please refer to the above method embodiment, which will not be repeated here.

[0072] like Figure 3 As shown, the embodiment of the present application also provides a spectrum control system of an LED plant lamp, including:

[0073] An image recognition unit, comprising a camera, through which image data of the plant is collected, and morphological data of the plant is determined according to the image data of the plant;

[0074] A core processor is communicatively connected with the image recognition unit, and is used to receive the morphological data of the plant sent by the image recognition unit, and compare the morphological data according to the preset model data corresponding to the plant, to obtain the growth state and growth stage of the plant, and to judge whether the growth state of the plant meets the preset expectation of the current growth stage of the plant, if not, to determine the spectrum requirement according to the growth state and growth stage of the plant, and then to fit a spectrum curve according to the spectrum requirement, and to generate a control instruction for controlling the LED plant lamp to irradiate the plant based on the spectrum curve;

[0075] A spectrum control unit, including an LED plant lamp, and in communication with the core processor, for receiving a control instruction sent by the core processor, and controlling the LED plant lamp to irradiate the plant based on the spectrum curve according to the control instruction;

[0076] The spectrum detection unit includes a spectrum detector and is communicatively connected to the core processor for detecting the spectrum information of the actual irradiation of the plants by the LED plant lamp and sending the spectrum information to the core processor so that the core processor can adjust the control instructions based on the spectrum information.

[0077] That is, the system consists of several key parts designed to optimize the lighting conditions of plants to promote their growth. First, the image recognition unit includes a camera for collecting image data of plants. By analyzing these image data, the image recognition unit can extract the morphological data of the plants to evaluate their growth conditions. Next, the core processor is connected to the image recognition unit and is responsible for receiving and analyzing these morphological data. The core processor compares these data with the preset model data to determine the growth status and growth stage of the plant. If the growth status of the plant does not meet expectations, the core processor will determine the required spectral requirements based on the actual condition and growth stage of the plant, and fit a spectral curve based on these requirements. Subsequently, the core processor generates control instructions to guide the irradiation method of the LED plant light.

[0078] The spectrum control unit includes an LED plant light, which receives control instructions from the core processor and adjusts its lighting settings according to the spectrum curve. To ensure that the plants receive accurate light, the system is also equipped with a spectrum detection unit, which includes a spectrum detector. The detector is responsible for detecting the actual spectrum information emitted by the LED plant light and feeding this data back to the core processor. The core processor adjusts the control instructions based on this feedback information to optimize the lighting conditions for the plants. Through this process, the system is able to adjust the spectrum output in real time to ensure that the plants receive the most suitable light at different growth stages to promote healthy growth.

[0079] Specifically, the spectrum detection unit uses a high-definition CMOS recognition device to detect the spectrum output by the LED plant lamp, and is connected to the core processor through a serial communication circuit to process the spectrum into a serial signal and feed it back to the core processor. The core processor compares and corrects the received spectrum serial signal data with the spectrum type data marked by the storage unit to determine and trigger the I / O port of the core processor to output different PWN signals.

[0080] The image recognition unit uses a high-definition camera to collect the morphological data of the plant, and processes it to form a serial port signal, which communicates with the core processing unit through the serial port communication circuit to form data interaction. The serial port data format is the same as the serial port data type formed by the spectrum detection unit; these morphological data include leaf size, leaf number, leaf hole data, leaf edge yellowing data, plant height, plant diameter at breast height, foreign matter on the leaf surface, petal size, petal number, flower number, fruit diameter, fruit number, and conduct point analysis of comprehensive factors. These analysis data are often based on a certain collection plane, and then The data is then transmitted to the core processor; the core processor stores the model data of the plant, which includes the judgment data of each stage of plant growth, such as the bud stage, growth stage, flowering stage, and fruiting stage; as well as data on pathological stages, such as insect pests; data on each standard growth period of the plant is also stored; the core processor receives the data transmitted by the image recognition unit, which is used to compare with the stored data to determine the life cycle status and health status of the plant; at the same time, these data can also be compared with the standard growth period data to determine whether the plant can effectively improve its own growth mechanism under the current spectrum.

[0081] The spectrum control unit is used to modulate light sources of different wavelengths such as white light, red light, far-red light, and ultraviolet light of LED plant lights, and adjust the intensity and illumination time of light sources of different wavelengths to achieve the function of forming different spectra; specifically, the I / O port of the core processor outputs multiple PWM signals with different duty cycles, which are compared and amplified by the operational amplifier and feedback adjusted to form a stable 0-10V dimming signal. This dimming signal is connected to the light source driver via a wired or wireless connection. Under the action of the signal, the light source driver outputs a changing current, and the light source changes brightness under the action of the current, generating photons of different numbers and wavelengths; after the photon energy curve is generated, different spectrum curves are formed by fitting.

[0082] The interactive display unit is the basis of human-computer interaction and is used to display operations, controls, spectra, parameters, images, such as Figure 4 As shown, it is composed of different display pages, each of which contains controls, vector graphics, numbers, images, etc. Each item has a corresponding code, and the code contains data information such as functions and coordinates.

[0083] The spectral pattern is formed as follows: On the spectral modulation page of the operation control unit, see Figure 5 After the control 1, control 2, control 3, and control 4 are touched and started, different PWM signals are formed in the spectrum control unit to adjust the output photon energy of light sources with different wavelengths respectively, and then different photosynthetic photon energy data tables are tested, such as Figure 6As shown; the adjustment ratio of control 1 and control 2 ranges from 0%-100%, with a step of 1%; control 3 and control 4 have two states: on and off; step 1%, multiple photon energy data tables will be obtained; the X coordinate of the photon energy data table is the wavelength data, the wavelength range is from 400-800nm, step 1nm, and the Y coordinate is the photon energy data, the unit is um / s / nm; through the change of X, Y values, the points are drawn to form an undulating curve, and this curve is the spectrum graph; after drawing points on multiple data tables, different curves are formed, and data statistics are used to find out the rules and form segmented function curves, which are reflected in the program with code; the curve fitting method is used to fit each segmented function curve, find out the rules, and reflect them with algorithm code, and generate different curve data, that is, X, Y coordinate information. In this way, when adjusting the controls of the spectrum modulation page, a changing spectrum is generated in the spectrum display area. After the data information generated by the spectrum detection unit is transmitted to the core processor through the serial port, the core processor uses data processing to generate different X, Y coordinate information to form a curve in the spectrum display area, that is, a spectrum graph.

[0084] The operation control unit is based on touch control and is used to set the parameters of the core processing unit, the display data of the interactive display unit, the output parameters of each light source part that forms the spectrum, the various trigger conditions and thresholds of automatic control, etc.; the touch control unit covers the interactive display unit, and its control logic is in the form of a coordinate dot matrix, and each dot matrix corresponds to a corresponding coordinate; its coordinates form an accurate correspondence with the controls, vector graphics, numbers, and images in the interactive display unit; pressing or touching can trigger the corresponding functions of each control, vector graphics, number, and image.

[0085] The core processor is the data processing and output center of this intelligent system. It can be a local server with data processing capabilities or a cloud server with cloud data processing capabilities. Its principle block diagram is as follows Figure 7 As shown. It mainly includes a core processing unit, an I / O port expansion circuit, a touch control circuit, a serial communication circuit, a reset circuit, an RTC clock circuit, a display circuit, a power supply circuit, a dimming control circuit, and a signal output port. Among them, the power supply circuit provides power to the core processing unit, the display circuit, the touch control circuit, the I / O port expansion circuit, the RTC clock circuit, etc.; its principle is: obtain 12V DC from the external circuit, after filtering by capacitors C33 and C34, enter the VIN input terminal of the U7 step-down voltage regulator chip, obtain 5V voltage from the OUT terminal of the voltage regulator chip, and obtain a stable 5V voltage after filtering by capacitor C35. The FB terminal is the feedback terminal, which is used to ensure the stability of the output 5V DC voltage; the 5V DC voltage is used to power the display circuit; the stable 5V DC voltage is then passed through the voltage regulator chip U8 to obtain a 3.3V DC voltage, which is used to provide power to the core processing unit, the touch control circuit, and the I / O port expansion circuit.

[0086] The dimming control circuit amplifies and feeds back the PWM signals PWM_W, PWM_W1, PWM_R, PWM_R1, PWM_U, PWM_U1, etc. output from the I / O port of the I / O port expansion circuit through an operational amplifier to form stable dimming signals AOUT_W, AOUT_W1, AOUT_R, AOUT_R1, AOUT_U, AOUT_U1, which are used to control the output parameters of the light source driver to obtain the set photon energy.

[0087] The function of the signal output port in this system is to output the dimming signal AOUT_W, AOUT_W1, AOUT_R, AOUT_R1, AOUT_U, AOUT_U1 generated by the dimming control circuit to the light source controller through a wired manner, which can be expressed as an RJ11 or RJ45 standard port.

[0088] The reset circuit is used to reset the core processing unit. When the core processor encounters data processing anomalies, it performs an emergency interrupt to prevent the core processor from crashing or entering an infinite loop. The principle is to use the reset IC U4 to output a reset signal, and use the switch SW1 to ground it. When the switch SW1 is turned on, a low-level signal is formed at the C3_EN terminal of the core processing unit, and the core processing unit is reset. Then the program is in standby state and runs again.

[0089] The serial communication circuit is the channel for data exchange between the core processor and the graphics processing unit, spectrum detection unit, and operation touch unit; it is used to receive and transmit serial port signals. In this system, the serial port communication circuit mainly transmits the serial port signals formed by the graphics processing unit, the serial port signals generated by the spectrum detection unit, and the serial port signals formed by the operation control unit.

[0090] The touch control circuit in this system is embodied as a multi-point capacitive sensing touch panel, which corresponds to the functional areas of the interactive display unit through different dot matrices marked by coordinates. Specifically, each vector icon in the interactive display unit area generates a code containing coordinate information and functional information through a special tool. This coordinate information coincides with the coordinate information of the touch panel contact area. In this way, after the coordinate information is generated in the touch area, it can be matched with the functional vector icon in the interactive display area; different serial port signals are generated in the touch control IC to exchange information with the core processing unit.

[0091] The RTC clock circuit provides kinetic energy to the core processing unit to ensure that the core processing unit can provide processing information in an orderly and unbiased manner. It is composed of clock chips, crystal oscillators, power supply regulators and other components, and communicates with the core processing unit through the serial port I²C.

[0092] The I / O port expansion circuit is used to expand the limited I / O port resources of the core processing unit, and also undertakes the Wifi Mesh networking communication function; its expanded I / O port provides control signals to the dimming control circuit, and also provides a serial communication port for the core processing unit to be compatible with touch control circuit and display circuit applications.

[0093] Wifi Mesh networking communication connects a group of nodes through wireless links to form a multi-hop network topology. Each node has the function of a wireless transceiver and can communicate with other nodes. When a node needs to send data, it sends the data to the nearest node, which then forwards it to the next node until the data reaches the destination. Compared with traditional wireless network technology, Wifi Mesh network technology adopts a multi-hop communication method, so each node needs to perform routing selection and data forwarding. The topology of the Wifi Mesh network changes dynamically, so there is no need for complex configuration and management; Wifi Mesh network technology is also highly reliable and flexible because it can establish multiple communication paths between nodes to ensure reliable data transmission. The key technologies of Wifi Mesh network technology include multi-channel negotiation, channel allocation, and network discovery. Multi-channel negotiation is a coordination mechanism used to ensure that both communicating nodes work on the same channel. Channel allocation technology is mainly used for the use and management of multiple channels in multi-channel wireless Mesh networks to reduce the probability of channel conflicts and improve network efficiency. Network discovery technology is used to discover new nodes and neighbor nodes and establish corresponding information lists.

[0094] Wifi Mesh very conveniently connects the image recognition unit, spectrum detection unit, and spectrum control unit to the core controller wirelessly for data interaction and transmission of logic control instructions.

[0095] The interactive display unit of this intelligent system is embodied as a display circuit, and various information of human-computer interaction is reflected in this circuit; the display circuit includes a backlight control circuit, a liquid crystal screen driving circuit and a liquid crystal screen; the backlight control circuit includes a backlight positive electrode LED_A and a backlight negative electrode LED_K, which are used to control the switch and brightness of the backlight source; the liquid crystal driving circuit is used to control the changes in the light and dark of each pixel of the liquid crystal screen, and the liquid crystal driving positive electrode LCD_VGH, the analog circuit driving LCD_AVDD, the external DC input VCOM, and the liquid crystal driving negative electrode VGL are used for imaging; the liquid crystal screen forms different graphic images according to the received driving signal, and the R, G, and B pins of the liquid crystal screen are electrically interconnected with the I / O port of the core processing unit. Through the different PWN signals output by the I / O port, different brightness and colors of each pixel are formed, thereby generating a true color image.

[0096] The core processing unit is the basis of this system, usually composed of a 32-bit or higher microprocessor. Its internal structure block diagram is as follows: Figure 8 As shown, it includes a microprocessor, JTAG, serial port, liquid crystal display structure, I / O port, timer, counter, digital-to-analog converter, read-only memory, read-write memory, internal clock circuit, etc. It is used to process external input and internal logic control data, calculate the results, and output the corresponding data to the display circuit, dimming control circuit, and serial communication circuit.

[0097] Reference Fig. 9 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Fig. 9 As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0098] Those skilled in the art will understand that Fig. 9 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0099] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0100] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0101] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0102] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A spectrum control method for LED plant lamps, characterized in that: The following steps are involved: The image data of the plant is collected by a camera, and the morphological data of the plant is determined according to the image data of the plant; wherein, the image data of the plant is collected by a camera, including: performing multi-frame image capture processing on the plant by a camera to obtain a plurality of frames of first images, wherein the plurality of frames of first images are images obtained by taking multiple shots of the plant at different angles at the same time; using Gaussian filtering to perform weighted averaging on each pixel in each of the first images and its adjacent pixels in turn, smoothing the fine noise in the first image, replacing the salt and pepper noise in the first image by median filtering, and obtaining a plurality of frames of second images; converting each frame of the second image into HSV space, and adjusting The color tone of the second image is filtered out, pixels in the background of the second image that do not match the color of the plant are filtered out, and the plant part in the second image is highlighted to obtain a third image; a background model is constructed through dynamic and static background features of the plant that do not appear in multiple frames of the first image, and each frame of the third image is compared with the background model one by one, and the area in the third image that is similar to the dynamic and static background features is again identified and removed to obtain a fourth image; Canny edge detection is used to perform boundary segmentation processing on the area with a large change in grayscale value in the fourth image, and the plant outline in the fourth image is extracted, and the image inside the plant outline is retained to obtain a fifth image of the plant; Acquiring preset model data corresponding to the plant, and performing comparison processing on the model data and the morphological data to obtain the growth state and growth stage of the plant; Determine whether the growth state of the plant meets the preset expectations of the current growth stage of the plant. If not, determine the spectral requirements according to the morphological data and growth stage of the plant. Determine the spectral requirements according to the morphological data and growth stage of the plant, including: retrieving the basic spectral requirements corresponding to the growth stage of the plant from a preset spectral requirement model; determine the spectral weights of each band corresponding to each morphological parameter of the plant, perform weighted accumulation on each morphological parameter of the plant and the spectral weights of each band corresponding to each morphological parameter to obtain the spectral correction requirements of each band; perform nonlinear mapping processing on the spectral correction requirements based on each morphological parameter of the plant, refine the influence of the morphological parameters on the spectral correction requirements, and obtain the final spectral distribution adjustment requirements; perform dynamic feedback adjustment on the basic spectral requirements based on the spectral distribution adjustment requirements to obtain the final spectral requirements, wherein the spectral requirements at least include: spectral energy distribution, spectral wavelength range, spectral intensity curve and photoperiod requirements; A spectral curve is fitted according to the spectral demand, and the spectral curve is used to control the LED plant lamp to irradiate the plant; wherein, fitting a spectral curve according to the spectral demand includes: normalizing the spectral data in the spectral demand, normalizing the demand parameters of different bands to the range of 0 to 1, and obtaining the normalized spectral demand; using cubic spline interpolation method to continuously fit each band in the spectral demand to obtain an initial spectral curve, wherein the initial spectral curve is used to show the energy distribution of different bands in the entire spectral range; using a Gaussian filter to remove high frequencies from the initial spectral curve to obtain a final spectral curve, wherein the stability value and continuity of the final spectral curve are higher than those of the initial spectral curve.

2. The spectrum control method according to claim 1, characterized in that: Determining the morphological data of the plant according to the image data of the plant includes: Calculating the gradient of each pixel point of the fifth image of the plant in the X and Y directions, and using the calculated gradient data to construct an autocorrelation matrix of each pixel point; Calculating a Harris response value of each pixel point according to the autocorrelation matrix of each pixel point, wherein the Harris response value represents a corner point intensity of each pixel point in the fifth image; According to the comparison result of the Harris response value and the preset threshold, a plurality of feature points are screened out from the fifth image, wherein the feature points can represent the most important corner points and structural information in the fifth image; Determine the position coordinates and structural information of the feature points, and classify the feature points into a plurality of categories according to the position coordinates and the structural information, wherein the categories at least include: leaves, flowers, and fruits; Converting the pixel coordinates of the feature points into physical coordinates in the image coordinate system of the fifth image, and calculating the geometric data of the plant according to the physical coordinates; Determining distribution information and arrangement information of the feature points of each category in the fifth image, and extracting structural information of the plant according to the distribution information and arrangement information; The geometric data and structural data of the plant are processed by data set to obtain the morphological data of the plant.

3. The spectrum control method according to claim 1, characterized in that: Comparing the model data and the morphological data to obtain the growth stage of the plant, includes: Integrate the morphological data into a data set containing multiple morphological parameters according to a preset arrangement order; Obtaining a preset standard morphological interval range set corresponding to different growth stages, and aligning the integrated morphological data set with the standard morphological interval range set; Compare each morphological parameter in the morphological data set with each standard in the standard morphological interval range set one by one, wherein the comparison process includes: determining whether each morphological parameter is within the corresponding standard interval range, if it falls within the standard interval range, marking it as meeting the standard, if it exceeds the standard interval range, marking it as not meeting the standard, and calculating the minimum distance value between the morphological parameter and the corresponding standard interval endpoint value; Based on the comparison result, determining whether there is a first growth stage in which all morphological parameters are marked as meeting the standards, and if so, determining the first growth stage as the growth stage of the plant; If it does not exist, based on the first weight value corresponding to each morphological parameter, weighted calculation is performed on the minimum distance value corresponding to each morphological parameter to obtain a final evaluation value; The second growth stage corresponding to the final evaluation value with the smallest value is determined, and the second growth stage is determined as the growth stage of the plant.

4. A spectrum control device for LED plant lamps, characterized in that: include: A determination unit is used to collect image data of the plant through a camera, and determine the morphological data of the plant according to the image data of the plant; wherein, collecting the image data of the plant through the camera includes: performing multi-frame image capture processing on the plant through the camera to obtain multiple frames of first images, wherein the multiple frames of first images are images obtained by multiple shots of the plant at different angles at the same time; using Gaussian filtering to perform weighted averaging on each pixel in each of the first images and its adjacent pixels in turn, smoothing the fine noise in the first image, replacing the salt and pepper noise in the first image by median filtering, and obtaining multiple frames of second images; converting each frame of the second image into HSV space, and By adjusting the hue of the second image, pixels in the background of the second image that do not match the color of the plant are filtered out, and the plant part in the second image is highlighted to obtain a third image; a background model is constructed through dynamic and static background features in which the plant does not appear in multiple frames of the first image, and each frame of the third image is compared with the background model one by one, and the area in the third image that is similar to the dynamic and static background features is again identified and removed to obtain a fourth image; Canny edge detection is used to perform boundary segmentation processing on the area in the fourth image with a large change in grayscale value, and the plant outline in the fourth image is extracted, and the image inside the plant outline is retained to obtain a fifth image of the plant; An acquisition unit, used for acquiring preset model data corresponding to the plant, and performing a comparison process on the model data and the morphological data to obtain the growth state and growth stage of the plant; A judgment unit, used for judging whether the growth state of the plant meets the preset expectation of the current growth stage of the plant, and if not, determining the spectral requirement according to the morphological data and growth stage of the plant; wherein, determining the spectral requirement according to the morphological data and growth stage of the plant includes: retrieving the basic spectral requirement corresponding to the growth stage of the plant from a preset spectral requirement model; determining the spectral weight of each band corresponding to each morphological parameter of the plant, performing weighted accumulation on each morphological parameter of the plant and the spectral weight of each band corresponding to each morphological parameter, and obtaining the spectral correction requirement of each band; performing nonlinear mapping processing on the spectral correction requirement based on each morphological parameter of the plant, refining the influence of the morphological parameter on the spectral correction requirement, and obtaining the final spectral distribution adjustment requirement; performing dynamic feedback adjustment on the basic spectral requirement based on the spectral distribution adjustment requirement, and obtaining the final spectral requirement, wherein the spectral requirement at least includes: spectral energy distribution, spectral wavelength range, spectral intensity curve and photoperiod requirement; An irradiation unit is used to fit a spectral curve according to the spectral demand, and use the spectral curve to control the LED plant lamp to irradiate the plant; wherein, fitting a spectral curve according to the spectral demand includes: normalizing the spectral data in the spectral demand, normalizing the demand parameters of different bands to the range of 0 to 1, and obtaining the normalized spectral demand; using cubic spline interpolation method to continuously fit each band in the spectral demand to obtain an initial spectral curve, wherein the initial spectral curve is used to show the energy distribution of different bands in the entire spectral range; using a Gaussian filter to remove high frequencies from the initial spectral curve to obtain a final spectral curve, wherein the stability value and continuity of the final spectral curve are higher than those of the initial spectral curve.

5. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 3 are implemented.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 3 are implemented.

Images