Plant growth state intelligent monitoring method and system

Abstract

The invention relates to the technical field of intelligent monitoring, and particularly discloses a plant growth state intelligent monitoring method and system, and the method comprises the following steps: S1, enabling an imaging device to make circular motion, and constructing a change function of the height of the imaging device along with the circular motion duration; s2, presetting a sampling period, enabling an imaging device to periodically shoot plant pictures, constructing a three-dimensional model of a plant, marking a detection area, and obtaining plant structure features in the detection area; s3, constructing a comparison matrix, and calculating a comprehensive color value of the comparison matrix; and S4, screening an abnormal matrix through the comprehensive color value, calculating the proportion of the abnormal matrix, marking first-level pictures according to the proportion of the abnormal matrix, and analyzing the growth condition based on the number of the first-level pictures and the number of the pictures of the same structure. According to the method, potential problems in a plant growth process are identified through construction of a comparison matrix, calculation of a comprehensive color value, screening of an abnormal matrix and an analysis process based on a first-level picture proportion.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent monitoring, in particular to a plant growth state intelligent monitoring method and system. BACKGROUND

[0002] In today's large-scale, fine fruit and vegetable planting industry, plant cultivation and growth are crucial, which directly relates to the final yield and quality. However, the traditional way of relying on visual observation to evaluate the growth state of plants has many limitations and has been difficult to meet the needs of modern agricultural production.

[0003] In the actual operation process, there are two defects, one is that the plant's nutrient demand may be incorrectly judged, resulting in improper fertilization, too much will cause fertilizer waste and environmental pollution, and too little will make the plant lack of necessary nutrients and hinder growth; on the other hand, the detection of large-scale fruit and vegetable planting industry by artificial is inefficient, and long-time observation of the growth state of plants by artificial is easy to cause fatigue, thereby increasing the misjudgment. SUMMARY

[0004] The purpose of the present application is to provide a plant growth state intelligent monitoring method and system to solve the above technical problems.

[0005] The purpose of the present application can be achieved by the following technical solutions: A plant growth state intelligent monitoring method and system, comprising the following steps: S1: Let the imaging device make circular motion around the detection circle with a preset angular velocity ω, and the detection circle has the plant monitoring platform center point as the center and a preset length r as the radius; Constructing a change function of the imaging device height H with the circular motion duration t Wherein, H max represents the maximum value of the imaging device height, and λ represents the preset angle, λ>0; S2: Pre-set the sampling period T, and let the imaging device periodically shoot plant pictures with the sampling period T, construct a three-dimensional model of the plant and mark the detection area in the three-dimensional model, and the detection area represents the corresponding area of the plant picture in the three-dimensional model; Obtaining plant structure features in the detection area, the plant structure features including branches, leaves and fruits of the plant, and classifying the corresponding plant pictures based on the plant structure features; S3: Marking the plant pictures belonging to the same classification as the same structure pictures, and performing the following operations on the same structure pictures: Constructing an N×N comparison matrix, wherein N represents a preset number of pixel points, and dividing the same structure pictures into a plurality of comparison matrices; Calculating the comprehensive color value of the comparison matrix wherein, R i represents the red channel value of the i-th pixel point in the comparison matrix, G i represents the green channel value of the i-th pixel point in the comparison matrix, and B i represents the blue channel value of the i-th pixel point in the comparison matrix. S4: if the comprehensive color value C is outside the interval [C sta -γ, C sta +γ], the comparison matrix is marked as an abnormal matrix, wherein C sta represents a preset standard color value, and γ represents a preset fluctuation interval. The proportion of abnormal matrices B=a / A is calculated, wherein a represents the number of abnormal matrices, N represents the total number of comparison matrices, and if the proportion of abnormal matrices B is greater than or equal to a preset abnormal threshold B max , the corresponding same-structure picture is marked as a first-level picture. The growth status of the corresponding plant structure characteristics is analyzed based on the number of first-level pictures and the number of same-structure pictures.

[0006] As a further scheme of the present application, in the step S4, the method for analyzing the growth status of the corresponding plant structure characteristics based on the number of first-level pictures and the number of same-structure pictures comprises: calculating the ratio C of the number of first-level pictures to the number of same-structure pictures, and presetting a first gradient ratio C1; if C=0, the plant growth status is marked as good; if 0 if C>C1, the plant growth status is marked as abnormal, and the three-dimensional image is uploaded to a background database for analysis and evaluation by a staff.

[0007] As a further scheme of the present application, in the step S1, the wrap-around time TH=2π / ω and the period TB=2π / |λ| of the change function are calculated, and TH≠TB is ensured.

[0008] As a further scheme of the present application, in the step S2, the sampling period T of the imaging device for periodically shooting the plant picture is ensured to be less than 0.25×TH.

[0009] As a further scheme of the present application, in the step S3, the respective byte number K of the three primary color channels is obtained, and if the byte number K≠1, the formula for calculating the comprehensive color value of the comparison matrix is adjusted, and the adjusted formula is .

[0010] As a further aspect of the present invention: in step S2, a minimum number of plant images is preset. When the number of plant images taken equals the minimum number of plant images, the taking of images is stopped and subsequent steps are started.

[0011] As a further aspect of the present invention: in step S3, when dividing the image with the same structure into a number of comparison matrices, the pixels in the remaining rows and columns are removed and do not participate in subsequent operations.

[0012] A smart plant growth status monitoring system, including: Rotation module: Makes the imaging device rotate around the detection circle at a preset angular velocity ω, wherein the detection circle is centered on the center point of the plant monitoring platform and has a preset length r as its radius; Construct a function of the height H of the imaging device as a function of the duration t of the circular motion. , where H max The maximum value representing the height of the imaging device, λ represents the preset angular frequency, λ>0; Sampling module: A pre-set sampling period T is set so that the imaging device periodically takes pictures of the plant at the sampling period T, constructs a three-dimensional model of the plant, and marks the detection area in the three-dimensional model. The detection area represents the area of ​​the plant picture in the three-dimensional model. The plant structural features in the detection area are obtained, including the plant's branches, leaves and fruits. The corresponding plant images are then classified based on the plant structural features. Comparison module: Images of plants belonging to the same category are categorized as images with the same structure, and the following operations are performed on images with the same structure: Construct an N×N alignment matrix, where N represents the preset number of pixels, and divide images with the same structure into several alignment matrices; Calculate the overall color value of the comparison matrix , where R i G represents the red channel value of the i-th pixel in the alignment matrix. i B represents the green channel value of the i-th pixel in the comparison matrix. i This represents the blue channel value of the i-th pixel in the alignment matrix; Analysis module: If the comprehensive color value C is located in the interval [C sta -γ,C sta The matrix outside of [+γ] is marked as an anomaly matrix, where C sta γ represents the preset standard color value, and γ represents the preset fluctuation range; The percentage of anomaly matrices is calculated as B = a / A, where a represents the number of anomaly matrices and N represents the total number of comparison matrices. If the percentage of anomaly matrices B is greater than or equal to a preset anomaly threshold B0, then the percentage is considered positive. maxMark the corresponding images with the same structure as level 1 images; The growth status of corresponding plant structural features was analyzed based on the number of primary images and the number of images with the same structure.

[0013] The beneficial effects of this invention are as follows: First, the imaging device is made to move in a circle around the detection circle at a preset angular velocity, and the height of the imaging device is controlled to change according to a sine function. This not only allows for comprehensive image capture of the plant, but also reduces some of the workload and avoids the need for comprehensive monitoring of the plant.

[0014] The imaging device then periodically captures images of the plant at a sampling period T, constructs a three-dimensional model of the plant, marks the detection area in the three-dimensional model, and obtains the plant structural features in the detection area. The plant structural features include the plant's branches, leaves, and fruits. Based on the plant structural features, the corresponding plant images are classified. The purpose of this step is to subdivide the different structures of the plant, because different plant structures have different judgment criteria. This subdivision is beneficial for a more detailed analysis of the plant's growth status.

[0015] Plant images belonging to the same category are denoted as images with the same structure, and a number of comparison matrices are created. The purpose of creating comparison matrices is twofold: firstly, to facilitate subsequent comparisons, and secondly, to reduce workload. If all pixels in an image are analyzed, it will not only consume a lot of computing power, but also greatly increase the analysis time, thereby reducing real-time performance.

[0016] Subsequently, a comprehensive color value is calculated based on the three primary color values ​​of each pixel in the comparison matrix. It should be noted that in this invention, each of the three primary color channels occupies only one byte by default. If there are device differences or if each of the three primary color channels does not occupy only one byte, adjustments are required, but this invention does not specify otherwise. Anomaly matrices are filtered out using the comprehensive color values. When the number of anomaly matrices in images with the same structure exceeds a certain threshold, they are marked as Level 1 images. Finally, the growth status of the corresponding plant structural features is analyzed based on the number of Level 1 images and the number of images with the same structure. Attached Figure Description

[0017] The invention will now be further described with reference to the accompanying drawings.

[0018] Figure 1 This is a flowchart illustrating an intelligent monitoring method and system for plant growth status according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1 As shown, this invention is an intelligent monitoring method and system for plant growth status, comprising the following steps: S1: The imaging device is made to move in a circle around the detection circle at a preset angular velocity ω, wherein the detection circle is centered on the center point of the plant monitoring platform and has a preset length r as its radius; Construct a function of the height H of the imaging device as a function of the duration t of the circular motion. , where H max The maximum value representing the height of the imaging device, λ represents the preset angular frequency, λ>0; S2: Set a sampling period T, and have the imaging device periodically take pictures of the plant at the sampling period T, construct a three-dimensional model of the plant, and mark the detection area in the three-dimensional model. The detection area represents the area of ​​the plant picture in the three-dimensional model. The plant structural features in the detection area are obtained, including the plant's branches, leaves and fruits. The corresponding plant images are then classified based on the plant structural features. S3: Record plant images belonging to the same category as images with the same structure, and perform the following operations on images with the same structure: Construct an N×N alignment matrix, where N represents the preset number of pixels, and divide images with the same structure into several alignment matrices; Calculate the overall color value of the comparison matrix , where R i G represents the red channel value of the i-th pixel in the alignment matrix. i B represents the green channel value of the i-th pixel in the comparison matrix. i This represents the blue channel value of the i-th pixel in the alignment matrix; S4: If the comprehensive color value C is located in the interval [C sta -γ,C sta The matrix outside of [+γ] is marked as an anomaly matrix, where C sta γ represents the preset standard color value, and γ represents the preset fluctuation range; The percentage of anomaly matrices is calculated as B = a / A, where a represents the number of anomaly matrices and N represents the total number of comparison matrices. If the percentage of anomaly matrices B is greater than or equal to a preset anomaly threshold B0, then the percentage is considered positive. max Mark the corresponding images with the same structure as level 1 images; The growth status of corresponding plant structural features was analyzed based on the number of primary images and the number of images with the same structure.

[0021] It should be noted that the imaging device is activated to perform a smooth and precise circular motion along the trajectory of the detection circle at a pre-set angular velocity. This ensures that the imaging device will not lose critical details due to excessive speed, nor will it suffer from overall inefficiency due to excessively slow speed.

[0022] Meanwhile, the height of the imaging device is dynamically controlled, changing according to a sine function. This height variation pattern has unique advantages. When the imaging device is at the peak of the sine curve, it can take a high-angle shot of the plant, obtaining macroscopic information about the top of the plant and its overall shape, including the plant's height and the approximate distribution of the canopy. At the trough, it can take a low-angle shot close to the plant, capturing details that are easily overlooked, such as those near the roots and on the undersides of the leaves.

[0023] Each image captured at any given moment contains information about the plant's condition from different perspectives. By integrating these images, a complete and rich 3D model of the plant can be constructed. Furthermore, this imaging method can significantly reduce unnecessary workload. Compared to traditional methods of comprehensive, indiscriminate monitoring of plants, it avoids wasting excessive resources and time on repetitive or non-critical areas. Therefore, it provides an efficient solution for plant growth monitoring and research.

[0024] After completing the initial data collection using imaging equipment in a specific manner, the next stage is data processing and analysis. The imaging equipment is then instructed to perform periodic imaging operations according to a predetermined sampling period T. It is important to note that if the sampling period is too short, the data volume will be too large, increasing the cost and difficulty of storage and processing; conversely, if the sampling period is too long, key changes in the plant's growth process may be missed, affecting the accuracy of subsequent analysis.

[0025] These two-dimensional images will be converted into detailed and realistic three-dimensional plant models. Once the 3D model is successfully constructed, the next step is to mark the detection areas. Key areas of interest will be identified, such as the main stem, lateral branches, areas with dense foliage, and fruit attachment points. After the detection areas are defined, the system will automatically extract the plant structural features within those areas. These features cover all components of the plant, including branches, leaves, and fruits. For branches, detailed information such as length, thickness, branching angle, and growth direction will be recorded; these parameters reflect the plant's skeletal structure and support capacity. For leaves, not only will geometric features such as area, perimeter, and shape index be measured, but also appearance attributes such as leaf color, texture, and gloss will be analyzed; these details are often closely related to photosynthetic efficiency and nutrient status. As for fruits, characteristics such as size, weight, color, and surface blemishes will be considered; these are important criteria for evaluating fruit quality.

[0026] The purpose of classifying plant images in this step is far-reaching. Different plant structures require drastically different criteria for assessment. The growth pattern of branches determines the overall rationality of the plant's structure, the health of the leaves directly affects the efficiency of photosynthesis, and the quality of the fruit is crucial to its final economic value. By subdividing the plant's different structures and breaking them down into independent units for study, it is possible to conduct a more detailed and in-depth analysis of the plant's growth status.

[0027] After classifying plant images based on structural features, the next step is a more refined processing stage: constructing comparison matrices. Specifically, all plant images belonging to the same category are considered as a set of images with the same structure, and a number of comparison matrices are drawn from this set. This approach has two significant implications: firstly, it greatly facilitates subsequent data comparison; secondly, by selectively focusing on key areas rather than scanning the entire image, it significantly reduces computational complexity and workload. Performing a detailed analysis of every pixel in every image would not only consume massive amounts of computing resources but also drastically increase processing time, severely impairing the system's real-time responsiveness.

[0028] Once the comparison matrix is ​​established, the next step is to calculate the comprehensive color value using the primary color values ​​of the pixels in these matrices. It's important to note that, under the current invention framework, each primary color channel is defaulted to occupying only one byte of space. However, considering potential device differences in practical applications, adjustments are necessary to adapt to different hardware configurations. Nevertheless, to maintain the universality and simplicity of the solution, this invention does not impose a rigid requirement on this; users are free to handle the space flexibly according to their specific needs.

[0029] The calculated composite color value becomes the key indicator for screening anomaly matrices. In this process, any color combination deviating from the normal range may be considered a potential anomaly signal. When the number of anomaly matrices in a set of images with the same structure exceeds a preset threshold, that batch of images is specially marked as "Level 1 Images." This indicates that the plant parts reflected in these images may have a special physiological state or have been affected by some external factor.

[0030] Finally, by analyzing the ratio of Level 1 images to the total number of images with the same structure, important conclusions can be drawn regarding the growth status of the corresponding plant structural characteristics. A higher proportion of Level 1 images within a particular structure may indicate that the growth of that part has been affected by environmental stress, pests and diseases, or nutrient deficiencies. Conversely, a lower proportion suggests that the structure's growth is relatively healthy and stable. This statistical analysis-based method provides a new way to quantitatively assess plant health, enabling researchers to more objectively judge the differences between different structures and formulate more precise management measures accordingly.

[0031] By constructing a comparison matrix, calculating comprehensive color values, filtering out anomaly matrices, and conducting analysis based on primary image ratios, we can not only efficiently identify potential problems in plant growth, but also provide a scientific basis for precision agriculture, promoting the development of modern agriculture towards greater intelligence and precision.

[0032] In another preferred embodiment of the present invention, the method for analyzing the growth status of corresponding plant structural features based on the number of primary images and the number of images with the same structure includes: Calculate the ratio C of the number of first-level images to the number of images with the same structure, and pre-set the first gradient ratio C1; If C=0, the plant's growth status is marked as good; If 0 < C < C1, the plant growth status is recorded as normal, and the time interval for monitoring the plant growth status is adjusted to 50% of the original interval; If C > C1, the plant growth status is recorded as abnormal, and the 3D image is uploaded to the backend database for analysis and evaluation by staff.

[0033] It is worth noting that this method scientifically quantifies plant growth status and uses a precise grading management system based on the ratio C: no abnormalities indicate optimal growth, minor abnormalities require intensive monitoring, and severe abnormalities trigger manual intervention. This improves efficiency while ensuring accuracy, enabling dynamic control and optimal resource allocation, and supporting smart agriculture decision-making.

[0034] In another preferred embodiment of the invention, the orbital time TH = 2π / ω and the period of the variation function TB = 2π / |λ| are calculated to ensure that TH ≠ TB.

[0035] Understandably, this method cleverly calculates the orbital time and the period of change, strictly controlling for any discrepancies between the two. It accurately grasps the differences in movement rhythm, providing a clear temporal framework for analysis, effectively avoiding the superposition of interference, improving data accuracy, facilitating efficient research on plant dynamic characteristics, and optimizing monitoring strategies.

[0036] In another preferred embodiment of the present invention, the sampling period T for periodically capturing plant images by the imaging device is ensured to be < 0.25 × TH.

[0037] It is important to note that the sampling period T should be set to be less than 0.25 times the orbital time TH, and plant images should be acquired at a high frequency. This is done to ensure a dense and abundant dataset, accurately capture growth details, solidify the analytical foundation, and allow subsequent research to rely on massive amounts of high-quality data to deeply explore plant patterns and improve the reliability of conclusions.

[0038] In another preferred embodiment of the present invention, the number of bytes K corresponding to each of the three primary color channels is obtained. If the number of bytes K ≠ 1, the formula for calculating the comprehensive color value of the comparison matrix is ​​adjusted. The adjusted formula is as follows: .

[0039] It should be noted that, in this invention, each of the three primary color channels occupies only one byte by default. If there are device differences or if each of the three primary color channels does not occupy only one byte, adjustments are required. Here, a method is specified for handling cases where the number of bytes is not 1.

[0040] In another preferred embodiment of the present invention, a minimum number of plant images is preset. When the number of plant images taken equals the minimum number of plant images, the shooting stops and subsequent steps are started.

[0041] Understandably, a key parameter—the minimum number of plant images—is scientifically pre-set in the plant monitoring process. The imaging equipment continuously captures images, and the number of images acquired is counted in real time. When the cumulative number of plant images reaches exactly the preset minimum, the system automatically stops capturing images. This control mechanism ensures sufficient basic data is collected to support subsequent analysis, preventing insufficient data from affecting research accuracy, while also preventing resource waste from excessive capturing. Subsequently, the system seamlessly initiates subsequent processing steps, such as image recognition and feature extraction, orderly driving the entire monitoring and analysis process to operate efficiently and achieving a smooth transition from data acquisition to in-depth analysis.

[0042] In another preferred embodiment of the present invention, when dividing images with the same structure into a number of comparison matrices, the pixels in the remaining rows and columns are removed and do not participate in subsequent operations.

[0043] It is worth noting that when dividing images with the same structure into an alignment matrix, pixels in the remaining rows and columns are removed to simplify the data size and focus on the core regions. This reduces irrelevant interference, improves processing efficiency and accuracy, allows the analysis to focus more on key information, optimizes algorithm performance, ensures that the results are driven by valid data, and enhances reliability.

[0044] A smart plant growth status monitoring system, including: Rotation module: Makes the imaging device rotate around the detection circle at a preset angular velocity ω, wherein the detection circle is centered on the center point of the plant monitoring platform and has a preset length r as its radius; Construct a function of the height H of the imaging device as a function of the duration t of the circular motion. , where H max The maximum value representing the height of the imaging device, λ represents the preset angular frequency, λ>0; Sampling module: A pre-set sampling period T is set so that the imaging device periodically takes pictures of the plant at the sampling period T, constructs a three-dimensional model of the plant, and marks the detection area in the three-dimensional model. The detection area represents the area of ​​the plant picture in the three-dimensional model. The plant structural features in the detection area are obtained, including the plant's branches, leaves and fruits. The corresponding plant images are then classified based on the plant structural features. Comparison module: Images of plants belonging to the same category are categorized as images with the same structure, and the following operations are performed on images with the same structure: Construct an N×N alignment matrix, where N represents the preset number of pixels, and divide images with the same structure into several alignment matrices; Calculate the overall color value of the comparison matrix , where R i G represents the red channel value of the i-th pixel in the alignment matrix. i B represents the green channel value of the i-th pixel in the comparison matrix. i This represents the blue channel value of the i-th pixel in the alignment matrix; Analysis module: If the comprehensive color value C is located in the interval [C sta -γ,C sta The matrix outside of [+γ] is marked as an anomaly matrix, where C sta γ represents the preset standard color value, and γ represents the preset fluctuation range; The percentage of anomaly matrices is calculated as B = a / A, where a represents the number of anomaly matrices and N represents the total number of comparison matrices. If the percentage of anomaly matrices B is greater than or equal to a preset anomaly threshold B0, then the percentage is considered positive. max Mark the corresponding images with the same structure as level 1 images; The growth status of corresponding plant structural features was analyzed based on the number of primary images and the number of images with the same structure.

[0045] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A method for intelligent monitoring of plant growth status, characterized in that, Includes the following steps: S1: The imaging device is made to move in a circle around the detection circle at a preset angular velocity ω, wherein the detection circle is centered on the center point of the plant monitoring platform and has a preset length r as its radius; Construct a function of the height H of the imaging device as a function of the duration t of the circular motion. , where H max The maximum value representing the height of the imaging device, λ represents the preset angular frequency, λ>0; S2: Set a sampling period T, and have the imaging device periodically take pictures of the plant at the sampling period T, construct a three-dimensional model of the plant, and mark the detection area in the three-dimensional model. The detection area represents the area of ​​the plant picture in the three-dimensional model. The plant structural features in the detection area are obtained, including the plant's branches, leaves and fruits. The corresponding plant images are then classified based on the plant structural features. S3: Record plant images belonging to the same category as images with the same structure, and perform the following operations on images with the same structure: Construct an N×N alignment matrix, where N represents the preset number of pixels, and divide images with the same structure into several alignment matrices; Calculate the overall color value of the comparison matrix , where R i G represents the red channel value of the i-th pixel in the alignment matrix. i B represents the green channel value of the i-th pixel in the comparison matrix. i This represents the blue channel value of the i-th pixel in the alignment matrix; S4: If the comprehensive color value C is located in the interval [C sta -γ,C sta The matrix outside of [+γ] is marked as an anomaly matrix, where C sta γ represents the preset standard color value, and γ represents the preset fluctuation range; The percentage of anomaly matrices is calculated as B = a / A, where a represents the number of anomaly matrices and N represents the total number of comparison matrices. If the percentage of anomaly matrices B is greater than or equal to a preset anomaly threshold B0, then the percentage is considered positive. max Mark the corresponding images with the same structure as level 1 images; The growth status of corresponding plant structural features was analyzed based on the number of primary images and the number of images with the same structure.

2. The intelligent monitoring method for plant growth status according to claim 1, characterized in that, In step S4, the method for analyzing the growth status of corresponding plant structural features based on the number of primary images and the number of images with the same structure includes: Calculate the ratio C of the number of first-level images to the number of images with the same structure, and pre-set the first gradient ratio C1; If C=0, the plant's growth status is marked as good; If 0 < C < C1, the plant growth status is recorded as normal, and the time interval for monitoring the plant growth status is adjusted to 50% of the original interval; If C > C1, the plant growth status is recorded as abnormal, and the 3D image is uploaded to the backend database for analysis and evaluation by staff.

3. The intelligent monitoring method for plant growth status according to claim 1, characterized in that, In step S1, the orbital time TH = 2π / ω and the period of the variation function TB = 2π / |λ| are calculated, ensuring that TH ≠ TB.

4. The intelligent monitoring method for plant growth status according to claim 1, characterized in that, In step S2, ensure that the sampling period T for the imaging device to periodically capture plant images is less than 0.25 × TH.

5. The intelligent monitoring method for plant growth status according to claim 1, characterized in that, In step S3, the number of bytes K corresponding to each of the three primary color channels is obtained. If the number of bytes K ≠ 1, the formula for calculating the comprehensive color value of the comparison matrix is ​​adjusted. The adjusted formula is as follows: .

6. The intelligent monitoring method for plant growth status according to claim 1, characterized in that, In step S2, a minimum number of plant images is preset. When the number of plant images taken equals the minimum number of plant images, the shooting stops and subsequent steps are started.

7. The intelligent monitoring method for plant growth status according to claim 1, characterized in that, In step S3, when dividing the image with the same structure into several comparison matrices, the pixels in the remaining rows and columns are removed and do not participate in subsequent operations.

8. An intelligent monitoring system for plant growth status, characterized in that, include: Rotation module: Makes the imaging device rotate around the detection circle at a preset angular velocity ω, wherein the detection circle is centered on the center point of the plant monitoring platform and has a preset length r as its radius; Construct a function of the height H of the imaging device as a function of the duration t of the circular motion. , where H max The maximum value representing the height of the imaging device, λ represents the preset angular frequency, λ>0; Sampling module: A pre-set sampling period T is set so that the imaging device periodically takes pictures of the plant at the sampling period T, constructs a three-dimensional model of the plant, and marks the detection area in the three-dimensional model. The detection area represents the area of ​​the plant picture in the three-dimensional model. The plant structural features in the detection area are obtained, including the plant's branches, leaves and fruits. The corresponding plant images are then classified based on the plant structural features. Comparison module: Images of plants belonging to the same category are categorized as images with the same structure, and the following operations are performed on images with the same structure: Construct an N×N alignment matrix, where N represents the preset number of pixels, and divide images with the same structure into several alignment matrices; Calculate the overall color value of the comparison matrix , where R i G represents the red channel value of the i-th pixel in the alignment matrix. i B represents the green channel value of the i-th pixel in the comparison matrix. i This represents the blue channel value of the i-th pixel in the alignment matrix; Analysis module: If the comprehensive color value C is located in the interval [C sta -γ,C sta The matrix outside of [+γ] is marked as an anomaly matrix, where C sta γ represents the preset standard color value, and γ represents the preset fluctuation range; The percentage of anomaly matrices is calculated as B = a / A, where a represents the number of anomaly matrices and N represents the total number of comparison matrices. If the percentage of anomaly matrices B is greater than or equal to a preset anomaly threshold B0, then the percentage is considered positive. max Mark the corresponding images with the same structure as level 1 images; The growth status of corresponding plant structural features was analyzed based on the number of primary images and the number of images with the same structure.

Images