A method for predicting leaf photosynthetic rate based on images

Abstract

A leaf photosynthetic rate prediction method based on an image relates to the technical field of plant photosynthesis evaluation. In order to solve the technical problems existing in the prior art, the existing photosynthetic rate measurement technology has the following problems: high cost and difficulty in large-scale application, and lack of effective analysis of the comprehensive characteristics of leaf images, resulting in low prediction accuracy. The technical solution provided by the present invention is: a leaf photosynthetic rate prediction model construction method based on an image, the method comprising: preprocessing the collected training image, including the step of dividing it into multiple intervals; counting the number of pixels in the partitions within a preset measurement area; obtaining the quantitative relationship between the number of pixels and the photosynthetic rate based on the least squares method; and constructing a photosynthetic rate prediction model based on the quantitative relationship. It can be applied to the work of analyzing the photosynthetic rate of leaves.

Description

Technical Field

[0001] The invention relates to the technical field of plant photosynthesis evaluation, and in particular to the prediction of leaf photosynthesis rate by using image processing technology. Background Art

[0002] Photosynthesis is a vital physiological process in the growth and development of plants, which directly affects the growth rate and yield of plants. Therefore, accurately evaluating the photosynthetic rate of plants is of great significance to agricultural production, crop breeding, and environmental monitoring. At present, the measurement method of plant photosynthetic rate mainly relies on physical measurement equipment, such as the LI-6800 photosynthetic meter. These devices can usually provide accurate photosynthetic rate data, but they are expensive, complicated to operate, and not suitable for large-scale measurement, which limits their wide application in agricultural production.

[0003] With the continuous development of image processing technology, some researchers have begun to explore methods to predict photosynthetic rate through leaf image features. This image-based technology has attracted widespread attention due to its low cost and convenience. Existing image-based prediction methods mainly focus on analyzing single features such as leaf color and morphology, and usually use simple statistical models or empirical formulas to predict photosynthetic rate. However, these methods have large errors when processing complex physiological data, making it difficult to achieve high-precision predictions of photosynthetic rates, especially in field environments or large-scale agricultural production applications. The reliability and accuracy of existing methods are still insufficient.

[0004] In summary, the existing technology has the following main problems in photosynthetic rate measurement: traditional physical measurement equipment is expensive and difficult to apply on a large scale; image-based prediction methods lack effective analysis of the comprehensive characteristics of leaf images, resulting in low prediction accuracy. Therefore, a new method that can reduce costs and achieve high-precision photosynthetic rate prediction is urgently needed to solve these problems in the existing technology. Summary of the invention

[0005] In order to solve the technical problems existing in the prior art, the existing photosynthetic rate measurement technology has the following problems: high cost and difficulty in large-scale application, and lack of effective analysis of the comprehensive characteristics of leaf images, resulting in low prediction accuracy. The technical solution provided by the present invention is:

[0006] A method for constructing a leaf photosynthetic rate prediction model based on an image, the method comprising:

[0007] Preprocessing the acquired training image, including the step of dividing it into multiple intervals;

[0008] A step of counting the number of pixels in the interval within the preset measurement area;

[0009] Based on the least square method, a step of obtaining a quantitative relationship between the number of pixels and the photosynthetic rate;

[0010] The step of constructing a photosynthetic rate prediction model based on the quantitative relationship.

[0011] Furthermore, a preferred embodiment is provided, wherein the preprocessing further includes extracting a green channel of the acquired image to be tested.

[0012] Furthermore, a preferred implementation is provided, in which a weight coefficient between the number of pixels and the photosynthetic rate is obtained based on the least squares method, and the weight coefficient with the smallest error is used as the quantized relationship.

[0013] Furthermore, a preferred embodiment is provided, in which a weight coefficient with the minimum fitting error is obtained through a linear regression data processing method.

[0014] Based on the same inventive concept, the present invention also provides a device for constructing a leaf photosynthetic rate prediction model based on an image, the device comprising:

[0015] Preprocessing the collected image to be tested, including a module for dividing it into multiple intervals;

[0016] A module for counting the number of pixels in the interval within a preset measurement area;

[0017] A module for obtaining a quantitative relationship between the number of pixels and the photosynthetic rate based on the least squares method;

[0018] According to the quantitative relationship, a module of a photosynthetic rate prediction model is constructed.

[0019] Based on the same inventive concept, the present invention also provides an image-based leaf photosynthetic rate prediction method, the method comprising:

[0020] The step of predicting the photosynthetic rate of the collected image to be measured according to the photosynthetic rate prediction model.

[0021] Based on the same inventive concept, the present invention also provides an image-based leaf photosynthetic rate prediction device, the device comprising:

[0022] A module for predicting the photosynthetic rate of the collected image to be measured according to the photosynthetic rate prediction model.

[0023] Based on the same inventive concept, the present invention also provides a computer storage medium for storing a computer program. When the computer program is read by a computer, the computer executes the method described.

[0024] Based on the same inventive concept, the present invention also provides a computer, including a processor and a storage medium. When the processor reads a computer program stored in the storage medium, the computer executes the method described.

[0025] Based on the same inventive concept, the present invention also provides a computer program product, which is a computer program. When the computer program is read by a computer, the method described above is implemented.

[0026] Compared with the prior art, the technical solution provided by the present invention is beneficial in that:

[0027] The present invention uses an image-based leaf photosynthetic rate prediction method, utilizes green channel extraction and pixel value interval division, and combines the least squares method to build a mathematical model to predict the photosynthetic rate and optimize the model. Compared with the prior art, the extraction of the green channel of the image in this scheme effectively reduces the data dimension, making feature extraction simpler and more efficient; the interval division of pixel values ​​further refines the image features, allowing the model to capture more detailed information related to the photosynthetic rate; the use of the least squares method ensures the optimality of the model parameters and reduces the prediction error. Overall, the combination of these steps enables the scheme to significantly improve the accuracy and stability of photosynthetic rate prediction while maintaining a low cost, and is particularly suitable for large-scale applications in agricultural production. Compared with traditional physical measurement methods and simple image analysis methods, the present invention not only has a significant improvement in prediction accuracy, but also has greater advantages in ease of operation and scope of application.

[0028] By extracting the green channel and dividing the pixel values ​​into intervals, the data features are made more targeted and concise, thereby improving the efficiency of model construction. At the same time, the least squares method is combined to perform data fitting analysis, which effectively reduces the prediction error value and improves the accuracy and robustness of the model. Compared with existing studies based on simple color features, this scheme has shown significant advantages in adaptability and prediction accuracy in complex environments through a multi-level data processing method, providing a more economical and efficient solution for photosynthetic status assessment in agricultural production.

[0029] In this scheme, the effect of greatly reducing the measurement cost is achieved through the image-based leaf photosynthetic rate prediction method. Traditional photosynthetic rate measurement requires the use of expensive photosynthetic instruments (such as LI-6800), and requires experienced operators to perform accurate measurements. The present invention uses a camera for image acquisition, combines image processing technology and mathematical models to replace traditional physical equipment, greatly reducing equipment and labor costs, making photosynthetic rate measurement more popular and low-cost, suitable for large-scale promotion.

[0030] The present invention improves the accuracy of photosynthetic rate prediction by extracting the green channel and dividing the pixel intervals in image preprocessing. Compared with the existing research that usually relies on a single leaf color or morphological feature, this solution divides the pixel values ​​into multiple intervals through in-depth analysis of the green channel of the image, and counts the number of pixels in each interval, ensuring the comprehensive capture of leaf features. This refined feature extraction method effectively improves data quality and ultimately improves the accuracy of the prediction model, making the evaluation of leaf photosynthetic rate more accurate.

[0031] The use of linear regression and least squares method to establish mathematical models significantly improves the applicability and simplicity of the model. Compared with traditional machine learning methods, linear regression models are easier to implement and have lower computational overhead. By calculating the weight coefficients using the least squares method, the optimal fit relationship can be quickly found while maintaining sufficient accuracy, allowing the solution to quickly build and optimize mathematical models, which is helpful for application in field conditions or large-scale agricultural production environments, and has greater practicality and ease of use.

[0032] This solution achieves continuous optimization of the model by dynamically adjusting the weight coefficients and intercept terms in the mathematical model, thereby further reducing the error in photosynthetic rate prediction. Compared with the static model used in the prior art, this solution allows dynamic correction of model parameters based on different leaf samples, improving the model's adaptability to different leaves and environmental conditions. This continuous optimization approach makes the model stable and reliable in long-term applications, and is suitable for real-time monitoring and guidance in actual agricultural management.

[0033] It can be used to analyze the photosynthetic rate of leaves. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Schematic diagram of the process of an image-based leaf photosynthetic rate prediction method. DETAILED DESCRIPTION

[0035] In order to make the advantages and benefits of the technical solution provided by the present invention more clearly reflected, the technical solution provided by the present invention is now further described in detail with reference to the accompanying drawings, specifically:

[0036] Embodiment 1: This embodiment provides a method for constructing a leaf photosynthetic rate prediction model based on an image, the method comprising:

[0037] Preprocessing the acquired training image, including the step of dividing it into multiple intervals;

[0038] A step of counting the number of pixels in the interval within the preset measurement area;

[0039] Based on the least square method, a step of obtaining a quantitative relationship between the number of pixels and the photosynthetic rate;

[0040] The step of constructing a photosynthetic rate prediction model based on the quantitative relationship.

[0041] Embodiment 2: This embodiment further limits the method for constructing an image-based leaf photosynthetic rate prediction model provided in Embodiment 1, and the preprocessing further includes extracting the green channel of the collected image to be measured.

[0042] Implementation method three: This implementation method is a further limitation of the method for constructing an image-based leaf photosynthetic rate prediction model provided in implementation method one. Based on the least squares method, the weight coefficient between the number of pixels and the photosynthetic rate is obtained, and the weight coefficient with the smallest error is used as the quantitative relationship.

[0043] Implementation method 4: This implementation method further limits the method for constructing an image-based leaf photosynthetic rate prediction model provided in implementation method 3. A weight coefficient with the smallest fitting error is obtained through a linear regression data processing method.

[0044] Embodiment 5: This embodiment provides a device for constructing a leaf photosynthetic rate prediction model based on an image, the device comprising:

[0045] Preprocessing the collected image to be tested, including a module for dividing it into multiple intervals;

[0046] A module for counting the number of pixels in the interval within a preset measurement area;

[0047] A module for obtaining a quantitative relationship between the number of pixels and the photosynthetic rate based on the least squares method;

[0048] According to the quantitative relationship, a module of a photosynthetic rate prediction model is constructed.

[0049] Embodiment 6: This embodiment provides a leaf photosynthetic rate prediction method based on an image, the method comprising:

[0050] According to the photosynthetic rate prediction model provided in the first embodiment, a step of predicting the photosynthetic rate of the collected image to be measured.

[0051] Specifically, this embodiment discloses an image-based leaf photosynthetic rate prediction method, which obtains the image data of the leaf and performs image preprocessing, extracts effective features, and then establishes a mathematical model to finally achieve the prediction of the photosynthetic rate. This method uses a camera to collect the RGB image data of the leaf, combines the actual net photosynthetic rate measured by the photosynthetic instrument, and establishes a mathematical relationship model between the net photosynthetic rate of the leaf and the image pixel value through the least squares method to achieve non-destructive prediction of the net photosynthetic rate of the leaf. This embodiment aims to improve the efficiency of evaluating the photosynthetic state of plants in an accurate, fast, and low-cost manner.

[0052] Specific steps description:

[0053] Step 1: Collection of leaf samples

[0054] Green leaves were selected as the measurement objects, and the samples were divided into a calibration set and a test set. The calibration set included samples with known net photosynthetic rates, and the test set included samples with net photosynthetic rates to be measured. When preparing the samples, ensure that there are no obvious stains on the leaf surface. If necessary, the leaf surface can be cleaned. The collection time should be selected in the morning on a sunny day, avoiding the photosynthesis lunch break (11:00-14:00) to obtain the best photosynthetic state of the leaves.

[0055] Step 2: Image acquisition and photosynthetic rate measurement

[0056] Build a dark room in the experimental environment to avoid interference from external light. Use blackout cloth to cover the dark room, and place shadowless lamps around the experimental table to eliminate reflections. A Canon 700D camera and a LI-6800 photosynthetic instrument are fixed in front of the desktop, and the distance, angle, and focal length between the camera and the leaves are kept fixed to ensure that the captured images are clear and consistent. During the measurement process, the leaves are first photographed with a camera to obtain the RGB image data of the leaves, and the net photosynthetic rate of the leaves is measured using the photosynthetic instrument. After the data displayed by the photosynthetic instrument is stable, the measured value is recorded.

[0057] Step 3: Image data preprocessing

[0058] The acquired RGB image data is processed. First, the measurement area is marked and the region of interest (ROI) is intercepted, and the area is saved as a PNG format image. The intercepted image is processed using Photoshop or similar software to extract the green channel of the image, because the pixel value of the green channel is closely related to the photosynthetic activity of the leaves. The extracted green channel image pixel values ​​are divided into five intervals, namely 0-50, 51-100, 101-150, 151-200, and 201-255, for further analysis.

[0059] Step 4: Feature extraction and statistics

[0060] Perform pixel statistics on the extracted green channel image and count the pixel values ​​in each interval within the 2cm² measurement area. Specifically, count the number of pixels in the five pixel intervals and record the number of pixels belonging to each interval. These data will be used to establish a mathematical relationship model between the net photosynthetic rate of leaves and pixel values.

[0061] Step 5: Build a mathematical model

[0062] The least squares method was used to process the acquired net photosynthetic rate and pixel value data. Assuming that there is a linear relationship between the net photosynthetic rate of the leaf and the pixel value of each interval, a linear relationship equation system was constructed. The regression coefficient was set to quantify the influence of each pixel interval value on the net photosynthetic rate. These coefficients were solved by minimizing the fitting error to establish a mathematical model. The mathematical model represents the linear quantitative relationship between the net photosynthetic rate of the leaf and the pixel value of the green channel.

[0063] Step 6: Photosynthetic rate prediction

[0064] The green channel pixel value of the leaf to be measured is input into the established mathematical model to predict the net photosynthetic rate of the leaf. For each leaf sample in the test set, its pixel data is calculated and substituted into the mathematical model to obtain the predicted value of the net photosynthetic rate. In this way, the net photosynthetic rate of leaves that have not been measured can be predicted without loss.

[0065] Step 7: Error evaluation and model optimization

[0066] Based on the actual net photosynthetic rate and the predicted net photosynthetic rate, the prediction error value is calculated. The prediction accuracy of the model is evaluated using the coefficient of determination and relative error. The coefficient of determination is used to measure how well the model fits the data, while the relative error provides a standardized measure of the deviation between the predicted value and the true value. According to the error analysis results, the weight coefficients and intercept terms in the mathematical model are adjusted accordingly to optimize the mathematical model to further improve the accuracy and stability of the prediction.

[0067] Step 8: Practical Application

[0068] After the optimized model is established, it can be used to quickly and non-destructively measure the photosynthetic rate of leaves, which is suitable for laboratory research, agricultural production management, and other scenarios that require rapid assessment of photosynthetic status. By continuously monitoring the photosynthetic rate of leaves, it can help accurately manage the growth of crops, optimize the input of resources such as water and fertilizer, and realize real-time monitoring and early warning of plant health.

[0069] Embodiment 7: This embodiment provides a leaf photosynthetic rate prediction device based on an image, the device comprising:

[0070] A module for predicting the photosynthetic rate of the collected image to be measured according to the photosynthetic rate prediction model provided in the first embodiment.

[0071] Embodiment 8: This embodiment provides a computer storage medium for storing a computer program. When the computer program is read by a computer, the computer executes the method provided in embodiment 1.

[0072] Embodiment 9: This embodiment provides a computer, including a processor and a storage medium. When the processor reads a computer program stored in the storage medium, the computer executes the method provided in embodiment 1.

[0073] Embodiment 10: This embodiment provides a computer program product, which is a computer program. When the computer program is read by a computer, the method provided in embodiment 1 is implemented.

[0074] Implementation Method XI: Combination Figure 1 This embodiment is described in detail by using specific examples to further describe the technical solution provided above. Specifically:

[0075] The purpose of this embodiment is to address the technical problem that the existing measurement technology completely relies on photosynthetic instrument equipment, and propose a leaf photosynthetic rate prediction method based on images, which not only saves time and cost, but also maintains the scientificity and accuracy of the measurement results.

[0076] To achieve the above objectives, the present application provides, in a first aspect, a method for predicting leaf photosynthetic rate based on an image, the method comprising:

[0077] Acquire photosynthetic rate data and leaf image data of leaves;

[0078] When acquiring data from the leaves, choose to do it in the morning on a sunny day, avoiding the photosynthetic lunch break time of the leaves from 11:00 to 14:00.

[0079] The obtained image data is processed and photos are taken before and after the measurement.

[0080] The obtained photosynthetic data were screened and outliers were eliminated.

[0081] Photoshop is used to process the obtained image, a PNG image of the measurement area is intercepted, and the green channel of the intercepted PNG image is extracted.

[0082] The pixel values ​​are divided into five intervals: 0-50 is the A interval, 51-100 is the B interval, 101-150 is the C interval, 151-200 is the D interval, and 201-255 is the E interval.

[0083] The number of pixel values ​​in each interval of the intercepted 2cm² green channel image is counted.

[0084] The obtained photosynthetic data and pixel value data are processed by least square method to obtain weight coefficients. The weight coefficients of pixel value and photosynthetic rate represent the quantitative relationship between pixel value and photosynthetic rate.

[0085] The mathematical model between the pixel value of the measured area of ​​the leaf and the net photosynthetic rate:

[0086] Assuming that there is a linear relationship between the actual net photosynthetic rate of the leaf and the pixel value of each interval, a linear relationship equation group is established, and the linear relationship equation group is solved according to the actual net photosynthetic rate and the pixel value of each interval to obtain the mathematical model.

[0087] In one embodiment, the linear relationship equation system includes:

[0088] ;

[0089] In the formula, is the actual value of net photosynthetic rate, which is the dependent variable and is the net photosynthetic rate data measured by the LI-6800 photosynthetic instrument; is the regression coefficient, which represents the weight coefficient in the interval 0-50. represents the weight coefficient in the interval 51-100, represents the weight coefficient of the interval 101-150, represents the weight coefficient of the interval 151-200, Indicates the weight coefficient in the interval 201-255; is the independent variable, indicating the number of pixel values ​​in the range of 0-50. is the independent variable, indicating the number of pixel values ​​in the range of 51-100. is the independent variable, indicating the number of pixel values ​​in the interval 101-150, is the independent variable, indicating the number of pixel values ​​in the interval 151-200, is an independent variable, indicating the number of pixel values ​​in the range of 201-255. is the intercept term.

[0090] Obtaining a predicted value of the net photosynthetic rate of the sample through the mathematical model;

[0091] The prediction error value of the mathematical model calculated based on the predicted value of net photosynthetic rate and the actual net photosynthetic rate includes:

[0092] Based on the predicted value of net photosynthetic rate pn i and the actual net photosynthetic rate Pn i Calculate the coefficient of determination R 2:

[0093]

[0094] The prediction error value of the mathematical model is obtained through the determination coefficient.

[0095] This embodiment provides an image-based leaf net photosynthetic rate prediction method, which saves time and cost while maintaining the scientificity and accuracy of the measurement results; and, by using the constructed mathematical model, the net photosynthetic rate of unknown leaves can be predicted quickly and losslessly, greatly improving the measurement efficiency and realizing the rapid evaluation of a large number of samples, which saves time and cost while maintaining the scientificity and accuracy of the measurement results. At the same time, the method can realize real-time monitoring, which helps to timely understand the physiological condition of the leaves and guide precision agricultural management. It can be used for growth monitoring, quality evaluation, and early warning of pests and diseases during the growth process, and has important value.

[0096] In a specific embodiment:

[0097] Embodiment 1

[0098] This embodiment provides a method for predicting the net photosynthetic rate of leaves based on images. Figure 1 As shown, the method includes:

[0099] Selecting leaves as samples, dividing the samples into a calibration set and a test set, wherein the calibration set includes samples with known net photosynthetic rates, and the test set includes samples with net photosynthetic rates to be measured;

[0100] Optionally, several suitable leaves are selected and their surface stains are cleaned as samples.

[0101] A darkroom with blackout cloth was built on all sides. A table was placed in the middle to place the experimental leaves. Four shadowless lamps were placed around the table to prevent the leaves from reflecting light. A Canon 700D camera and a LI-6800 photosynthetic instrument were placed in front of the table. The camera had a fixed shooting angle, fixed distance between the camera and the leaves, fixed camera focal length, and fixed camera shooting height to ensure a clear image.

[0102] The experiment uses a Canon 700D camera to shoot the blades to be measured and obtain the initial blade image data.

[0103] When acquiring data from the leaves, choose to do it in the morning on a sunny day, avoiding the photosynthetic lunch break time of the leaves from 11:00 to 14:00.

[0104] Use Li-6800 photosynthetic instrument to measure the leaves to be tested, and record the data after it shows stable.

[0105] The measured area of ​​the blade is marked and photographed to obtain the measured image data.

[0106] The obtained net photosynthetic rate data were screened and outliers were eliminated.

[0107] The obtained image was processed using Photoshop, the positions of the image marks after measurement were compared, a PNG image of the measurement area was intercepted from the initial image, and the green channel of the intercepted PNG image was extracted.

[0108] The pixel values ​​are divided into five intervals: 0-50 is the A interval, 51-100 is the B interval, 101-150 is the C interval, 151-200 is the D interval, and 201-255 is the E interval.

[0109] The number of pixels in each interval of the intercepted 2cm² image is counted.

[0110] The obtained net photosynthetic rate data and pixel value data are processed by least square method to obtain weight coefficients. The weight coefficients of net photosynthetic rate and pixel value represent the quantitative relationship between net photosynthetic rate and pixel value.

[0111] The mathematical model between the pixel value of the measured area of ​​the leaf and the net photosynthetic rate:

[0112] Assuming that there is a linear relationship between the actual net photosynthetic rate of the leaf and the pixel value of each interval, a linear relationship equation group is established, and the linear relationship equation group is solved according to the actual net photosynthetic rate and the pixel value of each interval to obtain the mathematical model.

[0113] In one embodiment, the linear relationship equation system includes:

[0114]

[0115] In the formula, is the actual value of the net photosynthetic rate, which is the dependent variable and is the net photosynthetic rate data measured by LI-6800; is the regression coefficient, which represents the weight coefficient in the interval 0-50. represents the weight coefficient in the interval 51-100, represents the weight coefficient of the interval 101-150, represents the weight coefficient of the interval 151-200, Indicates the weight coefficient in the interval 201-255; is the independent variable, indicating the number of pixel values ​​in the range of 0-50. is the independent variable, indicating the number of pixel values ​​in the range of 51-100. is the independent variable, indicating the number of pixel values ​​in the interval 101-150, is the independent variable, indicating the number of pixel values ​​in the interval 151-200, is an independent variable, indicating the number of pixel values ​​in the range of 201-255. is the intercept term.

[0116] Obtaining a predicted value of the net photosynthetic rate of the sample through the mathematical model;

[0117] The prediction error value of the mathematical model calculated based on the predicted value of net photosynthetic rate and the actual net photosynthetic rate includes:

[0118] Based on the predicted value of net photosynthetic rate pn i The actual net photosynthetic rate Pn i , by calculating the coefficient of determination R 2 Calculate the prediction error value of the mathematical model, the calculation formula is as follows:

[0119]

[0120] Among them, R 2 is the coefficient of determination, which is a key indicator to measure the degree of fit of the model to the data and reflects the ability of the model to explain the changes in the dependent variable;

[0121] The prediction error value calculated based on the above mathematical model is based on the weight coefficient

[0122] With the intercept term The mathematical model is adjusted to obtain an optimized mathematical model.

[0123] In addition, the error value of the model is predicted by calculating the relative error. The calculation formula is as follows:

[0124]

[0125] Where n is the number of samples in the calibration set. The relative error RE provides a standardized measure of the deviation between the predicted value and the true value, which is not affected by the data scale and is convenient for comparison across data sets of different scales.

[0126] This embodiment establishes an adjustment process for the mathematical model by calculating the determination coefficient and the relative error, which can evaluate the credibility and practicality of the model prediction from different levels, and then make necessary optimization and adjustments to the model to reduce the prediction error and improve the prediction accuracy. At the same time, a reasonable error tolerance range can be set by analyzing the two to help determine whether the model meets the needs of actual applications and to confirm the stability and effectiveness of the model.

[0127] The pixel value data of the test set samples are input into the mathematical model to obtain the predicted value of the net photosynthetic rate of the test set samples.

[0128] In one embodiment, the pixel value data of the leaf to be measured is input into the optimized mathematical model to predict its net photosynthetic rate, and the predicted value of the net photosynthetic rate of the leaf to be measured, pn i .

[0129] The technical solution provided by the present invention is further described in detail above through several specific implementation modes in order to highlight the advantages and benefits of the technical solution provided by the present invention. However, the several specific implementation modes described above are not intended to be used as limitations on the present invention. Any reasonable modification and improvement of the present invention, combination of implementation modes and equivalent substitution within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for constructing a leaf photosynthetic rate prediction model based on an image, characterized in that: Methods include: Preprocessing the acquired training image, including the step of dividing it into multiple intervals; Specifically, the blade to be measured is measured, the measurement area of ​​the blade after measurement is marked, and a photo is taken to obtain the image data after measurement as a training image, the training image is processed by Photoshop, a png image of the measurement area is intercepted in the initial image, and the green channel of the intercepted png image is extracted; the pixel value is divided into five intervals, 0-50 is the A interval, 51-100 is the B interval, 101-150 is the C interval, 151-200 is the D interval, and 201-255 is the E interval, and the number of pixels in each interval in the intercepted png image is counted; A step of counting the number of pixels in the subareas within a preset measurement area; Based on the least square method, a step of obtaining a quantitative relationship between the number of pixels and the photosynthetic rate; The step of constructing a photosynthetic rate prediction model based on the quantitative relationship.

2. The method for constructing a leaf photosynthetic rate prediction model based on an image according to claim 1, characterized in that: The preprocessing further includes extracting the green channel of the acquired image to be tested.

3. The method for constructing a leaf photosynthetic rate prediction model based on an image according to claim 1, characterized in that: Based on the least square method, a weight coefficient between the number of pixels and the photosynthetic rate is obtained, and the weight coefficient with the smallest error is used as the quantization relationship.

4. The method for constructing a leaf photosynthetic rate prediction model based on an image according to claim 3, characterized in that: Through the linear regression data processing method, the weight coefficient with the smallest fitting error is obtained.

5. A device for constructing a leaf photosynthetic rate prediction model based on an image, characterized in that: The device includes: Preprocessing the collected image to be tested, including a module for dividing it into multiple intervals; Specifically, the blade to be measured is measured, the measurement area of ​​the blade after measurement is marked, and a photo is taken to obtain the image data after measurement as a training image, the training image is processed by Photoshop, a png image of the measurement area is intercepted in the initial image, and the green channel of the intercepted png image is extracted; the pixel value is divided into five intervals, 0-50 is the A interval, 51-100 is the B interval, 101-150 is the C interval, 151-200 is the D interval, and 201-255 is the E interval, and the number of pixels in each interval in the intercepted png image is counted; A module for counting the number of pixels in the subareas within a preset measurement area; A module for obtaining a quantitative relationship between the number of pixels and the photosynthetic rate based on the least squares method; According to the quantitative relationship, a module of a photosynthetic rate prediction model is constructed.

6. A leaf photosynthetic rate prediction method based on an image, characterized in that: Methods include: According to the photosynthetic rate prediction model of claim 1, a step of predicting the photosynthetic rate of the collected image to be measured.

7. A leaf photosynthetic rate prediction device based on an image, characterized in that: The device includes: According to the photosynthetic rate prediction model of claim 1, a module for predicting the photosynthetic rate of the collected image to be measured.

8. A computer storage medium for storing a computer program, characterized in that: When the computer program is read by a computer, the computer executes the method of claim 1 .

9. A computer, comprising a processor and a storage medium, characterized in that: When the processor reads the computer program stored in the storage medium, the computer executes the method of claim 1 .

10. A computer program product, being a computer program, characterized in that When the computer program is read by a computer, the method according to claim 1 is implemented.

Images