Pesticide application effect evaluation method and system based on plant characteristic analysis

Abstract

The invention relates to the technical field of pesticide evaluation, in particular to a pesticide application effect evaluation method and system based on plant feature analysis, to-be-evaluated pesticide is applied to a plant of a preset disease spot type, and after a preset time period, an actual state feature image of a disease spot area after the to-be-evaluated pesticide is applied to the plant is obtained; obtaining the recovery state of the plant after the to-be-evaluated pesticide is applied according to the actual state feature image; if the plant is in a healthy growth state after the to-be-evaluated pesticide is applied, acquiring a health state grade of the plant after the to-be-evaluated pesticide is applied; if the plant is still in a germ infection state after the to-be-evaluated pesticide is applied, acquiring the scab expansion rate in the scab area; and evaluating the application pesticide effect of the to-be-evaluated pesticide according to the recovery state, the health state grade, the scab expansion rate and the actual growth rate of the plant after the to-be-evaluated pesticide is applied to the plant. According to the invention, the accuracy and objectivity of the evaluation result can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of pesticide evaluation, and in particular to a method and system for evaluating the effect of pesticide application based on plant characteristic analysis. Background Art

[0002] Traditional methods for evaluating the effects of pesticide application usually rely on subjective observation and empirical judgment, lack objective quantitative indicators, and are difficult to accurately reflect the actual effects of pesticides. For example, manual observation of changes in the area of ​​lesions is greatly affected by human factors, making it difficult to ensure the accuracy and consistency of the results. In addition, traditional methods often only focus on changes in the size of lesions, while ignoring other growth indicators of plants, and are unable to fully evaluate the effects of pesticides on plant growth.

[0003] In order to overcome the limitations of traditional methods, in recent years, pesticide application effect evaluation methods based on plant feature analysis have gradually emerged. This type of method uses modern image processing technology and machine learning algorithms to objectively and quantitatively analyze the growth status of plants, thereby more accurately evaluating the application effect of pesticides. For example, by analyzing the color, shape, size and texture characteristics of lesions, the expansion rate of lesions and the control effect of pesticides can be more accurately judged. At the same time, the effect of pesticides on plant growth can also be evaluated by analyzing the size, color, shape and other characteristics of plant leaves. In view of this, the present application proposes a method and system for evaluating the effect of pesticide application based on plant feature analysis. Summary of the invention

[0004] The present invention overcomes the shortcomings of traditional pesticide application effect evaluation methods that rely on subjective observation and empirical judgment, lack objective quantitative indicators, and are difficult to accurately reflect the actual effects of pesticides. A pesticide application effect evaluation method and system based on plant feature analysis is proposed.

[0005] To achieve the above object, the present invention proposes a method for evaluating the effect of pesticide application based on plant characteristic analysis, comprising the following steps:

[0006] Applying the pesticide to be evaluated to a plant with a preset disease spot type, obtaining an actual state characteristic image of the disease spot area of ​​the plant after the application of the pesticide to be evaluated after a preset time period, performing Haar wavelet transform processing on the actual state characteristic image, and obtaining a Haar wavelet coefficient matrix of the actual state characteristic image;

[0007] Constructing a health status knowledge base, importing the Haar wavelet coefficient matrix of the actual status feature image into the health status knowledge base for matching analysis, and obtaining the recovery status of the plant after the application of the pesticide to be evaluated;

[0008] If the plant is in a healthy growth state after the application of the pesticide to be evaluated, the health status level of the plant after the application of the pesticide to be evaluated is obtained; if the plant is still in a pathogen infection state after the application of the pesticide to be evaluated, the disease spot expansion rate in the disease spot area is obtained;

[0009] Obtaining a pre-application growth characteristic image of the plant before the application of the pesticide to be evaluated and a post-application growth characteristic image after the application of the pesticide to be evaluated, and calculating the actual growth rate of the plant after the application of the pesticide to be evaluated based on the pre-application growth characteristic image and the post-application growth characteristic image;

[0010] The application efficacy of the pesticide to be evaluated is evaluated based on the recovery state, health state level, lesion expansion rate and actual growth rate of the plant after the application of the pesticide to be evaluated, and an evaluation result is generated.

[0011] Preferably, the actual state feature image is subjected to Haar wavelet transform processing to obtain the Haar wavelet coefficient matrix of the actual state feature image, specifically:

[0012] Graying the actual state characteristic image of the lesion area to convert the color image into a grayscale image, normalizing the grayscale image to scale the pixel value to between 0 and 1, and obtaining the actual state characteristic image after preprocessing;

[0013] The preprocessed actual state feature image is decomposed step by step in a recursive manner until a predetermined number of decomposition layers is reached, and the decomposition level results after each step of decomposition are obtained; each decomposition divides the image into two sub-bands, including a low-frequency sub-band and a high-frequency sub-band; wherein the low-frequency sub-band corresponds to the overall information of the image, and the high-frequency sub-band corresponds to the detailed information of the image;

[0014] Performing Haar wavelet transform processing on the decomposition level results after each level of decomposition to obtain the Haar wavelet coefficients of each decomposition level; wherein the Haar wavelet coefficients represent the characteristic information of the image at different scales;

[0015] The Haar wavelet coefficient matrix of the actual state feature image is constructed according to the Haar wavelet coefficients of each decomposition level.

[0016] Preferably, a health status knowledge base is constructed, and the Haar wavelet coefficient matrix of the actual status feature image is imported into the health status knowledge base for matching analysis to obtain the recovery status of the plant after the application of the pesticide to be evaluated, specifically:

[0017] Obtain health status characteristic images corresponding to various health status levels of plants after they recover from infection with preset lesion types through a big data network; and perform Haar wavelet transform processing on the health status characteristic images corresponding to various health status levels to obtain the Haar wavelet coefficient matrix of each health status characteristic image;

[0018] Constructing a knowledge base, dividing the knowledge base into a plurality of storage spaces, and assigning a corresponding health status level label code to each storage space according to various health status levels;

[0019] The Haar wavelet coefficient matrices of the characteristic images of each health state are stored in the storage space assigned with the corresponding health state level label code to obtain a health state knowledge base;

[0020] Obtaining the Haar wavelet coefficient matrix of the actual state feature image, and calculating the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in each storage space in the health state knowledge base;

[0021] If the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in each storage space in the health state knowledge base is not greater than the preset threshold, it means that the plant is still in a state of pathogen infection after the application of the pesticide to be evaluated;

[0022] If the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in one or more storage spaces in the health state knowledge base is greater than a preset threshold, it means that the plant is in a healthy growth state after the application of the pesticide to be evaluated.

[0023] Preferably, if the plant is in a healthy growth state after the application of the pesticide to be evaluated, the health state level of the plant after the application of the pesticide to be evaluated is obtained, specifically:

[0024] If the plant is already in a healthy state after the application of the pesticide to be evaluated, the Haar wavelet coefficient matrix of the healthy state feature image with a similarity greater than a preset threshold is sorted to obtain the Haar wavelet coefficient matrix of the healthy state feature image with the greatest similarity;

[0025] And extract the health status grade label code of the storage space corresponding to the Haar wavelet coefficient matrix of the health status feature image with the greatest similarity, and determine the health status grade of the plant after applying the pesticide to be evaluated according to the extracted health status grade label code.

[0026] Preferably, if the plant is still in a state of being infected by pathogens after the application of the pesticide to be evaluated, the lesion expansion rate in the lesion area is obtained, specifically:

[0027] Acquire a first disease spot characteristic image of the disease spot area of ​​the plant before applying the pesticide to be evaluated; and acquire a second disease spot characteristic image of the plant after applying the pesticide to be evaluated;

[0028] Performing feature extraction processing on the first lesion feature image and the second lesion feature image based on the Canny edge detection algorithm to obtain a first lesion contour and a second lesion contour;

[0029] Calculate the areas enclosed by the first lesion outline and the second lesion outline respectively to obtain a first area value and a second area value;

[0030] The lesion expansion rate in the lesion area of ​​the plant after the pesticide to be evaluated is calculated based on the first area value, the second area value and the preset time period.

[0031] Preferably, the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in each storage space in the health state knowledge base is analyzed, specifically:

[0032] Randomly access the Haar wavelet coefficient matrix of the health state characteristic image in a certain storage space, and calculate the absolute value of the element difference between the Haar wavelet coefficient matrix of the health state characteristic image in the storage space and the Haar wavelet coefficient matrix of the actual state characteristic image at the same matrix position;

[0033] The absolute values ​​of the element differences at each matrix position are summed to obtain the Manhattan distance between the Haar wavelet coefficient matrix of the healthy state feature image in the storage space and the Haar wavelet coefficient matrix of the actual state feature image;

[0034] Determine the similarity between the Haar wavelet coefficient matrix of the healthy state feature image in the storage space and the Haar wavelet coefficient matrix of the actual state feature image according to the Manhattan distance;

[0035] Then randomly access the Haar wavelet coefficient matrix of the health status characteristic image in any storage space that has not been accessed, and repeat the above steps until the Haar wavelet coefficient matrices of the health status characteristic images in all storage spaces are ranged, and the similarity between the Haar wavelet coefficient matrix of the actual state characteristic image and the Haar wavelet coefficient matrix of the health status characteristic image in each storage space in the health status knowledge base is obtained.

[0036] Preferably, the pre-application growth characteristic image of the plant before the application of the pesticide to be evaluated and the post-application growth characteristic image after the application of the pesticide to be evaluated are obtained, and the actual growth rate of the plant after the application of the pesticide to be evaluated is calculated based on the pre-application growth characteristic image and the post-application growth characteristic image, specifically:

[0037] Obtaining a pre-pesticide application growth characteristic image corresponding to the plant before the pesticide to be evaluated is applied, and constructing a pre-pesticide application growth characteristic model diagram of the plant according to the pre-pesticide application growth characteristic image;

[0038] and obtaining a post-pesticide-application growth characteristic image of the plant after the pesticide to be evaluated is applied, and constructing a post-pesticide-application growth characteristic model diagram of the plant after the pesticide to be evaluated is applied according to the post-pesticide-application growth characteristic image;

[0039] Obtaining the pre-pesticide growth characteristic index of the plant before the application of the pesticide to be evaluated according to the pre-pesticide growth characteristic model diagram; and obtaining the post-pesticide growth characteristic index of the plant after the application of the pesticide to be evaluated according to the post-pesticide growth characteristic model diagram;

[0040] The actual growth rate of the plant after the application of the pesticide to be evaluated is calculated according to the growth characteristic index before application and the growth characteristic index after application;

[0041] Wherein, the growth characteristic indicators include leaf length, stem height and leaf area.

[0042] In addition, to achieve the above-mentioned purpose, the present invention also proposes a pesticide application effect evaluation system based on plant feature analysis, the pesticide application effect evaluation system includes a memory and a processor, the memory stores a pesticide application effect evaluation method program, when the pesticide application effect evaluation method program is executed by the processor, any one of the pesticide application effect evaluation method steps is implemented.

[0043] In addition, to achieve the above-mentioned purpose, the present invention also proposes a computer-readable storage medium, which includes a pesticide application effect evaluation method program. When the pesticide application effect evaluation method program is executed by a processor, any one of the pesticide application effect evaluation method steps is implemented.

[0044] In addition, to achieve the above-mentioned purpose, the present invention also proposes a pesticide application effect evaluation device based on plant feature analysis, which is applied to any of the pesticide application effect evaluation methods described above, comprising:

[0045] An image acquisition module, used to acquire the actual state characteristic image, the growth characteristic image before medication administration, and the growth characteristic image after medication administration;

[0046] An image processing module, used for processing the actual state characteristic image, the growth characteristic image before medication administration, and the growth characteristic image after medication administration;

[0047] The evaluation module is used to import the Haar wavelet coefficient matrix of the actual state characteristic image into the health state knowledge base for matching analysis.

[0048] The present invention solves the technical defects existing in the background technology, and has the following beneficial effects: applying the pesticide to be evaluated to a plant of a preset disease spot type, obtaining the actual state characteristic image of the disease spot area of ​​the plant after the application of the pesticide to be evaluated after a preset time period, and obtaining the recovery state of the plant after the application of the pesticide to be evaluated according to the actual state characteristic image; if the plant is in a healthy growth state after the application of the pesticide to be evaluated, then obtaining the health state grade of the plant after the application of the pesticide to be evaluated; if the plant is still in a pathogen infection state after the application of the pesticide to be evaluated, then obtaining the disease spot expansion rate in the disease spot area; and obtaining the actual growth rate of the plant after the application of the pesticide to be evaluated; evaluating and processing the application efficacy of the pesticide to be evaluated according to the recovery state, health state grade, disease spot expansion rate and actual growth rate of the plant after the application of the pesticide to be evaluated. The pesticide application effect evaluation method based on plant feature analysis can not only improve the accuracy and objectivity of the evaluation result, but also provide a more scientific basis for the research and development and application of pesticides, promote the rational use of pesticides, improve agricultural production efficiency, and protect the ecological environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0050] Figure 1 This is a flow chart of the first method of the pesticide application effect evaluation method;

[0051] Figure 2 This is a flow chart of the second method of the pesticide application effect evaluation method;

[0052] Figure 3 This is the system block diagram of the pesticide application effect evaluation system. DETAILED DESCRIPTION

[0053] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0054] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0055] The present invention proposes a method for evaluating the effect of pesticide application based on plant characteristic analysis, such as Figure 1 As shown, the following steps are included:

[0056] S102: applying the pesticide to be evaluated to a plant of a preset disease spot type, obtaining an actual state feature image of the disease spot area of ​​the plant after the application of the pesticide to be evaluated after a preset time period, performing Haar wavelet transform processing on the actual state feature image, and obtaining a Haar wavelet coefficient matrix of the actual state feature image;

[0057] S104: constructing a health status knowledge base, importing the Haar wavelet coefficient matrix of the actual status feature image into the health status knowledge base for matching analysis, and obtaining the recovery status of the plant after the application of the pesticide to be evaluated;

[0058] S106: If the plant is in a healthy growth state after the application of the pesticide to be evaluated, then the health state level of the plant after the application of the pesticide to be evaluated is obtained; if the plant is still in a pathogen infection state after the application of the pesticide to be evaluated, then the disease spot expansion rate in the disease spot area is obtained;

[0059] S108: Obtaining a pre-application growth characteristic image of the plant before the application of the pesticide to be evaluated and a post-application growth characteristic image after the application of the pesticide to be evaluated, and calculating the actual growth rate of the plant after the application of the pesticide to be evaluated according to the pre-application growth characteristic image and the post-application growth characteristic image;

[0060] S110: Evaluate the efficacy of the pesticide to be evaluated according to the recovery state, health state level, lesion expansion rate and actual growth rate of the plant after the application of the pesticide to be evaluated, and generate an evaluation result.

[0061] It should be noted that, first, plant plants with preset disease spot types are selected and the pesticide to be evaluated is applied to them. After applying the pesticide, wait for a preset time period to ensure that the pesticide fully acts on the plant and affects its disease spot condition. Next, obtain the actual state feature image of the disease spot area of ​​the plant after the application of the pesticide. These images contain the physical changes of the plant under the action of the pesticide, such as color, shape, size and other characteristics. Next, these actual state feature images are processed by Haar wavelet transform. Haar wavelet transform is a signal processing technology that can effectively extract key information from the image, such as edges, textures and other features. Through this process, a Haar wavelet coefficient matrix containing important information can be obtained. Then, a knowledge base of health status is established, and the obtained Haar wavelet coefficient matrix is ​​matched and analyzed with the data in the knowledge base. This knowledge base contains various feature information of healthy plants. By comparison, it can be judged whether the plant after the application of pesticides has recovered to a healthy state or whether there are still signs of bacterial infection. If the plant shows a healthy growth state after the application of pesticides, then its health status level is further evaluated to quantify the effect of the pesticide. On the contrary, if the plant is still in a state of pathogen infection, the evaluation will focus on the expansion rate of the lesion area, that is, the speed at which the lesions grow, which reflects the poor control effect of the pesticide. Finally, the growth characteristic images before and after the application of pesticides are obtained to calculate the actual growth rate of the plants. By comparing the changes in growth characteristics before and after application, the impact of pesticides on plant growth can be more comprehensively evaluated. By comprehensively considering the recovery state, health status level, lesion expansion rate and actual growth rate of the plants, the evaluated pesticide application efficacy is comprehensively evaluated and the final evaluation results are generated. Through quantitative analysis, this method can provide a scientific basis for the selection and application of pesticides, which helps to improve the efficiency of agricultural production and the health level of crops.

[0062] Among them, according to the recovery state, health status level, lesion expansion rate and actual growth rate of the plant after the application of the pesticide to be evaluated, the application efficacy of the pesticide to be evaluated is evaluated and processed, and the specific implementation steps for generating the evaluation result are as follows: First, according to the recovery state of the plant, judge whether the pesticide effectively controls the disease. If the plant recovers to a healthy state, the pesticide effect is good, and further quantitative evaluation can be performed according to the health status level. If the plant is still in a state of bacterial infection, it is necessary to judge the prevention and control effect of the pesticide according to the lesion expansion rate. Secondly, according to the lesion expansion rate, the inhibitory effect of the pesticide on the disease is quantitatively evaluated. The lower the lesion expansion rate, the better the prevention and control effect of the pesticide. Finally, according to the actual growth rate, the effect of the pesticide on plant growth is evaluated. The higher the growth rate, the more obvious negative impact the pesticide has on plant growth, and it may even promote growth. Taking the above indicators into consideration and combining with the preset evaluation criteria, the evaluation result of the application efficacy of the pesticide to be evaluated is finally generated.

[0063] The default evaluation criteria can be set in the following ways:

[0064] Disease spot recovery rate: If the plant recovers to health, the disease spot recovery rate is 100%; if the plant is still infected with bacteria, the disease spot recovery rate is 0%;

[0065] Health status grade scoring: According to the health status grade knowledge base, the health status grade is converted into a score, for example, the higher the health status grade, the higher the score;

[0066] Scoring of lesion expansion rate: according to the lesion expansion rate, set the scoring standard. For example, the lower the lesion expansion rate, the higher the score;

[0067] Growth rate scoring: Set scoring criteria based on the actual growth rate. For example, the higher the growth rate, the higher the score.

[0068] The comprehensive score is calculated based on the scores of each indicator and the weights assigned to it, for example:

[0069] Comprehensive score = (lesion recovery rate score * lesion recovery rate weight) + (health status grade score * health status grade weight) + (lesion expansion rate score * lesion expansion rate weight) + (growth rate score * growth rate weight)

[0070] According to the comprehensive score, the pesticide application effect is divided into different levels, such as:

[0071] Excellent: The overall score is higher than a certain threshold;

[0072] Good: The overall score is within a certain range;

[0073] Average: The overall score is within a certain range;

[0074] Poor: The overall score is below a certain threshold;

[0075] Present the evaluation results in a graphical or textual form, for example:

[0076] Pesticide name: certain pesticide;

[0077] Evaluation result: good;

[0078] Lesion recovery rate: 80%;

[0079] Health status rating: 90;

[0080] Lesion expansion rate score: 70;

[0081] Growth rate score: 95.

[0082] Preferably, the actual state feature image is subjected to Haar wavelet transform processing to obtain the Haar wavelet coefficient matrix of the actual state feature image, such as Figure 2 As shown, specifically:

[0083] S202: graying the actual state characteristic image of the lesion area to convert the color image into a grayscale image, normalizing the grayscale image to scale the pixel value to between 0 and 1, and obtaining a pre-processed actual state characteristic image;

[0084] S204: recursively decomposing the preprocessed actual state feature image level by level until a predetermined number of decomposition levels is reached, and obtaining decomposition level results after each level of decomposition; each decomposition divides the image into two sub-bands, including a low-frequency sub-band and a high-frequency sub-band; wherein the low-frequency sub-band corresponds to the overall information of the image, and the high-frequency sub-band corresponds to the detailed information of the image;

[0085] S206: performing Haar wavelet transform processing on the decomposition level results after each level of decomposition, and obtaining Haar wavelet coefficients of each decomposition level; wherein the Haar wavelet coefficients represent the characteristic information of the image at different scales;

[0086] S208: Constructing a Haar wavelet coefficient matrix of the actual state feature image according to the Haar wavelet coefficients of each decomposition level.

[0087] It should be noted that, first, the actual state feature image of the lesion area is grayed to convert the original color image into a single-channel gray image, which can simplify the subsequent processing steps and reduce the computational complexity. After that, the gray image is normalized, that is, all pixel values ​​are adjusted to between 0 and 1. This step helps to ensure the stability and consistency of the subsequent Haar wavelet transform. Then, the preprocessed gray image is decomposed step by step in a recursive manner until the predetermined number of decomposition layers is reached. In each decomposition process, the image is divided into two sub-bands, namely the low-frequency sub-band and the high-frequency sub-band. The low-frequency sub-band contains the overall information of the image, while the high-frequency sub-band contains the detailed information of the image, such as edges, textures, etc. This decomposition method can help us understand the structural characteristics of the image from different levels. Subsequently, the results of each decomposition level are processed by Haar wavelet transform to obtain the Haar wavelet coefficients of each decomposition level. The Haar wavelet coefficients represent the characteristic information of the image at different scales. They are the key values ​​for describing the image structure and are crucial for subsequent matching analysis and state evaluation. Finally, the Haar wavelet coefficient matrix of the actual state feature image is constructed based on the Haar wavelet coefficients of each decomposition level. This matrix is ​​a multidimensional array that contains the Haar wavelet coefficients of all decomposition levels and can be used for subsequent matching analysis and evaluation of plant recovery status, thus providing a scientific basis for the evaluation of pesticide application effects.

[0088] Preferably, a health status knowledge base is constructed, and the Haar wavelet coefficient matrix of the actual status feature image is imported into the health status knowledge base for matching analysis to obtain the recovery status of the plant after the application of the pesticide to be evaluated, specifically:

[0089] Obtain health status characteristic images corresponding to various health status levels of plants after they recover from infection with preset lesion types through a big data network; and perform Haar wavelet transform processing on the health status characteristic images corresponding to various health status levels to obtain the Haar wavelet coefficient matrix of each health status characteristic image;

[0090] Constructing a knowledge base, dividing the knowledge base into a plurality of storage spaces, and assigning a corresponding health status level label code to each storage space according to various health status levels;

[0091] The Haar wavelet coefficient matrices of the characteristic images of each health state are stored in the storage space assigned with the corresponding health state level label code to obtain a health state knowledge base;

[0092] Obtaining the Haar wavelet coefficient matrix of the actual state feature image, and calculating the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in each storage space in the health state knowledge base;

[0093] If the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in each storage space in the health state knowledge base is not greater than the preset threshold, it means that the plant is still in a state of pathogen infection after the application of the pesticide to be evaluated;

[0094] If the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in one or more storage spaces in the health state knowledge base is greater than a preset threshold, it means that the plant is in a healthy growth state after the application of the pesticide to be evaluated.

[0095] It should be noted that various health status characteristic images of plants with different health status levels in the recovery stage after being infected with a specific lesion type are collected and sorted out through the big data network. These images contain the growth characteristics of plants at different recovery stages, such as color, shape, size, etc. Then, these health status characteristic images are processed by Haar wavelet transform to extract key information in the image, such as edge, texture and other features, to form their own Haar wavelet coefficient matrices. These matrices represent the characteristic information of plants at different health status levels. Next, a health status knowledge base is constructed, and the Haar wavelet coefficient matrices obtained above are classified and stored according to different health status levels. Each storage space corresponds to a health status level label code, which is convenient for subsequent retrieval and matching analysis. When it is necessary to evaluate the recovery state of a plant after the application of the pesticide to be evaluated, the actual state characteristic image of the plant after the application of the pesticide is first obtained, and the Haar wavelet transform is also performed on it to obtain its Haar wavelet coefficient matrix. Next, the Haar wavelet coefficient matrix of this actual state characteristic image is similar to the Haar wavelet coefficient matrix in all storage spaces in the health status knowledge base. Here, a threshold is preset as a judgment standard. If the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix in all storage spaces in the health state knowledge base is lower than the preset threshold, it means that the plant is still in a state of pathogen infection after the application of pesticides, and its recovery is not ideal. On the contrary, if the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix in one or more storage spaces in the health state knowledge base is higher than the preset threshold, it indicates that the plant has entered a healthy state after the application of pesticides, and its recovery effect is good. This method can objectively and quantitatively evaluate the control effect of pesticides on plant diseases.

[0096] Preferably, if the plant is in a healthy growth state after the application of the pesticide to be evaluated, the health state level of the plant after the application of the pesticide to be evaluated is obtained, specifically:

[0097] If the plant is already in a healthy state after the application of the pesticide to be evaluated, the Haar wavelet coefficient matrix of the healthy state feature image with a similarity greater than a preset threshold is sorted to obtain the Haar wavelet coefficient matrix of the healthy state feature image with the greatest similarity;

[0098] And extract the health status grade label code of the storage space corresponding to the Haar wavelet coefficient matrix of the health status feature image with the greatest similarity, and determine the health status grade of the plant after applying the pesticide to be evaluated according to the extracted health status grade label code.

[0099] It should be noted that once it is detected that the plant has recovered to a healthy state after the application of the pesticide to be evaluated, the next operation is to find the health state characteristic image that is closest to the current plant state. This involves sorting the Haar wavelet coefficient matrix of the health state characteristic image to find the image with the highest similarity to the current plant state. Here, the degree of similarity is measured based on the comparison between the Haar wavelet coefficient matrix of each health state characteristic image and the Haar wavelet coefficient matrix of the current plant state characteristic image in the previously constructed health state knowledge base. After finding the health state characteristic image with the highest degree of similarity, the next step is to identify the storage space to which the image belongs and its corresponding health state level label code. These label codes are pre-set when constructing the health state knowledge base and are used to identify plant characteristic images of different health state levels. Finally, according to the extracted health state level label code, the specific health state level of the plant after the application of the pesticide to be evaluated can be determined. This method realizes the accurate quantitative evaluation of the health state of the plant, can quickly and accurately judge the application effect of the pesticide, and provides scientific guidance for the selection and use of pesticides, thereby promoting the improvement of agricultural production efficiency and the rational use of resources.

[0100] Preferably, if the plant is still in a state of being infected by pathogens after the application of the pesticide to be evaluated, the lesion expansion rate in the lesion area is obtained, specifically:

[0101] Acquire a first disease spot characteristic image of the disease spot area of ​​the plant before applying the pesticide to be evaluated; and acquire a second disease spot characteristic image of the plant after applying the pesticide to be evaluated;

[0102] Performing feature extraction processing on the first lesion feature image and the second lesion feature image based on the Canny edge detection algorithm to obtain a first lesion contour and a second lesion contour;

[0103] Calculate the areas enclosed by the first lesion outline and the second lesion outline respectively to obtain a first area value and a second area value;

[0104] The lesion expansion rate in the lesion area of ​​the plant after the pesticide to be evaluated is calculated based on the first area value, the second area value and the preset time period.

[0105] It should be noted that, first, a first lesion characteristic image of the lesion area of ​​the plant before the application of the pesticide to be evaluated and a second lesion characteristic image after the application of the pesticide are obtained. These two images record the state of the lesion area of ​​the same plant before and after the application of the pesticide, providing basic data for subsequent analysis. Then, the two lesion characteristic images are processed using the Canny edge detection algorithm, the purpose of which is to extract the edge information in the image, and then obtain the first lesion outline and the second lesion outline. The Canny algorithm is an efficient algorithm widely used in image edge detection, which can accurately identify the boundaries in the image. Then, the area surrounded by the first lesion outline and the second lesion outline is calculated to obtain the first area value and the second area value respectively. These two area values ​​reflect the size of the lesion area before and after the application of the pesticide. Finally, the lesion expansion rate is calculated based on the first area value, the second area value and the preset time period. The lesion expansion rate is an important indicator to measure the growth rate of the lesion area, which can be obtained by comparing the changes in the lesion area before and after the application of the pesticide, and the time span of the change. If the lesion expansion rate is low, it means that the applied pesticide has a good control effect on the pathogen; otherwise, it means that the control effect of the pesticide is poor. This method can intuitively and objectively evaluate the application effect of the pesticide to be evaluated by quantitatively analyzing the lesion expansion rate.

[0106] Preferably, the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in each storage space in the health state knowledge base is analyzed, specifically:

[0107] Randomly access the Haar wavelet coefficient matrix of the health state characteristic image in a certain storage space, and calculate the absolute value of the element difference between the Haar wavelet coefficient matrix of the health state characteristic image in the storage space and the Haar wavelet coefficient matrix of the actual state characteristic image at the same matrix position;

[0108] The absolute values ​​of the element differences at each matrix position are summed to obtain the Manhattan distance between the Haar wavelet coefficient matrix of the healthy state feature image in the storage space and the Haar wavelet coefficient matrix of the actual state feature image;

[0109] Determine the similarity between the Haar wavelet coefficient matrix of the healthy state feature image in the storage space and the Haar wavelet coefficient matrix of the actual state feature image according to the Manhattan distance;

[0110] Then randomly access the Haar wavelet coefficient matrix of the health status characteristic image in any storage space that has not been accessed, and repeat the above steps until the Haar wavelet coefficient matrices of the health status characteristic images in all storage spaces are ranged, and the similarity between the Haar wavelet coefficient matrix of the actual state characteristic image and the Haar wavelet coefficient matrix of the health status characteristic image in each storage space in the health status knowledge base is obtained.

[0111] It should be noted that, by randomly selecting a storage space in the health status knowledge base, the Haar wavelet coefficient matrix of a health status feature image in the storage space is accessed. Then, the absolute value of the difference between the elements of this matrix and the Haar wavelet coefficient matrix of the actual state feature image at the same position is calculated. Then, these differences are added to obtain the Manhattan distance between the Haar wavelet coefficient matrix of the health status feature image in the storage space and the Haar wavelet coefficient matrix of the actual state feature image. Manhattan distance is a measurement method for measuring the difference between two vectors, which is applicable to the sum of the absolute differences of the values ​​of each dimension of the vector. According to the Manhattan distance, the similarity between the Haar wavelet coefficient matrix of the health status feature image in the storage space and the actual state feature image can be judged. The smaller the Manhattan distance, the more similar the two are. Subsequently, another storage space that has not been accessed is randomly selected, and the above steps are repeated until the Haar wavelet coefficient matrices of the health status feature images in all storage spaces are accessed and compared. In this way, the similarity between the actual state feature image and all the health status feature images in the health status knowledge base can be comprehensively evaluated. By quantitatively analyzing the similarity between the actual state feature image and the health state feature image in each storage space in the health state knowledge base, the recovery state of the plant after the application of the pesticide to be evaluated can be objectively and accurately evaluated.

[0112] Preferably, the pre-application growth characteristic image of the plant before the application of the pesticide to be evaluated and the post-application growth characteristic image after the application of the pesticide to be evaluated are obtained, and the actual growth rate of the plant after the application of the pesticide to be evaluated is calculated based on the pre-application growth characteristic image and the post-application growth characteristic image, specifically:

[0113] Obtaining a pre-pesticide application growth characteristic image corresponding to the plant before the pesticide to be evaluated is applied, and constructing a pre-pesticide application growth characteristic model diagram of the plant according to the pre-pesticide application growth characteristic image;

[0114] and obtaining a post-pesticide-application growth characteristic image of the plant after the pesticide to be evaluated is applied, and constructing a post-pesticide-application growth characteristic model diagram of the plant after the pesticide to be evaluated is applied according to the post-pesticide-application growth characteristic image;

[0115] Obtaining the pre-pesticide growth characteristic index of the plant before the application of the pesticide to be evaluated according to the pre-pesticide growth characteristic model diagram; and obtaining the post-pesticide growth characteristic index of the plant after the application of the pesticide to be evaluated according to the post-pesticide growth characteristic model diagram;

[0116] The actual growth rate of the plant after the application of the pesticide to be evaluated is calculated according to the growth characteristic index before application and the growth characteristic index after application;

[0117] Wherein, the growth characteristic indicators include leaf length, stem height and leaf area.

[0118] It should be noted that by acquiring and analyzing the growth characteristic images of the plants before the application of the pesticide to be evaluated, a growth characteristic model diagram of the plants before the application of the pesticide is constructed. This step is intended to capture the growth state of the plants before the application of the pesticide and provide a benchmark for subsequent comparisons. Next, the growth characteristic images of the plants after the application of the pesticide to be evaluated are acquired and analyzed, and a growth characteristic model diagram of the plants after the application of the pesticide is constructed. Through these images, the growth changes of the plants after the application of the pesticide can be understood. Then, the growth characteristic indicators of the plants before the application of the pesticide, such as leaf length, stem height, and leaf area, are extracted from the growth characteristic model diagram before the application of the pesticide. These indicators reflect the growth state of the plants before the application of the pesticide. Similarly, the growth characteristic indicators of the plants after the application of the pesticide are extracted from the growth characteristic model diagram after the application of the pesticide. By comparing the growth characteristic indicators before and after the application of the pesticide, the growth changes of the plants after the application of the pesticide can be evaluated. Finally, the actual growth rate of the plants after the application of the pesticide to be evaluated is calculated based on the growth characteristic indicators before and after the application of the pesticide. The growth rate is an important indicator to measure the growth rate of plants. By comparing the growth rates before and after the application of the pesticide, the impact of the pesticide on the growth of the plant can be intuitively evaluated. This method can objectively and accurately evaluate the growth changes of plants after the application of the pesticide to be evaluated by quantitatively comparing the growth characteristic indicators before and after the application of the pesticide.

[0119] The present invention also proposes a pesticide application effect evaluation system based on plant feature analysis, such as Figure 3 As shown, the pesticide application effect evaluation system includes a memory 20 and a processor 60. The memory 20 stores a pesticide application effect evaluation method program. When the pesticide application effect evaluation method program is executed by the processor 60, any one of the pesticide application effect evaluation method steps is implemented.

[0120] The present invention also proposes a computer-readable storage medium, which includes a pesticide application effect evaluation method program. When the pesticide application effect evaluation method program is executed by a processor, any one of the pesticide application effect evaluation method steps is implemented.

[0121] The present invention also proposes a pesticide application effect evaluation device based on plant feature analysis, which is applied to any of the pesticide application effect evaluation methods described above, comprising:

[0122] An image acquisition module, used to acquire the actual state characteristic image, the growth characteristic image before medication administration, and the growth characteristic image after medication administration;

[0123] An image processing module, used for processing the actual state characteristic image, the growth characteristic image before medication administration, and the growth characteristic image after medication administration;

[0124] The evaluation module is used to import the Haar wavelet coefficient matrix of the actual state characteristic image into the health state knowledge base for matching analysis.

[0125] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0126] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0127] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0128] Those skilled in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), disks or optical disks, and other media that can store program codes.

[0129] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0130] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for evaluating the effect of pesticide application based on plant characteristic analysis, characterized in that: The following steps are involved: Applying the pesticide to be evaluated to a plant with a preset disease spot type, obtaining an actual state characteristic image of the disease spot area of ​​the plant after the application of the pesticide to be evaluated after a preset time period, performing Haar wavelet transform processing on the actual state characteristic image, and obtaining a Haar wavelet coefficient matrix of the actual state characteristic image; Constructing a health status knowledge base, importing the Haar wavelet coefficient matrix of the actual status feature image into the health status knowledge base for matching analysis, and obtaining the recovery status of the plant after the application of the pesticide to be evaluated; If the plant is in a healthy growth state after the application of the pesticide to be evaluated, the health status level of the plant after the application of the pesticide to be evaluated is obtained; if the plant is still in a pathogen infection state after the application of the pesticide to be evaluated, the disease spot expansion rate in the disease spot area is obtained; Obtaining a pre-application growth characteristic image of the plant before the application of the pesticide to be evaluated and a post-application growth characteristic image after the application of the pesticide to be evaluated, and calculating the actual growth rate of the plant after the application of the pesticide to be evaluated based on the pre-application growth characteristic image and the post-application growth characteristic image; The application efficacy of the pesticide to be evaluated is evaluated based on the recovery state, health state level, lesion expansion rate and actual growth rate of the plant after the application of the pesticide to be evaluated, and an evaluation result is generated.

2. A method for evaluating the effect of pesticide application based on plant characteristic analysis according to claim 1, characterized in that: Performing Haar wavelet transform processing on the actual state feature image to obtain the Haar wavelet coefficient matrix of the actual state feature image, specifically: Graying the actual state characteristic image of the lesion area to convert the color image into a grayscale image, normalizing the grayscale image to scale the pixel value to between 0 and 1, and obtaining the actual state characteristic image after preprocessing; Decomposing the preprocessed actual state feature image level by level in a recursive manner until a predetermined number of decomposition levels is reached, and obtaining decomposition level results after each level of decomposition; Each decomposition divides the image into two sub-bands, including a low-frequency sub-band and a high-frequency sub-band; the low-frequency sub-band corresponds to the overall information of the image, and the high-frequency sub-band corresponds to the detailed information of the image; Performing Haar wavelet transform processing on the decomposition level results after each level of decomposition to obtain the Haar wavelet coefficients of each decomposition level; wherein the Haar wavelet coefficients represent the characteristic information of the image at different scales; The Haar wavelet coefficient matrix of the actual state feature image is constructed according to the Haar wavelet coefficients of each decomposition level.

3. A method for evaluating the effect of pesticide application based on plant characteristic analysis according to claim 1, characterized in that: A health status knowledge base is constructed, and the Haar wavelet coefficient matrix of the actual status feature image is imported into the health status knowledge base for matching analysis to obtain the recovery status of the plant after the application of the pesticide to be evaluated, specifically: Obtain health status characteristic images corresponding to various health status levels of plants after they recover from infection with preset lesion types through a big data network; and perform Haar wavelet transform processing on the health status characteristic images corresponding to various health status levels to obtain the Haar wavelet coefficient matrix of each health status characteristic image; Constructing a knowledge base, dividing the knowledge base into a plurality of storage spaces, and assigning a corresponding health status level label code to each storage space according to various health status levels; The Haar wavelet coefficient matrices of the characteristic images of each health state are stored in the storage space assigned with the corresponding health state level label code to obtain a health state knowledge base; Obtaining the Haar wavelet coefficient matrix of the actual state feature image, and calculating the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in each storage space in the health state knowledge base; If the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in each storage space in the health state knowledge base is not greater than the preset threshold, it means that the plant is still in a state of pathogen infection after the application of the pesticide to be evaluated; If the similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in one or more storage spaces in the health state knowledge base is greater than a preset threshold, it means that the plant is in a healthy growth state after the application of the pesticide to be evaluated.

4. A method for evaluating the effect of pesticide application based on plant characteristic analysis according to claim 3, characterized in that: If the plant is in a healthy growth state after the application of the pesticide to be evaluated, the health state level of the plant after the application of the pesticide to be evaluated is obtained, specifically: If the plant is already in a healthy state after the application of the pesticide to be evaluated, the Haar wavelet coefficient matrix of the healthy state feature image with a similarity greater than a preset threshold is sorted to obtain the Haar wavelet coefficient matrix of the healthy state feature image with the greatest similarity; And extract the health status grade label code of the storage space corresponding to the Haar wavelet coefficient matrix of the health status feature image with the greatest similarity, and determine the health status grade of the plant after applying the pesticide to be evaluated according to the extracted health status grade label code.

5. The method for evaluating the effect of pesticide application based on plant characteristic analysis according to claim 3, characterized in that: If the plant is still infected with the pathogen after the application of the pesticide to be evaluated, the lesion expansion rate in the lesion area is obtained, specifically: Acquire a first disease spot characteristic image of the disease spot area of ​​the plant before applying the pesticide to be evaluated; and acquire a second disease spot characteristic image of the plant after applying the pesticide to be evaluated; Performing feature extraction processing on the first lesion feature image and the second lesion feature image based on the Canny edge detection algorithm to obtain a first lesion contour and a second lesion contour; Calculate the areas enclosed by the first lesion outline and the second lesion outline respectively to obtain a first area value and a second area value; The lesion expansion rate in the lesion area of ​​the plant after the pesticide to be evaluated is calculated based on the first area value, the second area value and the preset time period.

6. The method for evaluating the effect of pesticide application based on plant characteristic analysis according to claim 3, characterized in that: The similarity between the Haar wavelet coefficient matrix of the actual state feature image and the Haar wavelet coefficient matrix of the health state feature image in each storage space in the health state knowledge base is analyzed, specifically: Randomly access the Haar wavelet coefficient matrix of the health state characteristic image in a certain storage space, and calculate the absolute value of the element difference between the Haar wavelet coefficient matrix of the health state characteristic image in the storage space and the Haar wavelet coefficient matrix of the actual state characteristic image at the same matrix position; The absolute values ​​of the element differences at each matrix position are summed to obtain the Manhattan distance between the Haar wavelet coefficient matrix of the healthy state feature image in the storage space and the Haar wavelet coefficient matrix of the actual state feature image; Determine the similarity between the Haar wavelet coefficient matrix of the healthy state feature image in the storage space and the Haar wavelet coefficient matrix of the actual state feature image according to the Manhattan distance; Then randomly access the Haar wavelet coefficient matrix of the health status characteristic image in any storage space that has not been accessed, and repeat the above steps until the Haar wavelet coefficient matrices of the health status characteristic images in all storage spaces are ranged, and the similarity between the Haar wavelet coefficient matrix of the actual state characteristic image and the Haar wavelet coefficient matrix of the health status characteristic image in each storage space in the health status knowledge base is obtained.

7. The method for evaluating the effect of pesticide application based on plant characteristic analysis according to claim 1, characterized in that: Obtain the pre-application growth characteristic image of the plant before the application of the pesticide to be evaluated and the post-application growth characteristic image after the application of the pesticide to be evaluated, and calculate the actual growth rate of the plant after the application of the pesticide to be evaluated according to the pre-application growth characteristic image and the post-application growth characteristic image, specifically: Obtaining a pre-pesticide application growth characteristic image corresponding to the plant before the pesticide to be evaluated is applied, and constructing a pre-pesticide application growth characteristic model diagram of the plant according to the pre-pesticide application growth characteristic image; and obtaining a post-pesticide-application growth characteristic image of the plant after the pesticide to be evaluated is applied, and constructing a post-pesticide-application growth characteristic model diagram of the plant after the pesticide to be evaluated is applied according to the post-pesticide-application growth characteristic image; Obtaining the pre-pesticide growth characteristic index of the plant before the application of the pesticide to be evaluated according to the pre-pesticide growth characteristic model diagram; and obtaining the post-pesticide growth characteristic index of the plant after the application of the pesticide to be evaluated according to the post-pesticide growth characteristic model diagram; The actual growth rate of the plant after the application of the pesticide to be evaluated is calculated according to the growth characteristic index before application and the growth characteristic index after application; Wherein, the growth characteristic indicators include leaf length, stem height and leaf area.

8. A pesticide application effect evaluation system based on plant feature analysis, characterized in that: The pesticide application effect evaluation system includes a memory and a processor, wherein a pesticide application effect evaluation method program is stored in the memory, and when the pesticide application effect evaluation method program is executed by the processor, the pesticide application effect evaluation method steps as described in any one of claims 1 to 7 are implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a pesticide application effect evaluation method program. When the pesticide application effect evaluation method program is executed by a processor, the pesticide application effect evaluation method steps as described in any one of claims 1 to 7 are implemented.

10. A pesticide application effect evaluation device based on plant feature analysis, applied to the pesticide application effect evaluation method according to any one of claims 1 to 7, characterized in that: include: An image acquisition module, used to acquire the actual state characteristic image, the growth characteristic image before medication administration, and the growth characteristic image after medication administration; An image processing module, used for processing the actual state characteristic image, the growth characteristic image before medication administration, and the growth characteristic image after medication administration; The evaluation module is used to import the Haar wavelet coefficient matrix of the actual state characteristic image into the health state knowledge base for matching analysis.

Images