A disease early warning method and early warning system for blueberries

Abstract

The application discloses a disease early warning method and system for blueberries, which comprises the following steps: firstly, setting a first standard judgment parameter, a second standard judgment parameter and a disease judgment function for disease early warning; then, acquiring an infrared temperature image and a visible light image of a blueberry leaf to be detected; acquiring a temperature abnormal area coordinate set according to the infrared temperature image and a color abnormal area coordinate set according to the visible light image; then, screening the color abnormal area coordinate set with the temperature abnormal area coordinate set to obtain a disease area coordinate set; finally, outputting early warning information according to the disease area coordinate set and the disease judgment model. When leaf spot disease occurs, it will cause abnormal changes in the blueberry leaf and metabolic temperature. The application realizes early warning of the blueberry leaf spot disease through quantitative calculation and cross-validation, and the determination result is more accurate and objective compared with the subjective judgment mode which depends on the experience of workers.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent agriculture, in particular to a disease early warning method and system for blueberries. BACKGROUND

[0002] Blueberry is a plant of the genus Vaccinium in the subfamily Vaccinioideae of the family Ericaceae, also known as bilberry or blueberry. It is a perennial deciduous or evergreen shrub. Blueberries not only contain conventional sugars, proteins, fruit acids and Vc, but also are rich in Ve, Va, Vb, SOD, arbutin, anthocyanins, dietary fiber and rich mineral elements such as K, Fe, Zn and Ca. Studies have shown that regular consumption of blueberries and their products can help prevent the occurrence of colon cancer, significantly enhance vision, heart function, skin nutrition, delay brain aging, prevent Alzheimer's disease, and have good therapeutic effect on capillary lesions caused by diabetes. It is a characteristic fruit with economic and nutritional value.

[0003] Blueberry leaf spot is a common fungal or bacterial disease in blueberry cultivation, mainly harming leaves, and can affect photosynthesis of the plant, leading to tree weakness, yield reduction, and even inducing other diseases, which is one of the important factors restricting the high quality and high yield of blueberries.

[0004] In the prior art, the judgment of blueberry leaf spot is generally carried out by manual observation. For a large-scale blueberry plantation, the above-mentioned method has low early warning efficiency, which is not conducive to the early detection and prevention of leaf spot. At the same time, the manual observation method is highly dependent on the work experience of workers, and the judgment result is the subjective feeling of workers, which may lead to misjudgment, and the recognition accuracy is low. SUMMARY

[0005] The main purpose of the present application is to provide a disease early warning method and system for blueberries, which aims to solve the problem of low recognition accuracy in the prior art.

[0006] The above-mentioned purpose is achieved by the following technical solutions:

[0007] A disease early warning method for blueberries, comprising the following steps:

[0008] Setting a first standard judgment parameter, a second standard judgment parameter and a disease judgment function for blueberry disease early warning;

[0009] Obtaining an infrared temperature image and a visible light image of a blueberry leaf to be detected;

[0010] Converting the infrared temperature image into a gray-scale image, and extracting a temperature anomaly region coordinate set S T from the gray-scale image according to the first standard judgment parameter;

[0011] Based on the second standard judgment parameter, the coordinate set S of the color anomaly region is extracted from the visible light image. C ;

[0012] According to the coordinate set S of the temperature anomaly zone T For the coordinate set S of the color anomaly region C Screening is performed to obtain the coordinate set S of the diseased area. D ;

[0013] According to the coordinate set S of the diseased area D The disease determination model outputs early warning information.

[0014] Optionally, the first standard judgment parameter includes the standard judgment grayscale domain, and the second standard judgment parameter includes the standard judgment RGB domain.

[0015] Optionally, the infrared temperature image is converted into a grayscale image, and the coordinate set S of the temperature anomaly area is extracted from the grayscale image according to the first standard judgment parameter. T This includes the following steps:

[0016] The infrared temperature image is converted into a grayscale image, and a first standard coordinate system is constructed for the grayscale image;

[0017] The grayscale image is divided into several pixels, and the grayscale parameters (x) of each pixel are obtained respectively. i y i G i ); where x i y i This represents the first coordinate of a pixel, where i represents the pixel number, and G... i Indicates the actual grayscale value;

[0018] Obtain the first reference coordinates (x, y) of each calibration point in the first coordinate system. ok y ok The number of standard points is no less than three, and k represents the number of the calibration point.

[0019] Each of the grayscale parameters is compared with the first standard judgment parameter to screen out abnormal grayscale values.

[0020] The first coordinates corresponding to each abnormal gray value are aggregated to generate the coordinate set S of the temperature anomaly zone. T .

[0021] Optionally, extracting the coordinate set of color anomaly regions from the visible light image according to the second standard determination parameter includes the following steps:

[0022] The visible light image is converted into an RGB image, and a second standard coordinate system is constructed for the RGB image;

[0023] segmenting the RGB image into a plurality of pixels, and acquiring RGB parameters (x j , y j , r j , g j , b j ) of each pixel, wherein x i , y i represent the second coordinates of the pixel, j represents the pixel number, and r j , g j , b j are actual RGB values;

[0024] acquiring second reference coordinates (x' ok , y' ok ) of each calibration point in the second coordinate system; wherein the standard points are not less than 3;

[0025] comparing each actual RGB value with the second standard determination parameter, and screening to obtain abnormal RGB values;

[0026] collecting the second coordinates corresponding to each abnormal RGB value to generate a color abnormal area coordinate set S C .

[0027] Optionally, the color abnormal area coordinate set S T is screened according to the temperature abnormal area coordinate set S C to obtain a disease area coordinate set S D , including the following steps:

[0028] coordinate registration is performed according to the first reference coordinates (x ok , y ok ) and the second reference coordinates (x' ok , y' ok ), and a position conversion relationship between the temperature abnormal area coordinate set and the color abnormal area coordinate set is obtained;

[0029] the temperature abnormal area coordinate set S T is converted into a temperature determination coordinate set S' T according to the position conversion relationship;

[0030] the color abnormal area coordinate set S C is screened according to the temperature determination coordinate set S' T for the first time to obtain a first disease coordinate set S D1 and a to-be-determined disease coordinate set S' D , wherein the expression of the first disease coordinate set S D1 is , and the expression of the to-be-determined disease coordinate set S' D is ;

[0031] screening the first lesion coordinate set S D to obtain a second lesion coordinate set S D2 ;

[0032] merging the first lesion coordinate set S D1 and the second lesion coordinate set S D2 to obtain a disease area coordinate set.

[0033] Optionally, a position conversion relationship between the temperature anomaly area coordinate set and the color anomaly area coordinate set is obtained by performing coordinate registration according to the first reference coordinates and the second reference coordinates, including the following steps:

[0034] constructing a standard expression of the position conversion relationship;

[0035] constructing a one-to-one mapping between each of the first reference coordinates and each of the second reference coordinates to generate a plurality of fitting calculation groups {(x ok , y ok ), (x' ok , y' ok )};

[0036] respectively introducing each of the fitting calculation groups into the standard expression to obtain the position conversion relationship by fitting calculation, wherein an expression of the position conversion relationship is , wherein a, b, c, d, e, and f respectively represent constants.

[0037] Optionally, the first lesion coordinate set S D is screened again to obtain a second lesion coordinate set S D2 , including the following steps:

[0038] extracting any first lesion coordinate from the first lesion coordinate set S D ;

[0039] collecting coordinates of all adjacent pixel points to generate an adjacent pixel coordinate set with a pixel point corresponding to the first lesion coordinate as a center; wherein a calculation expression of the adjacent pixel coordinate set is

[0040] ;

[0041] comparing the adjacent pixel coordinate set with the color anomaly area coordinate set S C ; if there is an intersection, the first lesion coordinate meets the secondary screening, and the first lesion coordinate is collected into the second lesion coordinate set;

[0042] repeating the step of extracting any first lesion coordinate from the first lesion coordinate set S D to generate the second lesion coordinate set SD2 .

[0043] Optionally, according to the disease area coordinate set S D and the disease determination model outputs early warning information, comprising the following steps:

[0044] Obtaining a grayscale image, and extracting a leaf shape graph of the blueberry leaf to be detected from the grayscale image according to a background grayscale value;

[0045] According to the leaf shape graph, calculating a total number Q of pixel points of the blueberry leaf to be detected;

[0046] The total number of elements of the disease area coordinate set S D is taken as the number Q1 of diseased pixel points;

[0047] According to the total number of pixel points and the number of diseased pixel points, calculating a diseased proportion;

[0048] According to the diseased proportion and a disease determination function, outputting early warning information.

[0049] Optionally, the expression of the diseased proportion is q = 100% Q1 / Q, and the expression of the disease determination function is , wherein R represents the diseased proportion, and k1, k2 and k3 are constants.

[0050] Correspondingly, the application also discloses an early warning system based on the above test method, comprising:

[0051] A parameter setting module is configured to set a first standard determination parameter, a second standard determination parameter and a disease determination function for blueberry disease early warning;

[0052] A data acquisition module is configured to acquire an infrared temperature image and a visible light image of a blueberry leaf to be detected;

[0053] A first identification module is configured to convert the infrared temperature image into a grayscale image, and extract a temperature abnormal area coordinate set from the grayscale image according to the first standard determination parameter;

[0054] A second identification module is configured to extract a color abnormal area coordinate set from the visible light image according to the second standard determination parameter;

[0055] A screening calculation module is configured to screen the color abnormal area coordinate set according to the temperature abnormal area coordinate set, and acquire a disease area coordinate set;

[0056] An early warning information output module is configured to output early warning information according to the disease area coordinate set and the disease determination model.

[0057] Compared with the prior art, the application has the following beneficial effects:

[0058] The application first sets a first standard determination parameter, a second standard determination parameter and a disease determination function for disease early warning, then acquires an infrared temperature image and a visible light image of a blueberry leaf to be detected; acquires a temperature abnormal area coordinate set according to the infrared temperature image, acquires a color abnormal area coordinate set according to the visible light image, then screens the color abnormal area coordinate set with the temperature abnormal area coordinate set to obtain a disease area coordinate set, and finally outputs early warning information according to the disease area coordinate set and the disease determination model;

[0059] When the blueberry has leaf spot disease, the pathogen will destroy the cell structure and chlorophyll synthesis of the leaf, causing characteristic color changes in the diseased area: initially, it may be a small yellow spot, and as the disease progresses, it becomes brown, dark purple or gray-white spots, forming a clear color difference with the healthy area. This difference can be captured by an RGB image, showing that the R, G and B channel values deviate from the standard range of healthy leaves;

[0060] Pathogen infection can trigger local defense responses in leaves (such as increased cell respiration and changes in enzyme activity), leading to an increase in energy metabolism rate in the diseased area, with a higher temperature than the healthy area. In severe spot areas, cells are necrotic, lose normal physiological functions, transpiration is weakened, and heat loss is reduced, which may also appear as a local temperature rise (the temperature difference with the surrounding healthy tissue can be detected by infrared imaging). Therefore, the temperature of the diseased area will deviate significantly from the normal temperature range of healthy leaves, and this feature can be quantitatively captured by an infrared temperature image (converted to a grayscale image);

[0061] Based on the above principles, the application first searches for possible diseased areas on blueberry leaves through color changes, and then screens these areas through abnormal temperature changes, i.e. by screening pixel points that meet both temperature and color changes through coordinate overlap, thereby achieving rapid determination and positioning of leaf spot disease through cross-validation, and finally determining the disease stage through quantitative calculation of the diseased area and providing early warning information;

[0062] Compared with the prior art, the application realizes early warning of leaf spot disease through quantitative calculation and cross-validation, which is more accurate and objective than the subjective judgment method that relies on workers' experience;

[0063] Secondly, the application no longer relies on workers' experience, thus effectively reducing the judgment threshold of leaf spot disease, thereby facilitating timely detection and management of leaf spot disease in the early stages, improving the effect of blueberry disease control and increasing blueberry yield.

[0064] Finally, the application realizes automatic recognition of leaf spot disease, which can effectively improve the early prevention and control effect of leaf spot disease, especially for large blueberry plantations. BRIEF DESCRIPTION OF DRAWINGS

[0065] Figure 1 A flow chart of a disease early warning method for blueberries provided in Embodiment 1 of the present application;

[0066] Figure 2 A schematic diagram for obtaining the temperature abnormal region coordinate set;

[0067] Figure 3 A schematic diagram for obtaining the color abnormal region coordinate set;

[0068] Figure 4 A secondary screening schematic diagram;

[0069] Figure 5 A structural diagram of the early warning system provided in Embodiment 2 of the present application;

[0070] The object, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0071] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0072] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present application are only used to explain the relative positional relationship, movement condition, etc. between components in a certain specific posture (as shown in the drawings), and if the specific posture changes, the directional indications will also change accordingly.

[0073] In the present application, unless otherwise explicitly specified and limited, the terms "connection", "fixation" and the like should be understood in a broad sense, for example, "fixation" can be fixed connection, or detachable connection, or integral; can be mechanical connection, or electrical connection; can be direct connection, or indirect connection through an intermediate medium; can be internal connection of two elements or interaction relationship between two elements, unless otherwise explicitly limited. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0074] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the meaning of "and / or" throughout the text includes three parallel solutions. Taking "robot coordinate system and / or m" as an example, it includes the robot coordinate system solution, the m solution, or a solution where both the robot coordinate system and m are satisfied simultaneously. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0075] Example 1

[0076] Reference Figures 1 to 4 This embodiment, as an optional implementation of this application, discloses a method for early warning of diseases in blueberries, including the following steps:

[0077] S1. Set the first standard judgment parameter, the second standard judgment parameter, and the disease judgment function for blueberry disease early warning;

[0078] Specimens of normal blueberry leaves were collected, and infrared temperature images and visible light images of each leaf were acquired. After image conversion, the temperature and RGB values ​​of the normal green tissue were read.

[0079] When there are multiple possible values, a number field is used to ensure that all parameters are included, while also ensuring the accuracy of the parameters.

[0080] The first standard judgment parameter includes the standard judgment grayscale range, and its expression is [G min G max It should be noted that during the acquisition of the first standard judgment parameters, the shooting equipment needs to be calibrated to obtain the temperature-grayscale conversion function, thereby converting the actual temperature into a grayscale value.

[0081] The second standard judgment parameter includes the standard judgment RGB field, and the expression for the standard judgment RGB field is as follows: Where RGB0 represents the standard RGB value of a normal leaf, its expression is: ;

[0082] Because the color of the lesion area varies greatly, the abnormal color area can be quickly extracted by removing the normal area. This method not only has high extraction accuracy, but also effectively simplifies the calculation complexity and helps improve detection efficiency.

[0083] Because the color of the lesion area varies greatly, the abnormal color area can be quickly extracted by removing the normal area. This method not only has high extraction accuracy, but also effectively simplifies the calculation complexity and helps improve detection efficiency.

[0084] Secondly, it should be noted that since a background was set during shooting, the background area needs to be separated first through the RGB standard values ​​of the background area during the actual image processing.

[0085] S2. Acquire infrared temperature and visible light images of the blueberry leaf to be tested;

[0086] It should be noted that the testing device of this application includes a background plate, on which several markers of specific colors, such as red markers, are set; during measurement, the background plate is placed on the back of the blueberry leaf to be tested, and the corresponding images are acquired by an infrared camera and a visible light camera.

[0087] S3. Convert the infrared temperature image into a grayscale image, and extract the coordinate set S of the temperature anomaly area from the grayscale image according to the first standard judgment parameter. T ;

[0088] S31. Convert the infrared temperature image into a grayscale image and construct a first standard coordinate system for the grayscale image;

[0089] S32. Divide the grayscale image into several pixels and obtain the grayscale parameters (x) of each pixel. i y i G i ); where x i y i This represents the first coordinate of a pixel, where i represents the pixel number, and G... i Indicates the actual grayscale value;

[0090] After constructing the first coordinate system, coordinates are automatically generated for each pixel, which are the first coordinates. Simultaneously, the actual grayscale value of each pixel is bound to these first coordinates to form the grayscale parameter (x). i y i G i ); where x i y i This represents the first coordinate of a pixel, where i represents the pixel number, and G... i Indicates the actual grayscale value;

[0091] S33. Obtain the first reference coordinates (x, y) of each calibration point in the first coordinate system. ok y ok The number of standard points is no less than three, and k represents the number of the calibration point.

[0092] After the grayscale parameters are generated, the first reference coordinates of each calibration point in the first coordinate system are obtained. It should be noted that for the accuracy of subsequent calculations, it is preferable to have 3 calibration points.

[0093] S34. Compare each of the grayscale parameters with the first standard judgment parameter to screen out abnormal grayscale values.

[0094] Obtain all grayscale parameters and the first standard judgment parameter;

[0095] Obtain the actual gray value of any gray value parameter. If the actual gray value falls within the standard gray value range, the actual gray value is determined to meet the requirements and is an abnormal gray value.

[0096] By repeating the above steps and comparing each actual grayscale value, several abnormal grayscale values ​​can be obtained.

[0097] S35. Collect the first coordinates corresponding to each abnormal gray value to generate a coordinate set S for the temperature anomaly area. T ;

[0098] Extract the first coordinate corresponding to each abnormal gray value, and then aggregate all the first coordinates into the same set to generate the coordinate set S of the temperature anomaly area. T .

[0099] S4. Extract the coordinate set S of the color anomaly region from the visible light image according to the second standard judgment parameter. C ;

[0100] S41. Convert the visible light image into an RGB image, and construct a second standard coordinate system for the RGB image;

[0101] S42. Divide the RGB image into several pixels and obtain the RGB parameters (x, y, y) of each pixel. j y j r j g j b j ), where x i y i This represents the second coordinate of the pixel, j represents the pixel number, and r j g j b j These are the actual RGB values;

[0102] S43. Obtain the second reference coordinates (x') of each calibration point in the second coordinate system. ok y' ok The number of standard points mentioned above shall be no less than three.

[0103] S44. Compare each of the actual RGB values ​​with the second standard judgment parameter to screen out abnormal RGB values;

[0104] Get all RGB parameters (x j y j r j g j b j ) and standard determination of RGB domain, if r in the RGB parameter j g j b j All meet If the value is true, it indicates that the pixel corresponding to the RGB parameter is normal leaf tissue, and it should be removed; that is, it satisfies the condition. ;

[0105] Repeat the above steps to compare each RGB parameter, and classify all actual RGB values ​​that do not meet the standard RGB domain judgment as abnormal RGB values;

[0106] S45. Collect the second coordinates corresponding to each abnormal RGB value to generate a color anomaly area coordinate set S. C .

[0107] S5. Based on the coordinate set S of the temperature anomaly zone T For the coordinate set S of the color anomaly region C Screening is performed to obtain the coordinate set S of the diseased area. D ;

[0108] S51, based on the first reference coordinates (x) ok y ok ) and the second reference coordinate (x' ok y' ok Perform coordinate registration to obtain the positional conversion relationship between the coordinate set of the temperature anomaly area and the coordinate set of the color anomaly area;

[0109] S511, Constructing standard expressions for position conversion relationships;

[0110] The standard expression is: ;

[0111] S512. Construct a one-to-one mapping between each of the first reference coordinates and each of the second reference coordinates, and generate several fitting calculation groups {(x... ok y ok ), (x' ok y' ok )};

[0112] Since all calibration points are the same, the first and second reference coordinates are specific expressions of the same calibration point in different coordinate systems, and therefore they are slightly different.

[0113] By establishing a one-to-one mapping between different expressions at the same calibration point using the same numbering method, several fitting calculation groups {(x ok y ok ), (x' ok y' ok If there are 3 calibration points, then 3 sets of fitting calculations will be obtained;

[0114] S513. Import each fitting calculation group into the standard expression respectively, and perform fitting calculations to obtain the position conversion relationship, wherein the expression for the position conversion relationship is: , where a, b, c, d, e, and f represent constants;

[0115] Several sets of equations can be obtained by importing each fitted calculation group into the standard expression. Then, by solving the above equations using the least squares method, the constants can be calculated, and thus the position conversion relationship can be obtained.

[0116] S52. Based on the aforementioned location conversion relationship, convert the coordinate set S of the temperature anomaly zone... T Convert to temperature determination coordinate set S' T ;

[0117] The coordinate set S of the temperature anomaly region T Substituting each element into the position conversion relationship yields the corresponding converted coordinates, which is the temperature determination coordinate set S'. T ;

[0118] Since infrared temperature images and visible light images are generated based on different coordinate systems during post-processing, the above steps can normalize the coordinate systems, thereby establishing a precise transformation relationship between the coordinate sets of temperature anomaly areas and color anomaly areas, ensuring that subsequent calculations can strictly correspond and improving the accuracy of the calculations.

[0119] S53. Determine the coordinate set S' based on the temperature. T For the coordinate set S of the color anomaly region C The first screening was performed to obtain the first lesion coordinate set S. D1 and the coordinate set S' of the lesion to be determined D The first lesion coordinate set S D1 The expression is The lesion coordinate set S' to be determined D The expression is ;

[0120] The abnormal areas caused by blueberry leaf spot disease exhibit dual variability in temperature and color. Based on these characteristics, the temperature determination coordinate set S' T With the coordinate set S of the color anomaly regionC The intersection of these coordinates must be the lesion area; therefore, the first screening can be completed by solving for the intersection, i.e., the first lesion coordinate set S. D1 The expression is ;

[0121] In the coordinate normalization process, due to limitations of existing technology, the conversion of pixels in some edge transition areas still has errors and requires further judgment. The areas requiring further judgment are the lesion areas to be determined. The coordinate set S' of the lesion to be determined can be obtained through the difference set. D Its expression is ;

[0122] S54. Determine the lesion coordinate set S' D A second screening was performed to obtain the second lesion coordinate set S. D2 ;

[0123] S541, From the coordinate set S' of the lesion to be determined D Extract any coordinate of the lesion to be determined;

[0124] S542. Using the pixel corresponding to the lesion coordinates to be determined as the center, collect the coordinates of all its neighboring pixels to generate a set of neighboring pixel coordinates; wherein the calculation expression for the set of neighboring pixel coordinates is: ;

[0125] A 3x3 neighborhood is constructed centered on the pixel corresponding to the lesion coordinates to be determined. Within this neighborhood, the pixel corresponding to the lesion coordinates to be determined has a total of 8 adjacent pixels, meaning that the adjacent pixels satisfy the following condition: The coordinates of all adjacent pixels can be calculated using the formula above, thus obtaining the set of adjacent pixel coordinates; see details below. Figure 4 As shown;

[0126] S543. Compare the adjacent pixel coordinate set with the color anomaly region coordinate set S C If there is an intersection, then the second screening is satisfied, and the coordinates of the lesion to be determined are collected into the second lesion coordinate set;

[0127] If the set of adjacent pixel coordinates is the same as the set of coordinates of the color anomaly region S C If there is an intersection, it means that some adjacent pixels have been identified as color abnormality areas, and the coordinates of the lesion to be determined can be identified as the lesion area. Conversely, if there is no intersection, it means that the distance between it and the already identified color abnormality area is large, and it belongs to the normal area, so it can be excluded.

[0128] S544. Repeat the step of extracting any lesion coordinate from the lesion coordinate set S'D to generate a second lesion coordinate set S. D2 .

[0129] By constructing a 3*3 neighborhood, the tolerance screening range can be strictly limited to a reasonable registration error, so as not to miss the real diseased areas that are offset by the error, nor to include the non-diseased areas that are offset by too much, thereby further improving the accuracy of the judgment.

[0130] S55, Set the first lesion coordinate set S D1 With the second lesion coordinate set S D2 The coordinates of the diseased area are merged into a set.

[0131] When the first lesion coordinate set S D1 With the second lesion coordinate set S D2 Once determined, merge the first lesion coordinate set S. D1 With the second lesion coordinate set S D2 The coordinate set of the diseased area can then be obtained.

[0132] S6. Based on the coordinate set S of the diseased area D The disease determination model outputs early warning information.

[0133] S61. Obtain a grayscale image, and extract the leaf shape of the blueberry leaf to be detected from the grayscale image based on the background grayscale value;

[0134] The grayscale image obtained in step S3 is obtained. Since a background board is used during the shooting process, the background board has a single color and good consistency of its parameters. Moreover, it has significant differences from the relevant parameters of the leaves. Therefore, the leaf shape image can be quickly and accurately separated through the background board.

[0135] During separation, first extract all pixels with the same grayscale value as the background, and then subtract these pixels to obtain the leaf shape.

[0136] S62. Calculate the total number of pixels Q of the detected blueberry leaves based on the leaf shape diagram;

[0137] S63, Set the coordinates of the diseased area S D The total number of elements is taken as the number of diseased pixels, Q1;

[0138] S64. Calculate the lesion ratio based on the total number of pixels and the number of diseased pixels;

[0139] The expression for the proportion of lesions is q = 100%Q1 / Q;

[0140] S65. Output early warning information based on the lesion ratio and the lesion determination function.

[0141] The expression for the disease determination function is as follows: , where R represents the proportion of lesions, and k1, k2, and k3 are constants.

[0142] Example 2

[0143] Reference Figure 4 This embodiment, as another optional embodiment of this application, discloses an early warning system, including a parameter setting module and a data acquisition module. The parameter setting module is used to set a first standard judgment parameter, a second standard judgment parameter, and a disease judgment function.

[0144] The data acquisition module is used to acquire infrared temperature images and visible light images of the blueberry leaves to be tested;

[0145] The output end of the data acquisition module is communicatively connected to a first identification module and a second identification module. At the same time, the output ends of the first identification module and the second identification module are respectively connected to a screening calculation module. The screening calculation module obtains the coordinate set of the diseased area by receiving the relevant parameters from the first identification module and the second identification module.

[0146] The output of the screening calculation module is also connected to an early warning information output module.

[0147] Accordingly, this application also discloses an early warning device for blueberry diseases, including a housing, on which an infrared camera and a visible light camera are integrated;

[0148] The warning device also includes a background board with several calibration points marked on it. Preferably, the background board is white or black, and the calibration points are red or other colors that have a significant color difference between the background board and the leaves.

[0149] Compared with existing technologies, this application achieves early warning of leaf spot disease through quantitative calculation and cross-validation. Compared with subjective judgment methods that rely on worker experience, the judgment results are more accurate and more objective.

[0150] Secondly, this application no longer relies on workers' experience, thus effectively lowering the threshold for judging leaf spot disease, which is conducive to timely detection and treatment in the early stage of leaf spot disease, and helps to improve the control effect of blueberry diseases and increase blueberry yield.

[0151] Finally, this application enables automatic identification of leaf spot disease, which can effectively improve the early prevention and control of leaf spot disease, especially for large blueberry farms.

[0152] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A disease early warning method for blueberries, characterized by, The method comprises the following steps: Setting a first standard determination parameter, a second standard determination parameter, and a disease determination function for blueberry disease early warning; Obtaining an infrared temperature image and a visible light image of a blueberry leaf to be detected; Converting the infrared temperature image into a gray image, and constructing a first standard coordinate system for the gray image; The gray-scale image is divided into a plurality of pixel points, and gray-scale parameters (x i , y i , G i ) of each pixel point are acquired respectively; wherein x i , y i represent the first coordinates of the pixel point, i represents the pixel point number, and G i represents the actual gray-scale value. Obtaining first reference coordinates (x ok , y ok ) of each calibration point in the first coordinate system; wherein the calibration points are not less than three, x o represents the X-axis coordinate of the calibration point in the first coordinate system, y o represents the Y-axis coordinate of the calibration point in the first coordinate system, and k represents the number of the calibration point; Comparing each gray parameter with the first standard determination parameter respectively, and screening to obtain abnormal gray values; The first coordinates corresponding to each abnormal gray value are collected to generate a temperature abnormality region coordinate set S T ; Converting the visible light image into an RGB image, and constructing a second standard coordinate system for the RGB image; The RGB image is segmented into a plurality of pixel points, and RGB parameters (x j , y j , r j , g j , b j ) of each pixel point are acquired respectively, wherein x i , y i represent the second coordinates of the pixel point, j represents the pixel point number, and r j , g j , b j are actual RGB values; obtaining second reference coordinates (x o ' k , y o ' k ) of each calibration point in the second coordinate system; wherein the calibration points are not less than three; x o ' represents the X-axis coordinate of the calibration point in the second coordinate system, y o ' represents the Y-axis coordinate of the calibration point in the second coordinate system, and k represents the number of the calibration point; Comparing each actual RGB value with the second standard determination parameter respectively, and screening to obtain abnormal RGB values; The second coordinates corresponding to the abnormal RGB values are collected to generate a color anomaly region coordinate set S C ; According to the first reference coordinates (x ok 、 y ok ) and the second reference coordinates (x o ' k 、 y o ' k ), coordinate registration is performed to obtain a position conversion relationship between the temperature abnormal area coordinate set and the color abnormal area coordinate set. According to the position conversion relationship, the temperature abnormal region coordinate set S T is converted into a temperature determination coordinate set S' T ; According to the temperature determination coordinate set S T The color anomaly zone coordinate set S C First screening is carried out to obtain the first lesion coordinate set S D1 And the lesion coordinate set S' to be determined D Wherein the first lesion coordinate set S D1 The expression of is The expression of the lesion coordinate set S' to be determined D Is ; extracting any lesion coordinate to be determined from the set S' of lesion coordinates to be determined D to be determined from the set S' of lesion coordinates to be determined Taking the pixel point corresponding to the to-be-determined disease coordinate as the center, collecting the coordinates of all adjacent pixel points to generate an adjacent pixel coordinate set; wherein the calculation expression of the set of adjacent pixel coordinates is ; comparing the set of adjacent pixel coordinates with the set of color anomaly region coordinates S C If there is an intersection, the secondary screening is satisfied, and the to-be-determined lesion coordinates are collected into a second lesion coordinate set. repeating the step of extracting any lesion coordinate to be determined from the set S D of lesion coordinates to be determined, generating a second set S D2 of lesion coordinates merging the first lesion coordinate set S D1 with the second lesion coordinate set S D2 into a disease area coordinate set; According to the disease area coordinate set S D and the disease determination model outputs a warning information.

2. The disease early warning method for blueberries according to claim 1, characterized by, The first standard determination parameter comprises a standard determination gray domain, and the second standard determination parameter comprises a standard determination RGB domain.

3. The disease warning method for blueberries according to claim 1, characterized by, The coordinate registration according to the first reference coordinate and the second reference coordinate to obtain the position conversion relationship of the temperature abnormal area coordinate set and the color abnormal area coordinate set comprises the following steps: Constructing a standard expression of the position conversion relationship; A one-to-one mapping is constructed between each of the first reference coordinates and each of the second reference coordinates, generating a number of fitting calculation sets {(x ok , y ok ), (x o ' k , y o ' k )}; The respective fitting calculation sets are introduced into the standard expression, and a position conversion relationship is obtained by fitting calculation, wherein the expression of the position conversion relationship is wherein a, b, c, d, e, and f respectively represent constants.

4. The disease early warning method for blueberries according to claim 1, characterized by, The disease area coordinate set S D And the disease determination model outputs early warning information, including the following steps: Obtaining a gray image, and extracting a leaf shape image of the blueberry leaf to be detected from the gray image according to a background gray value; Calculating the total number Q of pixel points of the detected blueberry leaf according to the leaf shape image; The total number of elements of the disease area coordinate set S D is taken as the number of disease pixel points Q1; Calculating a disease proportion according to the total number of pixel points and the number of disease pixel points; Outputting early warning information according to the disease proportion and the disease determination function.

5. The disease early warning method for blueberries according to claim 4, characterized by, The expression of the lesion ratio is q = 100%Q1 / Q, and the expression of the disease judging function is ; Wherein R represents the disease proportion, R1, R2, and R3 are constants.

6. The warning system for the disease warning method for blueberry according to any one of claims 1 to 5, characterized by It comprises: A parameter setting module for setting a first standard determination parameter, a second standard determination parameter, and a disease determination function for blueberry disease early warning; A data acquisition module for obtaining an infrared temperature image and a visible light image of a blueberry leaf to be detected; A first identification module is configured to divide the gray-scale image into a plurality of pixel points, and acquire gray-scale parameters (x i , y i , G i ) of each pixel point respectively; wherein x i , y i represent the first coordinates of the pixel point, i represents the pixel point number, and G i represents the actual gray-scale value. obtaining first reference coordinates (x ok , y ok ) of each calibration point in the first coordinate system; wherein the calibration points are not less than three, x o represents the X-axis coordinate of the calibration point in the first coordinate system, y o represents the Y-axis coordinate of the calibration point in the first coordinate system, and k represents the number of the calibration point; Comparing each gray parameter with the first standard determination parameter respectively, and screening to obtain abnormal gray values; The first coordinates corresponding to each abnormal gray value are collected to generate a temperature abnormality region coordinate set S T ; A second identification module for converting the visible light image into an RGB image, and constructing a second standard coordinate system for the RGB image; The RGB image is segmented into a plurality of pixel points, and RGB parameters (x j , y j , r j , g j , b j ) of each pixel point are acquired respectively, wherein x i , y i represent the second coordinates of the pixel point, j represents the pixel point number, and r j , g j , b j are actual RGB values; obtaining second reference coordinates (x o ' k , y o ' k ) of each calibration point in the second coordinate system; wherein the calibration points are not less than three; x o ' represents the X-axis coordinate of the calibration point in the second coordinate system, y o ' represents the Y-axis coordinate of the calibration point in the second coordinate system, and k represents the number of the calibration point; Comparing each actual RGB value with the second standard determination parameter respectively, and screening to obtain abnormal RGB values; The second coordinates corresponding to the abnormal RGB values are collected to generate a color anomaly region coordinate set S C ; a screening calculation module configured to perform coordinate registration according to the first reference coordinates (x ok , y ok ) and the second reference coordinates (x o ' k , y o ' k ) to obtain a position conversion relationship between the temperature abnormal area coordinate set and the color abnormal area coordinate set; According to the position conversion relationship, the temperature abnormal region coordinate set S T is converted into a temperature determination coordinate set S' T ; According to the temperature determination coordinate set S' T The color anomaly zone coordinate set S C First screening is performed to obtain a first lesion coordinate set S D1 And a to-be-determined lesion coordinate set S' D Wherein the first lesion coordinate set S D1 The expression is The expression of the to-be-determined lesion coordinate set S' D ;​ extracting any lesion coordinate to be determined from the set S' of lesion coordinates to be determined D to be determined from the set S' of lesion coordinates to be determined Taking the pixel point corresponding to the to-be-determined disease coordinate as the center, collecting the coordinates of all adjacent pixel points to generate an adjacent pixel coordinate set; wherein the calculation expression of the set of adjacent pixel coordinates is ; comparing the set of adjacent pixel coordinates with the set of color anomaly region coordinates S C If there is an intersection, the secondary screening is satisfied, and the to-be-determined lesion coordinates are collected into a second lesion coordinate set. repeating the step of extracting any lesion coordinate to be determined from the set S D of lesion coordinates to be determined, generating a second set S D2 of lesion coordinates merging the first lesion coordinate set S D1 with the second lesion coordinate set S D2 into a disease area coordinate set; An early warning information output module for outputting early warning information according to the disease area coordinate set and the disease determination model.

Images