A monitoring method and system for Castanopsis tung oil plant diseases and pests

Abstract

The present invention relates to the technical field of pest and disease monitoring, and solves the technical problems of poor segmentation performance, over-segmentation and loss of edge details in the prior art. In particular, it relates to a monitoring method and system for Castanopsis chinensis pests and diseases, comprising collecting remote sensing images of Castanopsis chinensis, and preprocessing the remote sensing images to obtain preprocessed images; each preprocessed image is corrected and spliced ​​to obtain a range image of Castanopsis chinensis. The present invention avoids the problem of dividing adjacent pixels into the same category in the prior art through an improved fuzzy clustering method, which leads to poor segmentation performance, and also causes problems such as over-segmentation and loss of edge details. The present invention not only avoids the problem of dividing adjacent pixels into the same category, but also improves segmentation performance, reduces the problem of over-segmentation, and ensures the retention of edge details, thereby improving the accuracy and efficiency of the classification method.

Description

Technical Field

[0001] The present invention relates to the technical field of pest and disease monitoring, and in particular to a method and system for monitoring plant diseases and pests. Background Art

[0002] Idesia polycarpa is a deciduous tree widely distributed in the Yangtze River Basin and southern my country. It has strong adaptability and high ecological, economic and energy value. It can be used for ecological greening, landscape improvement, soil conservation and water conservation. Its fruit has a high oil content and a high proportion of unsaturated fatty acids. It is a high-quality woody oil and biofuel raw material and has received widespread attention in recent years. With the national grain and oil security strategic layout and the development of the bioenergy industry, the planting area of ​​Idesia polycarpa has continued to expand, and its disease and insect pest problems have become increasingly prominent, which not only affects the growth of Idesia polycarpa, resulting in a reduction in fruit yield, but also reduces the fruit quality. The main diseases of Idesia polycarpa include root rot, anthracnose, powdery mildew, rust, and sooty mold. The existing classification methods often use the A-IT2FCM method to classify and analyze images. Although this method is simple, it only considers the relationship between local neighborhood pixels. When the noise influence of the image is large, it is easy to classify adjacent pixels into the same class, resulting in poor segmentation performance. At the same time, there are also problems such as over-segmentation and loss of edge details. Summary of the Invention

[0003] In response to the shortcomings of the existing technology, the present invention provides a monitoring method and system for Castanopsis caryophyllus diseases and pests, which solves the technical problems of poor segmentation performance, over-segmentation and loss of edge details in the existing technology, thereby achieving the purpose of improving segmentation performance and avoiding over-segmentation and loss of edge details.

[0004] To solve the above technical problems, the present invention provides the following technical solution: a method for monitoring diseases and insect pests of Castanopsis truncatula, the method comprising the following steps:

[0005] S1. collecting remote sensing images of Castanopsis sylvestris and preprocessing the remote sensing images to obtain multiple noise-free preprocessed images;

[0006] S2. performing correction and splicing processing on each pre-processed image to obtain a range image containing the Castanopsis chinensis plant;

[0007] S3. Calculate any pixel in the range image The invariant eigenvalues ​​of , and based on the invariant eigenvalues Establish feature matrix;

[0008] S4. Calculate and optimize eigenvalues ​​based on the characteristic matrix and the health eigenvalues ​​of the healthy image of the mountain ash , based on health eigenvalues Screening to obtain the pest and disease area set;

[0009] S5. Determine the initial center in the pest and disease area by random method And calculate the nonlocal value used to reduce the local noise pixel by the nonlocal pixel , based on non-local values Calculate the first category weight and the second type of weights ;

[0010] S6. According to the first type of weight and the second type of weights Calculate reduced membership values ​​for defuzzifying a set of pest and disease areas , and according to the reduced membership value Compute new centers for classifying sets of pest and disease areas ;

[0011] S7. Use standard images containing the levels of Castanopsis tung oil plant diseases and pests as a comparison standard to divide the disease and pest set into different levels of mild, moderate and severe diseases and pests.

[0012] Furthermore, in step S1, the specific implementation steps are as follows:

[0013] S11, split the remote sensing image into R image, G image and B image using RGB channels, and obtain the pixel points at the same position in the R image, G image and B image respectively RGB pixel values 、 and ;

[0014] S12, according to RGB pixel value 、 and Calculate pixel points Grayscale pixel value , the calculation formula is:

[0015] ;

[0016] in, Indicates the Grayscale pixel values;

[0017] S13, according to the grayscale pixel value Calculate pixel points Denoising value , the calculation formula is:

[0018] ;

[0019] in, Represents grayscale pixel value The standard deviation of

[0020] S14, repeat steps S12 and S13 until the denoising values ​​of all pixels are completed The processed pixels are synthesized into a pre-processed image.

[0021] Furthermore, in step S2, the specific implementation steps are as follows:

[0022] S21, establish a side length on the pre-processed image And the step length is The divided window is used to obtain two adjacent corrected images in the preprocessed image. and ;

[0023] S22, respectively obtain the corrected images and Pixels at the same position in and , and calculate the deviation angle , the calculation formula is:

[0024] ;

[0025] in, Represents pixel points and the number of

[0026] S23, obtain the central pixel point in the preprocessed image , according to the deviation angle Calculate corrected pixels , the calculation formula is:

[0027] ;

[0028] in, and Indicates the The pixel value of a pixel in the corrected image;

[0029] S24, correct the pixel points by pixel coordinate conversion method Convert to corrected coordinates , calculate the ground coordinate values ​​of the pixel points in the preprocessed image , the calculation formula is:

[0030] ;

[0031] in, represents the rotation matrix of the camera extrinsic parameters, Represents the coordinate value of the camera, They represent the external parameters of the camera, represents the focal length of the camera, Indicates the height of the camera;

[0032] S25, based on the ground coordinate values ​​in the multiple pre-processed images Calculate image fusion value and , the calculation formula is:

[0033] ;

[0034] in, and Represent the weight coefficients, and Respectively represent the horizontal coordinate values ​​of the first and second pre-processed images, and Represent the vertical coordinate values ​​of the first and second pre-processed images respectively;

[0035] S26. Repeat step S25 to fuse the multiple pre-processed images into a range image.

[0036] Furthermore, in step S3, the specific implementation steps are as follows:

[0037] S31. Obtain the pixel value of any pixel in the range image As the center pixel point, and get the center pixel radius within Field pixels, calculate the reference value used to obtain the binary result , the calculation formula is:

[0038] ;

[0039] in, represents the symbolic function, Represents the grayscale value of the center pixel, Represents the grayscale value of the neighborhood pixel;

[0040] S32, change the radius of the central pixel to , repeat step S31 to obtain multiple reference values , based on multiple reference values Calculate invariant eigenvalues , the calculation formula is:

[0041] ;

[0042] in, Indicates that the reference value Circular right shift of the binary result Second-rate;

[0043] S33, obtaining the grayscale of the range image by grayscale quantization method and , according to the gray level and Calculate the ash value , the calculation formula is:

[0044] ;

[0045] in, Represents the pixel value of the pixel in the range image, and Respectively represent the offset of the pixel at different angles, and Indicates the length and width of the range image;

[0046] S34, according to the ash value Calculate the contrast value of the range image separately , characteristic entropy and related values , the calculation formula is:

[0047] ;

[0048] in, express the number of Indicates the gray value The average of and Respectively and The standard deviation of

[0049] S35, compare the value , characteristic entropy and related values Combine into multiple vectors ;

[0050] S36, according to the pixel value of any pixel point in the range image Calculate specific filter values , the calculation formula is:

[0051] ;

[0052] in, Respectively represent specific filter values The number of directions and scales, Indicates the standard deviation of the grayscale value of the range image, Represents pixel value The wavelength, Represents pixel value Phase shift;

[0053] S37, according to a specific filter value Calculate filter characteristics , the calculation formula is:

[0054] ;

[0055] in, Represents the grayscale value of the original image and the specific filter value Convolution calculation between;

[0056] S38, the unchanged eigenvalue , multiple vectors and filtering characteristics The combination becomes the feature matrix of the range image.

[0057] Furthermore, in step S4, the specific implementation steps are as follows:

[0058] S41. Use the random forest model to screen the importance of the feature matrix of the range image and obtain multiple optimized eigenvalues , and obtain the health feature value from the healthy image of the mountain ash through step S3 and the random forest model ;

[0059] S42, according to the optimized characteristic value Calculate partition value , the calculation formula is:

[0060] ;

[0061] in, Represents the optimized eigenvalue The gradient value of

[0062] S43, according to the division value Divide the range image into multiple areas to be detected , according to the area to be detected The partition value Get the segmentation region set , the expression is:

[0063] ;

[0064] in, Indicates the a set of segmented regions;

[0065] S44. According to health characteristic values Calculating Fluctuation Value , the calculation formula is:

[0066] ;

[0067] in, Indicates the The health feature value of the health image of each Castanopsis chinensis;

[0068] S45. Obtain the differential threshold by detrending analysis , based on the difference threshold In the segmented area set Screen out pest and disease areas;

[0069] like , it means that the optimized eigenvalue is abnormal and marked as a pest and disease area;

[0070] like , it means that the optimized eigenvalue is normal and marked as a healthy area;

[0071] S46. Grouping multiple pest and disease areas into a pest and disease area set.

[0072] Furthermore, in step S5, the specific implementation steps are as follows:

[0073] S51. Randomly select a pixel point in the pest and disease area as the initial center , define the initial center The cluster radius is ,exist Randomly select two pixel windows without duplication within the range and , and calculate the window difference , the calculation formula is:

[0074] ;

[0075] in, represents the smoothing parameter that controls the decay of the exponential function, Represents a pixel window and The logarithm of ;

[0076] S52, according to the window difference Computing non-local values , the calculation formula is:

[0077] ;

[0078] in, Represents a pixel window The center pixel value of

[0079] S53, get pixel window Gray value of inner pixel , according to the gray value Calculate local variation value , the calculation formula is:

[0080] ;

[0081] in, Represents a pixel window Gray value of inner pixel The average of Represents grayscale value and the average Number of groups;

[0082] S54, based on local variation value Calculating association weights , the calculation formula is:

[0083] ;

[0084] in, Indicates the association weights;

[0085] S55. According to the association weight Calculate the first regularization factor and the second regularization factor , the calculation formula is:

[0086] ;

[0087] in, and represent random constants and non-local values ​​of the pixels in the pest and disease area set The random constant, Represents non-local values The mean of

[0088] S56. According to the first regularization factor and the second regularization factor Calculate the first category weight and the second type of weights , the calculation formula is:

[0089] ;

[0090] in, represents a constant greater than zero, Represents a very small number greater than zero.

[0091] Furthermore, in step S6, the specific implementation steps are as follows:

[0092] S61. According to the first type of weight and the second type of weights Calculate the first fuzzy factor and the second fuzzy factor , the calculation formula is:

[0093] ;

[0094] in, represents the average spatial Euclidean distance between pixels, and represent the initial lower limit and initial upper limit respectively, and represents the fuzzy index, and ， ， Relative to the initial center spatial distance;

[0095] S62, according to the first fuzzy factor and the second fuzzy factor Calculate the updated upper and lower limits to update the membership and , the calculation formula is:

[0096] ;

[0097] in, Indicates update point Distance from initial center The spatial distance, and Indicates the gray value is The pixel point and initial center The fuzzy factor, and Indicates update point and initial center The fuzzy factor, Indicates the number of calculations;

[0098] S63. Update the membership degree according to the upper and lower limits and Constructing updated membership ranges , in the updated membership range Randomly determine the initial updated membership value ;

[0099] S64, based on the initial updated membership value Calculate the update factor , the calculation formula is:

[0100] ;

[0101] in, and represents a random number, and , , represents the cardinality of the pest and disease area set;

[0102] S65, according to the update factor Get the reduced membership value ;

[0103] S66, according to the reduced membership value Calculate the updated new center , the calculation formula is:

[0104] ;

[0105] in, Represents the grayscale value of the pixel the number of Indicates the The new center of the generation;

[0106] S67. Calculate the iterative threshold by the error square method , according to the iteration threshold Determine whether the iteration stops;

[0107] like , then the iteration ends and according to the new center Classify the pest and disease area set to obtain the pest and disease set;

[0108] like , then let And return to step S61.

[0109] Furthermore, the reduced membership value The expression is:

[0110] ;

[0111] in, Represents the reduction factor.

[0112] The technical solution also provides a system for the above-mentioned monitoring method of the tung tree pests and diseases, the system comprising:

[0113] A preprocessing module is used to collect remote sensing images of Castanopsis sylvestris and preprocess the remote sensing images to obtain multiple noise-free preprocessed images;

[0114] A correction and splicing module is used to perform correction and splicing processing on each pre-processed image to obtain a range image containing the Castanopsis chinensis plant;

[0115] Texture module, used to calculate the range of any pixel in the image The invariant eigenvalues ​​of , and based on the invariant eigenvalues Establish feature matrix;

[0116] Preliminary screening module, used to calculate and optimize eigenvalues ​​based on the feature matrix and the health eigenvalues ​​of the healthy image of the mountain ash , based on health eigenvalues Screening to obtain the pest and disease area set;

[0117] Pre-module, used to determine the initial center in the pest and disease area by random method And calculate the nonlocal value used to reduce the local noise pixel by the nonlocal pixel , based on non-local values Calculate the first category weight and the second type of weights ;

[0118] Fuzzy module, used to calculate the weight of the first category and the second type of weights Calculate reduced membership values ​​for defuzzifying a set of pest and disease areas , and according to the reduced membership value Compute new centers for classifying sets of pest and disease areas , get the pest and disease collection;

[0119] The classification module is used to use standard images containing the levels of tung tree pests and diseases as comparison standards to divide the pest and disease set into different levels of mild, moderate and severe pests and diseases.

[0120] By means of the above technical solution, the present invention provides a method and system for monitoring diseases and insect pests of Castanopsis truncatula, which has at least the following beneficial effects:

[0121] 1. The present invention can avoid the problems of offset angle and coordinate position deviation after image shooting by correcting and stitching the pre-processed images. Due to the wide viewing angle of the camera, multiple images need to be stitched. After correcting the stitching, overlapping or duplication can be avoided during the stitching process, which not only improves the accuracy of image processing, but also improves the processing efficiency of subsequent steps.

[0122] 2. The present invention extracts texture features from preprocessed images, which not only reduces the problem of image over-segmentation through improved steps, improves the accuracy of segmentation and the robustness of image processing, but also improves the accuracy of texture processing, enabling rapid classification of pest and disease images and healthy images, thereby improving the accuracy, robustness and work efficiency of image processing.

[0123] 3. The present invention uses an improved fuzzy clustering method to avoid the problem of dividing adjacent pixels into the same category in the prior art, which will lead to poor segmentation performance, and there will also be problems such as over-segmentation and loss of edge details. It can not only avoid the problem of dividing adjacent pixels into the same category, but also improve segmentation performance, reduce the problem of over-segmentation, and ensure the retention of edge details, thereby improving the accuracy and efficiency of the classification method. BRIEF DESCRIPTION OF THE DRAWINGS

[0124] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0125] Figure 1 This is a flow chart of a method for monitoring diseases and insect pests of Castanopsis chinensis according to the present invention;

[0126] Figure 2 This is a structural block diagram of a monitoring system for Castanopsis chinensis diseases and insect pests according to the present invention. DETAILED DESCRIPTION

[0127] To make the above-mentioned objectives, features, and advantages of the present invention more clearly understood, the present invention is further described below in detail with reference to the accompanying drawings and specific embodiments. This will enable a full understanding of how this application uses technical means to solve technical problems and achieve technical effects, and to implement the invention accordingly.

[0128] Since the existing technology has poor segmentation performance and has technical problems such as over-segmentation and loss of edge details, this embodiment proposes a monitoring method and system for tung oil plant diseases and pests, which can improve segmentation performance and avoid over-segmentation and loss of edge details. Figure 1 As shown, the method includes the following steps:

[0129] S1. Collect remote sensing images of Castanopsis sylvestris and preprocess them to obtain multiple noise-free preprocessed images. During the actual image capture process, due to the influence of airflow and obstruction by high-altitude fog, the captured images are prone to image offset or image occlusion. To solve this problem, image preprocessing is required. The specific implementation steps are as follows:

[0130] S11, split the remote sensing image into R image, G image and B image using RGB channels, and obtain the pixel points at the same position in the R image, G image and B image respectively RGB pixel values 、 and ;

[0131] S12, according to RGB pixel value 、 and Calculate pixel points Grayscale pixel value , the calculation formula is:

[0132] ;

[0133] in, Indicates the Grayscale pixel values;

[0134] S13, according to the grayscale pixel value Calculate pixel points Denoising value , the calculation formula is:

[0135] ;

[0136] in, Represents grayscale pixel value The standard deviation of

[0137] S14, repeat steps S12 and S13 to remove noise values ​​of all pixels The calculation is completed, and the processed pixels are synthesized into a preprocessed image. By correcting and stitching the preprocessed image, the offset angle and coordinate position deviation problems after the image is taken can be avoided. Due to the wide viewing angle of the camera, multiple images need to be stitched together. After correcting the stitching, overlap or duplication can be avoided during the stitching process, which not only improves the accuracy of image processing, but also improves the processing efficiency of subsequent steps.

[0138] S2. Correct and stitch each pre-processed image to obtain an image of the area containing the Castanopsis chinensis plant. The shaking of the drone easily causes the angle of the image to deviate, resulting in a relative tilt between the pre-processed image and the actual image. At the same time, the pre-processed image area is very small, so not only does it need to be corrected, but multiple pre-processed images also need to be stitched together without overlap. To solve this problem, the specific implementation steps are as follows:

[0139] S21, establish a side length on the pre-processed image And the step length is The divided window is used to obtain two adjacent corrected images in the preprocessed image. and ; By dividing the window, two adjacent corrected images in the window are obtained and .

[0140] S22, respectively obtain the corrected images and Pixels at the same position in and , and calculate the deviation angle , the calculation formula is:

[0141] ;

[0142] in, Represents pixel points and the number of

[0143] S23, obtain the central pixel point in the preprocessed image , according to the deviation angle Calculate corrected pixels , the calculation formula is:

[0144] ;

[0145] in, and Indicates the The pixel value of a pixel in the corrected image;

[0146] S24, correct the pixel points by pixel coordinate conversion method Convert to corrected coordinates , calculate the ground coordinate values ​​of the pixel points in the preprocessed image , the calculation formula is:

[0147] ;

[0148] in, represents the rotation matrix of the camera extrinsic parameters, Represents the coordinate value of the camera, They represent the external parameters of the camera, represents the focal length of the camera, Indicates the height of the camera. The pixel coordinate conversion method is a method that converts the pixels in the image into world coordinates based on the coordinates calibrated by the camera. The coordinates calibrated by the camera can use the Zhang Zhengyou calibration method. Therefore, the pixel coordinate conversion method is a commonly used coordinate conversion method and is not described in detail here.

[0149] S25, based on the ground coordinate values ​​in the multiple pre-processed images Calculate image fusion value and , the calculation formula is:

[0150] ;

[0151] in, and Represent the weight coefficients, and Respectively represent the horizontal coordinate values ​​of the first and second pre-processed images, and Respectively represent the vertical coordinate values ​​of the first and second pre-processed images; weight coefficient and It is obtained through the hierarchical analysis method. The hierarchical analysis method is a commonly used method for obtaining weight coefficients and will not be described in detail here.

[0152] S26. Repeat step S25 to fuse multiple pre-processed images into a range image. By correcting and stitching the pre-processed images, the offset angle and coordinate position deviation problems after the image is taken can be avoided. Due to the wide viewing angle of the camera, multiple images need to be stitched. After correcting the stitching, overlapping or duplication can be avoided during the stitching process, which not only improves the accuracy of image processing, but also improves the processing efficiency of subsequent steps.

[0153] S3. Calculate any pixel in the range image The invariant eigenvalues ​​of , and based on the invariant eigenvalues Establish a feature matrix; the features of the images of the Chinese tallow trees in different health states in the range image will be different. If no preliminary screening is performed, it is easy to increase the processing steps and increase the computational complexity of the subsequent steps. In order to preliminarily screen out healthy Chinese tallow trees, it is necessary to determine the features of the diseased and insect-infested Chinese tallow trees. The specific implementation steps are as follows:

[0154] S31. Obtain the pixel value of any pixel in the range image As the center pixel point, and get the center pixel radius within Field pixels, calculate the reference value used to obtain the binary result , the calculation formula is:

[0155] ;

[0156] in, represents the symbolic function, Represents the grayscale value of the center pixel, Represents the grayscale value of the neighborhood pixel;

[0157] S32, change the radius of the central pixel to , repeat step S31 to obtain multiple reference values , based on multiple reference values Calculate invariant eigenvalues , the calculation formula is:

[0158] ;

[0159] in, Indicates that the reference value Circular right shift of the binary result Second-rate;

[0160] S33, obtaining the grayscale of the range image by grayscale quantization method and , according to the gray level and Calculate the ash value , the calculation formula is:

[0161] ;

[0162] in, Represents the pixel value of the pixel in the range image, and Respectively represent the offset of the pixel at different angles, and Indicates the length and width of the range image;

[0163] S34, according to the ash value Calculate the contrast value of the range image separately , characteristic entropy and related values , the calculation formula is:

[0164] ;

[0165] in, express the number of Indicates the gray value The average of and Respectively and The standard deviation of

[0166] S35, compare the value , characteristic entropy and related values Combine into multiple vectors ;

[0167] S36, according to the pixel value of any pixel point in the range image Calculate specific filter values , the calculation formula is:

[0168] ;

[0169] in, Respectively represent specific filter values The number of directions and scales, Indicates the standard deviation of the grayscale value of the range image, Represents pixel value The wavelength, Represents pixel value Phase shift;

[0170] S37, according to a specific filter value Calculate filter characteristics , the calculation formula is:

[0171] ;

[0172] in, Represents the grayscale value of the original image and the specific filter value Convolution calculation between;

[0173] S38, the unchanged eigenvalue , multiple vectors and filtering characteristics The feature matrix of the range image is combined. The feature matrix contains multiple features of the range image. By extracting the texture features of the preprocessed image, the improved steps can not only reduce the problem of image over-segmentation, improve the accuracy of segmentation and the robustness of image processing, but also improve the accuracy of texture processing, so that pest and disease images and healthy images can be quickly classified, thereby improving the accuracy, robustness and work efficiency of image processing.

[0174] S4. Calculate and optimize eigenvalues ​​based on the characteristic matrix and the health eigenvalues ​​of the healthy image of the mountain ash , based on health eigenvalues The pest and disease area set is obtained by screening. After obtaining the feature matrix, healthy tung trees and pest and diseased tung trees are screened out based on the feature matrix. The specific implementation steps are as follows:

[0175] S41. Use the random forest model to screen the importance of the feature matrix of the range image and obtain multiple optimized eigenvalues , and obtain the health feature value from the healthy image of the mountain ash through step S3 and the random forest model ; Random forest model is a commonly used model for screening features, which will not be described in detail here. This is the image feature of a healthy Castanopsis serrata. Since the method is the same as step S3, it will not be repeated here.

[0176] S42, according to the optimized characteristic value Calculate partition value , the calculation formula is:

[0177] ;

[0178] in, Represents the optimized eigenvalue The gradient value of

[0179] S43, according to the division value Divide the range image into multiple areas to be detected , according to the area to be detected The partition value Get the segmentation region set , the expression is:

[0180] ;

[0181] in, Indicates the A set of segmented regions; since the area of ​​the range image is very large and the pre-processed image cannot analyze the whole tree fruit, the tree fruit is segmented on the basis of the range image, and then the health of each tree fruit is studied.

[0182] S44. According to health characteristic values Calculating Fluctuation Value , the calculation formula is:

[0183] ;

[0184] in, Indicates the The health feature value of the health image of each Castanopsis chinensis;

[0185] S45. Obtain the differential threshold by detrending analysis , based on the difference threshold In the segmented area set Screen out pest and disease areas;

[0186] like , it means that the optimized eigenvalue is abnormal and marked as a pest and disease area;

[0187] like , it means that the optimized eigenvalue is normal and marked as a healthy area;

[0188] S46. Group multiple pest and disease areas into a pest and disease area set. This step can preliminarily screen out the Castanopsis trees with pests and diseases. By extracting the texture features of the preprocessed image, the improved steps can not only reduce the problem of image over-segmentation, improve the accuracy of segmentation and the robustness of image processing, but also improve the accuracy of texture processing, so that pest and disease images and healthy images can be quickly classified, thereby improving the accuracy, robustness and work efficiency of image processing.

[0189] S5. Determine the initial center in the pest and disease area by random method And calculate the nonlocal value used to reduce the local noise pixel by the nonlocal pixel , based on non-local values Calculate the first category weight and the second type of weights The pest and disease area set consists of Tung trees with pests and diseases. In order to further classify Tung trees with pests and diseases, it is necessary to further divide the Tung trees with pests and diseases into detailed categories based on their characteristics. The specific implementation steps are as follows:

[0190] S51. Randomly select a pixel point in the pest and disease area as the initial center , define the initial center The cluster radius is ,exist Randomly select two pixel windows without duplication within the range and , and calculate the window difference , the calculation formula is:

[0191] ;

[0192] in, represents the smoothing parameter that controls the decay of the exponential function, Represents a pixel window and The logarithm of ;

[0193] S52, according to the window difference Computing non-local values , the calculation formula is:

[0194] ;

[0195] in, Represents a pixel window In order to solve the problem that the local pixel block contains too many noise pixels, the auxiliary image is reconstructed by calculating the pixels outside the local pixel block. More image information can be used to improve the accuracy and stability of image segmentation, that is, calculating the non-local value ;

[0196] S53, get pixel window Gray value of inner pixel , according to the gray value Calculate local variation value , the calculation formula is:

[0197] ;

[0198] in, Represents a pixel window Gray value of inner pixel The average of Represents grayscale value and the average Number of groups;

[0199] S54, based on local variation value Calculating association weights , the calculation formula is:

[0200] ;

[0201] in, Indicates the association weights;

[0202] S55. According to the association weight Calculate the first regularization factor and the second regularization factor , the calculation formula is:

[0203] ;

[0204] in, and represent random constants and non-local values ​​of the pixels in the pest and disease area set The random constant, Represents non-local values The mean of

[0205] S56. According to the first regularization factor and the second regularization factor Calculate the first category weight and the second type of weights , the calculation formula is:

[0206] ;

[0207] in, represents a constant greater than zero, Represents a very small number greater than zero. Through the improved fuzzy clustering method, it can avoid the problem of dividing adjacent pixels into the same category in the existing technology, which will lead to poor segmentation performance. At the same time, there will be problems such as over-segmentation and loss of edge details. It can not only avoid the problem of adjacent pixels being divided into the same category, but also improve the segmentation performance, reduce the problem of over-segmentation, and ensure the retention of edge details, thereby improving the accuracy and efficiency of the classification method.

[0208] S6. According to the first type of weight and the second type of weights Calculate reduced membership values ​​for defuzzifying a set of pest and disease areas , and according to the reduced membership value Compute new centers for classifying sets of pest and disease areas , obtain a set of pests and diseases; based on step S5, the pest and disease levels of the tung tree pests and diseases are also classified, and the specific implementation steps are as follows:

[0209] S61. According to the first type of weight and the second type of weights Calculate the first fuzzy factor and the second fuzzy factor , the calculation formula is:

[0210] ;

[0211] in, represents the average spatial Euclidean distance between pixels, and represent the initial lower limit and initial upper limit respectively, and represents the fuzzy index, and ， ， Relative to the initial center spatial distance;

[0212] S62, according to the first fuzzy factor and the second fuzzy factor Calculate the updated upper and lower limits to update the membership and , the calculation formula is:

[0213] ;

[0214] in, Indicates update point Distance from initial center The spatial distance, and Indicates the gray value is The pixel point and initial center The fuzzy factor, and Indicates update point and initial center The fuzzy factor, Indicates the number of calculations;

[0215] S63. Update the membership degree according to the upper and lower limits and Constructing updated membership ranges , in the updated membership range Randomly determine the initial updated membership value ;

[0216] S64, based on the initial updated membership value Calculate the update factor , the calculation formula is:

[0217] ;

[0218] in, and represents a random number, and , , represents the cardinality of the pest and disease area set;

[0219] S65, according to the update factor Get the reduced membership value , the expression is:

[0220] ;

[0221] in, Represents the reduction coefficient. In order to calculate the clear cluster center of the model, it is necessary to defuzzify the classification of the pest and disease area set and use the reduction method to calculate the reduced membership value. ;

[0222] S66, according to the reduced membership value Calculate the updated new center , the calculation formula is:

[0223] ;

[0224] in, Represents the grayscale value of the pixel the number of Indicates the The new center of the generation;

[0225] S67. Calculate the iterative threshold by the error square method , according to the iteration threshold Determine whether the iteration stops;

[0226] like , then the iteration ends and according to the new center Classify the pest and disease area set to obtain the pest and disease set;

[0227] like , then let And return to step S61. Through the improved fuzzy clustering method, the problem of dividing adjacent pixels into the same category in the prior art can be avoided, which will lead to poor segmentation performance, and there will also be problems such as over-segmentation and loss of edge details. It can not only avoid the problem of dividing adjacent pixels into the same category, but also improve the segmentation performance, reduce the problem of over-segmentation, and ensure the retention of edge details, thereby improving the accuracy and efficiency of the classification method.

[0228] S7. Use the standard image containing the level of tung tree pests and diseases as the comparison standard, and divide the pest and disease set into different levels of mild, moderate and severe pests and diseases. Through the classification of step S6, the pest and disease set can be classified according to the characteristics of mild, moderate and severe pests and diseases in the historical image. The feature method is a method of extracting image features, such as feature extraction based on grayscale values. The tung tree trees are still in the early stage of damage that is difficult to detect with the naked eye. The color presented in the infrared image will be different from the color presented by the normal tung tree, because the spectral reflectance of the tung tree in the near-infrared wavelength domain is different. The rate is high. When affected by pests and diseases, the water content of the leaves of Castanopsis chinensis decreases, the chlorophyll decreases, and the spectral reflectance of the leaves in the infrared wavelength domain will be significantly reduced. Therefore, the color tone of the damaged Castanopsis chinensis trees in the infrared image is darker than that of normal crops. As the degree of damage caused by pests and diseases of Castanopsis chinensis gradually deepens, the chlorophyll in its leaves disappears, and the image tone becomes darker, or even presents a green hue. Therefore, the damaged Castanopsis chinensis trees can be screened out based on the color tone difference generated by image contrast. This method can be used to extract features for further differentiation. The feature method is a commonly used feature extraction method and will not be described here.

[0229] Since the existing technology has poor segmentation performance and has technical problems such as over-segmentation and loss of edge details, this embodiment also proposes a monitoring system for tung tree pests and diseases, which can improve the segmentation performance and avoid over-segmentation and loss of edge details. Figure 2 As shown in FIG, the monitoring system includes a preprocessing module, a correction and splicing module, a texture module, a preliminary screening module, a front module, a fuzzy module, and a classification module.

[0230] The preprocessing module is used to collect remote sensing images of Castanopsis chinensis and preprocess the remote sensing images to obtain multiple noise-free preprocessed images; the correction and splicing module is used to correct and splice each preprocessed image to obtain a range image containing Castanopsis chinensis plants; the texture module is used to calculate the texture of any pixel in the range image. The invariant eigenvalues ​​of , and based on the invariant eigenvalues Establish a feature matrix; preliminary screening module, used to calculate and optimize eigenvalues ​​based on the feature matrix and the health eigenvalues ​​of the healthy image of the mountain ash , based on health eigenvalues Screening to obtain the pest and disease area set; the front module is used to determine the initial center in the pest and disease area set by random method And calculate the nonlocal value used to reduce the local noise pixel by the nonlocal pixel , based on non-local values Calculate the first category weight and the second type of weights ; Fuzzy module, used to calculate the weight of the first category and the second type of weights Calculate reduced membership values ​​for defuzzifying a set of pest and disease areas , and according to the reduced membership value Compute new centers for classifying sets of pest and disease areas ; The classification module is used to use standard images containing the levels of tung tree pests and diseases as comparison standards to divide the pest and disease set into different levels of mild, moderate and severe pests and diseases.

[0231] Those skilled in the art will appreciate that all or part of the steps in the above-described embodiment methods can be accomplished by programming related hardware. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0232] Each embodiment in this specification is described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the same or similar parts between the embodiments. For the above embodiments, since they are basically similar to the method embodiments, the description is relatively simple. For relevant parts, refer to the partial description of the method embodiments.

[0233] The above embodiments provide a detailed introduction to the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core ideas. At the same time, for those skilled in the art, according to the ideas of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

Claims

1. A method for monitoring diseases and insect pests of Castanopsis sylvestris, characterized in that: The method comprises the following steps: S1. collecting remote sensing images of Castanopsis sylvestris and preprocessing the remote sensing images to obtain multiple noise-free preprocessed images; S2. performing correction and splicing processing on each pre-processed image to obtain a range image containing the Castanopsis chinensis plant; S3. Calculate any pixel in the range image The invariant eigenvalues ​​of , and based on the invariant eigenvalues Establish feature matrix; S4. Calculate and optimize eigenvalues ​​based on the characteristic matrix and the health eigenvalues ​​of the healthy image of the mountain ash , based on health eigenvalues The pest and disease area set is screened and the specific implementation steps are as follows: S41. Use the random forest model to screen the importance of the feature matrix of the range image and obtain multiple optimized eigenvalues , and obtain the health feature value from the healthy image of the mountain ash through step S3 and the random forest model ; S42, according to the optimized characteristic value Calculate partition value , the calculation formula is: ； ； in, Represents the optimized eigenvalue The gradient value of S43, according to the division value Divide the range image into multiple areas to be detected , according to the area to be detected The partition value Get the segmentation region set , the expression is: ； in, Indicates the a set of segmented regions; S44. According to health characteristic values Calculating Fluctuation Value , the calculation formula is: ； S45. Obtain the differential threshold by detrending analysis , based on the difference threshold In the segmented area set Screen out pest and disease areas; like , it means that the optimized eigenvalue is abnormal and marked as a pest and disease area; like , it means that the optimized eigenvalue is normal and marked as a healthy area; S46, grouping multiple pest and disease areas into a pest and disease area set; S5. Determine the initial center in the pest and disease area by random method And calculate the nonlocal value used to reduce the local noise pixel by the nonlocal pixel , based on non-local values Calculate the first category weight and the second type of weights ; S6. According to the first type of weight and the second type of weights Calculate reduced membership values ​​for defuzzifying a set of pest and disease areas , and according to the reduced membership value Compute new centers for classifying sets of pest and disease areas , get the pest and disease set, the specific implementation steps are as follows: S61. According to the first type of weight and the second type of weights Calculate the first fuzzy factor and the second fuzzy factor , the calculation formula is: ； in, represents the average spatial Euclidean distance between pixels, and represent the initial lower limit and initial upper limit respectively, and represents the fuzzy index, and ， ， Relative to the initial center spatial distance; 、 are the first regularization factor and the second regularization factor respectively; S62, according to the first fuzzy factor and the second fuzzy factor Calculate the updated upper and lower limits to update the membership and , the calculation formula is: ； in, Indicates update point Distance from initial center The spatial distance, and Indicates the gray value is The pixel point and initial center The fuzzy factor, and Indicates update point and initial center The fuzzy factor, Indicates the number of calculations; S63. Update the membership degree according to the upper and lower limits and Constructing updated membership ranges , in the updated membership range Randomly determine the initial updated membership value ; S64, based on the initial updated membership value Calculate the update factor , the calculation formula is: ； in, and represents a random number, and , , represents the cardinality of the pest and disease area set; S65, according to the update factor Get the reduced membership value ; S66, according to the reduced membership value Calculate the updated new center , the calculation formula is: ； in, Represents the grayscale value of the pixel the number of Indicates the The new center of the generation; S67. Calculate the iterative threshold by the error square method , according to the iteration threshold Determine whether the iteration stops; like , then the iteration ends and according to the new center Classify the pest and disease area set to obtain the pest and disease set; like , then let And return to step S61; S7. Use standard images containing the levels of Castanopsis tung oil plant diseases and pests as a comparison standard to divide the disease and pest set into different levels of mild, moderate and severe diseases and pests.

2. The monitoring method according to claim 1, characterized in that: In step S1, the specific implementation steps are as follows: S11, split the remote sensing image into R image, G image and B image using RGB channels, and obtain the pixel points at the same position in the R image, G image and B image respectively RGB pixel values 、 and ; S12, according to RGB pixel value 、 and Calculate pixel points Grayscale pixel value , the calculation formula is: ； in, Indicates the Grayscale pixel values; S13, according to the grayscale pixel value Calculate pixel points Denoising value , the calculation formula is: ； in, Represents grayscale pixel value The standard deviation of S14, repeat steps S12 and S13 until the denoising values ​​of all pixels are completed The processed pixels are synthesized into a pre-processed image.

3. The monitoring method according to claim 1, characterized in that In step S2, the specific implementation steps are as follows: S21, establish a side length on the pre-processed image And the step length is The divided window is used to obtain two adjacent corrected images in the preprocessed image. and ; S22, respectively obtain the corrected images and Pixels at the same position in and , and calculate the deviation angle , the calculation formula is: ； in, Represents pixel points and the number of S23, obtain the central pixel point in the preprocessed image , according to the deviation angle Calculate corrected pixels , the calculation formula is: ； in, and Indicates the The pixel value of a pixel in the corrected image; S24, correct the pixel points by pixel coordinate conversion method Convert to corrected coordinates , calculate the ground coordinate values ​​of the pixel points in the preprocessed image , the calculation formula is: ； in, represents the rotation matrix of the camera extrinsic parameters, Represents the coordinate value of the camera, They represent the external parameters of the camera, represents the focal length of the camera, Indicates the height of the camera; S25, based on the ground coordinate values ​​in the multiple pre-processed images Calculate image fusion value and , the calculation formula is: ； ； in, and Represent the weight coefficients, and Respectively represent the horizontal coordinate values ​​of the first and second pre-processed images, and Represent the vertical coordinate values ​​of the first and second pre-processed images respectively; S26. Repeat step S25 to fuse the multiple pre-processed images into a range image.

4. The monitoring method according to claim 1, characterized in that: In step S3, the specific implementation steps are as follows: S31. Obtain the pixel value of any pixel in the range image As the center pixel point, and get the center pixel radius within Field pixels, calculate the reference value used to obtain the binary result , the calculation formula is: ； ； in, represents the symbolic function, Represents the grayscale value of the center pixel, Represents the grayscale value of the neighborhood pixel; S32, change the radius of the central pixel to , repeat step S31 to obtain multiple reference values , based on multiple reference values Calculate invariant eigenvalues , the calculation formula is: ； in, Indicates that the reference value Circular right shift of the binary result Second-rate; S33, obtaining the grayscale of the range image by grayscale quantization method and , according to the gray level and Calculate the ash value , the calculation formula is: ； in, Represents the pixel value of the pixel in the range image, and Respectively represent the offset of the pixel at different angles, and Indicates the length and width of the range image; S34, according to the ash value Calculate the contrast value of the range image separately , characteristic entropy and related values , the calculation formula is: ； ； ； in, express the number of Indicates the gray value The average of and Respectively and The standard deviation of S35, compare the value , characteristic entropy and related values Combine into multiple vectors ; S36, according to the pixel value of any pixel point in the range image Calculate specific filter values , the calculation formula is: ； in, Respectively represent specific filter values The number of directions and scales, Indicates the standard deviation of the grayscale value of the range image, Represents pixel value The wavelength, Represents pixel value Phase shift; S37, according to a specific filter value Calculate filter characteristics , the calculation formula is: ； in, Represents the grayscale value of the original image and the specific filter value Convolution calculation between; S38, the unchanged eigenvalue , multiple vectors and filtering characteristics The combination becomes the feature matrix of the range image.

5. The monitoring method according to claim 1, characterized in that: In step S5, the specific implementation steps are as follows: S51. Randomly select a pixel point in the pest and disease area as the initial center , define the initial center The cluster radius is ,exist Randomly select two pixel windows without duplication within the range and , and calculate the window difference , the calculation formula is: ； in, represents the smoothing parameter that controls the decay of the exponential function, Represents a pixel window and The logarithm of ; S52, according to the window difference Computing non-local values , the calculation formula is: ； in, Represents a pixel window The center pixel value of S53, get pixel window Gray value of inner pixel , according to the gray value Calculate local variation value , the calculation formula is: ； in, Represents a pixel window Gray value of inner pixel The average of Represents grayscale value and the average Number of groups; S54, based on local variation value Calculating association weights , the calculation formula is: ； in, Indicates the association weights; S55. According to the association weight Calculate the first regularization factor and the second regularization factor , the calculation formula is: ； in, and represent random constants and non-local values ​​of the pixels in the pest and disease area set The random constant, Represents non-local values The mean of S56. According to the first regularization factor and the second regularization factor Calculate the first category weight and the second type of weights , the calculation formula is: ； ； in, represents a constant greater than zero, Represents a very small number greater than zero.

6. The monitoring method according to claim 1, characterized in that: The reduced membership value The expression is: ； in, Represents the reduction factor.

7. A system for use in the method for monitoring plant diseases and insect pests of any one of claims 1 to 6, characterized in that: The system includes: A preprocessing module is used to collect remote sensing images of Castanopsis sylvestris and preprocess the remote sensing images to obtain multiple noise-free preprocessed images; A correction and splicing module is used to perform correction and splicing processing on each pre-processed image to obtain a range image containing the Castanopsis chinensis plant; Texture module, used to calculate the range of any pixel in the image The invariant eigenvalues ​​of , and based on the invariant eigenvalues Establish feature matrix; Preliminary screening module, used to calculate and optimize eigenvalues ​​based on the feature matrix and the health eigenvalues ​​of the healthy image of the mountain ash , based on health eigenvalues Screening to obtain the pest and disease area set; Pre-module, used to determine the initial center in the pest and disease area by random method And calculate the nonlocal value used to reduce the local noise pixel by the nonlocal pixel , based on non-local values Calculate the first category weight and the second type of weights ; Fuzzy module, used to calculate the weight of the first category and the second type of weights Calculate reduced membership values ​​for defuzzifying a set of pest and disease areas , and according to the reduced membership value Compute new centers for classifying sets of pest and disease areas , get the pest and disease collection; The classification module is used to use standard images containing the levels of tung tree pests and diseases as comparison standards to divide the pest and disease set into different levels of mild, moderate and severe pests and diseases.

Images