Jujube tree disease and insect pest identification method and system combined with visual technology

Abstract

The invention relates to the technical field of computer vision, in particular to a jujube tree disease and insect pest recognition method and system combined with a vision technology, and the method comprises the steps: extracting closed edge contours of a disease leaf gray image, and obtaining a contour skeleton of each closed edge contour; based on the graphic features of each contour skeleton, obtaining a to-be-recognized region, analyzing the distribution of gradient amplitudes of the skeleton contour edge and inner pixel points and the gradient amplitude distribution of the skeleton contour outer pixel points, and obtaining an edge transition value of each to-be-recognized region; based on the overall distribution characteristics of the gradient amplitudes of all edge pixel points of the skeleton contour of each to-be-identified area, obtaining an edge energy value of each to-be-identified area; and determining an edge distinguishing value of each to-be-identified area, and obtaining a disease identification result of each to-be-identified area. The invention aims to improve the identification capability of the jujube brown spot and the jujube gray leaf spot and improve the identification precision of disease detection.

Description

Technical Field

[0001] This application relates to the field of computer vision technology, and particularly to a method and system for identifying jujube tree diseases and pests by combining vision technology. Background Art

[0002] Jujube tree diseases and pests have a significant negative impact on the growth and development of jujube trees and the industrial economy. They not only directly damage the tree body tissues and interfere with physiological metabolism, but also reduce the fruit yield and quality. For example, when infected with jujube brown spot disease, small brown spots initially appear on the leaves, and then expand into irregular dark brown disease spots with a yellow halo at the edge. When it is severe, the disease spots merge, and the leaves wither and fall early; spider mites mainly damage the leaves, stems, flowers, roots, etc. of plants with adult mites and nymph mites. They mainly damage the leaves. After the leaves are damaged, they will show chlorosis and turn white, with dense pale small spots on the leaf surface, curling and turning yellow. When it is severe, the plants will have yellow leaves, scorched leaves, curled leaves, defoliation and death. After the jujube tree is damaged by diseases and pests, photosynthesis is blocked, and it is easy to shed leaves and fruits earlier, which will lead to a serious reduction in agricultural production. Therefore, the comprehensive management of diseases and pests is crucial, and timely prevention and control technologies for jujube tree diseases are needed to ensure the sustainable development of the jujube tree industry.

[0003] By combining computer vision to conduct real-time detection of diseases and pests in the jujube tree area, it is a commonly used auxiliary technical means for the prevention and control of jujube tree diseases and pests at present. Combining the high efficiency and real-time nature of computer vision detection can reduce the use of human resources as much as possible. However, in the current method for identifying jujube tree diseases and pests by combining vision technology, different diseases and pests with similar characteristics will affect the detection ability of the vision algorithm, resulting in incorrect detection results, which may affect the prevention and control of corresponding diseases and pests. Summary of the Invention

[0004] In view of the above, it is necessary to provide a method and system for identifying jujube tree diseases and pests by combining vision technology to solve the above problems.

[0005] The first aspect of this application provides a method for identifying jujube tree diseases and pests by combining vision technology, and the method includes: Collect the grayscale images of the diseased leaves of the jujube tree; Extract the closed edge contours of the grayscale images of the diseased leaves to obtain the contour skeletons of each closed edge contour; based on the second-order central moment of each contour skeleton, obtain the area to be identified; For the area to be identified, analyze the distribution of the gradient amplitudes of the edge and inner pixel points of the skeleton contour and the distribution of the gradient amplitudes of the pixel points outside the skeleton contour to obtain the edge transition value of each area to be identified; Based on the overall distribution characteristics of the gradient magnitudes of all edge pixel points of the skeleton contours of each region to be recognized, the edge energy value of each region to be recognized is obtained. Combining with the edge transition value, the edge discrimination value of each region to be recognized is determined, and the disease recognition result of each region to be recognized is obtained; Based on the disease recognition results of each region to be recognized, the deep learning model is trained to identify and detect diseases and pests of jujube trees.

[0006] Among them, the obtaining of the region to be recognized is specifically as follows: The eccentricity of each contour skeleton is calculated using the second central moment, and threshold segmentation is performed on the eccentricities of all contour skeletons to obtain the optimal threshold. The region with an eccentricity higher than the optimal threshold is determined as the region to be recognized.

[0007] Among them, the obtaining of the edge transition value of each region to be recognized is specifically as follows: For each region to be recognized, the set composed of the gradient magnitudes of all pixel points on the edge of the skeleton contour and the pixel points within the first preset number of pixel units adjacent to the skeleton contour inside the skeleton contour is denoted as the first set; The set composed of the gradient magnitudes of the pixel points of the second preset number of pixel units adjacent to the skeleton contour outside the edge of the skeleton contour is denoted as the second set; The dispersion degree of all elements in the first set is compared with the dispersion degree of all elements in the second set to obtain the edge transition value of each region to be recognized.

[0008] Among them, the edge transition value is specifically the result of positive fusion of the negative correlation mapping result of the dispersion degree of all elements in the second set and the dispersion degree of all elements in the first set.

[0009] Among them, the dispersion degree is calculated using variance.

[0010] Among them, the specific formula for obtaining the edge energy value of each region to be recognized is: ; In the formula, Z represents the edge energy value of the region to be recognized; represents the set of edge pixel points of the skeleton contour of the region to be recognized; is a preset threshold; represents the gradient magnitude of pixel point a on the skeleton contour of the region to be recognized; e represents the natural constant.

[0011] Among them, the preset threshold is determined by the preset quantile of the gradient magnitude histogram of the edge pixel points of the skeleton contour of the region to be recognized.

[0012] Among them, the edge discrimination value of each region to be recognized is specifically the result of positive fusion of the edge energy value and the edge transition value.

[0013] Among them, the specific process of obtaining the disease recognition result of each area to be recognized is as follows: Perform threshold segmentation on the edge discrimination values of all areas to be recognized to obtain a segmentation threshold. Identify the area where the edge discrimination value is greater than the segmentation threshold as the jujube gray spot area, otherwise, identify it as the jujube brown spot area.

[0014] In a second aspect, an embodiment of the present application further provides a jujube tree pest and disease recognition system combined with vision technology, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the method described in any one of the above.

[0015] The present application has at least the following beneficial effects: The present application first extracts the closed edge contour of the disease leaf grayscale image and obtains the contour skeleton of each closed edge contour; based on the second-order central moment of each contour skeleton, the area to be recognized is obtained, which helps to extract the shape features and structural information of the disease area and enhances the subsequent accurate recognition of the disease area; the gradient amplitude of the edge pixel points can usually reflect the intensity and clarity of the edge, and the gradient distribution inside and outside helps to distinguish different stages or types of diseases. Analyze the gradient amplitude distribution of the edge of the skeleton contour and the inner and outer pixel points to obtain the edge transition value, which indicates how the edge of the disease area transitions from the normal area, helps to identify the ambiguity or clarity of the area boundary, and captures the significant differences between the disease area and the healthy area; based on the overall distribution characteristics of the gradient amplitudes of all edge pixel points of the skeleton contour of each area to be recognized, obtain the edge energy value of each area to be recognized, which quantifies the edge intensity and structural complexity, thereby helping to determine the significance of the disease area; determine the edge discrimination value of each area to be recognized to obtain the disease recognition result of each area to be recognized. The present application constructs the edge discrimination value by combining the edge transition value and the edge energy value. Considering that there are jujube brown spot disease and jujube gray spot disease with similar shapes and colors in the diseases that harm jujube tree leaves, conventional vision algorithms may make misjudgments because the characteristics of the two diseases are similar during detection. The present application distinguishes the two diseases from the edge feature distribution by combining the edge transition value and the edge energy value, uses the edge discrimination value to determine different disease areas, and assists subsequent algorithms to perform targeted learning on the two different diseases, which can effectively improve the difference detection ability of the algorithm for similar diseases and improve the recognition accuracy for disease detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flowchart of the steps of a method for recognizing jujube tree pests and diseases combined with vision technology provided by an embodiment of the present application; Figure 2 It is a schematic diagram for calculating the edge discrimination value provided by an embodiment of the present application. Detailed implementation manners

[0017] In the description of the embodiments of the present application, words such as "exemplary", "or", and "for example" are used to represent examples, illustrations, or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of words such as "exemplary", "or", and "for example" is intended to present relevant concepts in a specific manner.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs. The terms used in the description of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0019] In addition, it should be noted that the terms "first" and "second" in this application and the accompanying drawings are used to distinguish similar objects and are not used to describe a specific order or sequence. The methods disclosed in the embodiments of this application or shown in the flowcharts of the methods include one or more steps for implementing the methods. Without departing from the scope of protection of this application, the execution orders of multiple steps can be interchanged with each other, and some steps can also be deleted.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs.

[0021] The following specifically describes the specific solutions of a method and system for identifying jujube tree diseases and pests combined with vision technology provided by this application with reference to the accompanying drawings.

[0022] Please refer to Figure 1 , which shows a flowchart of the steps of a method for identifying jujube tree diseases and pests combined with vision technology provided by an embodiment of this application. The method includes the following steps: The first step: Collect the grayscale images of the diseased leaves of the jujube tree.

[0023] During the process of jujube tree planting, jujube trees are extremely vulnerable to diseases and pests, which seriously affect the yield and quality. During the growth and fruiting process of jujube trees, jujube brown spot disease, as the main leaf disease of jujube trees, can cause a significant reduction in the photosynthetic efficiency of leaves, a decrease in the commercial rate of fruits, and a yield loss after being infected by its pathogenic bacteria. Currently, it has become an important biological factor restricting the sustainable development of the jujube industry. When detecting and controlling diseases of jujube trees combined with vision technology, since there is jujube gray spot disease, which has a similar color and shape to jujube brown spot disease, among the diseases that harm jujube tree leaves, but the control methods for the two diseases are completely different, this application mainly distinguishes and judges the two similar diseases.

[0024] First, it is necessary to obtain images of the leaves of jujube trees with diseases such as jujube brown spot disease and jujube gray spot disease through a professional camera configured with a high-definition camera. Since there may be a lot of environmental noise interference when obtaining images, it is necessary to denoise the obtained images. In this embodiment, non-local means denoising is used to denoise the obtained images. Non-local means denoising can retain texture details and effectively remove Gaussian noise and speckle noise by weighted averaging of similar pixel blocks. At the same time, in order to reduce the computational complexity of subsequent algorithms, the image is grayscale processed. In this embodiment of the present application, the grayscale processing method can be the maximum value method. In other embodiments, the grayscale processing method can also adopt any one of the minimum value method and the weighted average method to obtain the diseased leaf image.

[0025] The second step: Extract the closed edge contours of the grayscale image of the diseased leaf and obtain the contour skeletons of each closed edge contour; based on the second-order central moments of each contour skeleton, obtain the area to be recognized.

[0026] When detecting and obtaining the jujube brown spot disease that may occur during the growth and fruiting process of jujube trees, since the characteristics of jujube gray spot disease are similar to those of jujube brown spot disease, misdetection is likely to occur. And for the different causes of jujube gray spot disease and jujube brown spot disease, the prevention and control means are completely different, and different drugs need to be used for targeted disease prevention and control. Therefore, when detecting the jujube brown spot disease of jujube trees, it is necessary to avoid misdetection as much as possible.

[0027] When detecting and obtaining diseases, first, it is necessary to use the Canny edge detection algorithm to extract the edge contours of the preprocessed diseased leaf image, obtain all the closed contour areas in the diseased leaf image, and convert the images of all the closed contour areas into binary images, that is, the grayscale values of all edge pixel points are 255, which is the white foreground, and the grayscale values of the remaining pixel points are 0, which is the black background. Then, use the Guo-Hall algorithm to iteratively delete the edge pixels of the generated binary image and calculate and output the single-pixel-wide and topology-preserving contour skeleton. Among them, the Guo-Hall algorithm is a well-known technology, and the specific calculation process will not be elaborated too much.

[0028] Among the diseases and pests of jujube tree leaves, jujube brown spot disease is different from other common diseases, such as dew, jujube rust disease or red spider disease, etc. The latter usually shows lesion areas close to circular, while the lesions of jujube brown spot disease present irregular shapes. This irregular contour is significantly different from the circular lesions. The obtained contour skeleton coordinates are calculated using image moments. By calculating the output second-order central moment, the second-order central moment is used to calculate the eccentricity of the contour skeleton. The eccentricity of all obtained contour skeletons is calculated using the Otsu threshold method to obtain the optimal threshold. The area with eccentricity less than or equal to the optimal threshold is determined as the near-circular area, and the area with eccentricity higher than the optimal threshold is determined as the area to be recognized.

[0029] The third step: For the area to be recognized, analyze the distribution of the gradient magnitudes of the edge and inner pixel points of the skeleton contour and the distribution of the gradient magnitudes of the pixel points outside the skeleton contour to obtain the edge transition value of each area to be recognized.

[0030] Among the diseases of jujube trees, jujube brown spot disease has similarities with jujube gray spot disease in terms of color and morphological characteristics. To avoid confusion in detection, it is necessary to further consider the distinguishing features between the two diseases. Jujube brown spot disease is caused by the spread and infection of pathogenic bacteria (such as Alternaria), the spread of toxins, and a high-humidity environment, resulting in unclear demarcation between diseased and healthy tissues. In the initial stage, the lesions are in a halo shape, and the edges are more blurred when the spread is fast; while the edges of jujube gray spot disease are obvious mainly because the pathogenic bacteria (such as Phyllosticta ziziphi-vulgaris) quickly kill the surrounding cells after infection, forming a clear necrosis zone. The lesions expand slowly, and the host plant produces defensive pigment deposition at the junction of diseased and healthy tissues, making the boundary sharp, and the symptoms are more typical when the humidity is low. According to the differences in the edges of the two disease areas caused by different infected bacteria, the Sobel operator is used to obtain the gradient magnitudes of the pixel points of the edge skeleton contour, the pixel points of the first preset number of pixel units adjacent to the inner side of the skeleton contour and the skeleton, and the pixel points of the second preset number of pixel units adjacent to the outer side of the skeleton contour and the skeleton. In this embodiment, the first preset number is 1; the second preset number is 2.

[0031] Furthermore, the set composed of the gradient magnitudes of all pixel points on the edge of the skeleton contour and all pixel points of the first preset number of pixel units adjacent to the inner side of the skeleton contour and the skeleton contour is denoted as the first set; the set composed of the gradient magnitudes of all pixel points of the second preset number of pixel units adjacent to the outer side of the skeleton contour and the skeleton contour is denoted as the second set; compare the dispersion degree of all elements in the first set with the dispersion degree of all elements in the second set to obtain the edge transition value of each area to be recognized.

[0032] In this embodiment, the dispersion degree of the set elements is calculated using variance; the dispersion degree of all elements in the first set is denoted as A, and the dispersion degree of all elements in the second set is denoted as B, then the formula form of the edge transition value is: ; wherein, is a preset parameter to prevent the denominator from being 0, and its value is 0.01.

[0033] It should be understood that in the lesions of jujube brown spot disease, due to the infection of the pathogen, the edges of the diseased areas show a halo-like effect. The diffusion of toxins makes the brightness change on both sides of the lesion edges gradual rather than abrupt. Therefore, the variance of the brightness gradient amplitude of the pixels at the edges and inside of the lesions is small, while for the pixels outside the lesions, due to the transition from the diseased area to the healthy leaves, the brightness change is more obvious, resulting in a relatively large variance in the outer area. Here, it means that the edge of the grayish-brown disease is clear, the gradient amplitude of the edge contour is extremely high, while the inner necrosis is relatively uniform and the amplitude is low. The gradient amplitudes of the two circles of pixels are relatively discrete, so the numerator is large. And for the two circles of pixels outside the edge, because they are the healthy parts of the leaves, the overall gradient amplitude is extremely low and close, and the overall gradient amplitude is relatively concentrated, so the denominator is small, and the edge transition value is much greater than 1.

[0034] The fourth step: Based on the overall distribution characteristics of the gradient amplitudes of all edge pixels of the skeleton contour of each region to be recognized, obtain the edge energy value of each region to be recognized, combine with the edge transition value, determine the edge discrimination value of each region to be recognized, and obtain the disease recognition result of each region to be recognized.

[0035] Considering that the core difference between jujube brown spot disease and jujube gray spot disease lies in the infection mechanism of the pathogen and the pathological reaction of the host plant, these biological differences directly lead to significant differences in the image edge performance of the two types of diseases. The toxins secreted by the brown spot pathogen need time to diffuse, resulting in "semi-necrotic" cells at the junction of diseased and healthy tissues, showing a gradual change in brightness, with a medium overall gradient amplitude and a dispersed distribution. While the gray spot pathogen directly destroys the cell wall, forming a "clean cut" necrosis, with an abrupt change in brightness between diseased and healthy tissues, and a very high and concentrated edge gradient amplitude. Obtain the gradient amplitudes of all pixels in the entire region and form a histogram, and construct the edge energy value Z by calculating the edge energy, and combine with the edge transition index to form a disease discrimination index to distinguish the two diseases: ; where, represents the set of edge pixels of the skeleton contour; is a preset threshold. To normalize the gradient intensity, in this embodiment, a high weight is assigned to high gradients, and is the 90th percentile of the edge gradient amplitude histogram; represents the gradient amplitude of pixel a on the skeleton contour. The gradient amplitude is converted into an energy term through an exponential term, mapping the gradient amplitude to the interval (0, 1]. High gradient amplitudes output close to 1, and low gradient amplitudes output close to 0; e represents the natural constant.

[0036] It should be understood that when the skeleton contour is brown spot of jujube, the energy dispersion caused by the slow diffusion of the pathogen results in a lower calculated Z value; when the skeleton contour is gray spot of jujube, the energy concentration caused by the rapid necrosis of the area results in a higher calculated Z value.

[0037] Furthermore, by positively fusing the edge energy value and the edge transition value, an edge discrimination value is generated. In this embodiment, the positive fusion of multiple variables adopts a multiplication calculation method. It should be understood that when the area is brown spot of jujube, since both the edge energy value and the edge transition value are relatively low, the obtained edge discrimination value will be further reduced; calculate the edge discrimination values of all suspected brown spot of jujube and gray spot of jujube areas, and use the Otsu threshold method to obtain the segmentation threshold, and determine that the area with an edge discrimination value higher than the segmentation threshold is the gray spot of jujube area, and determine that the area with an edge discrimination value lower than or equal to the segmentation threshold is the brown spot of jujube area.

[0038] Among them, the calculation schematic diagram of the edge discrimination value is as Figure 2 shown.

[0039] The fifth step: Train the deep learning model based on the disease identification results of each area to be identified, and identify and detect the diseases and pests of jujube trees.

[0040] After distinguishing the two disease areas, convert the image of the obtained area contour into a single-channel mask image with a background of 0, a brown spot of jujube area of 1, and a gray spot of jujube area of 2, and use DeepLabv3+ for model training. Add the CBAM attention mechanism to the skip connection layer of the DeepLabv3+ network. By simultaneously optimizing the channel and spatial dimensions, the saliency of the disease area is improved, and the discriminative features of the disease area can be better learned. The loss function part uses a weighted mixed loss of "main loss + auxiliary loss" to balance class imbalance and edge accuracy. The main loss function selects the cross-entropy loss function, and the auxiliary loss function selects the Dice loss function. The optimizer is selected as AdamW, and overfitting is suppressed through adaptive momentum and regularization, which is suitable for small-scale jujube tree leaf disease datasets. The number of training epochs is set to 300 times to prevent overfitting. Obtain the weight model through training, and combine the trained weight model with the mobile monitoring device to identify and detect the corresponding diseases and pests of jujube trees through the model.

[0041] Based on the same inventive concept as the above method, the embodiment of the present application also provides a jujube tree disease and pest identification system combined with vision technology, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above methods of a jujube tree disease and pest identification method combined with vision technology.

[0042] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a portion of code that contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions noted in the blocks may occur in an order different from that noted in the drawings. For example, two consecutive blocks may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. In the description corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in an order different from that disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. Each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0043] For those skilled in the art, it is obvious that the present application is not limited to the details of the above-described exemplary embodiments, and that the present application can be implemented in other specific forms without departing from the basic characteristics of the present application. Therefore, from any point of view, the above-described embodiments of the present application should be regarded as exemplary and non-limiting; modifications to the technical solutions recorded in the foregoing embodiments, or equivalent replacements of some of the technical features, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included within the protection scope of the present application.

Claims

1. A jujube tree pest and disease identification method combined with vision technology, characterized in that, The method includes the following steps: Collect the grayscale images of the diseased leaves of jujube trees; Extract the closed edge contours of the grayscale images of the diseased leaves, and obtain the contour skeletons of each closed edge contour; Based on the second-order central moments of each contour skeleton, obtain the regions to be recognized; For the regions to be recognized, analyze the distribution of the gradient magnitudes of the edge and inner pixel points of the skeleton contour, and the distribution of the gradient magnitudes of the pixel points outside the skeleton contour, to obtain the edge transition value of each region to be recognized; Based on the overall distribution characteristics of the gradient magnitudes of all edge pixel points of the skeleton contour of each region to be recognized, obtain the edge energy value of each region to be recognized, and combine with the edge transition value to determine the edge discrimination value of each region to be recognized, and obtain the disease recognition result of each region to be recognized; Based on the disease recognition results of each region to be recognized, train the deep learning model to identify and detect pests and diseases of jujube trees.

2. The method for identifying jujube tree diseases and pests combined with vision technology according to claim 1, wherein The obtaining of the regions to be recognized is specifically: Calculate the eccentricity of each contour skeleton using the second-order central moment, perform threshold segmentation on the eccentricities of all contour skeletons to obtain the optimal threshold, and determine the regions with eccentricities higher than the optimal threshold as the regions to be recognized.

3. The jujube tree pest and disease identification method combining vision technology according to claim 1, characterized in that, The obtaining of the edge transition value of each region to be recognized is specifically: For each region to be recognized, denote the set composed of the gradient magnitudes of all pixel points on the edge of the skeleton contour and the first preset number of pixel units adjacent to the inner side of the skeleton contour as the first set; Denote the set composed of the gradient magnitudes of all pixel points adjacent to the outside of the skeleton contour by the second preset number of pixel units as the second set; Compare the dispersion degrees of all elements in the first set with the dispersion degrees of all elements in the second set to obtain the edge transition value of each region to be recognized.

4. The method for identifying jujube tree diseases and pests combined with vision technology according to claim 3, wherein, The edge transition value is specifically the result of the positive fusion of the negative correlation mapping result of the dispersion degree of all elements in the second set and the dispersion degree of all elements in the first set.

5. The method for identifying jujube tree diseases and pests combined with vision technology according to claim 4, characterized in that, The dispersion degree is calculated using variance.

6. The method for identifying jujube tree diseases and pests combined with vision technology according to claim 1, characterized in that, The edge energy value of each area to be recognized is obtained, and the specific formula is as follows: ; where Z represents the edge energy value of the area to be recognized; represents the set of edge pixels of the skeleton contour of the area to be recognized; is a preset threshold; represents the gradient amplitude of the pixel a on the skeleton contour of the area to be recognized; e represents the natural constant.

7. The method for identifying jujube tree diseases and pests combining vision technology according to claim 6, characterized in that, The preset threshold is determined by the preset quantile of the gradient magnitude histogram of the edge pixel points of the skeleton contour of the region to be recognized.

8. The method for identifying jujube tree diseases and pests combining vision technology according to claim 1, characterized in that, The edge discrimination value of each region to be recognized is specifically the result of the positive fusion of the edge energy value and the edge transition value.

9. The method for identifying jujube tree diseases and pests combining vision technology according to claim 1, wherein The specific process of obtaining the disease recognition result of each region to be recognized is: Perform threshold segmentation on the edge discrimination values of all regions to be recognized to obtain the segmentation threshold, and identify the regions with edge discrimination values greater than the segmentation threshold as the jujube gray spot regions, otherwise, identify them as the jujube brown spot regions.

10. A jujube tree pest and disease identification system combined with vision technology, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-9.

Images