Mulberry growth period identification method and system based on image analysis

Abstract

The invention belongs to the technical field of image analysis, and particularly relates to a mulberry growth period identification method and system based on image analysis, and the method comprises the steps: obtaining image data of each growth period of a mulberry, preprocessing the image data, and inputting the preprocessed image data into a feature extraction module; performing feature extraction on the image data in each growth period, and establishing a feature comparison table of the images in each growth period of the mulberry trees; mulberry image data of each mulberry planting area to be identified is collected in real time, and after preprocessing, the mulberry image data is input into the feature extraction module for real-time image feature extraction; inputting the real-time image features into a feature comparison module, calling a mulberry growth period image feature comparison table, and matching the real-time image features with the mulberry growth period image feature comparison table; obtaining the mulberry growth period of each mulberry planting area to be identified according to the matching result of the real-time image feature and the mulberry growth period image feature comparison table; a preset mulberry growth period management measure table is called, management measures of the mulberries in the mulberry planting areas are obtained, mulberry planting management is facilitated, and the economic benefits of mulberry planting are improved.

Description

Technical Field

[0001] The invention belongs to the technical field of image analysis, and in particular relates to a method and system for identifying the growth period of a mulberry tree by image analysis. Background Art

[0002] The growth period of mulberry trees refers to the various growth periods that mulberry trees go through from seed germination to maturity, including the germination period, vigorous growth period, slow growth period and dormancy period. Mulberry trees have high economic value, but they need to match corresponding management measures in different growth periods to achieve good management in each growth period, thereby obtaining higher economic value.

[0003] Different management measures need to be provided for mulberry trees in different growth periods. For example, during the vigorous growth period of mulberry trees, managers need to provide mulberry trees with fertilization methods and fertilizer amounts that match the vigorous growth period. At the same time, they also need to provide accurate disease and pest monitoring to avoid diseases and pests causing mulberry tree production reduction, thereby reducing the economic benefits of mulberry trees.

[0004] Therefore, how to identify the various growth stages of mulberry trees and then implement different management measures for different growth stages, such as fertilization management measures, pest and disease management measures, and pruning management measures, in order to improve the economic benefits of mulberry tree planting, is a technical problem that urgently needs to be solved. Summary of the invention

[0005] The object of the present invention is to provide a mulberry tree growth period recognition method and system based on image analysis, so as to identify the various growth periods of mulberry trees, and then carry out different management measures for different growth periods, such as fertilization management measures, pest and disease management measures and pruning management measures, so as to improve the economic benefits of mulberry tree planting.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows: In a first aspect, a method for identifying the growth period of a mulberry tree by image analysis is provided, comprising the following steps: S1: acquiring image data of each growth stage of the mulberry tree, preprocessing the image data of each growth stage, and inputting the preprocessed image data of each growth stage into a feature extraction module; S2: The feature extraction module performs feature extraction on the image data of each growth period respectively, extracts the image features of the image data of each growth period of the mulberry tree, and establishes a comparison table of image features of each growth period of the mulberry tree based on the extracted image features of each growth period of the mulberry tree; S3: collecting mulberry tree image data of each mulberry tree planting area to be identified in real time through the data acquisition module, pre-processing the mulberry tree image data of each mulberry tree planting area to be identified, and then inputting the pre-processed data into the feature extraction module for real-time image feature extraction; S4: inputting the real-time image features extracted from the mulberry tree image data of each mulberry tree planting area to be identified into a feature comparison module, the feature comparison module calling up the image feature comparison table of each growth stage of the mulberry tree, and matching the real-time image features with the image feature comparison table of each growth stage of the mulberry tree; S5: Obtaining the growth period of mulberry trees in each mulberry planting area to be identified according to the matching result of the real-time image feature and the image feature comparison table of each growth period of mulberry trees; S6: Retrieve a preset management measures table for the growing period of mulberry trees, match the growing period of mulberry trees in each mulberry tree planting area to be identified with the management measures in the management measures table for the growing period of mulberry trees, and respectively implement the matched management measures for the mulberry trees in each mulberry tree planting area.

[0007] Preferably, the specific process of preprocessing the image data of each growth period in step S1 is as follows: S11: deleting duplicate images, invalid images and abnormal images in the image data of each growth period, and converting the image data of the remaining growth periods into grayscale images; S12: putting the grayscale image into the frequency domain for image enhancement processing; S13: Perform image segmentation processing on the grayscale image after image enhancement processing, and divide the image into regions with similar features.

[0008] Preferably, the specific process of putting the image data of the remaining growth period into the frequency domain for processing in step S12 is as follows: S121: Preset a specified basis function and initialize the scale factor and translation factor of the specified basis function; S122: transforming the image data of each growth period using the initialized basis function, obtaining the coefficients of the image data of each growth period at the specified scale and signal frequency, and moving the translation factor at a uniform speed until reaching the end of the signal; S123: changing the scale factor in the specified direction, and iteratively executing step S122; S124: Perform inverse transformation on the image data of each growth stage by deconvolution and upsampling based on the obtained coefficients to obtain a processed image.

[0009] Preferably, the specific process of performing image segmentation processing on the grayscale image subjected to image enhancement processing in step S13 is as follows: S131: randomly selecting a specified number of initial central pixels in the grayscale image to be segmented; S132: assigning each data point to the nearest initial central pixel point to obtain a neighborhood of each initial central pixel point; S133: calculating the mean of each initial central pixel point based on each initial central pixel point and its neighborhood, assigning the calculated mean value to the corresponding central pixel point, and updating the central pixel point; S134: Iterate steps S132 to S133 until the central pixel does not change or the maximum number of iterations is reached, output the cluster label of each data point, and obtain different feature spaces.

[0010] Preferably, the specific process of the feature extraction module in step S2 performing feature extraction on the image data of each growth period is as follows: S21: Detect key image feature points in the different feature spaces, randomly select a pixel point in each feature space, and record its grayscale value as G; S22: setting a threshold t, and determining as key pixels the pixels whose pixel values ​​of N consecutive pixels in the neighborhood of the randomly selected pixel are less than Gt or greater than G+t; S23: If at least three of the randomly selected pixels in the right, left, top, and bottom directions satisfy the pixel values ​​less than Gt or greater than G+t, then continue to detect all 16 neighboring pixels of the pixel, otherwise the pixel is determined to be a non-feature point; S24: Calculate the gradient direction of pixels in the neighborhood of each key pixel as the direction of each key pixel, divide the image area around each key pixel into blocks, calculate the gradient histogram in each image block, and generate a feature vector as a descriptor of the key pixel.

[0011] Preferably, the specific process of matching the real-time image features with the image feature comparison table of each growth stage of mulberry trees in step S4 is as follows: Calculate the Euclidean distance between the descriptors of the key pixel points of the images in the real-time image of the mulberry tree and the image feature comparison table of the mulberry tree at each growth stage to perform feature matching; The calculation formula of the Euclidean distance is as follows: ; in, Dis ij is the Euclidean distance between the descriptors of key pixels, X ik For the real-time image of the mulberry tree i The descriptor of the key pixel k elements, X jk The image features of mulberry trees in different growth stages are compared in the image j The descriptor of the key pixel k elements, n is the dimension of the feature vector.

[0012] Preferably, the mulberry tree growth period management measures table preset in step S6 includes each mulberry tree growth period and management measures corresponding to each mulberry tree growth period, including fertilization measures, pest and disease management measures and pruning measures; The fertilization measures include fertilization type, fertilization node and fertilization amount; The pest and disease management measures include controlling the humidity of the mulberry planting environment, and the spraying time and density of the pest and disease control agent.

[0013] In a second aspect, a mulberry tree growth period recognition system based on image analysis is provided, which is used to implement the mulberry tree growth period recognition method based on image analysis, and comprises a data acquisition module, an image preprocessing module, a feature extraction module, a feature comparison module and a management measure matching module, wherein the data acquisition module is connected to the image preprocessing module, the image preprocessing module is connected to the feature extraction module, the feature extraction module is connected to the feature comparison module, and the feature comparison module is connected to the management measure matching module; The data acquisition module is used to acquire image data of mulberry trees at various growth stages and to collect image data of mulberry trees in various mulberry planting areas to be identified in real time; The image preprocessing module is used to preprocess the image data of each growth period and the mulberry tree image data of each mulberry tree planting area to be identified, and input the preprocessed image data of each growth period into the feature extraction module; The feature extraction module is used to extract features from the image data of each growth period, extract image features of the image data of each growth period of the mulberry tree, and establish a comparison table of image features of each growth period of the mulberry tree based on the extracted image features of each growth period of the mulberry tree; The feature comparison module is used to retrieve the image feature comparison table of each growth period of the mulberry tree, match the real-time image feature with the image feature comparison table of each growth period of the mulberry tree, and obtain the growth period of the mulberry tree in each mulberry tree planting area to be identified; The management measure matching module is used to retrieve a preset mulberry tree growth period management measure table, match the mulberry tree growth period of each mulberry tree planting area to be identified with the management measures in the mulberry tree growth period management measure table, and respectively execute the matched management measures on the mulberry trees in each mulberry tree planting area.

[0014] The beneficial effects of the present invention include: The image analysis method and system for identifying the growth period of mulberry trees provided by the present invention obtain image data of each growth period of mulberry trees, and input the data into a feature extraction module after preprocessing; perform feature extraction on the image data of each growth period, and establish an image feature comparison table of each growth period of mulberry trees; collect image data of mulberry trees in each mulberry tree planting area to be identified in real time, and input the data into a feature extraction module after preprocessing to perform real-time image feature extraction; input the real-time image feature into a feature comparison module, retrieve the image feature comparison table of each growth period of mulberry trees, and match the real-time image feature with the image feature comparison table of each growth period of mulberry trees; obtain the growth period of mulberry trees in each mulberry tree planting area to be identified according to the matching result of the real-time image feature and the image feature comparison table of each growth period of mulberry trees; retrieve a preset management measure table of mulberry tree growth period, and obtain the management measures of mulberry trees in each mulberry tree planting area, so as to facilitate the management of mulberry tree planting and improve the economic benefits of mulberry tree planting.

[0015] Firstly, by deleting duplicate images, invalid images and abnormal images in the image data of each growth period, the influence of duplicate images, invalid images and abnormal images on the subsequent growth period recognition is avoided, and the image processing complexity is effectively reduced by converting them into grayscale images, a specified basis function is preset, and the scale factor and translation factor of the specified basis function are initialized and transformed to obtain the coefficients of the image data of each growth period at the specified scale and signal frequency, and the translation factor is moved at a constant speed until the end of the signal is reached; the scale factor is changed in the specified direction, and the image data of each growth period is inversely transformed by deconvolution and upsampling to obtain the image enhancement processing of the processed image, so that the subsequent feature extraction can be more accurate, and the segmentation into areas with similar features can improve the feature extraction efficiency, thereby improving the recognition efficiency of the mulberry tree growth period.

[0016] Secondly, by detecting key image feature points in different feature spaces, a pixel point is randomly selected in each feature space, and the pixel point whose pixel values ​​of N consecutive pixels in the neighborhood of the randomly selected pixel point are less than the specified value or greater than the set value is determined as a key pixel point; if at least three of the pixels in the right, left, top and bottom directions of the randomly selected pixel point meet the requirement that the pixel values ​​are less than the specified value or greater than the set value, then the 16 neighborhood pixels of the pixel point are continuously detected, otherwise the pixel point is determined to be a non-feature point; the gradient direction of the pixels in the neighborhood of each key pixel point is calculated as the direction of each key pixel point, the image area around each key pixel point is divided into blocks, the gradient histogram in each image block is calculated, and a feature vector is generated as the descriptor of the key pixel point, thereby realizing the accurate extraction of image features, providing an accurate data basis for the subsequent identification of the growing period of mulberry trees, and effectively improving the accuracy of the identification of the growing period.

[0017] Thirdly, the feature matching process by calculating the Euclidean distance between the descriptors of the key pixel points of the images in the feature comparison table of the real-time image of the mulberry tree and the image of each growth period of the mulberry tree can effectively improve the accuracy of feature comparison and further improve the recognition accuracy of the growth period of the mulberry tree.

[0018] Finally, by setting up a management measures table for the growth period of mulberry trees based on big data, the growth period of each mulberry tree and the corresponding management measures for each growth period of the mulberry tree are preset, including fertilization measures, pest and disease management measures, and pruning measures. On the one hand, there is no need to manually set management measures after obtaining the growth period of mulberry trees in the corresponding area, but directly obtain standard management measures based on big data to promote the growth of mulberry trees. On the other hand, it can improve the management efficiency of mulberry planting. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The present invention is a schematic flow chart of the method for identifying the growth period of mulberry trees by image analysis.

[0020] Figure 2 The present invention is a schematic diagram of the feature extraction process in the method for identifying the growth period of mulberry trees by image analysis.

[0021] Figure 3 The schematic diagram of the architecture of the mulberry tree growth period recognition system based on image analysis of the present invention. DETAILED DESCRIPTION

[0022] The following is combined with Figure 1~Figure 3 The present invention is further described in detail: Example 1 See attached Figure 1 As shown, a method for identifying the growth period of mulberry trees by image analysis comprises the following steps: S1: Obtain image data of each growth stage of the mulberry tree, pre-process the image data of each growth stage, and input the pre-processed image data of each growth stage into a feature extraction module. The pre-processing eliminates duplicate images, invalid images and abnormal images to avoid affecting subsequent feature extraction results, and performs image conversion and image enhancement to facilitate subsequent efficient extraction of image features.

[0023] S2: The feature extraction module extracts features from the image data of each growth period, and obtains feature information of each growth period of the mulberry tree by extracting the image features of the image data of each growth period, so as to provide a reference standard for the subsequent identification of the growth period of the mulberry tree. The image features of the image data of each growth period of the mulberry tree are extracted, and a comparison table of image features of each growth period of the mulberry tree is established based on the extracted image features of each growth period of the mulberry tree, and the establishment of the comparison table of image features of each growth period provides convenience for subsequent feature comparison and matching.

[0024] S3: The mulberry tree image data of each mulberry tree planting area to be identified is collected in real time through the data collection module, and the mulberry tree image data of each mulberry tree planting area to be identified is pre-processed and then input into the feature extraction module for real-time image feature extraction. Through the real-time image feature extraction of mulberry trees, feature information of the mulberry tree images to be identified is obtained, which is convenient for subsequent feature comparison and matching of corresponding growth periods.

[0025] S4: The real-time image features extracted from the mulberry tree image data of each mulberry tree planting area to be identified are input into a feature comparison module, the feature comparison module retrieves the image feature comparison table of each growth period of the mulberry tree, and matches the real-time image features with the image feature comparison table of each growth period of the mulberry tree. The feature comparison table can achieve fast and accurate matching, thereby realizing fast identification of the growth period of the mulberry tree.

[0026] S5: The growth period of mulberry trees in each mulberry planting area to be identified is obtained according to the matching results of the real-time image features and the image feature comparison table of each growth period of mulberry trees, so as to facilitate the subsequent implementation of corresponding management measures for each mulberry planting area, so as to improve the management efficiency and management effect of each mulberry planting area and promote the increase of economic benefits of mulberry trees.

[0027] S6: Retrieve the preset mulberry tree growth period management measures table, match the mulberry tree growth period of each mulberry tree planting area to be identified with the management measures in the mulberry tree growth period management measures table, and implement the matching management measures for the mulberry trees in each mulberry tree planting area. By dividing different mulberry tree planting areas, identifying the mulberry tree growth period for different mulberry tree planting areas, and implementing more matching management measures according to the identification results of different mulberry tree growth periods, dynamic mulberry tree planting management is achieved.

[0028] Example 2 Based on Example 1, the specific process of preprocessing the image data of each growth stage in step S1 is as follows: S11: deleting duplicate images, invalid images and abnormal images in the image data of each growth period, and converting the image data of the remaining growth periods into grayscale images; S12: Put the grayscale image into the frequency domain for image enhancement. The main purpose of putting the grayscale image into the frequency domain is to simplify image processing tasks, extract features, enhance and denoise, and compress images. Through frequency domain conversion, the image can be converted from the spatial domain to the frequency domain, so as to better analyze and process the frequency components of the image; S13: Perform image segmentation processing on the grayscale image after image enhancement processing, and divide the image into regions with similar features.

[0029] In this embodiment, the specific process of putting the image data of the remaining growth period into the frequency domain for processing in step S12 is as follows: S121: Preset a specified basis function and initialize the scale factor and translation factor of the specified basis function; S122: transforming the image data of each growth period using the initialized basis function, obtaining the coefficients of the image data of each growth period at the specified scale and signal frequency, and moving the translation factor at a uniform speed until reaching the end of the signal; S123: changing the scale factor in the specified direction, and iteratively executing step S122; S124: Perform inverse transformation on the image data of each growth stage by deconvolution and upsampling based on the obtained coefficients to obtain a processed image.

[0030] By deleting duplicate images, invalid images and abnormal images in the image data of each growth period, avoiding the influence of duplicate images, invalid images and abnormal images on the subsequent growth period identification, and converting them into grayscale images to effectively reduce the complexity of image processing, a specified basis function is preset, and the scale factor and translation factor of the specified basis function are initialized and then transformed to obtain the coefficients of the image data of each growth period at the specified scale and signal frequency, and the translation factor is moved at a constant speed until the end of the signal is reached; the scale factor is changed in the specified direction, and the image data of each growth period is inversely transformed by deconvolution and upsampling to obtain image enhancement processing of the processed image, so that subsequent feature extraction can be performed more accurately, and segmentation into areas with similar features can improve the feature extraction efficiency, thereby improving the efficiency of mulberry tree growth period identification.

[0031] The specific process of performing image segmentation processing on the grayscale image processed by image enhancement in step S13 is as follows: S131: randomly selecting a specified number of initial central pixels in the grayscale image to be segmented; S132: assigning each data point to the nearest initial central pixel point to obtain a neighborhood of each initial central pixel point; S133: calculating the mean of each initial central pixel point based on each initial central pixel point and its neighborhood, assigning the calculated mean value to the corresponding central pixel point, and updating the central pixel point; S134: Iterate steps S132 to S133 until the central pixel does not change or the maximum number of iterations is reached, output the cluster label of each data point, and obtain different feature spaces.

[0032] The above process can effectively segment the image into different feature areas, reduce the complexity of subsequent image processing, and help identify and extract specific objects or areas in the image. The segmented image can provide clearer input for tasks such as feature extraction, classification, and recognition.

[0033] Example 3 Based on Example 1 or Example 2, see Figure 2 The specific process of the feature extraction module in step S2 for extracting features from the image data of each growth period is as follows: S21: Detect key image feature points in the different feature spaces, randomly select a pixel point in each feature space, and record its grayscale value as G; S22: setting a threshold t, and determining as key pixels the pixels whose pixel values ​​of N consecutive pixels in the neighborhood of the randomly selected pixel are less than Gt or greater than G+t; S23: If at least three of the randomly selected pixels in the right, left, top, and bottom directions satisfy the pixel values ​​less than Gt or greater than G+t, then continue to detect all 16 neighboring pixels of the pixel, otherwise the pixel is determined to be a non-feature point; S24: Calculate the gradient direction of pixels in the neighborhood of each key pixel as the direction of each key pixel, divide the image area around each key pixel into blocks, calculate the gradient histogram in each image block, and generate a feature vector as a descriptor of the key pixel.

[0034] The above process detects key image feature points in different feature spaces, randomly selects a pixel point in each feature space, and determines the pixel point whose pixel values ​​of N consecutive pixels in the neighborhood of the randomly selected pixel point are less than the specified value or greater than the set value as the key pixel point; if at least three of the pixels in the right, left, top and bottom directions of the randomly selected pixel point meet the requirement that the pixel values ​​are less than the specified value or greater than the set value, then continue to detect 16 neighborhood pixels of the pixel point, otherwise the pixel point is determined to be a non-feature point; calculate the gradient direction of the pixels in the neighborhood of each key pixel point as the direction of each key pixel point, divide the image area around each key pixel point into blocks, calculate the gradient histogram in each image block, generate a feature vector as the descriptor of the key pixel point, realize the accurate extraction of image features, provide an accurate data basis for the subsequent recognition of the growth period of mulberry trees, and effectively improve the accuracy of the recognition of the growth period.

[0035] Example 4 On the basis of Example 1 or Example 2 or Example 3, the specific process of matching the real-time image features with the image feature comparison table of each growth stage of mulberry trees in step S4 is as follows: Calculate the Euclidean distance between the descriptors of the key pixel points of the images in the real-time image of the mulberry tree and the image feature comparison table of the mulberry tree at each growth stage to perform feature matching; The calculation formula of the Euclidean distance is as follows: ; in, Dis ij is the Euclidean distance between the descriptors of key pixels, X ik For the real-time image of the mulberry tree i The descriptor of the key pixel k elements, X jk The image features of mulberry trees in different growth stages are compared in the image j The descriptor of the key pixel k elements, n is the dimension of the feature vector.

[0036] In this embodiment, the mulberry tree growth period management measures table preset in step S6 includes each mulberry tree growth period and the management measures corresponding to each mulberry tree growth period, including fertilization measures, pest and disease management measures and pruning measures. Among them, the fertilization measures include fertilization types, fertilization nodes and fertilization amounts; the pest and disease management measures include controlling the humidity of the mulberry tree planting environment, the spraying time and spraying density of the pest and disease control agent.

[0037] A mulberry tree growth period recognition system based on image analysis is used to implement the mulberry tree growth period recognition method based on image analysis, see Figure 3 , including a data acquisition module, an image preprocessing module, a feature extraction module, a feature comparison module and a management measure matching module, the data acquisition module is connected to the image preprocessing module, the image preprocessing module is connected to the feature extraction module, the feature extraction module is connected to the feature comparison module, and the feature comparison module is connected to the management measure matching module.

[0038] The data acquisition module is used to acquire image data of each growth period of mulberry trees and collect image data of mulberry trees in each mulberry planting area to be identified in real time; the image preprocessing module is used to preprocess the image data of each growth period and the image data of mulberry trees in each mulberry planting area to be identified, and input the preprocessed image data of each growth period into the feature extraction module; the feature extraction module is used to extract features of the image data of each growth period respectively, extract image features of the image data of each growth period of mulberry trees, and establish a comparison table of image features of each growth period of mulberry trees based on the extracted image features of each growth period of mulberry trees; the feature comparison module is used to call the comparison table of image features of each growth period of mulberry trees, match the real-time image features with the comparison table of image features of each growth period of mulberry trees, and obtain the growth period of mulberry trees in each mulberry planting area to be identified; the management measure matching module is used to call the preset management measure table of mulberry tree growth period, match the growth period of mulberry trees in each mulberry planting area to be identified with the management measures in the management measure table of mulberry tree growth period, and execute the matched management measures on the mulberry trees in each mulberry planting area respectively.

[0039] In summary, the image analysis method and system for identifying the growth period of mulberry trees provided by the present invention obtain image data of each growth period of mulberry trees, and input the data into a feature extraction module after preprocessing; perform feature extraction on the image data of each growth period, and establish an image feature comparison table of each growth period of mulberry trees; collect mulberry image data of each mulberry planting area to be identified in real time, and input the data into a feature extraction module after preprocessing to perform real-time image feature extraction; input the real-time image feature into a feature comparison module, retrieve the image feature comparison table of each growth period of mulberry trees, and match the real-time image feature with the image feature comparison table of each growth period of mulberry trees; obtain the growth period of mulberry trees in each mulberry planting area to be identified according to the matching result of the real-time image feature with the image feature comparison table of each growth period of mulberry trees; retrieve a preset management measure table for the growth period of mulberry trees, and obtain management measures for mulberry trees in each mulberry planting area, so as to facilitate mulberry planting management and improve the economic benefits of mulberry planting.

[0040] By deleting duplicate images, invalid images and abnormal images in the image data of each growth period to avoid affecting the subsequent growth period identification, and converting them into grayscale images to effectively reduce the complexity of image processing, a specified basis function is preset, and the scale factor and translation factor of the specified basis function are initialized and then transformed to obtain the coefficients of the image data of each growth period at the specified scale and signal frequency, and the translation factor is moved at a constant speed until the end of the signal is reached; the scale factor is changed in the specified direction, and the image data of each growth period is inversely transformed by deconvolution and upsampling to obtain image enhancement processing of the processed image, so that subsequent feature extraction can be performed more accurately, and segmentation into areas with similar features can improve the feature extraction efficiency, thereby improving the efficiency of mulberry tree growth period identification.

[0041] By detecting key image feature points in different feature spaces, randomly selecting a pixel point in each feature space, and determining the specified pixel points in the neighborhood of the randomly selected pixel point that meet specific conditions as key pixels or non-feature points; generating feature vectors as descriptors of key pixels, the accurate extraction of image features is realized, providing an accurate data basis for subsequent mulberry tree growth period identification, and effectively improving the accuracy of growth period identification. The process of feature matching by calculating the Euclidean distance between the descriptors of key pixel points in the real-time image of the mulberry tree and the image feature comparison table of each growth period of the mulberry tree can effectively improve the accuracy of feature comparison and further improve the recognition accuracy of the growth period of the mulberry tree. A management measure table for the growth period of the mulberry tree based on big data is set, and the growth period of each mulberry tree and the corresponding management measures for each growth period of the mulberry tree are preset, including fertilization measures, pest and disease management measures, and pruning measures. On the one hand, there is no need to manually set management measures after obtaining the growth period of the mulberry tree in the corresponding area, but directly obtain standard management measures based on big data to promote the growth of mulberry trees. On the other hand, it can improve the management efficiency of mulberry planting.

Claims

1. A method for identifying the growth period of mulberry trees by image analysis, characterized in that: The following steps are involved: S1: acquiring image data of each growth stage of the mulberry tree, preprocessing the image data of each growth stage, and inputting the preprocessed image data of each growth stage into a feature extraction module; S2: The feature extraction module performs feature extraction on the image data of each growth period respectively, extracts the image features of the image data of each growth period of the mulberry tree, and establishes a comparison table of image features of each growth period of the mulberry tree based on the extracted image features of each growth period of the mulberry tree; S3: collecting mulberry tree image data of each mulberry tree planting area to be identified in real time through the data acquisition module, pre-processing the mulberry tree image data of each mulberry tree planting area to be identified, and then inputting the pre-processed data into the feature extraction module for real-time image feature extraction; S4: inputting the real-time image features extracted from the mulberry tree image data of each mulberry tree planting area to be identified into a feature comparison module, the feature comparison module calling up the image feature comparison table of each growth stage of the mulberry tree, and matching the real-time image features with the image feature comparison table of each growth stage of the mulberry tree; S5: Obtaining the growth period of mulberry trees in each mulberry planting area to be identified according to the matching result of the real-time image feature and the image feature comparison table of each growth period of mulberry trees; S6: Retrieve a preset management measures table for the growing period of mulberry trees, match the growing period of mulberry trees in each mulberry tree planting area to be identified with the management measures in the management measures table for the growing period of mulberry trees, and respectively implement the matched management measures for the mulberry trees in each mulberry tree planting area.

2. The method for identifying the growth period of mulberry trees by image analysis according to claim 1, characterized in that: The specific process of preprocessing the image data of each growth period in step S1 is as follows: S11: deleting duplicate images, invalid images and abnormal images in the image data of each growth period, and converting the image data of the remaining growth periods into grayscale images; S12: putting the grayscale image into the frequency domain for image enhancement processing; S13: Perform image segmentation processing on the grayscale image after image enhancement processing, and divide the image into regions with similar features.

3. The method for identifying the growth period of mulberry trees by image analysis according to claim 2, characterized in that: The specific process of putting the grayscale image into the frequency domain for image enhancement processing in step S12 is as follows: S121: Preset a specified basis function and initialize the scale factor and translation factor of the specified basis function; S122: transforming the image data of each growth period using the initialized basis function, obtaining the coefficients of the image data of each growth period at the specified scale and signal frequency, and moving the translation factor at a uniform speed until reaching the end of the signal; S123: changing the scale factor in the specified direction, and iteratively executing step S122; S124: Perform inverse transformation on the image data of each growth stage by deconvolution and upsampling based on the obtained coefficients to obtain a processed image.

4. The method for identifying the growth period of mulberry trees by image analysis according to claim 2, characterized in that: The specific process of performing image segmentation processing on the grayscale image processed by image enhancement in step S13 is as follows: S131: randomly selecting a specified number of initial central pixels in the grayscale image to be segmented; S132: assigning each data point to the nearest initial central pixel point to obtain a neighborhood of each initial central pixel point; S133: calculating the mean of each initial central pixel point based on each initial central pixel point and its neighborhood, assigning the calculated mean value to the corresponding central pixel point, and updating the central pixel point; S134: Iterate steps S132 to S133 until the central pixel does not change or the maximum number of iterations is reached, output the cluster label of each data point, and obtain different feature spaces.

5. The method for identifying the growth period of mulberry trees by image analysis according to claim 4, characterized in that: The specific process of the feature extraction module in step S2 extracting features from the image data of each growth period is as follows: S21: Detect key image feature points in the different feature spaces, randomly select a pixel point in each feature space, and record its gray value as G; S22: setting a threshold t, and determining as key pixels the pixels whose pixel values ​​of N consecutive pixels in the neighborhood of the randomly selected pixel are less than Gt or greater than G+t; S23: If at least three of the randomly selected pixels in the right, left, top, and bottom directions satisfy the pixel values ​​less than Gt or greater than G+t, then continue to detect all 16 neighboring pixels of the pixel, otherwise the pixel is determined to be a non-feature point; S24: Calculate the gradient direction of pixels in the neighborhood of each key pixel as the direction of each key pixel, divide the image area around each key pixel into blocks, calculate the gradient histogram in each image block, and generate a feature vector as a descriptor of the key pixel.

6. The method for identifying the growth period of mulberry trees by image analysis according to claim 5, characterized in that: The specific process of matching the real-time image features with the image feature comparison table of each growth stage of mulberry trees in step S4 is as follows: Calculate the Euclidean distance between the descriptors of the key pixel points of the images in the real-time image of the mulberry tree and the image feature comparison table of the mulberry tree at each growth stage to perform feature matching; The calculation formula of the Euclidean distance is as follows: ； in, Dis ij is the Euclidean distance between the descriptors of key pixels, X ik For the real-time image of the mulberry tree i The descriptor of the key pixel k elements, X jk The image features of mulberry trees in different growth stages are compared in the image j The descriptor of the key pixel k elements, n is the dimension of the feature vector.

7. The method for identifying the growth period of mulberry trees by image analysis according to claim 1, characterized in that: The mulberry tree growth period management measures table preset in step S6 includes each mulberry tree growth period and the management measures corresponding to each mulberry tree growth period, including fertilization measures, pest and disease management measures and pruning measures; The fertilization measures include fertilization type, fertilization node and fertilization amount; The pest and disease management measures include controlling the humidity of the mulberry planting environment, and the spraying time and density of the pest and disease control agent.

8. A mulberry tree growth period recognition system based on image analysis, used to implement the mulberry tree growth period recognition method based on image analysis as claimed in any one of claims 1 to 7, characterized in that: It includes a data acquisition module, an image preprocessing module, a feature extraction module, a feature comparison module and a management measure matching module, wherein the data acquisition module is connected to the image preprocessing module, the image preprocessing module is connected to the feature extraction module, the feature extraction module is connected to the feature comparison module, and the feature comparison module is connected to the management measure matching module; The data acquisition module is used to acquire image data of mulberry trees at various growth stages and to collect image data of mulberry trees in various mulberry planting areas to be identified in real time; The image preprocessing module is used to preprocess the image data of each growth period and the mulberry tree image data of each mulberry tree planting area to be identified, and input the preprocessed image data of each growth period into the feature extraction module; The feature extraction module is used to extract features from the image data of each growth period, extract image features of the image data of each growth period of the mulberry tree, and establish a comparison table of image features of each growth period of the mulberry tree based on the extracted image features of each growth period of the mulberry tree; The feature comparison module is used to retrieve the image feature comparison table of each growth period of the mulberry tree, match the real-time image feature with the image feature comparison table of each growth period of the mulberry tree, and obtain the growth period of the mulberry tree in each mulberry tree planting area to be identified; The management measure matching module is used to retrieve a preset mulberry tree growth period management measure table, match the mulberry tree growth period of each mulberry tree planting area to be identified with the management measures in the mulberry tree growth period management measure table, and respectively execute the matched management measures on the mulberry trees in each mulberry tree planting area.

Images