Rice regional dry green stand disease early warning method and device based on unmanned aerial vehicle inspection

Abstract

This specification provides a method and apparatus for early warning of regional drought-induced green shoot disease in rice based on drone inspection. The method includes: inspecting a rice-growing area using a drone and acquiring an image sequence; acquiring the RGB three-channel image of each image in the image sequence and retaining the RG two-channel image matrix; calculating the mean of the R-channel matrix of each image in the RG two-channel image matrix, and calculating the mean and variance of the mean of the R-channel matrix; based on the mean and variance of the R-channel matrix, and according to a pre-set ω value, identifying abnormal growth areas; based on the abnormal growth areas, performing gradient segmentation on the green channel image of the identified abnormal growth areas according to a set gradient segmentation threshold t to obtain a labeling matrix; calculating the S value of the labeling matrix and determining whether drought-induced green shoot disease has occurred in the rice in the area according to a set θ value; and tracing back the drought-induced green shoot disease area according to the recorded index sequence and issuing an early warning to the client.

Description

Technical Field

[0001] This document relates to the field of computer technology, and in particular to a method and device for early warning of regional drought-induced green rice disease based on drone inspection. Background Technology

[0002] In recent years, the frequency and severity of rice stem rot in rice have been increasing across the country, often causing significant losses to farmers due to the disease's unique causes. The main symptoms of stem rot in rice are: after heading, the panicle stalks are bent and mostly upright instead of drooping; the rachis and branches are curved; the glumes are deformed, sometimes resembling an eagle's beak; the rice remains green during the yellow ripening stage; and grain filling is poor or the grains are empty. The causes of stem rot are low soil organic matter content, soil prone to compaction and sludge formation, poor physical and chemical properties, and insufficient active trace elements, which also contribute to the easy expansion of disease-affected areas. Stem rot typically only shows obvious symptoms near heading, making it difficult to detect in the early stages of rice growth, which is one of the reasons why it often causes huge losses for farmers.

[0003] Traditionally, the detection of rice seedling blight relies primarily on manual inspections. However, this method is labor-intensive, and in large paddy fields, manual inspections are prone to visual fatigue, leading to missed detection of infected rice. This delayed detection allows the disease-affected area to expand, increasing losses for farmers. Therefore, existing traditional methods suffer from the following drawbacks: high labor costs and low efficiency; limited effectiveness in inspecting large paddy fields; and the tendency for delayed inspections to lead to the expansion of the disease-affected area. Therefore, there is an urgent need for a regional rice seedling blight early warning method based on drone inspections to address the problems of existing traditional methods. Summary of the Invention

[0004] The purpose of this invention is to provide a method and device for early warning of regional drought-induced green shoot disease in rice based on unmanned aerial vehicle (UAV) inspection, aiming to solve the above-mentioned problems in the prior art.

[0005] This invention provides a method for early warning of regional drought-induced green shoot disease in rice based on unmanned aerial vehicle (UAV) inspection, comprising:

[0006] The rice-growing area is inspected by drones and image sequences are obtained. The RGB three-channel image of each image in the image sequence is obtained and the RG two-channel image matrix is ​​retained.

[0007] Calculate the mean of the R-channel matrix for each image in the RG two-channel image matrix, and calculate the mean and variance of the R-channel matrix mean;

[0008] Based on the mean and variance of the R channel matrix, and according to the preset ω value, the abnormal growth region is identified. Based on the abnormal growth region, the green channel image of the identified abnormal growth region is gradient segmented according to the preset gradient segmentation threshold t to obtain the label matrix.

[0009] The S value of the marker matrix is ​​calculated, and the θ value is used to determine whether the rice in the area has drought-induced green shoot disease. The drought-induced green shoot disease area is traced back according to the recorded index sequence and an early warning is issued to the client.

[0010] This invention provides a regional drought-induced green shoot disease early warning device for rice based on drone inspection, comprising:

[0011] The image processing module is used to inspect rice fields using drones and acquire image sequences, acquire the RGB three-channel image of each image in the image sequence and retain the RG two-channel image matrix;

[0012] The first calculation module is used to calculate the mean of the R-channel matrix of each image in the RG two-channel image matrix, and to calculate the mean and variance of the mean of the R-channel matrix.

[0013] The second calculation module is used to calculate and identify abnormal growth regions based on the mean and variance of the R channel matrix and according to a preset ω value. Based on the abnormal growth regions, gradient segmentation is performed on the green channel image of the identified abnormal growth regions according to a set gradient segmentation threshold t to obtain a label matrix.

[0014] The third calculation module is used to calculate the S value of the marker matrix and determine whether the rice in the area has drought-induced green shoot disease based on the set θ value. It also backtracks the drought-induced green shoot disease area based on the recorded subscript sequence and sends an early warning to the client.

[0015] This invention also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the above-described method for early warning of regional drought-induced green shoot disease in rice based on UAV inspection.

[0016] This invention also provides a computer-readable storage medium storing an information transmission implementation program. When the program is executed by a processor, it implements the steps of the above-described method for early warning of regional drought-induced green shoot disease in rice based on UAV inspection.

[0017] Using the embodiments of the present invention for the detection and early warning of drought-induced leaf spot disease can reduce labor costs; using drones for inspection and machine vision recognition avoids visual fatigue, has a good identification effect on drought-induced leaf spot disease, and greatly reduces the losses caused by the expansion of drought-induced leaf spot disease areas. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in one or more embodiments of this specification or in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a method for early warning of regional drought-induced green shoot disease in rice based on drone inspection, according to an embodiment of the present invention.

[0020] Figure 2 This is a detailed flowchart of the method for early warning of regional drought-induced green seedling disease in rice based on drone inspection, according to an embodiment of the present invention.

[0021] Figure 3 This is a schematic diagram of farmland and drones according to an embodiment of the present invention;

[0022] Figure 4 This is a schematic diagram of the drone inspection trajectory according to an embodiment of the present invention;

[0023] Figure 5 This is a schematic diagram of the eight elements surrounding an embodiment of the present invention;

[0024] Figure 6 This is a schematic diagram of a regional drought-induced green shoot disease early warning device for rice based on drone inspection, according to an embodiment of the present invention.

[0025] Figure 7 This is a schematic diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0026] To enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of this document.

[0027] Method Implementation Examples

[0028] According to embodiments of the present invention, a method for early warning of regional drought-induced green shoot disease in rice based on unmanned aerial vehicle (UAV) inspection is provided. Figure 1 This is a schematic diagram of a regional drought-induced green shoot disease early warning method for rice based on unmanned aerial vehicle (UAV) inspection, according to an embodiment of the present invention. Figure 1As shown, the method for early warning of regional drought-induced green shoot disease in rice based on drone inspection according to an embodiment of the present invention specifically includes:

[0029] Step S101: Inspect the rice paddy area using a drone and acquire an image sequence. Obtain the RGB three-channel image of each image in the image sequence and retain the RG two-channel image matrix; specifically:

[0030] The rice paddy area is defined as a rectangle with length m and width n. The camera area of ​​the drone at height h is k×k. Calculate the number of horizontal image acquisitions p and the number of vertical image acquisitions q during the drone's inspection using formulas 1 and 2 respectively. Then, determine the total number of image acquisitions required during the drone's inspection using formula 3.

[0031]

[0032]

[0033] num = p × q (Formula 3)

[0034] Where [·] indicates taking the integer part; num represents the total number of photos taken in the rice field area;

[0035] A drone inspection trajectory is set. Each time the drone moves, it collects one image and numbers them sequentially as P1, P2, P3, ..., P... num Obtain the image sequence P1, P2, P3, ..., P num .

[0036] For each image P i Where i = 1, 2, ..., num, read the RGB three-channel image, denoted as R. i G i B i Three data matrices, each k×k in size, representing the area captured by the drone's camera, with R retained. i G i Two-channel image matrix.

[0037] Step S102: Calculate the mean of the R-channel matrix for each image in the RG two-channel image matrix, and calculate the mean and variance of the R-channel matrix mean; specifically,

[0038] Calculate R for each image separately. i G i The mean of the data in the matrix:

[0039] Let operator For a k-dimensional element, all are If the vector is such that the mean is calculated according to Formula 4:

[0040] V Ri =QR i Q T Formula 4;

[0041] Where T represents transpose;

[0042] Calculate all R values ​​according to formulas 5 and 6. i The mean and variance of the means are as follows:

[0043]

[0044]

[0045] Where num represents the total number of photos taken in the rice-growing region.

[0046] Step S103: Based on the mean and variance of the R channel matrix, and according to a pre-set ω value, identify abnormal growth regions. Based on these abnormal growth regions, perform gradient segmentation on the green channel image of the identified abnormal growth regions according to a set gradient segmentation threshold t to obtain a labeling matrix; specifically, set V... R To represent the overall average maturity of rice in the paddy field, let ωσ be the value. R This represents the normal growth range, where the value of ω is determined by farmers' experience;

[0047] Based on Formula 7, determine the image sequence P1, P2, P3, ..., P... num Each V Ri Indicators for determining whether rice growth in the testing area is abnormal:

[0048]

[0049] Obtain the image sequence {P} that may show signs of drought. j |V Rj <V R -ωσ R}

[0050] For each image P j Gradient segmentation is performed using the green channel matrix Gj as shown in Formula 8.

[0051]

[0052] Among them, gj kk This represents the value in the k-th row and k-th column of this channel data matrix, where j represents the j-th image in the acquisition order.

[0053] For the green channel matrix G j Perform the traversal, and for each element encountered, gj ijTake the 8 elements surrounding it, and then combine each of the 8 elements surrounding it with gj. ij Calculate the difference; if the difference is greater than the gradient segmentation threshold t, then mark gj. ij Set to 1; otherwise, mark gj. ij If the value is 0, the tag matrix M is obtained. j .

[0054] Step S104: Calculate the S value of the marker matrix and determine whether drought-induced green shoot disease has occurred in the rice in the area based on the set θ value. Then, trace back the drought-induced green shoot disease area according to the recorded index sequence and issue an early warning to the client. Specifically,

[0055] Calculate the matrix M occupied by the elements marked as 1 according to Formula 9. j The proportions are as follows:

[0056] s j =QM j Q T Formula 9;

[0057] Set the drought warning threshold to θ and execute formula 10;

[0058]

[0059] Get the index sequence {j|s j ≥θ}, locate and backtrack according to the subscript sequence, and return the location where the warning occurred to the client.

[0060] In summary, the technical solution of this invention involves using drones for timed patrols to transmit images collected during the drone patrols; identifying drought-induced green leaf spot disease areas through a drought-induced green leaf spot disease identification module; and sending early warning information about areas in paddy fields where drought-induced green leaf spot disease may occur to farmers through user terminals.

[0061] The technical solutions of the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0062] like Figure 2 As shown, the technical solution of this invention uses a drone to inspect a large area of ​​paddy fields. First, some parameters of the farmland and the drone are set as follows:

[0063] For ease of understanding, such as Figure 3 As shown in the figure, in this embodiment of the invention, the paddy field (rice area) is set as a rectangle with a length of m and a width of n, and the camera area of ​​the drone at a height of h is k×k.

[0064] Calculate the number of images required for drone inspection using the following geometric relationships:

[0065] p: number of times the image is acquired laterally Where [·] indicates taking the integer part;

[0066] Number of times the image is acquired vertically, q: Where [·] indicates taking the integer part;

[0067] The number of times the CCP needs to acquire the image is p × q.

[0068] Setting the drone inspection trajectory, such as Figure 4 Example shown:

[0069] The distance the drone moves laterally each time is The distance moved vertically each time is

[0070] The drone collects one image each time it moves, and numbers them P1, P2, P3, ..., P6 according to the order in which they are acquired. num This facilitates the system's location of drought-stricken areas.

[0071] After obtaining the above image sequence P1, P2, P3, ..., P num Then, the initial excavation was carried out:

[0072] ① For each image P i Where i = 1, 2, ..., num, the three-channel colors are read and denoted as R. i G i B i Three data matrices, each k×k in size, representing the area captured by the camera, with R retained. i G i Two color matrices are used to calculate the R-value for each image. i G i The mean of the data in the matrix:

[0073] Let operator For a k-dimensional element, all are If the vector is given, the mean is calculated as follows:

[0074] V Ri =QR i Q T

[0075] Next, calculate all R values. i The mean and variance of the means are as follows:

[0076]

[0077]

[0078] First use V RThe overall average maturity of rice in a paddy field is expressed using ωσ. R This represents the normal growth range, where the value of ω is determined by farmers' experience and is usually set to 3.

[0079] Based on the image sequence P1, P2, P3, ..., P num Each V Ri The following are the criteria for determining whether rice growth in the monitored area is abnormal:

[0080]

[0081] ② Further image analysis of areas where drought-prone vegetation may occur:

[0082] From the initial excavation above, we obtained the image sequence {P} that may contain drought-resistant greening. j |V Rj <V R -3σ R}, then for each image P j Green channel matrix G j Gradient segmentation is performed, and the gradient segmentation scheme is as follows:

[0083] For matrix G j In the form of:

[0084]

[0085] Iterate through the matrix above, and for each element gj encountered... ij Take the 8 elements surrounding it, such as Figure 5 As shown:

[0086] The eight elements surrounding it are respectively associated with gj ij Calculate the difference; if the difference is greater than a given threshold t (t is the gradient segmentation threshold, which is obtained from historical experience and can be obtained after debugging the system in a test field), then mark gj. ij Set to 1; otherwise, mark gj. ij The value is 0. This yields the label matrix M. j .

[0087] Calculate the number of elements marked as 1 in matrix M. j The proportions are as follows:

[0088] s j =QM j Q T

[0089] Let the drought warning threshold be θ, where θ is set manually and represents the threshold at which a drought warning will be issued.

[0090]

[0091] ③ Return to the index sequence {j|s} recorded in the above steps. j ≥θ}, using the original number for location backtracking, returns the location of the drought-stricken area where the warning occurred to the client.

[0092] Device Example 1

[0093] According to an embodiment of the present invention, a regional drought-induced green shoot disease early warning device for rice based on unmanned aerial vehicle (UAV) inspection is provided. Figure 6 This is a schematic diagram of a regional drought-induced green shoot disease early warning device for rice based on drone inspection, according to an embodiment of the present invention. Figure 6 As shown, the regional drought-induced green shoot disease early warning device for rice based on drone inspection according to an embodiment of the present invention specifically includes:

[0094] Image processing module 60 is used to inspect rice fields by drone and acquire image sequences, acquire the RGB three-channel image of each image in the image sequence and retain the RG two-channel image matrix;

[0095] The first calculation module 62 is used to calculate the mean of the R channel matrix of each image in the RG two-channel image matrix, and to calculate the mean and variance of the mean of the R channel matrix.

[0096] The second calculation module 64 is used to calculate and identify abnormal growth regions based on the mean and variance of the R channel matrix and according to a preset ω value. Based on the abnormal growth regions, the green channel image of the identified abnormal growth regions is subjected to gradient segmentation according to a set gradient segmentation threshold t to obtain a label matrix.

[0097] The third calculation module 66 is used to calculate the S value of the marker matrix and determine whether the rice in the area has drought-induced green shoot disease based on the set θ value, and to backtrack the drought-induced green shoot disease area according to the recorded subscript sequence and issue an early warning to the client.

[0098] The embodiments of the present invention are device embodiments corresponding to the above method embodiments. The specific operation of each module can be understood with reference to the description of the method embodiments, and will not be repeated here.

[0099] Device Example 2

[0100] This invention provides an electronic device, such as... Figure 7 As shown, it includes: a memory 70, a processor 72, and a computer program stored in the memory 70 and executable on the processor 72, wherein the computer program, when executed by the processor 72, performs the steps as described in the method embodiment.

[0101] Device Example 2

[0102] This invention provides a computer-readable storage medium storing an information transmission implementation program, which, when executed by a processor 72, performs the steps described in the method embodiment.

[0103] The computer-readable storage media described in this embodiment include, but are not limited to, ROM, RAM, disk, or optical disk.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions 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 invention.

Claims

1. A rice regional dry green sprout disease early warning method based on unmanned aerial vehicle inspection, characterized in that, include: The rice-growing area is inspected by drones and image sequences are obtained. The RGB three-channel image of each image in the image sequence is obtained and the RG two-channel image matrix is ​​retained. Specifically, it includes: For each image ,in, Read the RGB three-channel image and denote them as follows: Three data matrices, the size of which is The area captured by the drone's camera retains Two-channel image matrix; Calculate the mean of the R-channel matrix for each image in the RG two-channel image matrix, and calculate the mean and variance of the R-channel matrix mean; specifically including: The mean value of data in the matrix is calculated as follows: Mean value of data in the matrix: Set operator For one The elements of a vector of dimension If all the elements of a vector of dimension are equal to, then the mean is calculated according to equation 4: Formula 4; Where T represents transpose; The mean and variance of the mean are calculated according to Equations 5 and 6 The mean and variance of the mean are calculated according to Equations 5 and 6 Equation 5; Equation 6; Where num represents the total number of photos taken in the rice-growing area; Based on the mean and variance of the R-channel matrix, and according to a pre-set ω value, abnormal growth regions are identified. Based on these abnormal growth regions, gradient segmentation is performed on the green channel image of the identified abnormal growth regions according to a set gradient segmentation threshold t, resulting in a labeling matrix. Specifically, the process of identifying abnormal growth regions based on the mean and variance of the R-channel matrix and according to a pre-set ω value includes: set up This indicates the overall average maturity of rice in the paddy field, and sets... This indicates the normal growth range, among which The value is determined by farmers' experience; Based on Formula 7, determine the image sequence. Each Indicators for determining whether rice growth in the testing area is abnormal: Equation 7; acquiring an image sequence in which a dry green stand can occur ; The S value of the marker matrix is ​​calculated, and the θ value is used to determine whether the rice in the area has drought-induced green shoot disease. The drought-induced green shoot disease area is traced back according to the recorded index sequence and an early warning is issued to the client.

2. The method of claim 1, wherein, The inspection of rice-growing areas and the acquisition of image sequences using drones specifically include: Set the rice paddy area as a length of Width The rectangle is located at a height of h, and the camera area of ​​the drone is [missing information]. Calculate the number of horizontal image acquisitions (p) and the number of vertical image acquisitions (q) during UAV inspection using Formulas 1 and 2 respectively, and determine the total number of image acquisitions required during UAV inspection using Formula 3: Official 1; Official 2; Formula 3: num = p × q; wherein, wherein represents the integer part; num represents the total number of photos acquired in the rice area; A drone inspection trajectory is set, and the drone collects one image each time it moves, and the images are numbered according to the order in which they are acquired. Obtain image sequences .

3. The method of claim 1, wherein, Based on the aforementioned abnormal growth region, gradient segmentation is performed on the green channel image of the region identifying the abnormal growth according to a set gradient segmentation threshold t, resulting in a labeling matrix that specifically includes: For each image the green channel matrix as in equation 8 Gradient segmentation is performed Official 8; Among them, gj kk This represents the value in the k-th row and k-th column of this channel data matrix, where j represents the j-th image in the acquisition order. Green channel matrix Perform the traversal, and each time an element is reached... Take the 8 elements surrounding it, and then combine each of the 8 elements surrounding it with... Calculate the difference; if the difference is greater than the gradient segmentation threshold t, then mark it. Set to 1; otherwise, mark as 1. If the value is 0, the tag matrix is ​​obtained. .

4. The method of claim 3, wherein, Calculating the S value of the marker matrix and determining whether drought-induced green shoot disease has occurred in the rice in the region based on the set θ value, and tracing back the drought-induced green shoot disease area based on the recorded index sequence and issuing an early warning to the client specifically includes: The ratio of the element marked 1 to the matrix is calculated according to equation 9 as follows: Official 9; The dry green stand early warning threshold is set as and formula 10 is executed; Official 10; Get the index sequence The system performs location backtracking based on the index sequence and returns the location of the drought-stricken area to the client.

5. A rice regional dry green sprout disease early warning device based on unmanned aerial vehicle inspection, characterized in that, include: The image processing module is used to inspect rice-growing areas using drones and acquire image sequences, obtaining the RGB three-channel image of each image in the sequence and retaining the RG two-channel image matrix; specifically, it is used to: process each image ,in, Read the RGB three-channel image and denote them as follows: Three data matrices, the size of which is The area captured by the drone's camera retains Two-channel image matrix; The first calculation module is used to calculate the mean of the R-channel matrix of each image in the RG two-channel image matrix, and to calculate the mean and variance of the R-channel matrix mean; specifically, it is used to calculate the mean and variance of each image. The mean of the data in the matrix: Set operator For one The elements of a vector of dimension The mean is calculated according to equation 4: Formula 4; Where T represents transpose; The mean and variance of the mean are calculated according to Equations 5 and 6 The mean and variance of the mean are calculated according to Equations 5 and 6 Equation 5; Equation 6; Where num represents the total number of photos taken in the rice-growing area; The second calculation module is used to calculate and identify abnormal growth regions based on the mean and variance of the R-channel matrix and according to a pre-set ω value. Based on the abnormal growth regions, gradient segmentation is performed on the green channel image of the identified abnormal growth regions according to a set gradient segmentation threshold t to obtain a label matrix; specifically, it is used to: set This indicates the overall average maturity of rice in the paddy field, and sets... This indicates the normal growth range, among which The value is determined by farmers' experience; Based on Formula 7, determine the image sequence. Each Indicators for determining whether rice growth in the testing area is abnormal: Equation 7; acquiring an image sequence in which a dry green stand can occur ; The third calculation module is used to calculate the S value of the marker matrix and determine whether the rice in the area has drought-induced green shoot disease based on the set θ value. It also backtracks the drought-induced green shoot disease area based on the recorded subscript sequence and sends an early warning to the client.

6. An electronic device, comprising: include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the method for early warning of regional drought-induced green shoot disease in rice based on unmanned aerial vehicle (UAV) inspection as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an information transmission implementation program, which, when executed by a processor, implements the steps of the method for early warning of regional drought-induced green shoot disease in rice based on unmanned aerial vehicle (UAV) inspection as described in any one of claims 1 to 4.

Images