A plant disease severity assessment method and device

Abstract

The present invention provides a method and device for evaluating the severity of plant diseases. The method includes: obtaining an actual plant unit image to be evaluated, extracting the actual plant unit from the actual plant unit image; determining the current ratio of the target disease spot area in the actual plant unit to the total area of ​​the actual plant unit; performing severity evaluation analysis based on the current ratio and the actual disease spot area ratio range corresponding to the severity level, and determining the severity level of the actual plant unit; the actual disease spot area ratio range is obtained by performing average value analysis based on the severity level and the sample plant unit images corresponding to it. The plant disease severity evaluation method disclosed by the present invention is simple and easy to operate, and evaluates the severity of plant diseases by performing an actual disease spot area ratio range obtained by performing an average value analysis based on the severity level and the sample plant unit images corresponding to it, thereby improving the accuracy of plant disease severity evaluation.

Description

Technical Field

[0001] The present invention relates to the field of plant protection technology, and in particular to a method and device for evaluating the severity of plant diseases. In addition, the present invention also relates to an electronic device and a non-transitory computer-readable storage medium. Background Art

[0002] Plant diseases are a major threat to agricultural production, affecting food security, food safety, ecological security, and even national security. Plant disease surveys and monitoring are an important way to understand the occurrence of diseases and the basis for disease management. They can provide a basis for disease prediction and forecasting, variety disease resistance identification and variety breeding, and disease prevention and control strategy formulation. In the process of plant disease surveys, the main survey indicators generally include prevalence, severity, disease index, etc. Severity is an indicator that describes the severity of the disease in a surveyed plant unit such as a single plant or organ. Generally, based on the ratio of the diseased area of ​​the surveyed plant unit to the total area of ​​the surveyed plant unit (expressed in the form of percentage or fraction, etc.), the severity is divided into different levels from light to heavy using a grading method to indicate the severity of plant diseases. There are often two specific ways to express the severity of plant diseases. One method is to use the representative values ​​of each level to express the severity of the disease, and the other method is to use the percentage of the diseased area of ​​the plant unit to the total area of ​​the surveyed plant unit to express the severity of the disease. Plant disease severity assessment is an important task in disease surveys and monitoring. Generally speaking, the investigation and assessment of disease severity should be carried out strictly in accordance with certain plant disease severity classification standards.

[0003] At present, the severity survey and assessment of plant diseases mainly rely on experienced personnel to observe with the naked eye. In addition to the naked eye observation method, the severity can be assessed by grid paper determination method, paper sample weighing method, etc. Sometimes the severity can be estimated by using the surveyed prevalence, or the severity level can be identified or estimated by using image processing technology, remote sensing technology, near-infrared spectroscopy technology, etc. The plant disease severity assessment based on remote sensing technology and near-infrared spectroscopy technology is still mainly in the research stage, and is rarely used in actual production. With the development of image acquisition and image processing technology, the plant disease severity assessment based on image processing has gradually increased in research and practical applications. There are two main methods for evaluating the severity of plant diseases using image processing technology. One method is to directly judge the severity of the disease based on the extracted disease image features; the other method is to obtain the ratio of the lesion area to the total area of ​​the plant unit through lesion segmentation, lesion area and plant unit total area (or lesion pixels and plant unit total pixels), and directly compare this ratio with the disease severity classification standard to evaluate the severity level of the corresponding diseased plant unit. However, the ratio of the lesion area to the total area of ​​the plant unit represented by the severity classification level of some plant diseases is not the ratio of the actual lesion area to the total area of ​​the plant unit, which brings great difficulties to the accurate assessment of the severity of such diseases. Therefore, when using the actual ratio of the lesion area to the total area of ​​the plant unit to compare with the ratio of the lesion area to the total area of ​​the plant unit in the disease severity classification standard to conduct severity assessment, a large error will occur, and even severity level assessment errors will occur, which will seriously affect the accurate assessment of disease severity. Therefore, how to provide a more accurate and efficient plant disease severity assessment scheme has become a technical issue that needs to be solved urgently by those skilled in the art. Summary of the invention

[0004] To this end, the present invention provides a plant disease severity assessment method and device to solve the defects of the prior art that plant disease severity assessment has high limitations, resulting in poor accuracy and applicability of plant disease severity assessment.

[0005] The present invention provides a method for assessing the severity of plant diseases, comprising:

[0006] Acquire an actual plant unit image to be evaluated, and extract actual plant units from the actual plant unit image;

[0007] determining a current ratio of the target diseased spot area in the actual plant unit to the total area of ​​the actual plant unit;

[0008] A severity assessment analysis is performed based on the current ratio and the actual diseased area ratio range corresponding to the preset severity level to determine the severity level of the actual plant unit; wherein the actual diseased area ratio range is obtained by performing an average value analysis based on the severity levels and their corresponding sample plant unit images.

[0009] Furthermore, before performing severity assessment analysis based on the current ratio and the actual lesion area ratio range corresponding to the preset severity level, the method further includes:

[0010] Obtaining sample plant images corresponding to the severity level;

[0011] Segmenting the sample plant unit region in the sample plant image to obtain a sample plant unit image; and segmenting the sample disease spot region in the sample plant unit image to obtain a sample disease spot image;

[0012] Determine the total area of ​​the sample plant unit in the sample plant unit image and the area of ​​the sample diseased spot in the sample diseased spot image;

[0013] Determining the ratio of the sample diseased spot area to the total area of ​​the sample plant unit;

[0014] Based on the ratio of the sample diseased spot area to the total area of ​​the sample plant unit, respectively determine the average value of the ratio of the sample diseased spot area to the total area of ​​the sample plant unit corresponding to each severity level;

[0015] Based on the average value of the ratio of the sample diseased spot area corresponding to each severity level to the total area of ​​the sample plant unit, the actual diseased spot area ratio range corresponding to each severity level is determined.

[0016] Further, based on the ratio of the sample diseased spot area to the total area of ​​the sample plant unit, the average value of the ratio of the sample diseased spot area to the total area of ​​the sample plant unit corresponding to each severity level is determined, specifically including:

[0017] Obtain the ratio of the sample lesion area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant units; and based on the ratio of the sample lesion area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant units, respectively determine the average value of the ratio of the sample lesion area corresponding to each severity level to the total area of ​​the sample plant units, and use the average value as the actual ratio corresponding to each severity level.

[0018] Further, based on the average value of the ratio of the sample diseased spot area corresponding to each severity level to the total area of ​​the sample plant unit, the actual diseased spot area ratio range corresponding to each severity level is determined, specifically including:

[0019] The middle value of the average value of the ratio of the diseased spot area of ​​the sample plant units corresponding to two adjacent severity levels in each severity level to the total area of ​​the sample plant units is taken as the dividing point; the dividing point is respectively taken as the upper limit of the actual ratio corresponding to the first severity level in the two adjacent severity levels, and as the lower limit of the actual ratio corresponding to the second severity level in the two adjacent severity levels, to determine the range of the actual diseased spot area ratio corresponding to each severity level; the first severity level is lower than the second severity level.

[0020] Furthermore, the plant disease severity assessment method further includes:

[0021] Determining the number of target diseased spot pixels in the actual plant unit and the number of total pixels in the actual plant unit;

[0022] The current ratio is obtained based on the number of target diseased spot pixels in the actual plant unit and the number of total pixels in the actual plant unit.

[0023] Furthermore, the plant disease severity assessment method further includes:

[0024] Determining the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit;

[0025] The ratio is obtained based on the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit.

[0026] The present invention also provides a plant disease severity assessment device, comprising:

[0027] An actual plant unit image acquisition module is used to acquire an actual plant unit image to be evaluated and extract an actual plant unit from the actual plant unit image;

[0028] a current ratio determination unit, used to determine a current ratio of the target diseased spot area in the actual plant unit to the total area of ​​the actual plant unit;

[0029] A severity assessment unit is used to perform severity assessment analysis based on the current ratio and the actual diseased area ratio range corresponding to the preset severity level to determine the severity level of the actual plant unit; wherein the actual diseased area ratio range is obtained by performing an average value analysis based on the severity levels and their corresponding sample plant unit images.

[0030] Furthermore, before performing severity assessment analysis based on the current ratio and the actual lesion area ratio range corresponding to the preset severity level, the method further includes:

[0031] A sample plant unit image acquisition unit, used to acquire a sample plant image corresponding to the severity level;

[0032] A sample image segmentation unit is used to segment a sample plant unit region in the sample plant image to obtain a sample plant unit image; and segment a sample disease spot region in the sample plant unit image to obtain a sample disease spot image;

[0033] A sample area determination unit, used to determine the total area of ​​the sample plant unit in the sample plant unit image and the sample disease spot area in the sample disease spot image;

[0034] A ratio determination unit, used to determine the ratio of the sample diseased spot area to the total area of ​​the sample plant unit;

[0035] an average value determination unit, for determining the average value of the ratio of the sample diseased spot area to the total area of ​​the sample plant unit corresponding to each severity level based on the ratio of the sample diseased spot area to the total area of ​​the sample plant unit;

[0036] The actual disease spot area ratio range determination unit is used to determine the actual disease spot area ratio range corresponding to each severity level based on the average value of the ratio of the sample disease spot area corresponding to each severity level to the total area of ​​the sample plant unit.

[0037] Furthermore, the average value determination unit is specifically used to:

[0038] Obtain the ratio of the sample lesion area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant units; and based on the ratio of the sample lesion area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant units, respectively determine the average value of the ratio of the sample lesion area corresponding to each severity level to the total area of ​​the sample plant units, and use the average value as the actual ratio corresponding to each severity level.

[0039] Furthermore, the actual lesion area ratio range determination unit is specifically used to:

[0040] The middle value of the average value of the ratio of the diseased spot area of ​​the sample plant units corresponding to two adjacent severity levels in each severity level to the total area of ​​the sample plant units is taken as the dividing point; the dividing point is respectively taken as the upper limit of the actual ratio corresponding to the first severity level in the two adjacent severity levels, and as the lower limit of the actual ratio corresponding to the second severity level in the two adjacent severity levels, to determine the range of the actual diseased spot area ratio corresponding to each severity level; the first severity level is lower than the second severity level.

[0041] Furthermore, the plant disease severity assessment device further comprises:

[0042] A first pixel number determination unit, used to determine the number of target diseased spot pixels in the actual plant unit and the number of total pixels in the actual plant unit;

[0043] The first current ratio determination unit is used to obtain the current ratio based on the number of target diseased spot pixels in the actual plant unit and the number of total pixels in the actual plant unit.

[0044] Furthermore, the plant disease severity assessment device further comprises:

[0045] A second pixel number determination unit is used to determine the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit;

[0046] The second current ratio determination unit is used to obtain the ratio based on the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit.

[0047] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the plant disease severity assessment method as described in any one of the above items is implemented.

[0048] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the plant disease severity assessment method as described in any one of the above items is implemented.

[0049] The plant disease severity assessment method of the present invention obtains the actual plant unit image to be assessed, extracts the actual plant unit from the actual plant unit image, and then calculates the current ratio of the target disease spot area in the actual plant unit to the total area of ​​the actual plant unit, and then performs severity assessment analysis based on the current ratio and the actual disease spot area ratio range corresponding to the severity level to determine the severity level of the actual plant unit. The method provided by the present invention utilizes and embodies the actual ratio of the disease spot area to the total area of ​​the plant unit, and the actual operation is simpler and faster. The plant disease severity is assessed by pre-analyzing the actual disease spot area ratio range obtained by the average value based on the severity level and the sample plant unit images corresponding to each of them, thereby improving the accuracy of plant disease severity assessment, thereby providing more reliable support for plant disease prediction and forecasting, variety disease resistance identification and variety breeding, and disease prevention and control strategy formulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 A schematic diagram of a process for evaluating plant disease severity provided by an embodiment of the present invention;

[0052] Figure 2 A complete flow chart of the plant disease severity assessment method provided by an embodiment of the present invention;

[0053] Figure 3 A schematic diagram of the structure of a plant disease severity assessment device provided by an embodiment of the present invention;

[0054] Figure 4 A schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] The following is a detailed description of an embodiment of the plant disease severity assessment method according to the present invention. Figure 1 and 2 As shown in the figure, they are respectively a flow chart and a complete flow chart of a method for evaluating the severity of plant diseases provided by an embodiment of the present invention, and the specific process includes the following steps:

[0057] Step 101: Acquire an actual plant unit image to be evaluated, and extract actual plant units from the actual plant unit image.

[0058] Specifically, the actual plant unit image may refer to an actual plant unit image of a certain plant disease acquired by an image acquisition device such as a digital camera, a mobile phone, or a camera with a camera function, that is, an image containing an actual plant unit to be evaluated. The actual plant unit may refer to a stem, a leaf, a root, a fruit, or a plant. In an embodiment of the present invention, the actual plant unit may be manually or automatically segmented from the background by a preset image processing software to obtain an actual plant unit segmentation image, that is, to extract the actual plant unit.

[0059] Step 102: Determine a current ratio of the target diseased spot area in the actual plant unit to the total area of ​​the actual plant unit.

[0060] Specifically, the target diseased spot refers to the diseased diseased spot to be evaluated. In the embodiment of the present invention, it is necessary to extract the area characteristics of the actual plant unit and the target diseased spot in advance through a preset image processing software, so as to obtain the total area of ​​the actual plant unit and the target diseased spot area, and then calculate the current ratio of the target diseased spot area in the actual plant unit to the total area of ​​the actual plant unit based on the total area of ​​the actual plant unit and the target diseased spot area.

[0061] In addition, the number of target disease spot pixels in the actual plant unit and the number of total pixels in the actual plant unit can be obtained by statistical calculation using the image processing software, and the ratio can be calculated based on the number of target disease spot pixels in the actual plant unit and the number of total pixels in the actual plant unit to obtain the current ratio.

[0062] Step 103: Perform severity assessment analysis based on the current ratio and the actual diseased area ratio range corresponding to the preset severity level to determine the severity level of the actual plant unit; wherein the actual diseased area ratio range is obtained by performing average value analysis based on the severity levels and their corresponding sample plant unit images.

[0063] In an embodiment of the present invention, before executing a severity assessment analysis based on the current ratio and the actual diseased spot area ratio range corresponding to the preset severity level, it is necessary to perform an average value analysis based on each severity level and its corresponding sample plant unit image in advance to obtain the severity level of the actual plant unit required by the application. The specific process includes: obtaining a sample plant image corresponding to the severity level; segmenting the sample plant unit area in the sample plant image to obtain a sample plant unit image; and segmenting the sample diseased spot area in the sample plant unit image to obtain a sample diseased spot image; determining the total area of ​​the sample plant unit in the sample plant unit image and the sample diseased spot area in the sample diseased spot image; determining the ratio of the sample diseased spot area to the total area of ​​the sample plant unit; based on the ratio of the sample diseased spot area to the total area of ​​the sample plant unit, determining the average value of the ratio of the sample diseased spot area to the total area of ​​the sample plant unit corresponding to each severity level; based on the average value of the ratio of the sample diseased spot area to the total area of ​​the sample plant unit corresponding to each severity level, determining the actual diseased spot area ratio range corresponding to each severity level.

[0064] Among them, the specific implementation process of obtaining the sample plant images corresponding to the severity levels includes: using a digital camera, a mobile phone or a camera with a camera function and other devices to obtain sample plant unit images of sample plant diseases corresponding to each severity level.

[0065] The sample plant unit region in the sample plant image is segmented to obtain a sample plant unit image; and the sample diseased spot region in the sample plant unit image is segmented to obtain a sample diseased spot image. The corresponding specific implementation process includes:

[0066] Based on image processing software, the sample plant unit is manually or automatically segmented from the background to obtain a sample plant unit segmentation image, i.e., a sample plant unit image, and the sample lesion is segmented from the sample plant unit to obtain a sample lesion segmentation image, i.e., a sample lesion image.

[0067] Determine the total area of ​​the sample plant unit in the sample plant unit image and the sample lesion area in the sample lesion image. The corresponding specific implementation process includes: extracting the area characteristics of the sample plant unit and the sample lesion based on image processing software to obtain the total area of ​​the sample plant unit and the sample lesion area.

[0068] Determine the ratio of the sample diseased spot area to the total area of ​​the sample plant unit, and the corresponding specific implementation process includes: according to the obtained total area of ​​the sample plant unit and the sample diseased spot area, divide the sample diseased spot area by the total area of ​​the sample plant unit to obtain the ratio of the sample diseased spot area to the total area of ​​the sample plant unit. In addition, in the process of determining the ratio, the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit obtained by statistical calculation using image processing software, and based on the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit, obtain the ratio. In the technical solution provided by the present invention, in addition to obtaining it by using image processing technology, the ratio of the diseased spot area to the total area of ​​the plant unit can also be estimated by methods such as naked eye observation, grid paper measurement, and paper sample weighing, which are not specifically limited here.

[0069] Based on the ratio of the sample lesion area to the total area of ​​the sample plant unit, the average value of the ratio of the sample lesion area to the total area of ​​the sample plant unit corresponding to each severity level is determined respectively. The corresponding implementation process includes: obtaining the ratio of the sample lesion area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant unit; and based on the ratio of the sample lesion area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant unit, the average value of the ratio of the sample lesion area corresponding to each severity level to the total area of ​​the sample plant unit is determined respectively, and the average value is used as the actual ratio corresponding to each severity level.

[0070] Specifically, in the process of calculating the average value of the ratio of the sample lesion area of ​​the sample plant units corresponding to each severity level to the total area of ​​the sample plant units, based on the obtained ratio of the sample lesion area of ​​the sample plant units corresponding to each severity level to the total area of ​​the sample plant units, the average value of the ratio of the sample lesion area of ​​the sample plant units corresponding to each level of severity to the total area of ​​the sample plant units is calculated respectively, so as to calculate the average value corresponding to each severity level as the actual ratio of the sample lesion area of ​​the corresponding severity level to the total area of ​​the sample plant units.

[0071] In addition, based on the average value of the ratio of the sample lesion area corresponding to each severity level to the total area of ​​the sample plant unit, the actual lesion area ratio range corresponding to each severity level is determined, and the corresponding implementation process includes: taking the middle value of the average value of the ratio of the lesion area of ​​the sample plant unit corresponding to two adjacent severity levels in each severity level to the total area of ​​the sample plant unit as the demarcation point; taking the demarcation point as the actual ratio upper limit corresponding to the first severity level of the two adjacent severity levels, and as the actual ratio lower limit corresponding to the second severity level of the two adjacent severity levels, respectively, to determine the actual lesion area ratio range corresponding to each severity level; the first severity level is lower than the second severity level. For example, each severity level may include eight severity levels: 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100%. The two adjacent severity levels may refer to 10% and 20%, and of course, may also refer to 40% and 60%, which are not specifically limited here. Correspondingly, the first severity level may refer to the 10% level, and the second severity level may refer to the 20% level; or, the first severity level may refer to the 40% level, and the second severity level may refer to the 60% level, without specific limitation herein.

[0072] Specifically, in the process of determining the actual lesion area ratio range corresponding to each severity level, the middle value of the average value of the ratio of the sample lesion area of ​​the sample plant units corresponding to two adjacent severity levels to the total area of ​​the sample plant units is used as the dividing point. This dividing point is used as the upper limit of the actual ratio corresponding to the lower severity level (such as 10% or 40%) of the two severity levels, and as the lower limit of the actual ratio corresponding to the higher severity level (such as 20% or 60%). For the lowest level of severity (such as 1%), the corresponding actual ratio is greater than 0. The actual lesion area ratio range corresponding to each severity level is established, and the severity of the diseased plant unit in actual application can be evaluated according to this range. That is, according to the established actual lesion area ratio range of the sample lesion area corresponding to each severity level to the total area of ​​the sample plant unit, the severity level of the actual plant unit to be evaluated is evaluated.

[0073] The following is a detailed explanation of the severity assessment of wheat stripe rust as an example. Follow the steps below to complete the severity assessment of wheat stripe rust:

[0074] 1) According to the severity grading standard of wheat stripe rust, diseased leaf samples (i.e., sample plant units) with typical symptoms of wheat stripe rust with severity levels of 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100% were collected, and image acquisition equipment was used to collect images. The image resolutions were 4256×2832 pixels, 3648×2736 pixels, and 4032×3024 pixels, respectively, and the formats were all JPEG, 24-bit images. 50 diseased leaf sample images corresponding to each severity level were obtained, and a total of 400 diseased leaf sample images, i.e., sample plant images, were obtained.

[0075] 2) Use image processing software (Adobe Photoshop) to complete the segmentation and pixel counting of the leaf area and the segmentation and pixel counting of the diseased spot area, respectively, to obtain the sample plant unit image and the sample diseased spot image.

[0076] 3) Calculate the ratio of the sample lesion area to the total leaf area (i.e., the total area of ​​the sample plant unit); or use the recorded total pixels of the leaf area (i.e., the total pixels of the sample plant unit) and the sample lesion pixels to calculate the ratio of the number of sample lesion pixels to the total pixels of the leaf area for the diseased leaf image (i.e., the sample plant image) corresponding to each severity level. Finally, based on the sample lesion area and the total leaf area, the ratio of the sample lesion area of ​​the corresponding diseased leaf to the total leaf area, i.e., the percentage, is obtained. The calculation formula is as follows:

[0077]

[0078] In the formula, S p A is the percentage of sample lesion area to the total leaf area; d is the number of pixels in the sample lesion area, A l is the total number of pixels in the leaf area.

[0079] 4) The UNIVARIATE procedure in SAS 9.4 software (SAS Institute Inc. Cary, NC, USA) was used to perform a normal distribution test on the percentage of the area of ​​diseased leaves in each severity level to the total leaf area. The test results showed that the 50 actual percentage data contained in each severity level all obeyed the normal distribution. The percentage of the area of ​​diseased leaves in each severity level to the total leaf area was sorted from large to small, and the ratio of the number of samples in the training set to the test set was 4:1 and 3:2, respectively. The systematic sampling method was used to sample from the 50 data of each severity level to establish the training set and the test set, which were named Train40 and Train41, respectively. S(indicates that when the severity level is S, the ratio of the number of samples in the training set to the test set is 4:1, and the training set consists of 40 samples), Test10 S (indicates that when the severity level is S, the test set consists of 10 samples obtained by systematic sampling when the ratio of the number of samples in the training set to the test set is 4:1) and Train30 S (indicates that when the severity level is S, the training set consists of 30 samples obtained by systematic sampling when the ratio of the number of samples in the training set to the test set is 3:2), Test20 S (represents the test set consisting of 20 samples obtained by systematic sampling when the ratio of the number of samples in the training set to the test set is 3:2 when the severity level is S), where S is the percentage corresponding to the different severity levels of wheat stripe rust, and its values ​​can be 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100% respectively. The UNIVARIATE procedure in SAS 9.4 software (SAS Institute Inc. Cary, NC, USA) was used to sample the training set Train40 for each severity level. S and Train30 S The actual percentage data contained in the normal distribution test shows that the training set Train40 for each severity level S and Train30 S The actual percentage data included all follow a normal distribution.

[0080] 5) Calculate the training set Train40 composed of samples from each severity level S and Train30 S The average of the actual percentage data included That is, determine the average value of the ratio of the sample lesion area corresponding to each severity level to the total area of ​​the sample plant unit; calculate the average value corresponding to each severity level as the actual percentage of the lesion area of ​​the corresponding severity level to the total leaf area, that is, take the average value as the actual ratio corresponding to each severity level. The specific results are shown in Table 1. The results show that based on the training set Train40 S The average actual percentages of the included data for severity levels 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100% are 0.40%, 1.27%, 2.50%, 4.92%, 9.89%, 16.61%, 21.23%, and 30.52%, respectively; based on the training set Train30 SThe average actual percentages of the included data for severity levels 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100% are 0.40%, 1.28%, 2.50%, 4.92%, 9.87%, 16.61%, 21.23%, and 30.53%, respectively. S and Train30 S The average values ​​of the actual percentage of the lesion area of ​​the same severity level to the total leaf area obtained by the data contained in the training set Train40 are not much different. Therefore, in the embodiment of the present invention, the training set Train40 can be used as the training set. S or Train30 S , an average value analysis was performed on each severity level and its corresponding sample plant unit image; at the same time, the results showed that there was a large gap between the actual percentages corresponding to different severity levels of wheat stripe rust and the percentages of the corresponding severity levels in the severity level classification standard. The actual lesion area ratio range obtained by the average value analysis of the present invention was more accurate than the preset severity levels of 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100%.

[0081] 6) For the training set Train40 respectively S and Train30 S The actual percentage data contained in the data is taken as the middle value of the average percentage of the diseased leaf area to the total leaf area corresponding to the two adjacent severity levels as the dividing point. This dividing point is used as the upper limit of the actual percentage corresponding to the lower severity level of the two severity levels and as the lower limit of the actual percentage corresponding to the higher severity level. For the lowest severity level, the corresponding actual percentage is greater than 0, thereby obtaining the training set Train40 and Train41 data respectively. S and Train30 S The actual lesion area ratio range corresponding to each severity level of the included data is shown in Table 1. The results show that based on the training set Train40 S The actual lesion area ratios of the severity levels 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100% obtained in the included data ranged from (0, 0.84%], (0.84%, 1.89%], (1.89%, 3.71%], (3.71%, 7.41%], (7.41%, 13.25%], (13.25%, 18.92%], (18.92%, 25.88%], (25.88%, 100%], respectively; based on the training set Train30 SThe actual lesion area ratios of the severity levels 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100% obtained from the included data ranged from (0, 0.84%], (0.84%, 1.89%], (1.89%, 3.71%], (3.71%, 7.40%], (7.40%, 13.24%], (13.24%, 18.92%], (18.92%, 25.88%], and (25.88%, 100%], respectively. The results show that based on the training set Train40 S The actual lesion area ratio ranges of the severity levels 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100% obtained from the included data are consistent with those based on the training set Train30. S The actual lesion area ratio ranges of the severity levels 1%, 5%, 10%, 20%, 40%, 60%, 80%, and 100% obtained by the included data are generally very small, so in the embodiment of the present invention, the training set Train40 can be used as the training set. S or Train30 S , the average value analysis was performed on each severity level and its corresponding sample plant unit images.

[0082] Table 1. Based on the training set Train40 S and Train30 S The range of actual lesion area ratios for each severity level of the included data

[0083]

[0084] 7) Using the training set Train40 determined in 6) above S The actual lesion area ratio range corresponding to each severity level of the included data, based on the training set Train40 S and test set Test10 S The actual percentage of the diseased area of ​​each diseased leaf to the total leaf area is used to evaluate the severity of each diseased leaf. S The actual lesion area ratio range corresponding to each severity level of the included data is based on the training set Train30 S and test set Test20 S The actual percentage of the diseased area of ​​each diseased leaf to the total leaf area was used to evaluate the severity of each diseased leaf. The accuracy of the severity evaluation of diseased leaves in each data set was calculated using the following formula:

[0085]

[0086] Where, Accuracy represents the accuracy of disease severity assessment; n p represents the number of samples correctly evaluated; n is the total number of samples evaluated. The results of severity evaluation of diseased leaves in each data set are shown in Table 2. It can be seen from the severity evaluation results that based on the training set Train40 S or Train30 S The actual lesion area ratio range corresponding to each severity level determined by the data included is for the training set Train40 S or Train30 S The accuracy of severity assessment is ≥85% for the corresponding test set Test10 S or Test20 S The accuracy of severity assessment was also ≥85%. The results show that the disease severity assessment method provided by the present invention can achieve a higher accuracy in the severity assessment of wheat stripe rust diseased leaves, and the method is suitable for severity assessment of wheat stripe rust diseased leaves.

[0087] Table 2. Evaluation results of wheat stripe rust disease severity on leaves

[0088]

[0089]

[0090] In this embodiment, the number of diseased leaves of each severity level of wheat stripe rust obtained is 50, and the systematic sampling method is used to sample samples according to the two ratios of 4:1 and 3:2, and the training set and the test set are established respectively. Under these two sampling ratios, the actual percentage of the lesion area of ​​each severity level obtained to the total leaf area and the actual lesion area ratio range are not much different, and after the severity evaluation test of the training set and the test set, a high evaluation accuracy rate is obtained at each severity level. In comparison, the severity evaluation results obtained by establishing the training set and the test set according to the ratio of 4:1 are slightly better than those of establishing the training set and the test set according to the ratio of 3:2. If more samples of each severity level can be obtained, according to the method provided by the present invention, better severity evaluation results can be obtained.

[0091] In the prior art, plant disease severity assessment using image processing is usually based on the ratio of the diseased area to the total area of ​​the plant unit obtained by segmentation, or a recognition model is established based on the extracted image features for classification and recognition. When the severity level is divided according to the ratio of the diseased area to the total area of ​​the plant unit, for some diseases, the actual ratio of the diseased area to the total area of ​​the plant unit will be significantly lower than the corresponding ratio in the severity classification standard, which will lead to large errors or complete errors in the severity assessment.

[0092] The method provided by the present invention is applicable to all plant diseases that are divided into severity levels based on the ratio of the diseased area to the total area of ​​the plant diseased unit, and completely solves the problem of inconsistency between the ratio of the diseased area to the total area of ​​the plant unit in the severity classification level and the actual ratio of the diseased area to the total area of ​​the plant unit when the severity of plant diseases is evaluated based on the ratio of the diseased area to the total area of ​​the plant unit, and provides a basis and methodological reference for accurately evaluating the severity of plant diseases.

[0093] The plant disease severity assessment method provided by the present invention is simple and easy to operate. The severity level division basis used reflects the actual ratio of the diseased area to the total area of ​​the plant unit, which is more conducive to application in practice and can improve the accuracy of severity assessment, which is of great significance for the investigation, monitoring, prediction, forecasting and prevention and control of plant diseases. The method provided by the present invention is applicable to all plant diseases that are divided into severity levels based on the ratio of the diseased area to the total area of ​​the plant diseased unit, and fundamentally solves the problems of difficulty in dividing the severity assessment levels of plant diseases and the problem of evaluating the severity of plant diseases based on the ratio of the diseased area to the total area of ​​the plant unit.

[0094] The plant disease severity assessment method provided by the present invention is a plant disease severity assessment method based on the actual ratio of the disease spot area to the total area of ​​the plant unit. The method utilizes the actual ratio of the disease spot area to the total area of ​​the plant unit, and provides a range of the actual disease spot area ratio of the disease spot area to the total area of ​​the plant unit when evaluating the disease severity. The disease severity assessment is performed based on this range, which is more direct and intuitive, more convenient for understanding the severity of plant diseases, and convenient for practical operation. The present invention provides the most basic basis and ideas for the realization of automatic plant disease severity assessment based on image processing technology, is conducive to the development of an automatic plant disease severity assessment system and improving the accuracy of plant disease severity assessment, is conducive to the realization of automation and intelligence of plant disease severity assessment, and can provide more reliable support for disease prediction and forecasting, variety disease resistance identification and variety breeding, and formulation of disease prevention and control strategies.

[0095] Corresponding to the plant disease severity assessment method provided above, the present invention also provides a plant disease severity assessment device. Since the embodiment of the device is similar to the above method embodiment, the description is relatively simple. For the relevant parts, please refer to the description of the above method embodiment. The embodiment of the plant disease severity assessment device described below is only illustrative.

[0096] Please refer to Figure 3 As shown, it is a schematic diagram of the structure of a plant disease severity assessment device provided by an embodiment of the present invention. The plant disease severity assessment device of the present invention specifically includes the following parts:

[0097] The actual plant unit image acquisition module 301 is used to acquire the actual plant unit image to be evaluated and extract the actual plant unit from the actual plant unit image;

[0098] A current ratio determination unit 302 is used to determine a current ratio of the target diseased spot area in the actual plant unit to the total area of ​​the actual plant unit;

[0099] The severity assessment unit 303 is used to perform severity assessment analysis based on the current ratio and the actual diseased area ratio range corresponding to the preset severity level to determine the severity level of the actual plant unit; wherein the actual diseased area ratio range is obtained by performing an average value analysis based on the severity levels and their corresponding sample plant unit images.

[0100] Furthermore, before performing severity assessment analysis based on the current ratio and the actual lesion area ratio range corresponding to the preset severity level, the method further includes:

[0101] A sample plant unit image acquisition unit, used to acquire a sample plant image corresponding to the severity level;

[0102] A sample image segmentation unit is used to segment a sample plant unit region in the sample plant image to obtain a sample plant unit image; and segment a sample disease spot region in the sample plant unit image to obtain a sample disease spot image;

[0103] A sample area determination unit, used to determine the total area of ​​the sample plant unit in the sample plant unit image and the sample disease spot area in the sample disease spot image;

[0104] A ratio determination unit, used to determine the ratio of the sample diseased spot area to the total area of ​​the sample plant unit;

[0105] an average value determination unit, for determining the average value of the ratio of the sample diseased spot area to the total area of ​​the sample plant unit corresponding to each severity level based on the ratio of the sample diseased spot area to the total area of ​​the sample plant unit;

[0106] The actual disease spot area ratio range determination unit is used to determine the actual disease spot area ratio range corresponding to each severity level based on the average value of the ratio of the sample disease spot area corresponding to each severity level to the total area of ​​the sample plant unit.

[0107] Furthermore, the average value determination unit is specifically used to:

[0108] Obtain the ratio of the sample lesion area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant units; and based on the ratio of the sample lesion area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant units, respectively determine the average value of the ratio of the sample lesion area corresponding to each severity level to the total area of ​​the sample plant units, and use the average value as the actual ratio corresponding to each severity level.

[0109] Furthermore, the actual lesion area ratio range determination unit is specifically used to:

[0110] The middle value of the average value of the ratio of the diseased spot area of ​​the sample plant units corresponding to two adjacent severity levels in each severity level to the total area of ​​the sample plant units is taken as the dividing point; the dividing point is respectively taken as the upper limit of the actual ratio corresponding to the first severity level in the two adjacent severity levels, and as the lower limit of the actual ratio corresponding to the second severity level in the two adjacent severity levels, to determine the range of the actual diseased spot area ratio corresponding to each severity level; the first severity level is lower than the second severity level.

[0111] Furthermore, the plant disease severity assessment device further comprises:

[0112] A first pixel number determination unit, used to determine the number of target diseased spot pixels in the actual plant unit and the number of total pixels in the actual plant unit;

[0113] The first current ratio determination unit is used to obtain the current ratio based on the number of target diseased spot pixels in the actual plant unit and the number of total pixels in the actual plant unit.

[0114] Furthermore, the plant disease severity assessment device further comprises:

[0115] A second pixel number determination unit is used to determine the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit;

[0116] The second current ratio determination unit is used to obtain the ratio based on the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit.

[0117] The plant disease severity assessment device of the present invention obtains the actual plant unit image to be assessed, extracts the actual plant unit from the actual plant unit image, and then calculates the current ratio of the target disease spot area in the actual plant unit to the total area of ​​the actual plant unit, and then performs severity assessment analysis based on the current ratio and the actual disease spot area ratio range corresponding to the severity level to determine the severity level of the actual plant unit. The device provided by the present invention utilizes and embodies the actual ratio of the disease spot area to the total area of ​​the plant unit, and the actual operation is simpler and faster. The plant disease severity is assessed by pre-analyzing the actual disease spot area ratio range obtained by the average value based on the severity level and the sample plant unit images corresponding to each of them, thereby improving the accuracy of plant disease severity assessment, thereby providing more reliable support for plant disease prediction and forecasting, variety disease resistance identification and variety breeding, and disease prevention and control strategy formulation.

[0118] Corresponding to the plant disease severity assessment method provided above, the present invention also provides an electronic device. Since the embodiment of the electronic device is similar to the above method embodiment, the description is relatively simple. For relevant parts, please refer to the description of the above method embodiment. The electronic device described below is only exemplary. Figure 4 As shown, it is a schematic diagram of the physical structure of an electronic device disclosed in an embodiment of the present invention. The electronic device may include: a processor 401, a memory 402 and a communication bus 403, wherein the processor 401 and the memory 402 communicate with each other through the communication bus 403. The processor 401 can call the logic instructions in the memory 402 to execute a plant disease severity assessment method, the method comprising: obtaining an actual plant unit image to be assessed, extracting an actual plant unit from the actual plant unit image; determining the current ratio of the target disease spot area in the actual plant unit to the total area of ​​the actual plant unit; performing severity assessment analysis based on the current ratio and the actual disease spot area ratio range corresponding to the preset severity level, and determining the severity level of the actual plant unit; wherein the actual disease spot area ratio range is obtained by performing average value analysis based on the severity level and the sample plant unit images corresponding to each of them.

[0119] In addition, the logic instructions in the above-mentioned memory 402 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk.

[0120] On the other hand, an embodiment of the present invention further provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the plant disease severity assessment method provided by the above-mentioned method embodiments, the method including: obtaining an actual plant unit image to be evaluated, and extracting the actual plant unit from the actual plant unit image; determining the current ratio of the target lesion area in the actual plant unit to the total area of ​​the actual plant unit; performing a severity assessment analysis based on the current ratio and the actual lesion area ratio range corresponding to the preset severity level, and determining the severity level of the actual plant unit; wherein the actual lesion area ratio range is obtained by performing an average value analysis based on the severity levels and their corresponding sample plant unit images.

[0121] On the other hand, an embodiment of the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the plant disease severity assessment method provided in the above-mentioned embodiments, the method comprising: obtaining an actual plant unit image to be evaluated, and extracting the actual plant unit from the actual plant unit image; determining the current ratio of the target disease spot area in the actual plant unit to the total area of ​​the actual plant unit; performing a severity assessment analysis based on the current ratio and the actual disease spot area ratio range corresponding to the preset severity level, and determining the severity level of the actual plant unit; wherein the actual disease spot area ratio range is obtained by performing an average value analysis based on the severity levels and their corresponding sample plant unit images.

[0122] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0123] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0124] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for assessing the severity of plant diseases, characterized in that: include: Acquire an actual plant unit image to be evaluated, and extract actual plant units from the actual plant unit image; determining a current ratio of the target diseased spot area in the actual plant unit to the total area of ​​the actual plant unit; Performing severity assessment analysis based on the current ratio and the actual diseased area ratio range corresponding to the preset severity level to determine the severity level of the actual plant unit; wherein the actual diseased area ratio range is obtained by performing average value analysis based on the severity levels and their corresponding sample plant unit images; Before performing severity assessment analysis based on the current ratio and the actual diseased spot area ratio range corresponding to the preset severity level, the method further includes: obtaining a sample plant image corresponding to the severity level; performing segmentation processing on the sample plant unit area in the sample plant image to obtain a sample plant unit image; and performing segmentation processing on the sample diseased spot area in the sample plant unit image to obtain a sample diseased spot image; determining the total area of ​​the sample plant unit in the sample plant unit image and the sample diseased spot area in the sample diseased spot image; and determining the ratio of the sample diseased spot area to the total area of ​​the sample plant unit; Obtaining the ratio of the sample diseased spot area of ​​all sample plant units corresponding to each severity level to the total area of ​​the sample plant unit; and based on the ratio of the sample diseased spot area of ​​all sample plant units corresponding to each severity level to the total area of ​​the sample plant unit, respectively determining the average value of the ratio of the sample diseased spot area corresponding to each severity level to the total area of ​​the sample plant unit, and using the average value as the actual ratio corresponding to each severity level; The middle value of the average value of the ratio of the diseased spot area of ​​the sample plant units corresponding to two adjacent severity levels in each severity level to the total area of ​​the sample plant units is taken as the dividing point; the dividing point is respectively taken as the upper limit of the actual ratio corresponding to the first severity level in the two adjacent severity levels, and as the lower limit of the actual ratio corresponding to the second severity level in the two adjacent severity levels, to determine the range of the actual diseased spot area ratio corresponding to each severity level; the first severity level is lower than the second severity level.

2. The plant disease severity assessment method according to claim 1, characterized in that: Also includes: Determining the number of target diseased spot pixels in the actual plant unit and the number of total pixels in the actual plant unit; The current ratio is obtained based on the number of target diseased spot pixels in the actual plant unit and the number of total pixels in the actual plant unit.

3. The plant disease severity assessment method according to claim 1, characterized in that: Also includes: Determining the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit; The ratio is obtained based on the number of sample diseased spot pixels in the sample plant unit and the number of total pixels in the sample plant unit.

4. A plant disease severity assessment device, characterized in that: include: An actual plant unit image acquisition module is used to acquire an actual plant unit image to be evaluated and extract an actual plant unit from the actual plant unit image; a current ratio determination unit, used to determine a current ratio of the target diseased spot area in the actual plant unit to the total area of ​​the actual plant unit; A severity assessment unit, configured to perform severity assessment analysis based on the current ratio and a preset range of actual diseased spot area ratios corresponding to the severity levels, and determine the severity level of the actual plant unit; wherein the actual diseased spot area ratio range is obtained by performing an average value analysis based on the severity levels and their corresponding sample plant unit images; Before performing severity assessment analysis based on the current ratio and the actual lesion area ratio range corresponding to the preset severity level, the method further includes: a sample plant unit image acquisition unit, configured to acquire a sample plant image corresponding to the severity level; An image segmentation unit, used to segment a sample plant unit region in the sample plant image to obtain a sample plant unit image; and segment a sample disease spot region in the sample plant unit image to obtain a sample disease spot image; An area determination unit, used to determine the total area of ​​the sample plant unit in the sample plant unit image and the area of ​​the sample diseased spot in the sample diseased spot image; A ratio determination unit, used to determine the ratio of the sample diseased spot area to the total area of ​​the sample plant unit; The average value determination unit is used to: obtain the ratio of the sample diseased spot area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant unit; and based on the ratio of the sample diseased spot area of ​​all the sample plant units corresponding to each severity level to the total area of ​​the sample plant unit, respectively determine the average value of the ratio of the sample diseased spot area corresponding to each severity level to the total area of ​​the sample plant unit, and use the average value as the actual ratio corresponding to each severity level; The actual lesion area ratio range determination unit is used to take the middle value of the average value of the ratio of the lesion area of ​​the sample plant units corresponding to two adjacent severity levels in each severity level to the total area of ​​the sample plant units as the dividing point; and take the dividing point as the upper limit of the actual ratio corresponding to the first severity level in the two adjacent severity levels, and as the lower limit of the actual ratio corresponding to the second severity level in the two adjacent severity levels, to determine the actual lesion area ratio range corresponding to each severity level; the first severity level is lower than the second severity level.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the plant disease severity assessment method according to any one of claims 1 to 3 are implemented.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the plant disease severity assessment method according to any one of claims 1 to 3 are implemented.

Images