A method for detecting and locating corn root rot based on UAV-based hyperspectral data

Abstract

The method for detecting and locating root rot based on the collection of corn hyperspectral by unmanned aerial vehicles belongs to the technical field of early prediction of agricultural diseases, and can timely and efficiently monitor the corn root rot that may occur in the whole process of corn production and management. The present invention is based on the difference in chlorophyll changes in the leaves of corn plants at different periods and different degrees of root rot, and establishes a root rot identification model. Through the root rot identification model, a correct diagnosis can be made about 14 days before the diseased corn plants have no apparent symptoms, and the diseased area can be located in the farmland at the same time, providing decision-making for the early prevention and control of corn root rot and timely treatment, thereby increasing corn yield. The present invention has the characteristics of accuracy and speed, can improve the level of agricultural production, reduce the cost of agricultural production management, and increase the yield, quality and benefits of agricultural products.

Description

Technical Field

[0001] The present invention belongs to the technical field of early prediction of agricultural diseases, and in particular relates to a method for detecting and locating corn root rot based on collecting corn hyperspectral images by unmanned aerial vehicles. Background Art

[0002] Corn is one of the main food crops in my country, and its planting area and total output are second only to wheat and rice. Corn is widely used. In addition to being eaten and used to prepare high-quality livestock feed, it is also one of the important raw materials for light industry and pharmaceutical industry. Therefore, corn plays an important role in the development of the national economy. However, due to climate change, changes in cultivation systems and changes in varieties, frequent corn diseases are one of the important factors restricting the development of corn. Therefore, preventing corn diseases has become a key link to ensure the current sustainable increase in corn production. Corn root rot is one of the serious fungal diseases in the seedling stage of corn. It is generally infected in the 2-leaf stage of corn, and the roots turn brown in the 4-leaf stage. Symptoms appear on the leaves of corn in the 8-leaf stage, gradually turning yellow and withering from bottom to top. Due to the strong concealment of corn root rot in the early stage of onset and the late appearance of surface symptoms, the incidence of this disease in some parts of the country is as high as 80%, which has a great impact on corn yield.

[0003] At present, traditional pest and disease monitoring mainly uses manual field surveys to diagnose through the morphology and symptoms of the occurrence and development of diseases. It relies on human sensory judgment, which is not only inefficient, inaccurate and difficult, but also requires strong professional knowledge or experience of the inspectors, and is difficult to promote on a large scale. There are also field sampling and chemical analysis for diagnosis, but this detection method has high requirements on the accuracy of the test samples and the operating skills of the inspectors, and is costly, time-consuming, and causes more damage to the samples, and is also prone to environmental pollution. Summary of the invention

[0004] In order to solve the problems existing in the prior art, the present invention provides a method for detecting and locating corn root rot based on collecting corn hyperspectral images by unmanned aerial vehicles, which can timely and efficiently monitor corn root rot that may occur in the entire process of corn production management.

[0005] The technical solution adopted by the present invention to solve the technical problem is as follows:

[0006] A method for detecting and locating corn root rot based on collecting corn hyperspectral images from drones includes the following steps:

[0007] Step 1: Induce corn root rot at the 2-leaf stage, and collect the high-spectral reflectance of the diseased corn roots and healthy corn plants;

[0008] Step 2: Performing first-order differential calculation on the hyperspectral reflectance of the diseased corn plants and the healthy corn plants described in step 1, and selecting bands with significant differences under different disease severity as characteristic bands;

[0009] Step 3: Calculate the normalized values ​​of hyperspectral reflectance of healthy corn plants and diseased corn roots according to the characteristic bands described in step 2, and establish a root rot identification model according to the normalized values ​​of hyperspectral reflectance and the characteristic bands;

[0010] Step 4: Use drones to collect the hyperspectral reflectance of corn at different leaf stages in the test area;

[0011] Step 5: Predict the corn disease condition in the test area based on the corn hyperspectral reflectance described in step 4 and the root rot identification model described in step 3, draw a diseased area map, and determine the extent of the disease through manual re-inspection.

[0012] Preferably, the inducing of corn root rot at the 2-leaf stage described in step 1 is achieved by artificial inoculation of pathogens.

[0013] Preferably, the diseased corn roots are mildly diseased at the 4-leaf stage and moderately diseased at the 6-leaf stage.

[0014] Preferably, in step one, the high spectral reflectance of the healthy corn plants and the diseased corn roots is collected by a portable spectrometer.

[0015] Preferably, the wavelength of the characteristic spectrum band is 550nm to 740nm as the characteristic band.

[0016] Preferably, the root rot identification model includes mildly affected, moderately affected, and severely affected areas.

[0017] Preferably, the specific steps of step 4 are: using a drone equipped with a hyperspectral camera to take orthophoto images of the area to be measured and collect data.

[0018] Preferably, the specific steps of step 4 also include: setting a route for the UAV before taking off, and placing a reflective panel for radiation calibration on the ground.

[0019] The beneficial effects of the present invention are as follows: based on the difference in chlorophyll changes in leaves of corn plants at different periods and different degrees of root rot, the present invention establishes a root rot identification model, through which a correct diagnosis can be made about 14 days before the diseased corn plants have no apparent symptoms, and the diseased area can be located in the farmland at the same time, providing decision-making for early prevention and control of corn root rot and timely treatment, thereby increasing corn yield. The present invention has the characteristics of accuracy and rapidity, can improve the level of agricultural production, reduce agricultural production management costs, and increase the yield, quality and benefits of agricultural products. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1The present invention is a flow chart of a method for detecting and locating corn root rot based on collecting corn hyperspectral images by drone.

[0021] Figure 2 This is a relationship diagram between the hyperspectral wavelength and the hyperspectral reflectivity of corn root rot in the method for detecting and locating corn root rot based on collecting corn hyperspectral data by drone.

[0022] Figure 3 This is a relationship diagram between the wavelength of the hyperspectral spectrum of corn root rot and the first-order differential of the corn hyperspectral spectrum in the method for detecting and locating corn root rot based on collecting corn hyperspectral spectrum by unmanned aerial vehicle of the present invention.

[0023] Figure 4 The invention discloses a corn root rot identification model based on a method for detecting and locating corn root rot using an unmanned aerial vehicle (UAV) to collect corn hyperspectral data.

[0024] Figure 5 This is the diseased location of corn in the experimental field research area in the method for detecting and locating root rot based on collecting corn hyperspectral images by drones in the present invention. DETAILED DESCRIPTION

[0025] The present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0026] like Figure 1 As shown, the method for detecting and locating corn root rot based on collecting corn root hyperspectral data from drones includes the following steps:

[0027] Step 1: Induce corn root rot at the 2-leaf stage by artificial inoculation of pathogens, and collect the high-spectral reflectance of diseased corn and healthy corn plants at the 4-leaf and 6-leaf stages using a portable spectrometer. Figure 2 shown.

[0028] Step 2: Perform first-order differential calculation on the hyperspectral reflectance of diseased corn plants and healthy corn plants.

[0029]

[0030] Where: p'(i) is the first-order differential at band i, p(i+1) is the reflectivity of the next sampling band, p(i-1) is the reflectivity of the previous sampling band, Δp is the wavelength sampling interval, and the result is as follows: Figure 3 As shown, the 550nm-740nm band with significant differences in mild, moderate and severe levels was selected as the characteristic band.

[0031] Step 3: Calculate the normalized value of hyperspectral reflectance.

[0032]

[0033] Wherein: z(i) is the normalized value of the hyperspectral reflectance at band i, p(i) is the hyperspectral reflectance at band i, and m(i) is the envelope value at band i; a root rot identification model is established based on the normalized value of the hyperspectral reflectance and the characteristic band described in step 2, and the normalized value curves of the hyperspectral reflectance of corn with mild disease at the 4-leaf stage, corn with moderate disease at the 6-leaf stage, and healthy corn plants are drawn, and the areas with mild disease (I), moderate disease (II), and severe disease (III) are determined in the characteristic band, such as Figure 4 shown.

[0034] Step 4: Use a hyperspectral drone to collect the hyperspectral reflectivity of corn in the test area. Use a drone equipped with a hyperspectral camera to take orthophotos of the test area and collect data. Set the route and flight parameters before the drone takes off, and place a calibration reflective panel on the ground for radiation calibration.

[0035] Step 5: Predict the corn disease condition in the test area based on the root rot identification model, determine the location of the disease, compare the spectral data of the test area collected by the drone with the root rot identification model, determine whether the corn is diseased, draw a map of the diseased area, and determine the extent of the disease through manual re-inspection.

[0036] Example 1: The experimental field of Jilin Agricultural Science and Technology College in Jilin Province was selected as the observation object. The hyperspectral data of corn at the 4-leaf stage was collected by drone. After analysis and calculation, it was compared with the corn root rot identification model. There were 2 areas where the hyperspectral reflectance curves of corn were in the mild disease area. The disease occurrence was determined by manual re-inspection. As shown in Table 1, the above 2 areas were determined to be "disease" through manual re-inspection, which was consistent with the spectral detection results. The diseased areas were as follows: Figure 5 shown.

[0037] Table 1 Disease situation in different regions

[0038]

[0039] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the attached claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved.

[0040] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A method for detecting and locating corn root rot based on collecting corn hyperspectral images using drones, characterized in that: The method comprises the following steps: Step 1: Induce corn root rot at the 2-leaf stage, and collect the high-spectral reflectance of diseased corn plants and healthy corn plants using a portable spectrometer. The diseased corn plants are mildly diseased at the 4-leaf stage and moderately diseased at the 6-leaf stage; Step 2: performing first-order differential calculation on the hyperspectral reflectance of the diseased corn plants and the healthy corn plants described in step 1, and selecting bands with significant differences under different disease severity as characteristic bands, wherein the wavelength of the characteristic band is 550nm-740nm; Step 3: According to the characteristic bands described in step 2, the normalized values ​​of the hyperspectral reflectance of healthy corn plants and diseased corn plants are calculated, and a root rot identification model is established according to the normalized values ​​of the hyperspectral reflectance and the characteristic bands, and the normalized value curves of the hyperspectral reflectance of mildly diseased corn at the 4-leaf stage, moderately diseased corn at the 6-leaf stage, and healthy corn plants are drawn, and the mildly diseased, moderately diseased, and severely diseased areas are determined in the characteristic bands; Step 4: using a drone to collect the hyperspectral reflectance of corn at different leaf stages in the area to be tested; the specific steps of step 4 are: using a drone equipped with a hyperspectral camera to take orthophoto images of the area to be tested and collect data, the drone sets a route before taking off, and places a reflective panel for radiation calibration on the ground in the area to be tested; Step 5: Predict the corn disease condition in the test area based on the corn hyperspectral reflectance described in step 4 and the root rot identification model described in step 3, draw a diseased area map, and determine the extent of the disease through manual re-inspection.

2. The method for detecting and locating corn root rot based on collecting corn hyperspectral images by drone according to claim 1 is characterized in that: The induction of corn root rot at the 2-leaf stage described in step 1 is achieved by artificially inoculating pathogens.

3. The method for detecting and locating corn root rot based on collecting corn hyperspectral images by drone according to claim 1, characterized in that: The root rot identification model includes mildly diseased, moderately diseased, and severely diseased areas.

Images