A method for representing soil composition using images

Abstract

The present invention discloses a method for representing soil components by using an image, comprising: obtaining geographic coordinate information of a target plot; establishing a plot image corresponding to a hyperspectral image; reorganizing the plot image into the hyperspectral image to form a temporary hyperspectral image on the hyperspectral image; correcting the temporary hyperspectral image by grouping according to plot information and acquisition point detection results; and reorganizing the corrected temporary hyperspectral image according to the plot information to form a corrected hyperspectral image. The present invention can finally form a corrected hyperspectral image by reorganizing the plot image into the hyperspectral image to form a temporary hyperspectral image on the hyperspectral image, and correcting the temporary hyperspectral image by grouping according to plot information and acquisition point detection results, thereby achieving the purpose of correcting the hyperspectral image with the plot as the center.

Description

Technical Field

[0001] The present invention relates to the technical field of soil components, and in particular to a method for representing soil components by using images. Background Art

[0002] Soil health is a key production factor in agriculture. Changes in soil composition directly affect crop yields and the ecological environment, especially the water environment and air quality. Understanding soil composition is the key to managing, maintaining and improving soil fertility and agricultural productivity. Effectively detecting soil composition and understanding soil change trends are necessary means of protecting and supervising cultivated land. At present, most soil testing methods are based on collecting sample points, conducting laboratory tests through chemical methods or physicochemical methods, and then issuing test reports. Changes in soil composition can only be analyzed through long-term monitoring of fixed collection points and the formation of an analysis of soil composition change trends based on previous test reports issued by the laboratory. This method makes it difficult to compare soil change trends in different regions, and it is also difficult to compare soil change trends between plots of land planted with different crops in the same area.

[0003] At present, hyperspectral remote sensing technology uses satellites, airplanes, drones, mobile vehicle-mounted equipment, etc. to scan the soil efficiently to form hyperspectral images, so there are many cases of using images to represent soil composition and its changing trends with points. Hyperspectral remote sensing technology usually scans a large area at one time, and cannot only scan a target plot with a limited area. Hyperspectral remote sensing technology determines the characteristic spectral band and pixel value according to the soil component to be represented, and finally analyzes the content of the component in the soil by combining energy comprehensive analysis. This technology is affected by factors such as season, weather, environment, and sunshine time during remote sensing, and the collection results need to be corrected for deviations. The most important basis for correction is calibration with the "standard value" of the detected soil component. The standard value of soil composition is obtained through traditional soil fixed-point collection and laboratory testing, and can also be obtained through other rapid detection methods. It should be emphasized that the "standard value" can only be based on the test value of the soil sample collected from the detected plot, and this operation is independent of the hyperspectral remote sensing process. A hyperspectral scanning area may contain multiple plots, and each plot has different characteristics due to different crops or farming operations, and the soil composition of adjacent plots is not necessarily the same. Therefore, how to effectively correct the hyperspectral remote sensing image to correctly reflect the soil composition of a specific plot, that is, dynamically correct the image according to the region of interest (ROI), is the key to using hyperspectral reflectance images to reflect soil composition. The currently widely used Geographic Information System (Geographic Information System) combines different data such as terrain, buildings, vegetation, etc. to generate and visualize images based on geographic location information (latitude, longitude and altitude). However, the system does not provide a method for dynamically correcting images based on regions of interest.

[0004] The chemical and physical components in the soil, such as nitrogen, phosphorus, potassium, organic matter, pH value, electrical conductivity (EC), and heavy metals (mercury, lead, cadmium, chromium, arsenic, zinc, copper, nickel, etc.) can be expressed by hyperspectral (specific band) reflectance and pixel value (DN, Digital Number) value. Therefore, hyperspectral reflectance images can reflect the chemical and physical components of the soil, which can usually be obtained by satellite, airplane, or drone aerial photography or scanning. Due to the use of natural light sources, hyperspectral images are easily affected by factors such as the collection season, environment, and sunshine time. It is necessary to correct the deviation of soil sample test data from a certain number of collection points on the surface to have practical application value. Since the collection method of hyperspectral images is independent of the plot boundaries and plot information, it is difficult to provide accurate deviation correction for the hyperspectral image containing the plot based on the detection results of the collection points provided by the plot. Summary of the invention

[0005] To this end, the present invention provides a method for representing soil components using images to solve the problem that the prior art cannot perform dynamic image correction according to the region of interest.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] According to the present invention, there is provided a method for representing soil composition using an image, comprising:

[0008] Get the geographic coordinate information of the target plot;

[0009] Establish a plot image corresponding to the hyperspectral image;

[0010] Recombining the land plot image into the hyperspectral image to form a temporary hyperspectral image on the hyperspectral image;

[0011] Correcting the temporary hyperspectral image in groups according to the plot information and the detection results of the collection points;

[0012] and recombining the corrected temporary hyperspectral image according to the plot information to form a corrected hyperspectral image.

[0013] Furthermore, there are two ways to obtain the geographic coordinate information of the target plot. One is to outline the plot based on high-definition satellite images using an editable polygon tool to obtain a set of polygon vertices with longitude and latitude; the other is to obtain the geographic coordinate information of the target plot from the local natural resources management department.

[0014] Furthermore, the geographic coordinate information of the target plot is obtained from the local natural resource management department based on administrative data. The geographic reference is linked to the location coordinates through a geocoding engine and then converted into polygons. After vector data correction, the plot location information corresponding to the longitude and latitude of the map is obtained.

[0015] Furthermore, the geographic reference includes an address and a zip code, and the location coordinates include longitude and latitude.

[0016] Furthermore, the corrected hyperspectral image includes a background hyperspectral image, and the background hyperspectral image is a portion that does not require deviation correction.

[0017] Furthermore, the temporary hyperspectral image includes a background image with a pixel value of 0 and an interested spectral image, and the interested spectral image is composed of all pixel points in the interested region.

[0018] Furthermore, the corrected temporary hyperspectral image is reorganized according to the land parcel information, and the corrected hyperspectral image is formed specifically according to the type of land parcel information.

[0019] Furthermore, the same type of plot information is determined based on the similarities and differences between the crops planted and the farming operations. If the crops planted and the farming operations are the same, the plot information is of the same type.

[0020] Further, after the hyperspectral image is mapped to the image of the land parcel information, the polygon describing the shape of the land parcel is directly mapped to the hyperspectral image to form a region of interest on the hyperspectral image.

[0021] Furthermore, a background image with a pixel value of 0 is a blank image.

[0022] The present invention has the following advantages:

[0023] The present invention reorganizes the plot image into the hyperspectral image to form a temporary hyperspectral image on the hyperspectral image, and corrects the temporary hyperspectral image by grouping according to the plot information and the detection results of the acquisition points, so as to finally form a corrected hyperspectral image, thereby achieving the purpose of correcting the hyperspectral image with the plot as the center. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0025] The structures, proportions, sizes, etc. illustrated in this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with the technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantial technical significance. Any structural modification, change in proportion or adjustment of size shall still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0026] Figure 1 A schematic flow chart of a method for using a hyperspectral image to reflect soil composition according to an exemplary embodiment is shown;

[0027] Figure 2 To combine the map to show the different types of soil structures between plots and plots due to different plant species;

[0028] Figure 3 Schematic diagram of the relationship between pixel coordinate systems after imaging using two different acquisition methods;

[0029] Figure 4 It is a schematic diagram of the graphical solution of the nonlinear two-dimensional difference method;

[0030] Figure 5 It is a schematic diagram of the region of interest (plot) in the hyperspectral image;

[0031] Figure 6 A schematic diagram of a temporary image with a background (pixel) of 0 reorganized according to the land parcel category. DETAILED DESCRIPTION

[0032] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. 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.

[0033] According to an embodiment of the present invention, a method for representing soil components using an image is provided, such as Figures 1 to 6 As shown, including:

[0034] S11, obtaining geographic coordinate information of the target plot;

[0035] S12, establishing a plot image corresponding to the hyperspectral image;

[0036] S13, recombining the land plot image into the hyperspectral image to form a temporary hyperspectral image on the hyperspectral image;

[0037] S14, correcting the temporary hyperspectral image in groups according to the plot information and the detection results of the collection points;

[0038] S15. Recombining the corrected temporary hyperspectral image according to the land parcel information to form a corrected hyperspectral image.

[0039] The present invention reorganizes the plot image into the hyperspectral image to form a temporary hyperspectral image on the hyperspectral image, and corrects the temporary hyperspectral image by grouping according to the plot information and the detection results of the acquisition points, so as to finally form a corrected hyperspectral image, thereby achieving the purpose of correcting the hyperspectral image with the plot as the center.

[0040] There are two ways to obtain the geographic coordinate information of the target plot. One is to outline the plot through the editable polygon tool based on high-definition satellite images to obtain a set of polygon vertices with longitude and latitude; the other is to obtain the geographic coordinate information of the target plot from the local natural resources management department. The specific way to obtain the geographic coordinate information of the target plot from the local natural resources management department is to use administrative data as the basis, link the geographic reference with the location coordinates through the geocoding engine, and then convert it into a polygon; after vector data correction, obtain the plot location information corresponding to the longitude and latitude of the map. The geographic reference includes the address and zip code, and the location coordinates include longitude and latitude.

[0041] In some optional embodiments, the corrected hyperspectral image includes a background hyperspectral image, which is a part that does not need to be corrected for deviation. The temporary hyperspectral image includes a background image with a pixel value of 0 and an interested spectral image, which is composed of all pixels in the region of interest. The background image with a pixel value of 0 is a blank image.

[0042] The corrected temporary hyperspectral image is reorganized according to the land parcel information, and the corrected hyperspectral image is formed specifically according to the type of land parcel information.

[0043] The same type of plot information is determined based on the similarities and differences of the crops planted. If the crops planted are the same, the plot information is of the same type.

[0044] After hyperspectral image mapping is aligned with the image of the plot information, the polygon describing the shape of the plot is directly mapped to the hyperspectral image to form the region of interest on the hyperspectral image.

[0045] exist Figure 2In the figure, it is shown that the soils of the plots to be tested and between the plots are of different types due to different plants and farming operations. From the perspective of the plot, the image representing the geographic location of the plot is just an information carrier. The type of characteristics of the plot information (soil composition, crop type) can be displayed by different colored boundaries. For convenience, they are marked as black boundaries here. For example, plots ①, ③, and ⑤ belong to the same type, plots ②, ⑦, and ⑧ belong to the same type, and plots ④ and ⑥ belong to the same type.

[0046] exist Figure 3 In the image representing the geographic location of the plot, each pixel is represented by the (μ, ν) coordinate system, and each pixel in the hyperspectral image is represented by the (x, y) coordinate system. The pixel (μ, ν) of the plot image is mapped to the four adjacent pixels in the hyperspectral image. The conversion relationship from (μ, ν) to (x, y) coordinate system is:

[0047] x=f 1 (μ, v)

[0048] y=f 2 (μ, v)

[0049] The conversion relationship from (μ, ν) to (x, y) coordinate system requires not only the plane geometric position differential parameters, but also the actual terrain changes. Taking the above two factors into consideration, the nonlinear two-dimensional difference method (Bi-cubic Interpolation) is used. The biggest feature of the nonlinear two-dimensional difference method is to expand (extend) the number of adjacent pixels, use parabola regression to fit the terrain to generate a differential function, and calculate the differential weight based on the differential function. By expanding the pixel points beyond the 4 directly adjacent pixel points and giving weights, the differential operation is performed.

[0050] Figure 5 It is the region of interest (ROI) in the hyperspectral image, generally composed of polygons. Each polygon corresponds to a plot. Each plot information contains a polygon vertex array. According to this array, it is mapped to the region of interest in the hyperspectral image, and all pixels in the region (ROI) can be extracted to reconstruct a new image or array, which is convenient for image or data processing.

[0051] Figure 6 It is a temporary image with a blank background image reorganized according to the land plot category. This operation is mainly to prepare for the correction (image processing) of the temporary image. Generally, there will be multiple adjacent or non-adjacent plots with similar characteristics (crops and farming) in the hyperspectral coverage area. They are extracted and reorganized into temporary images to facilitate independent deviation correction (image processing) according to the soil detection standard value of this type of plot. This figure shows three groups of different types of plots, with a blank background image (pixel value is 0).

[0052] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. A method for representing soil composition using an image, characterized in that: include: Get the geographic coordinate information of the target plot; Establish a plot image corresponding to the hyperspectral image; The hyperspectral image is an image formed by efficiently scanning the soil using a satellite, an airplane, an unmanned aerial vehicle or a mobile vehicle-mounted device using hyperspectral remote sensing technology; Recombining the land plot image into the hyperspectral image to form a temporary hyperspectral image on the hyperspectral image; Correcting the temporary hyperspectral image in groups according to the plot information and the detection results of the collection points; and recombining the corrected temporary hyperspectral image according to the plot information to form a corrected hyperspectral image; The temporary hyperspectral image includes a background image with a pixel value of 0 and an interested spectral image, and the interested spectral image is composed of all pixel points in the interested region; after the hyperspectral image is mapped to the image of the land parcel information, the polygon describing the shape of the land parcel is directly mapped to the hyperspectral image to form the interested region on the hyperspectral image; the reorganization of the land parcel image to the hyperspectral image to form the temporary hyperspectral image on the hyperspectral image is specifically: all pixel points in the interested region are taken out, and a new image or array is reorganized, and the reorganized image or array is the temporary hyperspectral image; The corrected temporary hyperspectral image is reorganized according to the plot information, and the corrected hyperspectral image is formed specifically according to the type of plot information; the plot information having the same planting crops and farming operations is of the same type.

2. A method for representing soil composition by using an image according to claim 1, characterized in that: There are two ways to obtain the geographic coordinate information of the target plot. One is to use the editable polygon tool to outline the plot based on the high-definition satellite image to obtain a set of polygon vertices with longitude and latitude. Another method is to obtain the geographic coordinate information of the target plot from the local natural resources management department.

3. A method for representing soil composition by using an image according to claim 2, characterized in that: The geographic coordinate information of the target plot is obtained from the local natural resources management department based on administrative data. The geographic reference is linked to the location coordinates through the geocoding engine and then converted into polygons. After vector data correction, the plot location information corresponding to the longitude and latitude of the map is obtained.

4. A method for representing soil composition by using an image according to claim 3, characterized in that: The geographic reference includes an address and a postal code, and the location coordinates include longitude and latitude.

5. The method for representing soil composition by using an image according to claim 1, characterized in that: The corrected hyperspectral image includes a background hyperspectral image, and the background hyperspectral image is a portion that does not require deviation correction.

6. A method for representing soil composition by using an image according to claim 1, characterized in that: The same type of plot information is determined based on the similarities and differences between the crops planted and the farming operations. The plot information of the same type has the same crops planted and the same farming operations.

7. A method for representing soil composition by using an image according to claim 1, characterized in that: A background image with a pixel value of 0 is a blank image.

Images