A method and system for crop water assessment and irrigation effect inference

Abstract

The application discloses a crop water content evaluation and irrigation effect inference method and system, and belongs to the technical field of precision agriculture and remote sensing image processing. The method comprises the following steps: acquiring a UAV remote sensing image (RGB or multispectral) of a to-be-detected region; receiving a user instruction through a graphical user interface, selecting or customizing a vegetation index calculation formula; the system automatically analyzes image channels and calculates corresponding vegetation index layers, and visualizes display through a heat map; based on a preset threshold, statistical analysis is performed on index mean values, and crop water content is automatically divided into different grades (High / Medium / Low); and finally, an analysis report containing regional statistical data and grading results is generated. The application can quickly and intuitively evaluate crop growth and water content, and provides a scientific basis for agricultural irrigation decision-making.

Description

Technical Field

[0001] This invention relates to the fields of computer image processing and agricultural remote sensing technology, specifically to a method and system for crop moisture assessment and irrigation effect inference. Background Technology

[0002] In the fields of modern precision agriculture and smart water management, real-time monitoring of crop moisture content is a core basis for making scientific irrigation decisions. Currently, methods for obtaining crop moisture status are mainly divided into ground-based measurement methods and remote sensing monitoring methods. Existing technologies have the following limitations: (1) Ground measurement method: It relies on manual handheld instruments for single-point sampling. Although it is highly accurate, it is time-consuming and laborious, cannot reflect the spatial distribution differences of large areas of farmland, and is destructive.

[0003] (2) Traditional remote sensing software: Although existing commercial remote sensing processing software is powerful, it is usually bulky, expensive, and has a complex operation process, requiring users to have professional remote sensing knowledge.

[0004] (3) Lack of specificity and flexibility: Most lightweight drone software can only calculate a few fixed vegetation indices and does not support users to customize calculation formulas according to specific crop or environmental needs. In addition, existing software usually only outputs index images and lacks automated interpretation and grading functions. Users still need to manually judge the moisture level corresponding to the values ​​based on experience, which is prone to subjective errors.

[0005] To address the aforementioned issues, this invention proposes a lightweight evaluation system developed using MATLAB. This system not only supports automatic adaptation of multispectral and RGB images but also innovatively introduces a dynamic formula parsing engine, allowing users to input arbitrary mathematical expressions for exponential calculations. Furthermore, it incorporates an automated grading algorithm based on statistical thresholds, capable of directly outputting "high / medium / low" moisture level reports, significantly reducing the professional barrier and improving the efficiency of irrigation decision-making. Summary of the Invention

[0006] The purpose of this invention is to solve the technical problems mentioned above and to propose a crop moisture assessment method based on UAV remote sensing images, comprising the following steps: S1. Acquire UAV remote sensing images of the farmland area to be tested, supporting RGB or multispectral format input, and identify and separate image channels; S2. Provides a preset vegetation index model and a user-defined formula interface; calculates the vegetation index matrix using the red, green, blue, and near-infrared band data of the image according to the mode selected by the user. S3. Extract the effective area from the vegetation index matrix, calculate the index mean of the effective area, and divide the water content into three levels: "High", "Medium" and "Low" according to the preset irrigation effect threshold. S4. Map the calculated vegetation index matrix to a pseudo-color heat map for display, and export the structured data of assessment level and statistical mean to a spreadsheet file.

[0007] In the preferred embodiment, in step S2, the preset vegetation index model includes: (1); GNDVI = (NIR - Green) / (NIR + Green) (2); (3); Among them, NDVI is the Normalized Difference Vegetation Index, GNDVI is the Green Normalized Difference Vegetation Index, SAVI is the Soil-Regulated Vegetation Index, NIR is infrared data, Red is red band data, and Green is green band data.

[0008] In a preferred embodiment, the hierarchical evaluation logic in step S3 is specifically as follows: calculate the mean index Avg of the effective region; if Avg > 0.6, determine the moisture content as "High"; if 0.4 < Avg < 0.6, determine the moisture content as "Medium"; if Avg ≤ 0.4, determine the moisture content as "Low"; the effective region is obtained after removing the background noise region with Index ≤ 0 through masking.

[0009] In the preferred embodiment, the user-defined formula interface in step S2 supports inputting any mathematical expression with red, green, blue, and near-infrared bands as variables. The system performs a validity check on the expression using a dynamic parsing algorithm. If the check passes, algebraic operations are performed on the image matrix to generate a vegetation index matrix. If the check fails, an error message is triggered.

[0010] In the preferred embodiment, the channel identification and separation of the UAV remote sensing image in step S1 is specifically as follows: if it is a three-channel RGB format image, extract the red, green, and blue band channels; if it is a four-channel or higher multispectral format image, automatically match the first three channels as red, green, and blue bands for visualization, and extract the near-infrared band from the remaining channels for vegetation index calculation.

[0011] In the preferred embodiment, the pseudo-color heat map in step S4 uses the Jet color mapping table to map the vegetation index values ​​from high to low to blue-red color gradients. The lower the value, the more blue it is displayed, and the higher the value, the more red it is displayed. The heat map is also equipped with color legends and numerical scales to achieve a visual display of the spatial distribution of crop moisture.

[0012] A crop moisture assessment system based on UAV remote sensing images is provided. The system runs on a computer and includes an image import module, an index calculation module, an analysis and evaluation module, and a data export module. The modules work together to process UAV remote sensing images, assess crop moisture, and output results.

[0013] In a preferred embodiment, the image import module supports loading remote sensing image files in tif, jpg, and png formats, can automatically determine the image channel dimension, adapt to the band separation and data display of three-channel RGB images or four-channel and above multispectral images, and has an image preview function that can display the original remote sensing image in a graphical user interface.

[0014] In a preferred embodiment, the index calculation module incorporates NDVI, GNDVI, and SAVI standard vegetation index calculation models, and is configured with a custom formula input box and a dynamic parsing engine. The eval function parses the valid formula strings input by the user and performs algebraic operations on the image matrix to generate an index distribution map. The module also has an anomaly capture mechanism to identify and display a calculation failure prompt when there are errors in formula variables or calculation logic.

[0015] In a preferred embodiment, the analysis and evaluation module can perform masking on the index layer, generate a logical matrix with index>0 to filter non-vegetation background noise, and statistically analyze the mean, maximum, minimum, and standard deviation of the vegetation index in the effective crop area; the data export module can encapsulate the region name, index mean, moisture rating results, and index statistical characteristic values ​​into a table object and write it into an Excel spreadsheet file, while also supporting the export and saving of the pseudo-color heatmap as an image.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The dynamic vegetation index analysis mechanism improves the adaptability of the water assessment model to different crops and remote sensing conditions; (2) By using the effective crop region extraction mechanism, the interference of non-crop background on the statistical results was reduced, and the stability of the water assessment results was improved; (3) By using an automated hierarchical decision-making mechanism, the involvement of human experience is reduced, and the objectivity and repeatability of the evaluation results are improved. Attached Figure Description

[0017] Figure 1 A schematic diagram of the main interface of the system of this invention.

[0018] Figure 2 This is a flowchart of the method of the present invention.

[0019] Figure 3 This is a schematic diagram illustrating the content derived from the present invention. Detailed Implementation

[0020] Example 1 A method for crop moisture assessment based on UAV remote sensing images includes the following steps: S1. Acquire UAV remote sensing images of the farmland area to be tested, supporting RGB or multispectral format input, and identify and separate image channels; S2. Provides a preset vegetation index model and a user-defined formula interface; calculates the vegetation index matrix using the red, green, blue, and near-infrared band data of the image according to the mode selected by the user. S3. Extract the effective area from the vegetation index matrix, calculate the index mean of the effective area, and divide the water content into three levels: "High", "Medium" and "Low" according to the preset irrigation effect threshold. S4. Map the calculated vegetation index matrix to a pseudo-color heat map for display, and export the structured data of assessment level and statistical mean to a spreadsheet file.

[0021] Preferably, in step S2, the preset vegetation index model includes: (1); GNDVI = (NIR - Green) / (NIR + Green) (2); (3); Among them, NDVI is the Normalized Difference Vegetation Index, GNDVI is the Green Normalized Difference Vegetation Index, SAVI is the Soil-Regulated Vegetation Index, NIR is infrared data, Red is red band data, and Green is green band data.

[0022] Preferably, the hierarchical evaluation logic in step S3 is as follows: calculate the index mean Avg of the effective region; if Avg > 0.6, determine the moisture content as "High"; if 0.4 < Avg < 0.6, determine the moisture content as "Medium"; if Avg ≤ 0.4, determine the moisture content as "Low"; the effective region is obtained after removing the background noise region with Index ≤ 0 through masking.

[0023] Preferably, in step S2, the user-defined formula interface supports inputting any mathematical expression with red, green, blue, and near-infrared bands as variables. The system performs a validity check on the expression using a dynamic parsing algorithm. If the check passes, algebraic operations are performed on the image matrix to generate a vegetation index matrix. If the check fails, an error message is triggered.

[0024] Preferably, the channel identification and separation of the UAV remote sensing image in step S1 is as follows: if it is a three-channel RGB format image, extract the red, green and blue band channels; if it is a four-channel or higher multispectral format image, automatically match the first three channels as red, green and blue bands for visualization, and extract the near-infrared band from the remaining channels for vegetation index calculation.

[0025] Preferably, in step S4, the pseudo-color heat map uses the Jet color mapping table to map the vegetation index values ​​from high to low to blue-red color gradients. The lower the value, the more blue it is displayed, and the higher the value, the more red it is displayed. The heat map is also equipped with color legends and numerical scales to achieve a visual display of the spatial distribution of crop moisture.

[0026] A crop moisture assessment system based on UAV remote sensing images is provided. The system runs on a computer and includes an image import module, an index calculation module, an analysis and evaluation module, and a data export module. The modules work together to process UAV remote sensing images, assess crop moisture, and output results.

[0027] Preferably, the image import module supports loading remote sensing image files in tif, jpg, and png formats, can automatically determine the image channel dimension, adapt to the band separation and data display of three-channel RGB images or four-channel and above multispectral images, and has an image preview function, which can display the original remote sensing images in the graphical user interface.

[0028] Preferably, the index calculation module has built-in NDVI, GNDVI, and SAVI standard vegetation index calculation models, and is also configured with a custom formula input box and a dynamic parsing engine. The eval function is used to parse the valid formula string input by the user and perform algebraic operations on the image matrix to generate an index distribution map. The module also has an anomaly capture mechanism to identify and display a calculation failure prompt when there are errors in formula variables or calculation logic.

[0029] Preferably, the analysis and evaluation module can perform masking on the index layer, generate a logical matrix with index>0 to filter non-vegetation background noise, and statistically analyze the mean, maximum, minimum, and standard deviation of the vegetation index in the effective crop area; the data export module can encapsulate the region name, index mean, moisture rating results, and index statistical feature values ​​into a table object and write it into an Excel spreadsheet file, while also supporting the export and saving of the pseudo-color heatmap as an image.

[0030] Example 2 To better understand the purpose, system architecture, and functional implementation of this embodiment, the embodiments and features in the embodiments of this application can be combined with each other without conflict. The exemplary embodiments disclosed in this embodiment will be described below with reference to the accompanying drawings. This embodiment is based on a Windows 10 / 11 64-bit operating system, developed in MATLAB R2018a, and utilizes the GUIDE tool to build a graphical user interface. It includes specific technical details disclosed in this embodiment to aid understanding, but these details should be considered exemplary rather than restrictive.

[0031] Figure 1 This is a schematic diagram of the main interface of the system of the present invention, which includes an image browsing area, a function control area, and a result display area. Figure 2 This is a flowchart of the method of the present invention. Figure 3 The output is shown below.

[0032] like Figure 2 As shown, steps S110 to S140 of the present invention consist of: In step S110, the user clicks "Import Image," and the system loads the UAV multispectral TIFF image using the imread function. The system detects that size(img,3)>3, and automatically uses channels 1, 2, and 3 for RGB display, while defining channel 4 as an NIR variable. In step S120, the user selects "NDVI" from the drop-down menu. The system backend calls the compute_index function to execute the formula (nir - red) . / (nir + red). The calculated index is a double-precision matrix. In step S130, the system uses imagesc(index) to draw an image on the right coordinate axis and applies colormap('jet') to map the high and low values ​​to a blue-red color gradient, visually displaying the differences in crop growth; In step S130, the user clicks "Analyze Moisture Content". The system executes the analyze_efficiency function: Mask: Generate a logical matrix mask = index>0 to exclude soil background.

[0033] Statistics: The mean of the effective region is calculated as avg = 0.75 (hypothetical value).

[0034] Verdict: Since 0.75 > 0.6, the system determines the level to be "High" and outputs "MoistureContent: High" on the interface.

[0035] In step S140, click the export button, and the system will write "Region 1", "0.75", and "High" into the result.xlsx file.

[0036] Example 3 For research experiments based on custom formulas, if users wish to test new indices, they can select "Custom" from the drop-down menu and enter "(nir-red). / (nir+red+0.5)" in the text box. "1.5" refers to the SAVI formula. After clicking "Calculate," the system dynamically executes this string by calling the `eval()` function through a try-catch block. If the formula variable is misspelled, such as entering the non-existent "blue2," the system will catch the exception and display a "Failed to calculate exponent" message to ensure system stability.

[0037] Hardware operating environment requirements: CPU: 12th Gen Intel Core i3-12100F or equivalent processor.

[0038] Hard drive: 20GB or more of available space.

[0039] Display: Supports resolutions of 1920x1080 and above to fully display comparison images.

[0040] Example 4 Comparative analysis of existing technologies and the method of this invention: In existing methods, the vegetation index value of the entire image is directly averaged without removing non-crop areas; in the method of this invention, non-crop areas are first removed by logical masking, and then the effective crop areas are statistically analyzed.

[0041] In a test area with a clearly exposed bare soil background, photographs were taken one day after rain. The average vegetation index calculated using traditional methods was 0.41, while the average vegetation index calculated using the method of this invention was 0.48. The results indicate that this invention, by introducing an effective crop area screening mechanism, can effectively reduce the impact of background noise on water assessment results, making the assessment results closer to the actual water status of crops.

[0042] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for assessing crop moisture and inferring irrigation effects, characterized in that, Includes the following steps: S1. Acquire UAV remote sensing images of the farmland area to be tested, supporting RGB or multispectral format input, and identify and separate image channels; S2. Provides a preset vegetation index model and a user-defined formula interface; calculates the vegetation index matrix using the red, green, blue, and near-infrared band data of the image according to the mode selected by the user. S3. Extract the effective area from the vegetation index matrix, calculate the index mean of the effective area, and divide the water content into three levels: "High", "Medium" and "Low" according to the preset irrigation effect threshold. S4. Map the calculated vegetation index matrix to a pseudo-color heat map for display, and export the structured data of assessment level and statistical mean to a spreadsheet file.

2. The method for crop moisture assessment and irrigation effect inference according to claim 1, characterized in that, In step S2, the preset vegetation index model includes: (1)； GNDVI = (NIR - Green) / (NIR + Green) (2)； (3)； Among them, NDVI is the Normalized Difference Vegetation Index, GNDVI is the Green Normalized Difference Vegetation Index, SAVI is the Soil-Regulated Vegetation Index, NIR is infrared data, Red is red band data, and Green is green band data.

3. The method for crop moisture assessment and irrigation effect inference according to claim 1, characterized in that, The specific hierarchical evaluation logic in step S3 is as follows: calculate the index mean Avg of the effective region; if Avg > 0.6, the moisture content is determined to be "High"; if 0.4 < Avg < 0.6, the moisture content is determined to be "Medium"; if Avg ≤ 0.4, the moisture content is determined to be "Low"; the effective region is obtained after removing the background noise region with Index ≤ 0 through masking.

4. The method for crop moisture assessment and irrigation effect inference according to claim 1, characterized in that, In step S2, the user-defined formula interface supports inputting any mathematical expression with red, green, blue, and near-infrared bands as variables. The system performs a validity check on the expression using a dynamic parsing algorithm. If the check passes, algebraic operations are performed on the image matrix to generate a vegetation index matrix. If the check fails, an error message is triggered.

5. The method for crop moisture assessment and irrigation effect inference according to claim 1, characterized in that the steps are as follows: The channel identification and separation of UAV remote sensing images in S1 are as follows: if it is a three-channel RGB format image, the red, green and blue band channels are extracted; if it is a four-channel or higher multispectral format image, the first three channels are automatically matched as red, green and blue bands for visualization, and the near-infrared band is extracted from the remaining channels for vegetation index calculation.

6. The method for crop moisture assessment and irrigation effect inference according to claim 1, characterized in that, In step S4, the pseudo-color heat map uses the Jet color mapping table to map the vegetation index values ​​from high to low to blue-red color gradients. The lower the value, the more blue it is displayed, and the higher the value, the more red it is displayed. The heat map is also equipped with color legends and numerical scales to visualize the spatial distribution of crop moisture.

7. A crop moisture assessment and irrigation effect inference system implementing the method of any one of claims 1-6, characterized in that, The system runs on a computer device and includes an image import module, an index calculation module, an analysis and evaluation module, and a data export module. These modules work together to process UAV remote sensing images, assess crop moisture, and output results.

8. The crop moisture assessment and irrigation effect inference system according to claim 7, characterized in that, The image import module supports loading remote sensing image files in tif, jpg, and png formats. It can automatically determine the image channel dimension, adapt to the band separation and data display of three-channel RGB images or four-channel and above multispectral images, and has an image preview function that can display the original remote sensing images in a graphical user interface.

9. The crop moisture assessment and irrigation effect inference system according to claim 7, characterized in that, The index calculation module has built-in NDVI, GNDVI, and SAVI standard vegetation index calculation models, and is also configured with a custom formula input box and a dynamic parsing engine. The eval function parses the valid formula strings input by the user and performs algebraic operations on the image matrix to generate an index distribution map. The module also has an exception capture mechanism, which can identify errors in formula variables and calculation logic and pop up a calculation failure prompt.

10. The crop moisture assessment and irrigation effect inference system according to claim 7, characterized in that, The analysis and evaluation module can perform masking on the index layer, generate a logical matrix with index>0 to filter non-vegetation background noise, and statistically analyze the mean, maximum, minimum and standard deviation of the vegetation index in the effective crop area. The data export module can encapsulate region names, index averages, moisture rating results, and index statistical characteristic values ​​into table objects and write them into an Excel spreadsheet file. It also supports exporting and saving pseudo-color heatmaps as image formats.

Images