A phenology-based approach to global rice mapping integrating optical and radar data

Abstract

The present invention provides a global rice mapping method based on phenology-based integration of optical and radar data, comprising: obtaining VH polarization backscatter coefficient time series data from synthetic aperture radar (SAR) and land surface moisture index and enhanced vegetation index (EVI) from optical images; obtaining phenological time windows for rice flooding and transplanting based on the VH time series data; obtaining a rice flooding signal based on the SAR according to the phenological period and normalizing it to obtain a normalized rice flooding signal; obtaining a rice flooding signal based on the optical image using the vegetation index; obtaining a rice flooding signal based on the integrated optical and radar data based on an integration formula; obtaining a rice rapid growth signal based on a vegetation rapid growth formula; and integrating the rice flooding signal and the rapid growth signal to obtain rice spatial distribution information. The present invention solves the problem in the prior art that rice planting information relies on sample point acquisition, which makes it difficult to extract rice spatial distribution on a large scale.

Description

Technical Field

[0001] The present invention relates to the technical field of rice mapping, and in particular to a global rice mapping method based on phenology integrating optical and radar data. Background Art

[0002] Rice paddies account for over 12% of global agricultural land and are crucial to global food security. Satellite remote sensing, due to its high timeliness and low cost, is considered an effective method for monitoring rice and has been widely used to extract crop planting areas.

[0003] Current rice mapping methods primarily rely on expert knowledge (such as phenology-based algorithms), machine learning (such as support vector machines and random forests), and deep learning. Machine learning and deep learning can achieve high classification accuracy in regions with good sample conditions, but they are difficult to apply at the national or even global scale.

[0004] Remote sensing extraction of rice planting information is based on optical data and synthetic aperture radar data. Traditional optical imaging algorithms are often restricted by the acquisition of phenological periods, and the acquisition of phenological periods mostly relies on station data, which makes it difficult to achieve rice mapping in areas without station data; secondly, it is inevitable that optical images will be subject to invalid observations in cloudy and rainy areas, which is insurmountable for optical imaging algorithms that need to capture irrigation signals. Synthetic aperture radar data has the ability to penetrate clouds and rain and is not affected by weather factors. However, synthetic aperture radar is always subject to speckle noise, and some dryland crops or wetlands exhibit backscattering characteristics similar to rice, which is difficult to overcome for rice mapping based on synthetic aperture radar. The existing technology flow chart is as follows. Figure 1 shown.

[0005] Fusion of two images is an effective method for overcoming the mapping difficulties associated with a single image. However, differences in the imaging principles of the two data types make effective fusion extremely challenging. Existing research often uses machine learning methods to fuse the two data types. However, machine learning relies on the acquisition of sample points. Using such methods to extract rice spatial distribution in different regions often requires reselecting training samples. As a result, such methods are not very generalizable and are not suitable for extracting rice spatial distribution over a large area. Summary of the Invention

[0006] In order to overcome the shortcomings of the existing technology, the purpose of the present invention is to provide a phenological global rice mapping method by integrating optical and radar data. The present invention solves the problems in the existing technology of rice planting information mapping that rely on the acquisition of sample points, resulting in the inability to carry out rice spatial distribution extraction on a large scale, and the difficulty in effectively fusing optical and radar data, resulting in poor accuracy of rice information extraction.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A phenology-based approach to global rice mapping integrating optical and radar data, including:

[0009] Acquiring effective VH polarization data of a synthetic aperture radar within a preset time based on rice information in a test area, and preprocessing the VH polarization data to obtain VH polarization backscatter coefficient time series data;

[0010] Acquire effective optical image data of the area to be measured within a preset time based on rice information, and pre-process the optical image data to obtain a land surface moisture index and an enhanced vegetation index;

[0011] Obtaining a phenological time window for rice flooding and transplanting based on the VH polarization backscatter coefficient time series data;

[0012] obtaining a rice flooding signal based on synthetic aperture radar according to the phenological period time window of rice flooding and transplanting;

[0013] Normalizing the rice flooding signal based on the synthetic aperture radar to obtain a normalized rice flooding signal of the synthetic aperture radar;

[0014] A rice flooding signal recognition formula based on optical images is used to obtain the rice flooding signal based on the optical image according to the land surface moisture index and the enhanced vegetation index;

[0015] Based on an integration formula, an integrated rice flooding signal is obtained according to the rice flooding signal of the synthetic aperture radar and the rice flooding signal based on the optical image;

[0016] Based on the rapid vegetation growth formula, a rice rapid growth signal is obtained according to the enhanced vegetation index;

[0017] The rice flooding transplanting signal and the rice rapid growth signal are integrated to obtain rice spatial distribution information and complete rice mapping.

[0018] Preferably, the preprocessing of VH polarization data includes:

[0019] Orbit correction operation, radiation correction operation, terrain correction operation, resampling operation, and spatial filtering denoising operation.

[0020] Preferably, the preprocessing of the optical image data includes:

[0021] Near infrared band calculation, infrared band calculation and red band calculation.

[0022] Preferably, obtaining the phenological time window of rice flooding and transplanting based on the VH polarization backscatter coefficient time series data includes:

[0023] Determine the minimum value of the VH polarization backscatter coefficient time series data during the rice growing season;

[0024] Determining a maximum value of the VH polarization backscatter coefficient time series data forward in the VH polarization backscatter coefficient time series data based on the minimum value;

[0025] determining a maximum value of the VH polarization backscatter coefficient time series data backward in the VH polarization backscatter coefficient time series data based on the minimum value;

[0026] determining a start date for rice flooding and transplanting based on the maximum value of the forward determined VH polarization backscatter coefficient time series data and the minimum value of the VH polarization backscatter coefficient time series data within the rice growing season;

[0027] determining an end date for rice flooding and transplanting based on the maximum value of the backward determined VH polarization backscatter coefficient time series data and the minimum value of the VH polarization backscatter coefficient time series data within the rice growing season;

[0028] The phenological period time window for rice flooding and transplanting is determined according to the start date and the end date of rice flooding and transplanting.

[0029] Preferably, the normalization formulas of the radar data flooding signal are:

[0030] Y Pre =T anh (ΔVH1-6);

[0031] Y Post =T anh (ΔVH2-6);

[0032] Among them, Y Pre With Y Post Respectively represent the results of the dynamic change range of the VH time series curve of rice before and after flooding, normalized to ±1, T anh Represents the hyperbolic tangent function, 6 represents the threshold value in the formula, ΔVH1 and ΔVH2 are the dynamic change ranges on the VH time series curves before and after the flooding period of rice, respectively.

[0033] Preferably, the rice flooding signal recognition formula of the optical image is:

[0034] LSWI+0.05≥EVI;

[0035] LSWI represents the land surface water index of the remote sensing pixel during the rice transplanting period, EVI represents the vegetation index of the remote sensing pixel during the rice transplanting period, and 0.05 represents the threshold of the formula.

[0036] Preferably, the integration formula is:

[0037]

[0038] Y Flood is the integrated rice flooding signal, Y FTP This is the result obtained by the rice flooding signal recognition formula based on optical imaging. Pre With Y Post The dynamic change range of the VH time series curves before and after the flooding period of rice was normalized to ±1.

[0039] Preferably, obtaining a rice rapid growth signal according to the enhanced vegetation index comprises:

[0040] calculating the rapid growth characteristics of rice according to the enhanced vegetation index;

[0041] Based on the rapid growth formula of rice, the rapid growth characteristics of rice are obtained.

[0042] Preferably, the vegetation rapid growth signal extraction formula is:

[0043]

[0044] Among them, Y RGP EVI is a rapid growth signal for rice. RGPmax EVI is the maximum value of the enhanced vegetation index during the rapid growth period of rice. max It is the maximum value of the Rice Enhanced Vegetation Index throughout the year.

[0045] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0046] The present invention provides a global rice mapping method based on phenology and integrating optical and radar data, comprising: obtaining effective VH polarization data of a synthetic aperture radar within a preset time based on rice information in a region to be measured, and preprocessing the VH polarization data to obtain VH polarization backscatter coefficient time series data; obtaining effective optical image data within a preset time based on rice information in a region to be measured, and preprocessing the optical image data to obtain a land surface moisture index and an enhanced vegetation index; obtaining the phenological period of rice flooding and transplanting based on the VH polarization backscatter coefficient time series data; obtaining the phenological period of rice flooding based on synthetic aperture radar based on the phenological period of rice flooding and transplanting signal; scaling the synthetic aperture radar-based rice flooding signal using a normalization formula to obtain a normalized synthetic aperture radar rice flooding signal; obtaining an optical image-based rice flooding signal based on the land surface moisture index and the enhanced vegetation index using a rice flooding signal identification formula based on optical imaging; obtaining an integrated rice flooding signal based on the synthetic aperture radar and optical image-based rice flooding signals using an integration formula; obtaining a rice rapid growth signal based on the enhanced vegetation index using a vegetation rapid growth formula; and integrating the rice flooding transplanting signal and the rice rapid growth signal to obtain rice spatial distribution information, thereby completing rice mapping. The present invention improves the accuracy of rice spatial distribution mapping by integrating optical and radar data into a phenological global rice mapping method. Furthermore, by integrating different satellite remote sensing data sources and applying a phenological algorithm, the present invention eliminates the need to rely on sample point acquisition, thereby expanding the scope of rice spatial distribution extraction. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] 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. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0048] Figure 1 A flow chart of the prior art provided for an embodiment of the present invention;

[0049] Figure 2 A flow chart of a method for phenological global rice mapping by integrating optical and radar data provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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 making creative efforts are within the scope of protection of the present invention.

[0051] The purpose of the present invention is to provide a phenological global rice mapping method by integrating optical and radar data. The present invention solves the problems in the existing technology of rice planting information extraction, which depends on the acquisition of sample points, resulting in the inability to carry out rice spatial distribution extraction on a large scale, and the defects of the method of fusing optical and radar data, resulting in poor accuracy.

[0052] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0053] like Figure 2 As shown, the present invention provides a phenological global rice mapping method by integrating optical and radar data, comprising:

[0054] Step 100: Acquire effective VH polarization data of a synthetic aperture radar within a preset time based on rice information in a test area, and preprocess the VH polarization data to obtain VH polarization backscatter coefficient time series data;

[0055] Step 200: Acquire effective optical image data of the area to be measured within a preset time based on rice information, and pre-process the optical image data to obtain a land surface moisture index and an enhanced vegetation index;

[0056] Step 300: Obtaining a phenological time window for rice flooding and transplanting based on the VH polarization backscatter coefficient time series data;

[0057] Step 400: obtaining a rice flooding signal based on synthetic aperture radar according to the phenological period of rice flooding and transplanting;

[0058] Step 500: normalizing the rice flooding signal based on the synthetic aperture radar to obtain a normalized rice flooding signal of the synthetic aperture radar;

[0059] Step 600: Obtaining a rice flooding signal based on an optical image according to the land surface moisture index and the enhanced vegetation index using a rice flooding signal recognition formula based on an optical image;

[0060] Step 700: Obtaining an integrated rice flooding signal based on an integration formula according to the scaled synthetic aperture radar rice flooding signal and the rice flooding signal based on the optical image;

[0061] Step 800: obtaining a rice rapid growth signal according to the enhanced vegetation index based on a rapid vegetation growth formula;

[0062] Step 900: obtaining rice spatial distribution information according to the rice rapid growth signal and the integrated rice flooding signal, and completing rice mapping.

[0063] Specifically, the extraction principle of rice is as follows:

[0064] Since rice is the only crop that requires large amounts of water for irrigation during its growth stage, water will briefly flood the fields during its irrigation and transplanting cycle. This is reflected in optical images as the LSWI index briefly rising above the NDVI or EVI index, and in the VH polarization of synthetic aperture radar, as a unique feature of rapidly decreasing backscatter values. These two features are considered key to extracting rice.

[0065] Furthermore, the preprocessing of VH polarization data includes:

[0066] Orbit correction operation, radiation correction operation, terrain correction operation, resampling operation, and spatial filtering denoising operation.

[0067] Specifically, all available VH polarization data of synthetic aperture radar in the corresponding area throughout the year are obtained, and the data are subjected to orbit correction, radiation correction, terrain correction, resampling, and spatial filtering to remove noise, so as to obtain the VH polarization backscatter coefficient time series data.

[0068] Furthermore, the preprocessing of optical image data includes:

[0069] Near infrared band calculation, infrared band calculation and red band calculation.

[0070] Specifically, the optical image data available in the corresponding area throughout the year is obtained and de-clouded, and the remote sensing vegetation index is calculated using near-infrared, infrared, red and other bands: the Land Surface Water Index (LSWI) and the Enhanced Vegetation Index (EVI).

[0071] Furthermore, obtaining the phenological period of rice flooding and transplanting based on the VH polarization backscatter coefficient time series data includes:

[0072] Determine the minimum value of the VH polarization backscatter coefficient time series data during the rice growing season;

[0073] Determining a maximum value of the VH polarization backscatter coefficient time series data forward in the VH polarization backscatter coefficient time series data based on the minimum value;

[0074] determining a maximum value of the VH polarization backscatter coefficient time series data backward in the VH polarization backscatter coefficient time series data based on the minimum value;

[0075] determining a start date for rice flooding and transplanting based on the maximum value of the forward determined VH polarization backscatter coefficient time series data and the minimum value of the VH polarization backscatter coefficient time series data within the rice growing season;

[0076] determining an end date for rice flooding and transplanting based on the maximum value of the backward determined VH polarization backscatter coefficient time series data and the minimum value of the VH polarization backscatter coefficient time series data within the rice growing season;

[0077] The phenological period of rice flooding and transplanting is determined according to the start date of rice flooding and transplanting and the end date of rice flooding and transplanting.

[0078] Specifically, first, the lowest value of the VH time series data within the rice growing season was extracted, recorded as P1. Second, the maximum VH value in the VH time series was determined forward from P1, recorded as P2. The point closest to the radar observation value and 80% of the difference between P2 and P1 was taken as the start date of the rice field flooding / transplanting stage. Third, the maximum VH value was determined backward from P1, recorded as P3. The point closest to the radar observation value and 20% of the difference between P1 and P3 was taken as the end date of the rice field flooding / transplanting stage.

[0079] Specifically, to extract the rice flooding signal based on synthetic aperture radar, the difference between points P2 and P1 and the difference between points P3 and P1 in the VH polarization time series must be calculated, that is, the following two mathematical formulas are used:

[0080] ΔVH (before flooding) = VH P2 –VH P1

[0081] ΔVH(after flooding period)=VH P3 –VH P1

[0082] Wherein, ΔVH (before flooding) represents the value obtained by subtracting the remote sensing pixel value from point P2 and point P1 in the time series, and ΔVH (after flooding) represents the value obtained by subtracting the remote sensing pixel value from point P3 and point P1 in the time series.

[0083] The ΔVH (before / after flooding) obtained in the previous step is scaled to ±1 using the following mathematical formula to calculate the probability of rice. The two mathematical formulas are:

[0084] Y Pre = Tanh(ΔVH(before flooding)-6)

[0085] Y Post = Tanh(ΔVH(after flooding period)-6)

[0086] Where Y Pre With Y Post represent the results of the dynamic change range of the VH time series curve of rice before and after the flooding period normalized to ±1, Tanh represents the hyperbolic tangent function, and 6 represents the threshold value in the formula (this threshold value can be changed according to actual conditions, including but not limited to 6);

[0087] Specifically, the rice flooding signal based on the optical image is extracted, and the pixel is determined to be a potential rice pixel if the following formula is satisfied. The mathematical formula is the rice flooding signal recognition formula based on the optical image:

[0088] LSWI+0.05≥EVI;

[0089] LSWI represents the land surface water index of the remote sensing pixel during the rice transplanting period, EVI represents the vegetation index of the remote sensing pixel during the rice transplanting period, and 0.05 represents the threshold of the formula.

[0090] Specifically, the integration formula is:

[0091]

[0092] Y Flood is the integrated rice flooding signal, Y FTP is the result obtained by the rice flooding signal recognition formula based on optical imaging, Y Pre With Y Post They represent the results of the dynamic change range of the VH time series curve of rice before and after the flooding period being scaled to ±1.

[0093] Furthermore, obtaining a rice rapid growth signal according to the enhanced vegetation index includes:

[0094] Calculating vegetation rapid growth characteristics according to the enhanced vegetation index;

[0095] Based on the rapid growth formula of vegetation, a rapid growth signal of rice is obtained according to the rapid growth characteristics of vegetation.

[0096] Specifically, the formula for rapid vegetation growth is:

[0097]

[0098] Among them, Y RGP To obtain the potential rice results, EVI RGPmax EVI is the maximum value of the enhanced vegetation index during the rapid growth period of rice. maxIt is the maximum value of the Rice Enhanced Vegetation Index throughout the year.

[0099] Furthermore, Y RGP With Y Flood Multiplying them together gives the final probability of rice distribution, and selecting 0.7 as the threshold to separate rice from non-rice to obtain the final result.

[0100] Furthermore, all data in this embodiment are obtained by the Bing-1 satellite and the Sentinel-2 satellite, and the flooded transplanting period of rice and its spatial distribution are captured by the above two satellite data.

[0101] The beneficial effects of the present invention are as follows:

[0102] The present invention provides a global rice mapping method based on phenology integrating optical and radar data, comprising: obtaining effective VH polarization data of a synthetic aperture radar within a preset time based on rice information in a region to be measured, and preprocessing the VH polarization data to obtain VH polarization backscatter coefficient time series data; obtaining effective optical image data within a preset time based on rice information in a region to be measured, and preprocessing the optical image data to obtain a land surface moisture index and an enhanced vegetation index; obtaining the phenological period of rice flooding and transplanting according to the VH polarization backscatter coefficient time series data; obtaining the phenological period of rice flooding and transplanting based on the synthetic aperture radar time series data according to the phenological period of rice flooding and transplanting Aperture radar rice flooding signal; normalizing the rice flooding signal based on synthetic aperture radar to obtain the normalized synthetic aperture radar rice flooding signal; based on the optical image rice flooding signal identification formula, obtaining the optical image rice flooding signal according to the land surface moisture index and the enhanced vegetation index; based on the integration formula, obtaining the integrated rice flooding signal according to the synthetic aperture radar rice flooding signal and the optical image rice flooding signal; based on the vegetation rapid growth formula, obtaining the rice rapid growth signal according to the enhanced vegetation index; obtaining rice spatial distribution information according to the rice flooding signal and the rice rapid growth signal. The present invention improves the accuracy of rice information extraction by integrating optical and radar data into a phenological global rice mapping method, and the present invention does not need to rely on the acquisition of sample points, thereby expanding the scope of conducting early rice spatial distribution.

[0103] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0104] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A global rice mapping method based on phenology integrating optical and radar data, characterized in that: include: Acquiring effective VH polarization data of a synthetic aperture radar within a preset time based on rice information in a test area, and preprocessing the VH polarization data to obtain VH polarization backscatter coefficient time series data; Acquire effective optical image data of the area to be measured within a preset time based on rice information, and pre-process the optical image data to obtain a land surface moisture index and an enhanced vegetation index; Obtaining a phenological time window for rice flooding and transplanting based on the VH polarization backscatter coefficient time series data; obtaining a rice flooding signal based on synthetic aperture radar according to the phenological period time window of rice flooding and transplanting; Normalizing the rice flooding signal based on the synthetic aperture radar to obtain a normalized rice flooding signal of the synthetic aperture radar; A rice flooding signal recognition formula based on optical images is used to obtain the rice flooding signal based on the optical image according to the land surface moisture index and the enhanced vegetation index; Based on the integration formula, an integrated rice flooding signal is obtained according to the normalized rice flooding signal of the synthetic aperture radar and the rice flooding signal based on the optical image; Based on the rapid vegetation growth formula, a rice rapid growth signal is obtained according to the enhanced vegetation index; The integrated rice flooding signal and the rice rapid growth signal are integrated to obtain rice spatial distribution information, thereby completing rice mapping.

2. The global rice mapping method based on phenology integrating optical and radar data according to claim 1, characterized in that: The preprocessing of VH polarization data includes: Orbit correction operation, radiation correction operation, terrain correction operation, resampling operation, and spatial filtering denoising operation.

3. The global rice mapping method based on phenology integrating optical and radar data according to claim 1, characterized in that: The preprocessing of optical image data includes: Near infrared band calculation, infrared band calculation and red band calculation.

4. The global rice mapping method based on phenology integrating optical and radar data according to claim 1, characterized in that: The method of obtaining the phenological period time window of rice flooding and transplanting according to the VH polarization backscatter coefficient time series data includes: Determine the minimum value of the VH polarization backscatter coefficient time series data during the rice growing season; Determining a maximum value of the VH polarization backscatter coefficient time series data forward in the VH polarization backscatter coefficient time series data based on the minimum value; determining a maximum value of the VH polarization backscatter coefficient time series data backward in the VH polarization backscatter coefficient time series data based on the minimum value; determining a start date for rice flooding and transplanting based on the maximum value of the forward determined VH polarization backscatter coefficient time series data and the minimum value of the VH polarization backscatter coefficient time series data within the rice growing season; determining an end date for rice flooding and transplanting based on the maximum value of the backward determined VH polarization backscatter coefficient time series data and the minimum value of the VH polarization backscatter coefficient time series data within the rice growing season; The phenological period time window for rice flooding and transplanting is determined according to the start date and the end date of rice flooding and transplanting.

5. The global rice mapping method based on phenology integrating optical and radar data according to claim 1, characterized in that: The normalization formulas of the rice flooding signal based on synthetic aperture radar are: Y Pre =T anh (ΔVH1-6); Y Post =T anh (ΔVH2-6); Among them, Y Pre With Y Post Respectively represent the results of the dynamic change range of the VH time series curve of rice before and after flooding, normalized to ±1, T anh Represents the hyperbolic tangent function, 6 represents the threshold value in the formula, ΔVH1 and ΔVH2 are the dynamic change ranges on the VH time series curves before and after the flooding period of rice, respectively.

6. The global rice mapping method based on phenology integrating optical and radar data according to claim 1, characterized in that: The rice flooding signal recognition formula of the optical image is: LSWI+0.05≥EVI; LSWI represents the land surface water index of remote sensing pixels during the rice transplanting period, EVI represents the enhanced vegetation index of remote sensing pixels during the rice transplanting period, and 0.05 represents the threshold of the formula.

7. The global rice mapping method based on phenology integrating optical and radar data according to claim 1, characterized in that: The integration formula is: Y Flood is the integrated rice flooding signal, Y FTP is the result obtained by the rice flooding signal recognition formula based on optical imaging; Y Pre With Y Post The dynamic change range of the VH time series curves before and after the flooding period of rice was normalized to ±1.

8. The global rice mapping method based on phenology integrating optical and radar data according to claim 1, characterized in that: Obtaining a rice rapid growth signal according to the enhanced vegetation index includes: calculating the rapid growth characteristics of rice according to the enhanced vegetation index; Based on the rapid growth formula of rice, the rapid growth characteristics of rice are obtained.

9. The global rice mapping method based on phenology integrating optical and radar data according to claim 1, characterized in that: The formula for extracting the vegetation rapid growth signal is: Among them, Y RGP EVI is a rapid growth signal for rice. RGPmax EVI is the maximum value of the enhanced vegetation index during the rapid growth period of rice. max It is the maximum value of the Rice Enhanced Vegetation Index throughout the year.

Images