Rice extraction method combining time sequence optics and SAR (Synthetic Aperture Radar) image

Abstract

The invention relates to a rice extraction method combining time sequence optics and SAR (Synthetic Aperture Radar) images, which comprises the following steps of: firstly, reconstructing time sequence data of a normalized difference vegetation index (NDVI) and an improved normalized difference water index (MNDWI); then quantifying the characteristics of the NDVI in the rice submerging-transplanting stage and the growth peak, the characteristics of the MNDWI in the rice submerging-transplanting stage and the difference value characteristics of the VH time sequence in the rice submerging-transplanting stage and the growth peak, constructing a multi-characteristic rice index MFRI, and determining a corresponding threshold value according to the index statistical result of a rice sample to accurately extract rice; spectral features, time sequence features and morphological features are combined during NDVI and MNDWI time sequence reconstruction, multi-modal long-time-sequence remote sensing data are fully used, rice planting distribution data of a target year can be rapidly and accurately obtained by using the method, spatial-temporal dynamic changes of rice planting distribution are obtained, data support and reference are provided for agricultural decision making, and the method is suitable for large-scale popularization and application. And the method has important practical application significance.

Description

Technical Field

[0001] The invention belongs to the technical field of classification extraction of remote sensing images, and in particular relates to a rice extraction method combining time series optics and SAR images. Background Art

[0002] Rice occupies more than 9% of the world's cultivated land area and is the staple food resource for more than half of the world's population, playing an important role in food security. In addition, rice is a major water consumer and one of the main sources of greenhouse gas emissions. Therefore, rapid and accurate identification of the distribution range of rice planting is of great significance for ensuring food security and evaluating environmental resources.

[0003] Currently, satellite data is the main data source for large-scale rice mapping. Sentinel-1 and Sentinel-2 data have high temporal and spatial resolution and are the main data source for rice mapping research in recent years. In rice mapping, detection based on key phenological stages is a common method. By analyzing multiple spectral bands and spectral indices, the corresponding features of rice can be screened out, and large-scale rice planting distribution maps can be quickly drawn. However, when using the key phenological period of rice for classification, it depends on the accuracy of phenological information extraction, and there is a high overlap between the characteristics of rice and wetland vegetation. Therefore, most current studies are based on remote sensing time series data and rice phenological characteristics, and combine machine learning or deep learning models to solve the above problems. Although traditional machine learning and deep learning algorithms can process large-scale data quickly and effectively, the models require a large amount of labeled sample data for training, and obtaining rice labeled sample data requires a lot of time and effort. Although many studies currently use transfer learning or use phenological characteristics to automatically generate rice samples, thereby reducing the need for ground truth samples, such methods still have problems such as limited model generalization ability, insufficient representativeness of generated samples, and computational complexity.

[0004] The essence of the index method is to explore the commonalities of the target objects, to adapt to target recognition in different scenarios, and to conduct more flexible mapping research on target objects than rigid classification. At present, a variety of indices have been successfully developed and applied in the mapping research of various types of landforms. For rice monitoring, there are also optical-based rice indices and SAR-based rice indices, but they cannot avoid their own data defects. Therefore, a new method based on the rice index is urgently needed to achieve rapid and large-scale rice mapping. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a rice extraction method combining time series optical and SAR images.

[0006] The rice extraction method combining time series optical and SAR images includes the following steps:

[0007] Step 1: Calculate the normalized vegetation index NDVI and the improved normalized water index MNDWI corresponding to the surface reflectance data of the remote sensing image, and construct the original time series of NDVI and MNDWI;

[0008] Step 2: Reconstruct the NDVI and MNDWI time series of each pixel to characterize the phenological temporal characteristics of the pixel;

[0009] Step 3, extract the minimum value P1 of NDVI in the rice flooding-transplanting stage, the maximum value P2 in the growth peak period, and the maximum value P3 of MNDWI in the rice flooding-transplanting stage, extract the minimum value P4 of VH time series in the rice flooding-transplanting stage and the maximum value P5 in the growth peak period, and calculate the difference h=P5-P4;

[0010] Step 4: transform P1 with f(n) to represent the similarity between each pixel and non-vegetation pixels; transform P2 with f(v) to represent the similarity between each pixel and vegetation pixels; transform P3 with f(w) to represent the similarity between each pixel and water pixels; transform the value range of h into 0-1 with f(h) to construct the multi-feature rice index MFRI = f(n)×f(v)×f(w)×f(h);

[0011] Step 5: Determine the corresponding MFRI threshold according to the MFRI statistical results of the sample to extract the rice planting range.

[0012] Preferably, in step 2, the NDVI and MNDWI time series of each pixel are reconstructed using a harmonic model, and the temporal features, spectral features and spatial texture features are combined.

[0013] Preferably, in step 4, f(n) is a quadratic function of P1, which is used to convert the distance between the value of P1 and 0 into a value between 0 and 1, thereby amplifying the difference between the pixel and the evergreen vegetation.

[0014] Preferably, in step 4, f(v) is a quadratic function of (1-P2), which is used to convert the distance between P2 and 1 into a value between -3 and 1, thereby amplifying the difference between the pixel and non-vegetation.

[0015] Preferably, in step 4, if P3 is greater than or equal to 0, the pixel point is determined to be a water body or a land feature with high humidity, f(w)=1; if P3 is less than 0, f(w)=1+P3, and the closer the value of f(w) is to 1, the greater the possibility that the pixel point is a water body.

[0016] Preferably, in step 4, f(h) uses trigonometric functions to convert h into a value between -1 and 1.

[0017] The beneficial effects of the present invention are:

[0018] 1) The present invention first reconstructs the time series data of the Normalized Difference Vegetation Index (NDVI) and the improved Normalized Difference Water Index (MNDWI), and then quantifies the characteristics of NDVI in the rice flooding-transplanting stage and the peak growth period, the characteristics of MNDWI in the rice flooding-transplanting stage, and the difference characteristics of the VH time series between the rice flooding-transplanting stage and the peak growth period, constructs a multi-feature rice index (MFRI), and determines the corresponding threshold value according to the index statistical results of the rice sample to accurately extract the rice.

[0019] 2) The present invention combines spectral features, time series features, and morphological features when reconstructing NDVI and MNDWI time series, and makes full use of multimodal long-time series remote sensing data. This method can quickly and accurately obtain rice planting distribution data for the target year, obtain the spatiotemporal dynamic changes of rice planting distribution, and provide data support and reference for agricultural decision-making, which has important practical application significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a flow chart of rice extraction combining time series optical and SAR images;

[0021] Figure 2 The original and reconstructed NDVI time series of rice pixels;

[0022] Figure 3 The original and reconstructed MNDWI time series of rice pixels;

[0023] Figure 4 Design schematic diagram for rice spectral index characteristics;

[0024] Figure 5 This is a schematic diagram of the backscattering characteristic coefficient of rice VH;

[0025] Figure 6 Statistical box plot of MFRI values ​​for rice and non-rice samples;

[0026] Figure 7 Schematic diagram of rice extraction results. DETAILED DESCRIPTION

[0027] The present invention is further described below in conjunction with embodiments. The description of the following embodiments is only used to help understand the present invention. It should be noted that for ordinary persons in the art, without departing from the principle of the present invention, the present invention can also be modified in some ways, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

[0028] Embodiment 1

[0029] As an embodiment, in view of the fact that the existing rice extraction methods use a single optical or SAR image, which cannot avoid the quality defects of the data itself, and the problem of low computational efficiency when using fused images, a rice extraction method combining time series optical and SAR images is proposed. This rice extraction method combining time series optical and SAR images, such as Figure 1 As shown in the figure, firstly, the time series data of the Normalized Difference Vegetation Index (NDVI) and the Improved Normalized Difference Water Index (MNDWI) are reconstructed, and then the difference between rice and other landforms is amplified by quantifying the characteristics of the rice flooding-transplanting stage and the peak growth stage. The corresponding threshold is determined according to the index statistical results of the sample to accurately extract rice, which specifically includes the following steps:

[0030] Step 1: Use the screening module to screen all available Sentinel-1 and Sentinel-2 remote sensing images of the target area and target year, and pre-process all remote sensing images;

[0031] Step 2: The spectral index calculation module calculates the normalized vegetation index NDVI image and the improved normalized water index MNDWI corresponding to all Sentinel-2 remote sensing images, and constructs the original time series of NDVI and MNDWI;

[0032] Step 3: The reconstruction module uses the harmonic model to reconstruct the NDVI and MNDWI time series of each pixel to obtain the reconstructed time series data, which are used to characterize the phenological timing characteristics of the pixel.

[0033] Step 4: extract the minimum value P1 of NDVI in the rice flooding-transplanting stage and the maximum value P2 in the growth peak period, extract the maximum value P3 of MNDWI in the rice flooding-transplanting stage, extract the minimum value P4 of VH time series in the rice flooding-transplanting stage and the maximum value P5 in the growth peak period, and calculate the difference h=P5-P4;

[0034] Step 5: The MFRI construction module constructs the multi-feature rice index MFRI = f(n) × f(v) × f(w) × f(h) by using the range expression of P1, the range expression of P2, the range expression of P3, and the range expression of h by using f(h);

[0035] Step 6: The extraction module determines the corresponding threshold according to the MFRI statistical results of the sample and extracts the rice planting area, such as Figure 7 shown.

[0036] Embodiment 2

[0037] As another embodiment, this embodiment 2 proposes a more specific rice extraction method combining time series optical and SAR images on the basis of embodiment 1. In step 5, the calculation formulas of f(n), f(v), f(w), f(h), and MFRI are respectively:

[0038] In f(n), a quadratic function is used to convert the distance between P1 and 0 into a value range of 0 to 1. A square operation is applied to P1 to magnify the difference with evergreen vegetation.

[0039] f(n)=1-P1 2

[0040] In f(v), a quadratic function is used to convert the distance between p2 and 1 into a range of values ​​from -3 to 1. A square operation is applied to (1-P2) to amplify the difference from non-vegetation.

[0041] f(v)=1-(1-P2) 2

[0042] In f(w), if P3 is greater than or equal to 0, it is considered to be a water body or a land feature with high humidity, and a value of 0 is entered; if P3 is less than 0, considering the influence of data quality or fitting smoothing, the closer it is to 0, the more likely it is a water body, so the original value is entered.

[0043]

[0044] In f(h), trigonometric functions are used to convert the dynamic changes of VH backscattering into a value range from -1 to 1.

[0045]

[0046] Expressing each feature by multiplication meets the requirements in MFRI.

[0047] MFRI=f(n)×f(v)×f(w)×f(h)

[0048] In step 6, the expression that rice extraction needs to satisfy is:

[0049] MFRI(x,y)≥TH MFRI

[0050] Among them, MFRI(x,y) is the MFRI value at position (x,y), TH MFRI is the threshold of MFRI. Figure 6 As shown, in this embodiment, TH MFRI It is 0.57.

[0051] In summary, the present invention combines time series optical and SAR images, and proposes an index MFRI for rice planting range extraction by quantifying the key phenological stage characteristics of rice. The index can be used to quickly obtain a large-scale rice distribution dataset for the target year, which helps to understand the spatiotemporal dynamics of rice planting and provide data support and reference for agricultural decision-making.

[0052] It should be noted that the parts in this embodiment that are the same or similar to those in the first embodiment can be referenced to each other and will not be described in detail in this application.

[0053] Embodiment 3

[0054] As another embodiment, this embodiment three proposes, on the basis of embodiments one and two, a more specific rice extraction method combining time series optical and SAR images, using Sentinel-1 and Sentinel-2 data as data sources and relying on the Google Earth Engine (GEE) platform.

[0055] A computer storage medium and a computer program product are provided. The computer storage medium stores a computer program. When the computer program is run on a computer, the computer executes the rice extraction method.

[0056] In step 1, all available Sentinel-1 and Sentinel-2 remote sensing images of the target area and target year are screened and preprocessed. Sentinel-1 remote sensing images are radar band data, and speckle noise reduction is performed using a Lee-sigma filter with a 5×5 window size. Sentinel-2 remote sensing images are surface reflectivity data, and preprocessing includes atmospheric correction and cloud removal.

[0057] In step 2, the calculation formulas for NDVI and MNDWI are:

[0058]

[0059] Among them, NIR is the near infrared band value, Red is the red light band value, Green is the green light band value, and SWIR1 is the first band value of short-wave infrared;

[0060] In this embodiment, NIR, Red, Green, and SWIR1 are B8, B4, B3, and B11 of the Sentinel-2 data, respectively.

[0061] In step three, if Figure 2 and Figure 3 As shown in the figure, the original time series of NDVI and MNDWI are reconstructed using 3-cycle harmonic fitting. The calculation formula for reconstructing the NDVI and MNDWI time series of each pixel is:

[0062]

[0063] in, is the fitted value on day t, t is the Julian day, a is the overall horizontal offset value, b is the linear coefficient, c is the n and d n are the coefficients of the sine and cosine functions of different frequencies, and T is the number of days in a year.

[0064] Step 4, such as Figure 4 and Figure 5 As shown, the reconstructed NDVI and MNDWI original time series and the denoised VH time series were used to design the characteristics of the key phenological period of rice.

[0065] It should be noted that the parts in this embodiment that are the same or similar to those in Embodiments 1 and 2 can be referenced to each other and will not be described in detail in this application.

[0066] 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.

Claims

1. A rice extraction method combining time series optical and SAR images, characterized in that: The following steps are involved: Step 1: Calculate the normalized vegetation index NDVI and the improved normalized water index MNDWI corresponding to the surface reflectance data of the remote sensing image, and construct the original time series of NDVI and MNDWI; Step 2: Reconstruct the NDVI and MNDWI time series of each pixel to characterize the phenological temporal characteristics of the pixel; Step 3, extract the minimum value P1 of NDVI in the rice flooding-transplanting stage, the maximum value P2 in the growth peak period, and the maximum value P3 of MNDWI in the rice flooding-transplanting stage, extract the minimum value P4 of VH time series in the rice flooding-transplanting stage and the maximum value P5 in the growth peak period, and calculate the difference h=P5-P4; Step 4: transform P1 with f(n) to represent the similarity between each pixel and non-vegetation pixels; transform P2 with f(v) to represent the similarity between each pixel and vegetation pixels; transform P3 with f(w) to represent the similarity between each pixel and water pixels; transform the value range of h into 0-1 with f(h) to construct the multi-feature rice index MFRI = f(n)×f(v)×f(w)×f(h); Step 5: Determine the corresponding MFRI threshold according to the MFRI statistical results of the sample to extract the rice planting range.

2. The rice extraction method combining time series optical and SAR images according to claim 1, characterized in that: In step 2, the harmonic model is used to reconstruct the NDVI and MNDWI time series of each pixel, combining the temporal features, spectral features, and spatial texture features.

3. The rice extraction method combining time series optical and SAR images according to claim 1, characterized in that: In step 4, f(n) is a quadratic function of P1, which is used to convert the distance between the value of P1 and 0 into a value between 0 and 1, thereby amplifying the difference between the pixel and the evergreen vegetation.

4. The rice extraction method combining time series optical and SAR images according to claim 1, characterized in that: In step 4, f(v) is a quadratic function of (1-P2), which is used to convert the distance between P2 and 1 into a value between -3 and 1, thereby amplifying the difference between the pixel and non-vegetation.

5. The rice extraction method combining time series optical and SAR images according to claim 1, characterized in that: In step 4, if P3 is greater than or equal to 0, the pixel is considered to be a water body or a land feature with high humidity, f(w)=1; if P3 is less than 0, f(w)=1+P3, and the closer the value of f(w) is to 1, the greater the possibility that the pixel is a water body.

6. The rice extraction method combining time series optical and SAR images according to claim 1, characterized in that: In step 4, f(h) uses trigonometric functions to convert h into a value between -1 and 1.

Images