NDVI product registration method, apparatus and computer system

Abstract

This invention discloses an NDVI product registration method, apparatus, and computer system, which solves the problem of stitching and registration of NDVI products in different times and spaces. The method includes: performing quality checks on spatiotemporal tile images, removing tiles with cloud cover exceeding a threshold, incomplete frames, or lacking near-infrared bands; calculating NDVI using the near-infrared and red bands of the quality-checked tile images to obtain NDVI products; layering the NDVI products according to time sequence, and stitching adjacent NDVI products in each layer; and registering the stitched NDVI products to obtain seamlessly aligned NDVI products.

Description

Technical Field

[0001] This invention relates to the field of vegetation index product applications, and more particularly to an NDVI registration method, apparatus and computer system. Background Technology

[0002] Vegetation information influences the balance of global ecosystems and plays an indispensable role in climate, hydrology, and ecological cycles. Given the crucial role of vegetation information in global ecosystems, countries worldwide pay close attention to changes in Earth's vegetation. Currently, satellite remote sensing is the most commonly used method for monitoring vegetation information. Satellites can acquire remote sensing data covering the largest area in the shortest time through images taken from space, and this data includes vegetation information. Therefore, extracting this vegetation information from remote sensing images has become a very worthwhile research area.

[0003] Based on the spectral characteristics of vegetation indices derived from satellite remote sensing imagery, vegetation indices are obtained by combining the spectral values ​​of the visible and near-infrared bands observed in remote sensing images. These indices are derived from a simple and effective empirical measurement of surface vegetation conditions. Through continuous research and exploration by scholars both domestically and internationally, more than 40 vegetation indices have been developed to date, and they are widely used in fields such as land vegetation cover change, crop yield forecasting, regional drought monitoring, calculation of surface biophysical parameters, and global ecosystem research. Among these vegetation indices, the Normalized Difference Vegetation Index (NDVI) is currently the most widely used. It is not only easy to calculate but also eliminates most of the influences related to sensor radiometric calibration, topography, atmospheric conditions, and observation angle, enhancing its sensitivity to vegetation. After more than 20 years of research, it has accumulated rich research results, making it highly effective in the study and application of ground vegetation cover. Summary of the Invention

[0004] This invention proposes an NDVI product registration method and system, which solves the problem of splicing and registration of NDVI products in different times and spaces.

[0005] Firstly, a method for registering NDVI products is proposed, including: performing quality checks on spatiotemporal tile images and removing tiles with cloud cover exceeding a threshold, incomplete frames, or no near-infrared bands; calculating NDVI using the near-infrared and red bands of the tile images to obtain NDVI products; layering the NDVI products according to time sequence and stitching adjacent NDVI products in each layer; and registering the stitched NDVI products to obtain seamlessly aligned NDVI products.

[0006] Secondly, an NDVI product registration device is proposed, comprising: a preprocessing module configured to perform quality checks on spatiotemporal tile images, removing tiles with cloud cover exceeding a threshold, incomplete image size, or lacking a near-infrared band; an NDVI calculation module configured to calculate NDVI using the near-infrared and red bands of the tile images to obtain NDVI products; a stitching module configured to layer the NDVI products in chronological order and stitch adjacent NDVI products in each layer; and a registration module configured to register the stitched NDVI products to obtain seamlessly aligned NDVI products.

[0007] Thirdly, a computer system is proposed, comprising: a processor; a memory including one or more computer program modules; wherein the one or more computer program modules are stored in the memory and configured to be executed by the processor, and the one or more computer program modules include instructions for implementing the NDVI product registration method.

[0008] Fourthly, a computer-readable storage medium is proposed for storing non-transitory computer-readable instructions, characterized in that the non-transitory computer-readable instructions can implement the NDVI product registration method when executed by a computer.

[0009] In some examples, the NDVI products after layering and stitching are registered based on whether a reference NDVI product is available. When no reference NDVI product is available, the information entropy of each layer of NDVI products after stitching is calculated, and the NDVI product with the highest information entropy is selected as the reference NDVI product for registration of the other NDVI products. When a reference NDVI product is available, it is used to register the other NDVI products.

[0010] The present invention has the following advantages: (1) it stitches multiple NDVI products into one NDVI product, which reduces the computational complexity and improves the registration accuracy; (2) even without reference data, the present invention can still register the NDVI products of spatiotemporal tiles. Attached Figure Description

[0011] Figure 1 This is a flowchart of an NDVI product registration method / system according to an embodiment of the present invention.

[0012] Figure 2 The diagram shows the NDVI results of an embodiment of the present invention; wherein, (a) the red light band and near-infrared band in the original tile, and (b) the NDVI product.

[0013] Figure 3The above are layered spliced ​​NDVI products according to embodiments of the present invention; wherein, (a) an uncontrolled NDVI product, (b) an NDVI product divided according to time sequence, and (c) an NDVI product spliced ​​from each layer.

[0014] Figure 4 This is the NDVI registration result when there is no reference data in the embodiment of the present invention.

[0015] Figure 5 This is the NDVI registration result when reference data is available in the embodiments of the present invention. Detailed Implementation

[0016] Figure 1 This is a flowchart of an NDVI product registration method. For example... Figure 1 As shown, this invention calculates NDVI using the near-infrared and red bands of time-series tiles, and registers NDVI products based on the availability of reference data: when reference data is available, the reference NDVI is used to register the NDVI products; when no reference data is available, the information entropy of the NDVI is calculated, and the NDVI with the highest information entropy is selected to register other NDVI products. The method is described in detail below.

[0017] Step 1: Perform quality checks on the spatiotemporal tile images, removing tiles with high cloud cover, incomplete image coverage, or lacking near-infrared band coverage. For example, tiles with cloud cover exceeding 10% can be removed.

[0018] Step 2: Calculate the Normalized Difference Vegetation Index (NDVI) using the near-infrared and red bands of the tiles. The calculation formula is as follows:

[0019]

[0020] Wherein, NIR represents the reflectivity of the near-infrared band, and RED represents the reflectivity of the red band.

[0021] Figure 2 This is an NDVI result image; where, Figure 2 (a) shows the red and near-infrared bands in the original tile. Figure 2 (b) is an NDVI product.

[0022] Step 3: Layer the NDVI according to time sequence, and splice adjacent NDVI products in each layer to obtain seamlessly spliced ​​NDVI products in each layer.

[0023] Figure 3 It is a layered, spliced ​​NDVI product, in which... Figure 3 (a) is an out-of-control NDVI product. Figure 3 (b) NDVI products stratified by time sequence Figure 3 (c) is the NDVI product after each layer is spliced ​​together. Figure 3 (b) and 3(c) have 5 NDVI products: (1), (2), (3), (4), and (5).

[0024] Step 4: Register the spliced ​​NDVI to obtain a seamless, high-precision aligned NDVI product.

[0025] Step 4.1: Register the layered NDVI according to whether there is reference data.

[0026] (1) No reference data

[0027] Calculate the information entropy of each NDVI layer after stitching. Information entropy represents the average amount of information contained in an image; the higher the entropy value, the more information it contains. The formula for calculating information entropy is as follows:

[0028]

[0029] Where p(x) i ) represents a random event X = x i The probability of.

[0030] For example, for Figure 3 The information entropy of the three NDVI images in (c) that are exactly 3×3 stitched together is calculated as follows:

[0031] (2) (3) (5) Information entropy 0.6409 1.6615 0.8903

[0032] Based on the entropy value calculated in the previous step, the NDVI with the largest information entropy is selected as the reference NDVI data, and other NDVIs are registered using this as a reference.

[0033] For example, Figure 3 In (c), using the NDVI product of the third image as a reference, the other four NDVIs are registered, and the registration results are as follows: Figure 4 As shown, it can be seen that the alignment effect of the same ground feature is greatly improved after registration.

[0034] (2) Reference data is available

[0035] Use the reference NDVI to register other NDVI products. Figure 5 This is an NDVI registration result when reference data is available.

[0036] The NDVI product registration process is as follows:

[0037] First, the SIFT method is used to extract and match the corresponding points of the reference NDVI and the NDVI to be registered;

[0038] Then, registration is achieved by performing geometric transformations on the NDVI feature points to be registered using corresponding points. The method is geometric polynomial correction, and the formula is as follows:

[0039]

[0040] Among them, a i and b i Here are the polynomial coefficients, (x,y) are the row and column coordinates of the image, i = 0, 1, 2, and (X,Y) represent the corrected coordinates.

[0041] The present invention also provides an embodiment of an NDVI product registration device. This device includes a preprocessing module, an NDVI calculation module, a stitching module, and a registration module.

[0042] The preprocessing module is configured to perform quality checks on spatiotemporal tile imagery, removing tiles with cloud cover exceeding a threshold, incomplete swath, or lacking near-infrared bands. For example, tiles with cloud cover exceeding 10% can be removed.

[0043] The NDVI calculation module is configured to calculate the NDVI using the near-infrared and red bands of the tile image to obtain the NDVI product. The calculation formula is the same as formula (1) in step 2.

[0044] The splicing module is configured to layer NDVI products in sequence and splice adjacent NDVI products in each layer.

[0045] The registration module is configured to register the stitched NDVI products to obtain seamlessly aligned NDVI products. Refer to step 4 for the specific registration method of the registration module.

[0046] The present invention also provides an embodiment of a computer system. The computer system includes a processor and a memory. The memory is used to store non-transitory computer-readable instructions (e.g., one or more computer program modules). The processor is used to execute the non-transitory computer-readable instructions, which, when executed by the processor, can perform one or more steps in the NDVI product registration method described above. The memory and the processor can be interconnected via a bus system and / or other forms of connection mechanisms.

[0047] For example, a processor can be a central processing unit (CPU), a graphics processing unit (GPU), or other form of processing unit with data processing and / or program execution capabilities. For instance, a CPU can be based on x86 or ARM architectures. A processor can be a general-purpose processor or a special-purpose processor, and it can control other components in the computer to perform desired functions.

[0048] For example, memory can include any combination of one or more computer program products, which can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory can include, for example, random access memory (RAM) and / or cache memory. Non-volatile memory can include, for example, read-only memory (ROM), hard disk, erasable programmable read-only memory (EPROM), compact optical disc read-only memory (CD-ROM), USB storage, flash memory, etc. One or more computer program modules can be stored on the computer-readable storage medium, and the processor can run one or more computer program modules to implement various functions of the computer.

[0049] This invention also provides a computer-readable storage medium for storing non-transitory computer-readable instructions. When executed by a computer, these instructions can implement one or more steps in the NDVI product registration method described above. When the NDVI product registration method and apparatus provided in this invention are implemented in software and sold or used as independent products, they can be stored in a computer-readable storage medium. For further details regarding the storage medium, please refer to the corresponding description of memory in the computer system above; further elaboration is not provided here.

Claims

1. A method for registering NDVI products, characterized in that, include: Perform quality checks on spatiotemporal tile images and remove tiles with cloud cover exceeding the threshold, incomplete image size, or no near-infrared band. NDVI is calculated using the near-infrared and red bands of the tile images after quality inspection, and the NDVI product is obtained. The NDVI products are layered according to time sequence, and adjacent NDVI products in each layer are spliced ​​together. The spliced ​​NDVI products are registered to obtain seamlessly aligned NDVI products. Specifically, the registration of the layered spliced ​​NDVI products is performed based on whether there is a reference NDVI product. When there is no reference NDVI product, the information entropy of each layer of NDVI products after splicing is calculated, and the NDVI product with the largest information entropy is selected as the reference NDVI product for registration of other NDVI products. When there is a reference NDVI product, the reference NDVI product is used to register other NDVI products.

2. The NDVI product registration method according to claim 1, characterized in that, The SIFT method is used to extract and match corresponding points between the reference NDVI product and the NDVI product to be registered; registration is then achieved by performing a geometric transformation on the feature points of the NDVI product to be registered using these corresponding points. The formula for the geometric transformation is as follows: in, a i and b i The coefficients are polynomial coefficients, i = 0, 1, 2, ( x , y ) represents the row and column coordinates of the image. X , Y () indicates the corrected coordinates.

3. An NDVI product registration device, characterized in that, include: The preprocessing module is configured to perform quality checks on spatiotemporal tile images, removing tiles with cloud cover exceeding a threshold, incomplete swaths, and no near-infrared bands. The NDVI calculation module is configured to calculate the NDVI using the near-infrared and red bands of the quality-checked tile image to obtain the NDVI product. The splicing module is configured to layer NDVI products in time sequence and splice adjacent NDVI products in each layer. The registration module is configured to register the stitched NDVI products to obtain seamlessly aligned NDVI products. Specifically, the registration of the layered NDVI products is performed based on whether there is a reference NDVI product. When there is no reference NDVI product, the information entropy of each layer of NDVI products after stitching is calculated, and the NDVI product with the largest information entropy is selected as the reference NDVI product for registration of other NDVI products. When there is a reference NDVI product, the reference NDVI product is used to register other NDVI products.

4. The NDVI product registration device according to claim 3, characterized in that, The SIFT method is used to extract and match corresponding points between the reference NDVI product and the NDVI product to be registered; registration is then achieved by performing a geometric transformation on the feature points of the NDVI product to be registered using these corresponding points. The formula for the geometric transformation is as follows: in, a i and b i For polynomial coefficients, i =0,1,2,( x , y ) represents the row and column coordinates of the image. X , Y () indicates the corrected coordinates.

5. A computer system, characterized in that, include: processor; Memory, including one or more computer program modules; The one or more computer program modules are stored in the memory and configured to be executed by the processor, and the one or more computer program modules include instructions for implementing the NDVI product registration method according to any one of claims 1-2.

6. A computer-readable storage medium for storing non-transitory computer-readable instructions, characterized in that, When the non-transitory computer-readable instructions are executed by a computer, the NDVI product registration method according to any one of claims 1-2 can be implemented.

Images