Remote sensing time sequence measurement technology for suspended sediment in middle and lower reaches of Yangtze River

Abstract

The invention discloses a remote sensing time sequence measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River. The technology comprises the following steps: S1, acquiring multi-source remote sensing image data covering the middle and lower reaches of the Yangtze River; s2, preprocessing the acquired remote sensing image data; s3, feature extraction is carried out on the preprocessed data, and spectral feature parameters related to the suspended sediment concentration are obtained; s4, constructing a neural network model suitable for the inversion of the suspended sediment concentration in the middle and lower reaches of the Yangtze River, and obtaining the suspended sediment concentration corresponding to each remote sensing image based on the inversion of the spectral characteristic parameters in S3; and S5, sorting the suspended sediment concentration data obtained by inversion according to a time sequence, and generating remote sensing time sequence data of the suspended sediment in the middle and lower reaches of the Yangtze River. Compared with the prior art, the remote sensing time sequence measurement technology for the suspended sediment in the middle and lower reaches of the Yangtze River is convenient to operate and use and ensures data accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental remote sensing monitoring, and in particular to a remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River. Background Art

[0002] Suspended sediment is one of the most important water quality parameters. Its concentration and the patterns of its deposition and movement directly impact estuary management and flood control safety. Conventional methods for measuring suspended sediment involve collecting water samples on-site, filtering them, drying them, weighing them, and calculating their concentration. This method is time-consuming and labor-intensive, and struggles to meet the requirements of large-scale monitoring. Remote sensing technology, with its macroscopic, large-scale monitoring capabilities, can provide simultaneous remote sensing images of large areas of water, offering unique advantages in water environment monitoring.

[0003] However, the current traditional single-source remote sensing inversion method is difficult to adapt to the complex water and sediment environment in the middle and lower reaches of the Yangtze River and fails to effectively integrate spatial characteristics and time series correlations, resulting in insufficient data accuracy. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the above technical defects and provide a remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River that is easy to operate and use and ensures accurate data.

[0005] To solve the above technical problems, the present invention provides a technical solution: a remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River, comprising the following steps: S1: acquiring multi-source remote sensing image data covering the middle and lower reaches of the Yangtze River; S2: preprocessing the acquired remote sensing image data; S3: performing feature extraction on the preprocessed data to obtain spectral characteristic parameters related to the suspended sediment concentration; S4: constructing a neural network model suitable for inverting the suspended sediment concentration in the middle and lower reaches of the Yangtze River, and inverting the spectral characteristic parameters in S3 to obtain the suspended sediment concentration corresponding to each remote sensing image; S5: arranging the inverted suspended sediment concentration data in chronological order to generate remote sensing time series data of suspended sediment in the middle and lower reaches of the Yangtze River.

[0006] Preferably, the image data in S1 includes remote sensing images of different platforms, different spectral resolutions and different temporal resolutions;

[0007] Said platforms include satellite impact, drone hyperspectral imagery, and drone hyperspectral imagery;

[0008] The time mentioned includes the flood season, dry season and flood season.

[0009] Preferably, the preprocessing in S2 includes radiation calibration, atmospheric correction, and geometric correction.

[0010] Preferably, the step S3 includes calculating the reflectivity ratio of the pre-processed remote sensing image in the red and near-infrared bands, the normalized difference vegetation index in the blue, green and red bands, and extracting the slope and curvature of the reflectivity curve of the remote sensing image in different bands;

[0011] The extracted parameters are used as spectral characteristic parameters related to suspended sediment concentration.

[0012] Preferably, the neural network model in S4 adopts a dual-channel deep learning architecture, including a spatial feature extraction channel and a temporal association analysis channel, and dynamically allocates the weight ratio of spatial and temporal features through an attention mechanism.

[0013] Preferably, the neural network model is pre-trained based on a public data set, and measured data from the Yangtze River Basin is used in the fine-tuning stage.

[0014] The advantages of the present invention compared with the existing technology are: constructing a technical system of collaborative observation-hybrid deep learning modeling-spatiotemporal dynamic reconstruction, solving the problem of insufficient spatiotemporal coverage of traditional single-source remote sensing based on satellite-UAV-ground observation, realizing full-time monitoring during the flood season, dry season and flood season, and ensuring data reliability through a dual-channel deep learning model. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a structural diagram of the patent name of this invention. DETAILED DESCRIPTION

[0016] The present invention will be described in further detail below with reference to the accompanying drawings.

[0017] Combined with attachment Figure 1 As shown in FIG, the remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River includes the following steps: S1: acquiring multi-source remote sensing image data covering the middle and lower reaches of the Yangtze River; S2: preprocessing the acquired remote sensing image data; S3: performing feature extraction on the preprocessed data to obtain spectral characteristic parameters related to the suspended sediment concentration; S4: constructing a neural network model suitable for inverting the suspended sediment concentration in the middle and lower reaches of the Yangtze River, and inverting the spectral characteristic parameters in S3 to obtain the suspended sediment concentration corresponding to each remote sensing image; S5: arranging the inverted suspended sediment concentration data in chronological order to generate remote sensing time series data of suspended sediment in the middle and lower reaches of the Yangtze River.

[0018] The image data in S1 include remote sensing images from different platforms, different spectral resolutions and different temporal resolutions; the platforms include satellite images, drone hyperspectral images and drone hyperspectral images; the time includes the wet season, dry season and flood season, and the preprocessing includes radiation calibration, atmospheric correction and geometric correction.

[0019] For the acquired preprocessed data, the reflectance ratio of the preprocessed remote sensing image in the red and near-infrared bands, the normalized difference vegetation index in the blue, green and red bands are calculated, and the slope and curvature of the reflectance curve of the remote sensing image in different bands are extracted. The extracted parameters are used as spectral characteristic parameters related to suspended sediment concentration.

[0020] The neural network model adopts a dual-channel deep learning architecture, including a spatial feature extraction channel and a temporal association analysis channel. It dynamically allocates the weight ratio of spatial and temporal features through the attention mechanism. The neural network model is pre-trained based on a public dataset, and measured data from the Yangtze River Basin is used in the fine-tuning stage.

[0021] In a specific implementation of the present invention, multi-source remote sensing image data covering the middle and lower reaches of the Yangtze River is acquired, the multi-source remote sensing image data including remote sensing images from different satellite platforms, with different spectral resolutions and different temporal resolutions. The acquired multi-source remote sensing image data is preprocessed, including radiometric calibration, atmospheric correction, and geometric correction. Based on the preprocessed remote sensing image data, spectral characteristic parameters related to suspended sediment concentration are extracted.

[0022] Among them, radiation calibration converts the pixel brightness value of the remote sensing image into apparent reflectance or radiance. Atmospheric correction adopts a method based on the radiation transfer model to remove the influence of scattering and absorption of atmospheric molecules and aerosols on the radiation information of the remote sensing image. Geometric correction uses high-precision digital elevation models and ground control points as a benchmark to perform geometric deformation correction on remote sensing images, so that remote sensing images from different sources have unified geographic coordinates and projections.

[0023] The measured suspended sediment concentration data were collected from multiple sampling points in the middle and lower reaches of the Yangtze River, and pre-processed remote sensing image data of the corresponding sampling points during the same period were obtained. The measured suspended sediment concentration data and the corresponding spectral characteristic parameters were combined into a training data set. The convolutional neural network was trained using the training data set and the parameters of the neural network were adjusted.

[0024] The inverted suspended sediment concentration data were sorted in chronological order to generate the suspended sediment remote sensing time series data for the middle and lower reaches of the Yangtze River.

[0025] The contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0026] The basic principles, main features and advantages of the present invention are shown and described above. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention, and such changes and modifications fall within the scope of the invention as claimed.

[0027] In the present invention, unless otherwise expressly specified or limited, a first feature being "above" or "below" a second feature may include the first and second features being in direct contact, or may include the first and second features being in contact not directly but through another feature between them. Moreover, a first feature being "above," "above," and "above" a second feature includes the first feature being directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. A first feature being "below," "below," and "below" a second feature includes the first feature being directly below and obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0028] The present invention and its embodiments are described above. This description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if a person skilled in the art is inspired by this and, without departing from the purpose of the present invention, designs structures and embodiments similar to this technical solution without inventiveness, they shall fall within the scope of protection of the present invention.

Claims

1. Remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River, characterized by: The following steps are involved: S1: Acquire multi-source remote sensing image data covering the middle and lower reaches of the Yangtze River; S2: Preprocess the acquired remote sensing image data; S3: Extract features from the preprocessed data to obtain spectral characteristic parameters related to suspended sediment concentration; S4: Build a neural network model suitable for the inversion of suspended sediment concentration in the middle and lower reaches of the Yangtze River. Based on the spectral characteristic parameters in S3, the suspended sediment concentration corresponding to each remote sensing image is obtained. S5: Arrange the inverted suspended sediment concentration data in chronological order to generate the remote sensing time series data of suspended sediment in the middle and lower reaches of the Yangtze River.

2. The remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River according to claim 1 is characterized by: The image data in S1 includes remote sensing images of different platforms, different spectral resolutions and different temporal resolutions; Said platforms include satellite impact, drone hyperspectral imagery, and drone hyperspectral imagery; The time mentioned includes the flood season, dry season and flood season.

3. The remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River according to claim 1 is characterized by: The preprocessing in S2 includes radiation calibration, atmospheric correction, and geometric correction.

4. The remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River according to claim 1 is characterized by: Said S3 includes calculating the reflectivity ratio of the pre-processed remote sensing image in the red and near-infrared bands, the normalized difference vegetation index in the blue, green and red bands, and extracting the slope and curvature of the reflectivity curve of the remote sensing image in different bands; The extracted parameters are used as spectral characteristic parameters related to suspended sediment concentration.

5. The remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River according to claim 1 is characterized by: The neural network model in S4 adopts a dual-channel deep learning architecture, including a spatial feature extraction channel and a temporal association analysis channel, and dynamically allocates the weight ratio of spatial and temporal features through the attention mechanism.

6. The remote sensing time series measurement technology for suspended sediment in the middle and lower reaches of the Yangtze River according to claim 5 is characterized by: The neural network model is pre-trained based on a public dataset, and measured data from the Yangtze River Basin is used in the fine-tuning stage.

Images