CN-121978038-A - River water quality dynamic identification method based on hyperspectral inversion

Abstract

The invention provides a river water quality dynamic identification method based on hyperspectral inversion, which comprises the steps of obtaining space base map data of a river reach, and extracting a hyperspectral acquisition region from the space base map data of the river reach so as to construct a space monitoring unit for navigation disturbance perception based on the hyperspectral acquisition region. And monitoring spectrum data of different vertical water body levels and the fluctuation range of the shipping water body in the river reach space monitoring unit, and generating vertical spectrum distribution according to the spectrum data and the fluctuation range of the shipping water body. And carrying out space recombination on spectrum data inside and outside the fluctuation range of the shipping water body according to the vertical spectrum distribution so as to construct a water body spectrum characterization unit corresponding to each river space monitoring unit. And mapping the water spectrum characterization unit into the space base map data of the river reach to generate a water quality space distribution map. In the air traffic disturbance river reach, a water quality identification mechanism capable of identifying the disturbance state and stripping the influence of the disturbance is established, so that the hyperspectral inversion result can truly reflect the space process of pollution diffusion.

Inventors

• CHEN GUIHONG
• LIU JIANGFENG

Assignees

• 江苏创威电子有限公司

Dates

Publication Date

20260505

Application Date

20260407

Claims (10)

1. 1. A river water quality dynamic identification method based on hyperspectral inversion, which is characterized by comprising the following steps: acquiring space base map data of a river reach, and extracting a hyperspectral acquisition region from the space base map data of the river reach to construct a space monitoring unit for sensing shipping disturbance based on the hyperspectral acquisition region; Monitoring spectrum data of different vertical water body levels and the fluctuation range of the shipping water body in the river reach space monitoring unit, and generating vertical spectrum distribution according to the spectrum data and the fluctuation range of the shipping water body; carrying out space recombination on spectrum data inside and outside a fluctuation range of a shipping water body according to the vertical spectrum distribution so as to construct a water body spectrum characterization unit corresponding to each river space monitoring unit; And mapping the water spectrum characterization unit into the space base map data of the river reach to generate a water quality space distribution map.
2. 2. The method for dynamically identifying river water quality based on hyperspectral inversion of claim 1, wherein the steps of obtaining the space base map data of the river reach and extracting hyperspectral collection areas from the space base map data of the river reach to construct a space monitoring unit for sensing the shipping disturbance based on the hyperspectral collection areas comprise: cutting the space base map data of the river reach to extract the hyperspectral acquisition region and determining the corresponding river channel center line in the hyperspectral acquisition region; Discretizing the center line of the river channel according to a set step length to obtain a plurality of discrete points, and expanding a river channel region between adjacent discrete points to generate a river channel plane grid; And acquiring a pre-stored ship passing path, ship draft and river channel water depth distribution map, and superposing the ship passing path, the ship draft and the river channel water depth distribution map into the corresponding river channel plane grids to determine the shipping activity sensitivity corresponding to each river channel plane grid.
3. 3. The method for dynamically identifying river water quality based on hyperspectral inversion of claim 2, wherein the steps of obtaining space base map data of a river reach and extracting hyperspectral acquisition areas from the space base map data of the river reach to construct a space monitoring unit for sensing navigation disturbance based on the hyperspectral acquisition areas, further comprise: When the shipping activity sensitivity exceeds a set sensitivity threshold, judging the corresponding river reach plane grid as hyperspectral acquisition areas, and vertically layering each hyperspectral acquisition area to obtain a plurality of vertical water body levels; Allocating a water spectrum acquisition identifier to each vertical water level, and setting spectrum acquisition priority for the water spectrum acquisition identifier according to the corresponding position of each vertical water level and shipping activity sensitivity; And nesting the vertical water body level and the corresponding river reach plane grid to construct the river reach space monitoring unit.
4. 4. The method for dynamically identifying river water quality based on hyperspectral inversion as recited in claim 3, wherein the monitoring of the spectral data and the fluctuation range of the shipping water body for different vertical water body levels in the river reach space monitoring unit and the generation of the vertical spectral distribution according to the spectral data and the fluctuation range of the shipping water body comprises the steps of: Sequencing the river reach space monitoring units according to the sequence of the spectrum acquisition priority, and setting a time sequence organization reference and different disturbance acquisition segments to construct a vertical acquisition time sequence frame based on the time sequence organization reference, the disturbance acquisition segments and the sequenced river reach space monitoring units; And carrying out continuous spectrum acquisition on each vertical water body level in each river reach space monitoring unit in the vertical acquisition time sequence frame, and generating a vertical spectrum acquisition sequence according to the sequence of acquisition time stamps.
5. 5. The method for dynamically identifying river water quality based on hyperspectral inversion of claim 4, wherein the monitoring of the spectral data and the fluctuation range of the shipping water body for different vertical water body levels in the space monitoring unit of the river reach and generating vertical spectral distribution according to the spectral data and the fluctuation range of the shipping water body further comprises: extracting spectral data corresponding to adjacent acquisition time stamps of the same river reach space monitoring unit from the vertical spectrum acquisition sequence, and identifying water body billowing boundaries before and after ship passing according to the spectral data so as to construct a billowing boundary time sequence based on the water body billowing boundaries; And converting the spectral data at two sides of the water body billowing boundary into water body transparency change ranges corresponding to different vertical water body levels at the same time, and carrying out space calibration on the water body transparency change ranges of each vertical water body level in the shipping water body fluctuation range according to the billowing boundary time sequence to obtain the vertical spectral distribution.
6. 6. The method for dynamically identifying river water quality based on hyperspectral inversion according to claim 1, wherein the spatial reorganization is performed on the spectral data inside and outside the fluctuation range of the shipping water body according to the vertical spectral distribution to construct a water body spectrum characterization unit corresponding to each river segment spatial monitoring unit, and the method comprises the following steps: The vertical spectrum distribution is collected to corresponding river reach space monitoring units, and spectrum data are divided into billowing area spectrum data and non-billowing area spectrum data in each river reach space monitoring unit, so that a level disturbance separation sequence corresponding to each river reach space monitoring unit is constructed; Performing piecewise iterative comparison on the spectral data of the billowing region and the spectral data of the non-billowing region in the hierarchical disturbance separation sequence, and determining a hierarchical disturbance stripping coefficient when the spectral data of the billowing region deviates from the spectral data of the non-billowing region; And carrying out space recombination on the spectral data of the billowing region and the spectral data of the non-billowing region according to the hierarchical disturbance stripping coefficient to obtain recombined hierarchical spectral fragments, and smoothly splicing the hierarchical spectral fragments corresponding to the same river reach space monitoring unit at a plurality of continuous moments to obtain the water body spectrum characterization unit.
7. 7. The method for dynamically identifying water quality in a river based on hyperspectral inversion of claim 1, wherein the mapping the water body spectrum characterization unit into the space base map data of the river reach to generate a water quality space distribution map comprises: writing the water spectrum characterization units into the corresponding positions of the space base map data of the river reach layer by layer, and calculating the spectrum intensity errors corresponding to the boundary areas between the adjacent river reach space monitoring units; When the spectrum intensity error exceeds a set error threshold, carrying out sectional smooth transition on the boundary region to obtain a layered spectrum reconstruction base map; Reconstructing vertical water body levels in each river reach space monitoring unit based on the layered spectrum reconstruction base map so as to establish an interlayer transition region between adjacent vertical water body levels, and performing smoothing treatment on the spectrum intensity change of the interlayer transition region to obtain a water quality level distribution map.
8. 8. The method for dynamically identifying water quality in a river based on hyperspectral inversion of claim 7, wherein the mapping the water body spectrum characterization unit into the space base map data of the river reach to generate a water quality space distribution map further comprises: acquiring prestored historical water quality distribution records, and constructing a water quality time sequence change chain according to the historical water quality distribution records corresponding to different times so as to determine a water quality diffusion candidate path between the water quality level distribution map and the historical water quality distribution records through the water quality time sequence change chain; calculating the credibility grade corresponding to the water quality diffusion candidate path, and screening the water quality diffusion candidate path according to the credibility grade to reserve the water quality diffusion candidate path with the credibility grade exceeding a set grade threshold as the water quality diffusion path; And superposing the water quality diffusion paths at corresponding positions in the space base map data of the river reach to generate the water quality space distribution map.
9. 9. A terminal comprises a processor and a storage medium, and is characterized in that: The storage medium is used for storing instructions; The processor being operative according to the instructions to perform the steps of the method according to any one of claims 1-8.
10. 10. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method according to any one of claims 1-8.

Description

River water quality dynamic identification method based on hyperspectral inversion Technical Field The disclosure belongs to the field of visual intelligence of river water quality, and more particularly relates to a river water quality dynamic identification method based on hyperspectral inversion. Background At present, the disturbance of the propeller of a large ship can cause the vertical strong mixing of the water body, so that the phenomenon of high turbidity layer billowing and transparent layer alternation occurs in the water body in a short time. The transient vertical structural change is easy to cause the spectrum reflection characteristic to be in a highly unstable state, so that the actual pollution diffusion process is difficult to be truly expressed by the spatial water quality distribution map. However, in the prior art, a water quality identification mechanism for identifying the disturbance state and stripping the disturbance influence is difficult to establish aiming at the water body vertical strong mixing and transient layered turning phenomenon caused by ship propellers in the navigation disturbance river reach, so that the spatial process of the diffusion evolution of water quality in the scene is difficult to identify. Disclosure of Invention In order to solve the defects in the prior art, the invention aims to solve the defects, and further provides a river water quality dynamic identification method based on hyperspectral inversion. The invention adopts the following technical scheme. The invention discloses a river water quality dynamic identification method based on hyperspectral inversion, which comprises the following steps: acquiring space base map data of a river reach, and extracting a hyperspectral acquisition region from the space base map data of the river reach to construct a space monitoring unit for sensing shipping disturbance based on the hyperspectral acquisition region; Monitoring spectrum data of different vertical water body levels and the fluctuation range of the shipping water body in the river reach space monitoring unit, and generating vertical spectrum distribution according to the spectrum data and the fluctuation range of the shipping water body; carrying out space recombination on spectrum data inside and outside a fluctuation range of a shipping water body according to the vertical spectrum distribution so as to construct a water body spectrum characterization unit corresponding to each river space monitoring unit; And mapping the water spectrum characterization unit into the space base map data of the river reach to generate a water quality space distribution map. Further, the obtaining the space base map data of the river reach and extracting a hyperspectral acquisition region from the space base map data of the river reach to construct a space monitoring unit for sensing the shipping disturbance based on the hyperspectral acquisition region comprises: cutting the space base map data of the river reach to extract the hyperspectral acquisition region and determining the corresponding river channel center line in the hyperspectral acquisition region; Discretizing the center line of the river channel according to a set step length to obtain a plurality of discrete points, and expanding a river channel region between adjacent discrete points to generate a river channel plane grid; And acquiring a pre-stored ship passing path, ship draft and river channel water depth distribution map, and superposing the ship passing path, the ship draft and the river channel water depth distribution map into the corresponding river channel plane grids to determine the shipping activity sensitivity corresponding to each river channel plane grid. Further, the method for obtaining the space base map data of the river reach, extracting a hyperspectral acquisition region from the space base map data of the river reach, and constructing a space monitoring unit for sensing shipping disturbance based on the hyperspectral acquisition region, further comprises the following steps: When the shipping activity sensitivity exceeds a set sensitivity threshold, judging the corresponding river reach plane grid as hyperspectral acquisition areas, and vertically layering each hyperspectral acquisition area to obtain a plurality of vertical water body levels; Allocating a water spectrum acquisition identifier to each vertical water level, and setting spectrum acquisition priority for the water spectrum acquisition identifier according to the corresponding position of each vertical water level and shipping activity sensitivity; And nesting the vertical water body level and the corresponding river reach plane grid to construct the river reach space monitoring unit. Further, the monitoring of the spectrum data and the fluctuation range of the shipping water body of different vertical water body levels in the river reach space monitoring unit, and generating vertical spectrum distribution accordi