Search

CN-121010755-B - Wind turbine SAR identification method for complex tidal flat underlying surface

CN121010755BCN 121010755 BCN121010755 BCN 121010755BCN-121010755-B

Abstract

The invention relates to a wind turbine SAR identification method aiming at a tidal flat complex underlying surface, which comprises the steps of generating an annual SAR time sequence synthesized image, calculating the mean value of each pixel in the synthesized image in a preset backscattering coefficient interval, generating a high backscattering value image, carrying out binarization segmentation on the high backscattering value image, carrying out morphological processing on the binarized image to strengthen a high backscattering target, establishing a geometric feature-based multi-scale screening mechanism, carrying out differential spatial resolution processing, and outputting wind turbine point vector data. The method has the beneficial effects that the identification accuracy of tidal beaches and open sea wind turbines is effectively improved, and reliable technical support is provided for the fine assessment and management of offshore wind power resources.

Inventors

  • WANG LIHUA
  • Bao Yanxiao
  • JIANG WENJING
  • SUN WEIWEI
  • YANG GANG

Assignees

  • 宁波大学

Dates

Publication Date
20260512
Application Date
20250811

Claims (8)

  1. 1. A method of wind turbine SAR identification for tidal flat complex underlying surfaces, comprising: Step 1, acquiring SAR image data of an annual time sequence, and preprocessing to generate SAR backward scattering coefficients, and screening VV polarized images of a target year from an SAR image set to generate an annual SAR time sequence synthesized image; Step 2, calculating the average value of each pixel in the synthesized image in a preset backscattering coefficient interval, and generating an image with a high backscattering value so as to eliminate a temporary moving target and a drifting target; step 3, performing binarization segmentation on the image with the high backscattering value based on a threshold T to generate a binarized image; Step 4, performing morphological processing on the binarized image to enhance the high back scattering target; step 5, establishing a multi-scale screening mechanism based on geometric features, and eliminating interference targets and noise points; in step 5, the pixel statistics of the connected areas is carried out on the binarized image after morphological processing to determine the size of each connected area, the maximum pixel statistics upper limit of a single connected area is set to be 256, if the pixel number of a certain area exceeds the value, the cut-off processing is carried out according to 256, in addition, an 8 neighborhood connection mode is adopted, namely, one pixel and adjacent pixels in the upper, lower, left and right directions and four 6 diagonal directions of the pixel are regarded as connected areas; The 8 neighborhood expression is as follows: For pixel coordinates The 8 neighborhood thereof contains pixels in the surrounding 8 directions, i.e. all adjacent pixels in the lateral, longitudinal, diagonal directions, expressed as: And 6, performing differential spatial resolution processing, and outputting wind turbine point vector data.
  2. 2. The method of identifying a wind turbine SAR for a complex tidal flat underlying surface of claim 1, further comprising: And 7, evaluating the performance of the SAR identification wind turbine by adopting indexes of miss score, correct identification, precision, recall rate and F1.
  3. 3. The method of wind turbine SAR identification for tidal flat complex undersea according to claim 2, wherein in step 1, said preprocessing comprises noise removal and radiometric scaling.
  4. 4. The method for identifying the SAR of the wind turbine for the tidal flat complex underlying surface according to claim 3, wherein in the step 3, a grid self-adaptive backscatter filter is adopted, a backscatter coefficient mean value in a preset backscatter coefficient interval range is compared with a threshold value T, a binarized image is generated, and the expression of the threshold value T is as follows: Wherein, the Is the maximum value of the backscattering coefficient in the grid, Is the minimum value of the backscattering coefficient in the grid.
  5. 5. The method of identifying a wind turbine SAR for a tidal flat complex underlying surface according to claim 4, wherein in step 4, the morphological processing comprises an expansion operation and a corrosion operation.
  6. 6. A wind turbine SAR identification system for tidal flat complex undersea, characterized by performing the method of any of claims 1 to 5, comprising: The acquisition module is used for acquiring SAR image data of the annual time sequence, preprocessing the SAR image data to generate SAR backward scattering coefficients, and screening VV polarized images of target years from the SAR image set to generate annual SAR time sequence synthesized images; The calculation module is used for calculating the average value of each pixel in the synthesized image in a preset backscattering coefficient interval and generating an image with a high backscattering value so as to eliminate a temporary moving target and a drifting target; the segmentation module is used for carrying out binarization segmentation on the high backscattering value image based on the threshold T to generate a binarized image; The processing module is used for carrying out morphological processing on the binarized image so as to enhance the high backscattering target; the establishing module is used for establishing a multi-scale screening mechanism based on geometric features and eliminating interference targets and noise points; The implementation module is used for implementing differential spatial resolution processing and outputting wind turbine point vector data.
  7. 7. A computer storage medium, characterized in that the computer storage medium has stored therein a computer program which, when run on a computer, causes the computer to perform the method of any of claims 1 to 5.
  8. 8. An electronic device, comprising: A memory for storing a computer program; a processor for executing the computer program to implement the method of any one of claims 1 to 5.

Description

Wind turbine SAR identification method for complex tidal flat underlying surface Technical Field The invention belongs to the technical field of remote sensing image processing, and particularly relates to a wind turbine SAR identification method for a tidal flat complex underlying surface. Background At present, an offshore wind turbine identification technology based on Synthetic Aperture Radar (SAR) image features mainly has shown remarkable effects in application in open sea areas by extracting physical features such as backscattering coefficients, polarization characteristics and the like of targets and analyzing the physical features in combination with texture statistical features. However, when the technology is applied to tidal flat areas, a plurality of technical challenges are faced, and the main reason is that the tidal flat environment has unique complexity, the underlying surface of the tidal flat environment is remarkably different from the open sea area, the target form presents diversified characteristics, and meanwhile, the tidal dynamic change, the sand migration, the vegetation coverage and other multiple factors are interfered, so that the misjudgment rate and the omission rate of the traditional identification method in the areas are higher. Disclosure of Invention The invention aims to overcome the defects in the prior art and provides a wind turbine SAR identification method aiming at complex tidal flat underlying surfaces. In a first aspect, a method for wind turbine SAR identification for tidal flat complex underlying surfaces is provided, comprising: Step 1, acquiring SAR image data of an annual time sequence, and preprocessing to generate SAR backward scattering coefficients, and screening VV polarized images of a target year from an SAR image set to generate an annual SAR time sequence synthesized image; Step 2, calculating the average value of each pixel in the synthesized image in a preset backscattering coefficient interval, and generating an image with a high backscattering value so as to eliminate a temporary moving target and a drifting target; step 3, performing binarization segmentation on the image with the high backscattering value based on a threshold T to generate a binarized image; Step 4, performing morphological processing on the binarized image to enhance the high back scattering target; step 5, establishing a multi-scale screening mechanism based on geometric features, and eliminating interference targets and noise points; And 6, performing differential spatial resolution processing, and outputting wind turbine point vector data. Preferably, the method further comprises: And 7, evaluating the performance of the SAR identification wind turbine by adopting indexes of miss score, correct identification, precision, recall rate and F1. Preferably, in step 1, the preprocessing includes noise removal and radiometric scaling. Preferably, in step 3, a grid adaptive backscatter filter is adopted, and a backscatter coefficient mean value in a preset backscatter coefficient interval range is compared with a threshold T to generate a binarized image, wherein the expression of the threshold T is as follows: Wherein σ max is the intra-grid backscattering coefficient maximum value, and σ min is the intra-grid backscattering coefficient minimum value. Preferably, in step 4, the morphological treatment includes an expansion operation and a corrosion operation. Preferably, in step 5, pixel statistics of connected regions is performed on the binarized image after morphological processing, so as to determine the size of each connected region. In a second aspect, there is provided a wind turbine SAR identification system for tidal flat complex underlying surfaces for performing the method of any of the first aspects, comprising: The acquisition module is used for acquiring SAR image data of the annual time sequence, preprocessing the SAR image data to generate SAR backward scattering coefficients, and screening VV polarized images of target years from the SAR image set to generate annual SAR time sequence synthesized images; The calculation module is used for calculating the average value of each pixel in the synthesized image in a preset backscattering coefficient interval and generating an image with a high backscattering value so as to eliminate a temporary moving target and a drifting target; the segmentation module is used for carrying out binarization segmentation on the high backscattering value image based on the threshold T to generate a binarized image; The processing module is used for carrying out morphological processing on the binarized image so as to enhance the high backscattering target; the establishing module is used for establishing a multi-scale screening mechanism based on geometric features and eliminating interference targets and noise points; The implementation module is used for implementing differential spatial resolution processing and outputting wind turb