Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Polarization SAR non-local noise reduction method based on dual feature cluster homogeneous pixel selection

A non-local, pixel-based technology, applied in image enhancement, image analysis, image data processing, etc., can solve the problem of insufficient maintenance of polarimetric SAR image structure information and scattering information, and achieve the effect of suppressing coherent speckle

Inactive Publication Date: 2017-06-20
HEFEI UNIV OF TECH
View PDF5 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a polarimetric SAR non-local denoising method based on dual-feature clustering homogeneous pixel selection, to improve the accuracy of homogeneous pixel selection, and to solve the problem of polarimetric SAR image structure information and scattering information in the prior art. problem of keeping insufficient

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Polarization SAR non-local noise reduction method based on dual feature cluster homogeneous pixel selection
  • Polarization SAR non-local noise reduction method based on dual feature cluster homogeneous pixel selection
  • Polarization SAR non-local noise reduction method based on dual feature cluster homogeneous pixel selection

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0043] A non-local noise reduction method for polarimetric SAR based on dual-feature clustering and homogeneous pixel selection, including the following steps:

[0044] (1), select homogeneous pixels for each pixel, the specific steps are as follows:

[0045] 1a), calculate the Span diagram of the polarimetric SAR covariance matrix data;

[0046] 1b), for the i-th image block vector containing the d-dimensional space, calculate the mean shift vector M containing the statistical distance (PPB) of the block on the Span map h,g (x i ):

[0047]

[0048] In the formula:

[0049]

[0050] P(x)=c p,d p(||x|| 2 ),

[0051] g(x)=-p'(x),

[0052] x i,k is the i-th image block vector x i The kth data of x j,k is the jth image block vector x j The kth data of p(||x|| 2 ) is a Gaussian kernel function, c p,d is a normalization constant, h is the bandwidth;

[0053]1c), in the kernel function window, calculate the mean shift vectors of all feature space data points sequ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a polarization SAR non-local noise reduction method based on dual feature cluster homogeneous pixel selection. The method comprises the steps that pixels with similar structural features are selected through mean shift clustering based on the statistical distance of blocks; a Wishart unsupervised classifier based on a Freeman-Durden scattering model is used to select pixels with similar scattering mechanisms; a homogeneous pixel set is composed of pixels containing similar structural features and scattering mechanisms; and according to the homogeneous pixel set, original polarization SAR data are combined to calculate non-local weight weighted sums to realize polarization SAR non-local mean value filtering. According to the invention, coherent speckle noise is effectively suppressed; image structure information and polarization scattering information can be further maintained; and the method can be used for polarization SAR image filtering.

Description

technical field [0001] The invention relates to the field of image filtering algorithms, in particular to a polarization SAR non-local noise reduction method based on double-feature clustering and homogeneous pixel selection. Background technique [0002] The coherence speckle is an intrinsic interaction between polarized electromagnetic waves and physical scatterers. This interplay induces pixel-to-pixel variations that appear as speckle noise on the intensity values. Therefore, speckle suppression is especially important for image segmentation and feature classification of polarimetric SAR data. [0003] In the past few decades, researchers have proposed a variety of filtering algorithms. The simple and commonly used boxcar filter does not work well because of the indiscriminate averaging within the window. The polarimetric whitening filter is only available for single-look polarimetric SAR scattering matrices. It was developed for multi-view data by Liu et al. Lee et...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06T5/00
CPCG06T2207/10044G06T5/70
Inventor 周芳刘留杨学志许开炜
Owner HEFEI UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com