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

SAR full-image deformation field estimation method based on multi-scale residual image regularization

A multi-scale, deformation field technology, applied in computing, scene recognition, computer components, etc.

Active Publication Date: 2020-05-01
WUHAN UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the deficiencies of the existing full-image deformation field estimation (point-to-point registration) technology, the present invention proposes a full-image SAR deformation field estimation algorithm based on multi-scale residual image energy minimum regularization

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
  • SAR full-image deformation field estimation method based on multi-scale residual image regularization
  • SAR full-image deformation field estimation method based on multi-scale residual image regularization
  • SAR full-image deformation field estimation method based on multi-scale residual image regularization

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0055] The technical scheme of the present invention will be further described below in conjunction with the accompanying drawings.

[0056] Such as figure 1 Shown, the technical scheme that the present invention adopts comprises following several key parts and technology:

[0057] The first part: Dense and evenly distributed control point pair generation and high-precision matching strategy.

[0058] The present invention designs a new method on the generation and matching of dense control point pairs. First, we use a large number of SIFT points as candidate control point pairs, so we use the SIFT operator to extract a large number of candidate point pairs. These feature points are relatively dense and relatively evenly distributed, and these points have the characteristics of rich local features.

[0059] The feature point extraction based on SIFT can be summarized as:

[0060] (1) Construction of scale space (2) Extreme point detection (3) Key point location (4) Main di...

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 provides an SAR full-image deformation field estimation method based on multi-scale residual image regularization. The method comprises the steps: firstly employing a large number of local feature points extracted through an SIFT algorithm as candidate control points, designing a control point generation strategy based on phase correlation and gradient refinement, so as to generate dense control point pairs which are relatively uniform in distribution; secondly, in a deformation field estimation stage, designing a distortion model based on regularization consistency point drift of the multi-scale residual image so as to realize point-to-point sub-pixel registration of the whole image; according to the method, a residual graph regularization item is introduced on the basis ofa consistency point drift algorithm, the defect that the consistency point drift algorithm is prone to falling into a local minimum value is overcome, and the method can be quickly converged to a globally optimal solution.

Description

technical field [0001] The invention relates to a high-precision (sub-pixel) full-image deformation field estimation (also called high-precision dense registration or full-image point-by-point registration) technology of a SAR image based on a reference image. This technology is a key technology in the field of remote sensing applications (image change detection, high-precision data fusion). In particular, it relates to a dense control point pair generation strategy (extraction, high-precision matching) and deformation field parameter estimation based on multi-scale residual map regularization consistency drift. Background technique [0002] With the development of remote sensing observation technology, high-resolution, multi-platform (satellite, near-Earth space, airborne), multi-source (SAR, optical, hyperspectral) remote sensing images are becoming more and more abundant. Accurate, multi-source spatio-temporal data fusion and change detection and other related technologi...

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): G06K9/00G06K9/46G06K9/62
CPCG06V20/13G06V10/462G06V10/751
Inventor 于秋则倪达文雷震吴鹏杰胡海波
Owner WUHAN UNIV
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