Unlock instant, AI-driven research and patent intelligence for your innovation.

Coarse-to-fine image registration method and system based on DFS, terminal and readable storage medium

An image registration and image technology, which is applied in image analysis, graphic image conversion, image data processing, etc., can solve problems such as matching needs to be further improved, high computational complexity, and feature islands, so as to improve registration accuracy and registration Efficiency, improved search accuracy, and the effect of improved accuracy

Pending Publication Date: 2022-05-06
CENT SOUTH UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, MLD features have feature islands, which have a certain impact on the effectiveness of the matching algorithm, and there is still room for improvement in the registration accuracy.
[0005] In summary, the feature point matching method based on graph theory combines global similarity to divide and pair features, which has high computational complexity, and there is a phenomenon of feature islands, which affects the effectiveness of the matching algorithm and leads to further matching. improve

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
  • Coarse-to-fine image registration method and system based on DFS, terminal and readable storage medium
  • Coarse-to-fine image registration method and system based on DFS, terminal and readable storage medium
  • Coarse-to-fine image registration method and system based on DFS, terminal and readable storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0063] Such as figure 1 As shown, a DFS-based coarse-to-fine image registration method provided by an embodiment of the present invention includes the following steps:

[0064] Step 1: Obtain the image feature points of the two images to be matched, and then generate the linear descriptor sMLD and divide long line segments and short line segments.

[0065] Shi et al. proposed that the line segment formed by connecting feature points in an image is the MLD feature. The present invention improves the MLD feature and proposes the sMLD feature. Compared with the MLD feature, it only retains the line length range (L min , L max )=features within (128,256), the present invention will line length range L∈(l 1 , l 2 )'s line type descriptor is called "long line segment", which is used to span the grid in the matching algorithm; search for matching points; line length range Lmin The line type descriptor of is called "short line segment", which is mainly used for in-grid search in th...

Embodiment 2

[0091] This embodiment provides a system based on an image registration method, which includes:

[0092] An image feature point acquisition module is used to obtain image feature points of two images to be matched;

[0093] A linear descriptor sMLD generation module generates a linear descriptor sMLD based on image feature points;

[0094] The long-short-line division module is used to divide the linear descriptor sMLD of long-line segments and short-line segments;

[0095] The rough matching module performs rough matching based on long line segments to obtain matching point pairs;

[0096] The fine matching module performs fine matching based on short line segments, and determines matching point pairs on the fine matching path;

[0097] The local homography matrix calculation module calculates the local homography matrix based on the matching point pairs determined by the precise matching, and the local homography matrix is ​​the registration information of the two images. ...

Embodiment 3

[0101] This embodiment provides a terminal, which includes one or more processors and a memory storing one or more computer programs, and the processor invokes the computer programs to implement:

[0102] Step S1: Obtain the image feature points of the two images to be matched, and then generate the linear descriptor sMLD and divide long line segments and short line segments;

[0103] Step S2: Perform rough matching based on long line segments to obtain matching point pairs, wherein, determine the initial matching point pairs; then use the initial matching point pairs as the starting point of the rough matching path, and determine rough matching based on similarity within the long line segment Other pairs of matching points on the path, the similarity represents the similarity of the two linear descriptors sMLD on the two images;

[0104] Step S3: Perform fine matching based on the short line segment, wherein the matching point pair obtained by the rough matching is used as th...

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 DFS-based coarse-to-fine image registration method and system, a terminal and a readable storage medium, and the method comprises the steps: obtaining image feature points of two to-be-matched images, generating a linear descriptor sMLD, and dividing the sMLD into a long line segment and a short line segment; performing rough matching based on the long line segment to obtain a matching point pair, namely determining an initial matching point pair; taking the initial matching point pair as a starting point of a rough matching path, and sequentially determining other pairs of matching point pairs on the rough matching path within a long line range by taking similarity as a basis; fine matching is carried out based on the short line segments, that is, matching point pairs obtained through rough matching serve as current point pairs, and matching point pairs on a fine matching path are sequentially determined within the range of the short line segments with the similarity as the basis; and calculating a local homography matrix based on the matching point pairs determined by the fine matching, wherein the local homography matrix represents registration information of the two images. Compared with other registration methods, the method provided by the invention has great advantages in registration precision and registration efficiency.

Description

technical field [0001] The invention belongs to the technical field of image splicing, and in particular relates to a coarse-to-fine image registration method, system, terminal and readable storage medium based on DFS. Background technique [0002] Image registration is generally divided into two processes: feature extraction and feature matching. Its feature extraction includes feature detection and generating descriptors. The classic feature extraction method is to count the neighborhood information of feature points. In order to effectively express information such as scale and direction, the feature description is usually complicated. [0003] In the stage of feature matching, it is usually carried out from two ideas of direct matching and indirect matching. Direct matching is to transform the problem of feature matching into the problem of corresponding two point sets, so as to estimate the corresponding relationship between points; indirect matching establishes the c...

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
IPC IPC(8): G06T7/32G06T3/40
CPCG06T7/32G06T3/4038G06T2200/32
Inventor 唐琎彭言松朱洪肖小月
Owner CENT SOUTH UNIV