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

Robust image matching method based on local features of super pixels

A local feature and matching method technology, applied in the field of image matching, can solve problems such as easy matching failure and applicability limitations

Active Publication Date: 2017-12-08
BEIJING DEEP AI INTELLIGENT TECH
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the basis for this method to extract local features is that the feature points have significant characteristics on the texture. Therefore, for images lacking significant features, such as walls, ground, sky, sea, etc., the extraction of local feature points is too few or wrong, so it is easy to match. Failed, with limited applicability

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
  • Robust image matching method based on local features of super pixels
  • Robust image matching method based on local features of super pixels

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0026] Example 1 figure 1 A robust image matching method based on superpixel local features is given, which mainly aims at two images and calculates the corresponding points between the two images. The calculation steps are as follows:

[0027] Step 1. Carry out superpixel segmentation on the two images respectively (for example, the SLIC superpixel segmentation method can be adopted), and record all superpixels (the so-called superpixel refers to the internal features obtained by the superpixel calculation method that are consistent with the surrounding features Inconsistent small areas) location and area.

[0028] Step 2. Divide the two images into several regions; each time only one region is taken from the two images, and superpixel matching between the two regions is performed. In the present invention, the entire image is not matched in units of pixels, but divided into superpixels and matched by regions. Since the number of superpixels is far smaller than the number of...

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 a robust image matching method based on the local features of super pixels. Corresponding points between two images are calculated. The method further includes the following steps: S01, carrying out super pixel segmentation on the two images, and recording the locations and regions of all the super pixels; S02, dividing each of the two images into a plurality of regions; S03, selecting any pair of regions from the two images as two regions to be matched; S04, getting corresponding super pixels in the two regions and the matching credibility thereof; and S05, selecting a result with the highest matching credibility as a final matching result between the two images.

Description

technical field [0001] The invention relates to an image matching method, in particular to a robust image matching method based on superpixel local features. Background technique [0002] Image matching refers to the problem of finding corresponding points between two images. This problem is one of the basic problems in the fields of computer vision and pattern recognition. The methods to solve this problem are in image recognition, stereo registration, panoramic stitching, object It is widely used in applications such as tracking and motion analysis. [0003] Existing image matching methods are mainly divided into two categories: [0004] 1) block-based image matching method; [0005] 2) Matching method based on local features. [0006] The block-based image matching method is to divide the image into blocks, and use the similarity before the blocks to match the blocks, or measure the similarity between blocks centered on pixels in units of pixels, so that Establish cor...

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): G06T7/11G06T7/33G06K9/62
CPCG06T7/11G06T7/33G06V10/758
Inventor 不公告发明人
Owner BEIJING DEEP AI INTELLIGENT TECH