A Minimal Branch Stereo Matching Method Based on Feature Fusion

A technology of feature fusion and stereo matching, which is applied in image analysis, image enhancement, instruments, etc., can solve the problems of weak texture area occlusion area mis-match and other problems, and achieve the effect of improving accuracy and improving mis-match problems

Active Publication Date: 2019-08-09
NORTHEASTERN UNIV LIAONING
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The semi-global algorithm has the advantages of strong robustness and insensitivity to lighting effects, but there are still mismatching problems in weak texture areas, discontinuous areas, and occluded areas

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
  • A Minimal Branch Stereo Matching Method Based on Feature Fusion
  • A Minimal Branch Stereo Matching Method Based on Feature Fusion
  • A Minimal Branch Stereo Matching Method Based on Feature Fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0039] An embodiment of the present invention will be further described below in conjunction with the accompanying drawings.

[0040] In the embodiment of the present invention, the minimum branch stereo matching method based on feature fusion, such as figure 1 shown, including the following steps:

[0041] Step 1. Obtain the left image and the right image of the image to be processed as the reference image and the target image respectively. In the example of the present invention, such as figure 2 The image to be processed is shown, and the fast median filter method is used to preprocess the two images;

[0042] Step 2. Calculate the initial matching cost of the two preprocessed images based on the feature fusion method. In the example of the present invention, given a pixel point p=(x, y) in the left picture, under the parallax d, calculate C AD (p,d),C GRD (p,d),C Census (p,d) and C LBP (p,d). C Census (p,d) is defined as the Hamming distance between the correspondi...

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 minimum branch stereo matching method based on feature fusion, can effectively solve the mismatching problems of a textureless region, a discontinuous region and an occlusion region, and improves stereo matching accuracy. A minimum branch structure firstly utilizes gradient information to construct directed graph aggregation matching cost; after a minimum branch is constructed, an image is segmented into a plurality of regions; and in a process that the image is segmented into the regions, no parameters need to be arranged, so that a segmentation process is natural, a texture region in the image can be effectively distinguished, and stereo matching accuracy is improved. Initial matching cost calculation based on the feature fusion solves the mismatching problem of the textureless region and the discontinuous region, the left and right consistency detection of non-occlusion points is searched on the basis of four directions, the mismatching problem of the occlusion region is effectively solved, and the accuracy of stereo matching is further improved.

Description

technical field [0001] The invention belongs to the technical field of stereo matching, and relates to a minimum branch stereo matching method based on feature fusion. Background technique [0002] Stereo matching technology is a very valuable and hot issue in the field of computer vision. Stereo matching usually includes four steps: (1) calculating the initial matching cost for each pixel; (2) aggregating the matching cost based on a window or a special structure; (3) calculating the disparity; (4) disparity optimization. [0003] There are three main types of stereo matching algorithms, local algorithms, global algorithms and semi-global algorithms. The local algorithm aggregates the matching cost based on a specific window, which runs fast but has poor accuracy. The appearance of the global algorithm improves the accuracy of stereo matching, but the slow real-time performance limits its application in practical scenes. The semi-global algorithm is proposed to effective...

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 Patents(China)
IPC IPC(8): G06T7/33
CPCG06T7/337G06T2207/10012
Inventor 张云洲刘及惟林淮佳楚好张珊珊商艳丽张凯
Owner NORTHEASTERN UNIV LIAONING
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap