pdf

Three-view image feature matching algorithm based on SIFT

  • 2013-09-22
  • 199.2KB
  • Points it Requires : 2

A SIFT-based three-view image feature matching algorithm is proposed. The SIFT algorithm is used to perform feature matching on the three-view images. The trilinear relationship in computer vision and the pixel transfer error calculation method based on the trifocal tensor are introduced to screen the matching results of the SIFT algorithm and eliminate the mismatched points generated during the matching process. Experiments on real image data show that the algorithm effectively eliminates more than 97% of the mismatched points generated by the SIFT algorithm and improves the accuracy of three-view image matching. Keywords: SIFT algorithm; trilinear relationship; trifocal tensor

unfold

You Might Like

Uploader
mamselc
 

Recommended ContentMore

Popular Components

Just Take a LookMore

EEWorld
subscription
account

EEWorld
service
account

Automotive
development
circle

About Us Customer Service Contact Information Datasheet Sitemap LatestNews


Room 1530, 15th Floor, Building B, No.18 Zhongguancun Street, Haidian District, Beijing, Postal Code: 100190 China Telephone: 008610 8235 0740

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京ICP证060456号 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号
×