An Airborne Remote Sensing Image Mosaic Algorithm Based on Feature Points

被引：3

作者：

Zhen, Yumei ^{[1
]}

Sun, Zhen ^{[1
]}

Li, Junbao ^{[1
]}

Peng, Yu ^{[1
]}

机构：

[1] Harbin Inst Technol, Harbin, Peoples R China

来源：

PROCEEDINGS OF 2016 SIXTH INTERNATIONAL CONFERENCE ON INSTRUMENTATION & MEASUREMENT, COMPUTER, COMMUNICATION AND CONTROL (IMCCC 2016) | 2016年

关键词：

image mosaic; image match; SIFT; Scale-space; image fusion; feature points;

D O I：

10.1109/IMCCC.2016.145

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

This paper focuses on a image mosaic method based on feature points. Image registration and stitching are the key technology of image mosaic. Based on analysis of research, this paper selected a mosaic algorithm based on SIFT feature points which have a good scale and rotation invariant feature for further research. For the wrong matches because of the algorithm restriction, occlusion, sensor moves, or the symmetric structure of the scene itself, this paper introduces the RANSAC algorithm for precise matching. During the process of image stitching, this paper reaches the goal of image mosaic by using the Homography Matrix solved by RANSAC.

引用

页码：202 / 205

页数：4

共 6 条

[1]

[Anonymous], 2009, Int. J. Image Process., DOI DOI 10.1007/S11270-006-2859-8

[2] RANDOM SAMPLE CONSENSUS - A PARADIGM FOR MODEL-FITTING WITH APPLICATIONS TO IMAGE-ANALYSIS AND AUTOMATED CARTOGRAPHY [J].