Computer Engineering and Applications ›› 2011, Vol. 47 ›› Issue (36): 187-189.

• 图形、图像、模式识别 • Previous Articles     Next Articles

Grid-affine image mosaic algorithm based on SIFT

WANG Cunrui1,2,GUO Changqing2,DUAN Xiaodong2,1,ZHANG Qingling1   

  1. 1.College of Information Science and Engineering,Northeastern University,Shenyang 110004,China
    2.The Research Institute of Nonlinear Information Technology,Dalian Nationalities University,Dalian,Liaoning 116600,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-12-21 Published:2011-12-21

一种基于SIFT的网格仿射拼接算法

王存睿1,2,郭长青2,段晓东2,1,张庆灵1   

  1. 1.东北大学 信息科学与工程学院,沈阳 110004
    2.大连民族学院 非线性信息技术研究所,辽宁 大连 116600

Abstract: Image mosaic in the satellite remote image,medical image processing has a wide range of application.This paper presents a modified algorithm based on SIFT feature of image,constructs a way to select points from the matched points to build affine transformation,and then builds a mosaic border based on the selected point.For the distortion of edge image by affine transformation,this paper presents an incomplete affine transformation to resolve this problem.Experimental results indicate that the method can overcome the problem of traditional affine by selecting points randomly to generate the corresponding formula,and has better stability and accuracy.

Key words: Scale Invariant Feature Transform(SIFT), affine transformation, image mosaic algorithm

摘要: 图像拼接在卫星图像遥感、医学图像处理都具有广泛的应用价值。利用SIFT作为图像局部特征,构建一种基于SIFT特征的仿射计算方法,利用网格覆盖匹配特征点,通过该方法在SIFT匹配特征点中选取仿射点,进而构建相应的仿射变换,通过仿射待选点建立拼接边缘,还给出了不同程度仿射变换的图像拼接方法来解决边缘图像仿射失真的问题。实验结果表明该方法可以克服传统方法的仿射不稳定问题,具有较好的稳定性和准确度。

关键词: 尺度不变特征变换(SIFT), 仿射变换, 拼接算法