计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (9): 171-173.DOI: 10.3778/j.issn.1002-8331.2009.09.049

• 图形、图像、模式识别 • 上一篇    下一篇

柱面全景图像自动拼接算法

杨艳伟,郭宝龙   

  1. 西安电子科技大学 智能控制与图像工程研究所,西安 710071
  • 收稿日期:2008-09-17 修回日期:2008-11-17 出版日期:2009-03-21 发布日期:2009-03-21
  • 通讯作者: 杨艳伟

Algorithm for cylindrical panoramic image mosaic

YANG Yan-wei,GUO Bao-long   

  1. Institute of ICIE,School of Electromechanical Engineering,Xidian University,Xi’an 710071,China
  • Received:2008-09-17 Revised:2008-11-17 Online:2009-03-21 Published:2009-03-21
  • Contact: YANG Yan-wei

摘要: 提出了一种基于特征点匹配的柱面全景图像拼接算法。首先将360°环绕拍摄的序列图像投影到柱面坐标系下;然后提取各图像的SIFT(Scale Invariant Feature Transform,尺度不变特征变换)特征点,通过特征点匹配完成两幅图像的配准;再根据配准结果计算出图像间的变换参数;最后采用加权平均的融合方法对两幅图像进行无缝拼接。实验表明,算法可以有效、快速地自动生成柱面全景图像。

关键词: 柱面全景图, 尺度不变特征变换, 图像配准, 图像拼接

Abstract: In this paper,an efficient cylindrical panoramic image mosaic algorithm based on SIFT(Scale Invariant Feature Transform) is proposed.Firstly,the 360 degrees of circle panned image sequences are projected to cylindrical separately.Then SIFT feature points are extracted,and images are registered by matching SIFT feature points.And then the transform parameters can be calculated by registration results.Finally,the weighted average method is used to achieve image mosaic.Experiments indicate that the proposed algorithm can show the cylindrical panoramic image automatically and effectively.

Key words: cylindrical panoramic image, Scale Invariant Feature Transform(SIFT) feature, image registration, image mosaics