计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (21): 200-204.

• 图形图像处理 • 上一篇    下一篇

图像多尺度配准的小波域SURF算法

卫建龙1,田  铮1,2,延伟东1,范  明1   

  1. 1.西北工业大学 理学院,西安 710129
    2.中国科学院 遥感应用研究所 遥感科学国家重点实验室,北京 100101
  • 出版日期:2014-11-01 发布日期:2014-10-28

Image multiscale registration in wavelet domain using SURF

WEI Jianlong1, TIAN Zheng1,2, YAN Weidong1, FAN Ming1   

  1. 1.School of Science, Northwestern Polytechnical University, Xi’an 710129, China
    2.State Key Laboratory of Remote Sensing Science, Institute of Remote Sensing Applications, Chinese Academy of Sciences, Beijing 100101, China
  • Online:2014-11-01 Published:2014-10-28

摘要: 针对SURF描述子仅利用特征点的局部邻域信息而对于局部场景发生变化的图像容易产生误匹配的现象,提出了图像多尺度配准的小波域SURF算法。该方法对图像进行小波分解,利用低分辨率上得到的变换参数剔除高分辨率上的错误匹配,得到精确匹配点对,对图像进行配准。实验结果表明,该方法能有效地剔除误匹配点,提高图像配准的精度。

关键词: 图像配准, 加速稳健特征(SURF), 小波域, 多尺度

Abstract: A method to correct Speed-Up Robust Features(SURF) mismatching combining wavelet transform with SURF is proposed to solve the mismatching in the registration of the images broken in the locate structure. This paper applies wavelet decomposition to the images. It utilizes the transformation parameters derived from images of lower resolution, and a number of mismatching points of higher resolution are eliminated. Then images are registrated using the exact matching points. The experimental results on optical-remote sensing images and SAR images show that the proposed algorithm can eliminate the mismatching points effectively and enhance the registration accuracy.

Key words: image registration, Speed-Up Robust Features(SURF), wavelet domain, multiscale