基于改进FAST特征匹配的电子稳像算法
DOI:
CSTR:
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TP391.41

基金项目:

吉林省科技发展计划资助项目(20130522107 JH)


Electronic image stabilization based on improved FAST feature matching
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对机载摄像机视频抖动问题,提出了基于改进FAST的特征匹配方法,去除视频序列的帧间抖动,实现视频序列的电子稳像。提出了尺度空间的FAST特征提取方法,结合了优化的SURF描述算子,利用相邻帧的特征点进行匹配,消除视频的帧间旋转量,得到帧间的局部运动矢量,然后通过RANSAC算法求解出全局运动矢量,最后利用二次线性插值算法逐帧补偿。实验结果表明,稳像后的视频帧间PSNR提高5 dB,提出的稳像算法去除了视频帧间复杂抖动,且快速准确并具有一定的鲁棒性。

    Abstract:

    In order to overcome image shakes in the video of the camera mounted on the aerial platform, an improved feature estimation was proposed to remove effectively the inter frame jitter of the video sequences and to realize the electronic image stability. An algorithm combining scalespace FAST corner detector and improved SURF descriptor , is proposed in this paper. The video rotation is removed ,and the stable feature points of neighbor frames was extracted and the estimation of inter frame motion vectors was implemented. RANSAC was employed to solve the global motion vectors. Quadratic interpolation algorithm is adopted to get the stable video. The experimental result indicates that the PSNR of video sequence has improved about 5 db, which verifies that the proposed stabilized algorithm removes interframe jitter under complicated movement ,higher accuracy and stronger robustness.

    参考文献
    相似文献
    引证文献
引用本文

FAST ;RANSAC;全局运动估计;电子稳像.基于改进FAST特征匹配的电子稳像算法[J].电子测量技术,2015,38(11):42-45

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2016-03-01
  • 出版日期:
文章二维码
×
《电子测量技术》
财务封账不开票通知