通信学报 ›› 2017, Vol. 38 ›› Issue (1): 177-186.doi: 10.11959/j.issn.1000-436x.2017020

• 学术通信 • 上一篇    下一篇

基于改进SURF算法的移动目标实时图像配准方法研究

巨刚,袁亮,刘小月,岳昊恩   

  1. 新疆大学机械工程学院，新疆 乌鲁木齐 830047
  • 修回日期:2016-08-06 出版日期:2017-01-01 发布日期:2017-01-23
  • 作者简介:巨刚（1989-），男，陕西咸阳人，新疆大学硕士生，主要研究方向为图像处理、视觉图像跟踪。|袁亮（1972-），男，新疆乌鲁木齐人，博士，新疆大学教授、博士生导师，主要研究方向为机器人视觉、图像处理。|刘小月（1990-），女，陕西咸阳人，新疆大学硕士生，主要研究方向为流体动力学。|岳昊恩（1994-），男，湖北武穴人，新疆大学硕士生，主要研究方向为图像处理、视觉图像跟踪。
  • 基金资助:
    国家自然科学基金资助项目(31460284);国家自然科学基金资助项目(61262059);国家自然科学基金资助项目(61662075);新疆自治区科技支疆基金资助项目(201591102);乌鲁木齐市人才工程计划基金资助项目(P151010006)

Study on mobile target real-time image registration based on improved SURF algorithm

Gang JU,Liang YUAN,Xiao-yue LIU,Hao-en YUE   

  1. School of Mechanical Engineering,Xinjiang University,Urumqi 830047,China
  • Revised:2016-08-06 Online:2017-01-01 Published:2017-01-23
  • Supported by:
    The National Natural Science Foundation of China(31460284);The National Natural Science Foundation of China(61262059);The National Natural Science Foundation of China(61662075);Technology Branch Project of Xinjiang(201591102);Outstanding Talent Training Project of Urumqi(P151010006)

在线阅读

3

PDF下载

924

被引次数

4

摘要：

针对目标在移动过程中实时视觉图像特征点提取与配准的不稳定性，提出一种多算法融合的改进配准方法。首先，采用双边滤波、Canny边缘检测及形态学处理方法得到具有较强顽健性特征的边缘周边检测区域并基于离散Gaussian-Hermite矩对SURF算法中的Hessian矩阵进行修正，重新定义特征描述向量，同时采用肯德尔系数对配准的特征点进行约束。其次，通过融合光谱辐射颜色不变量模型及I_SURF算法对实时视觉彩色图像进行配准。最后，将改进算法与目标自适应更新算法相结合，实现了移动目标在室内环境中的实时匹配。实验结果表明，在不同旋转尺度下，改进算法的静态图像配准较SURF算法具有较高配准精度，移动图像特征点提取及配准数量的稳定性达到97%以上。

关键词: 改进SURF, 移动目标, 图像配准, 实时性, Gaussian-Hermite矩

Abstract:

For the unstability of the real-time visual image characteristic point extraction and matching for the mobile target,an improved registration method of the multiple algorithm-fusion was introduced.Firstly,the method of bilateral filtering,Canny edge detection and morphological processing,was adopted to get the more robust image's edge map.And the founded points were limited in this edge map,then the Hessian matrix of SURF based on the discrete Gaussian-Hermite moment was modified.The character description vector was redefined in the algorithm.Following the above analysis,the Kendall coefficient constraint was discussed in image matching characteristic points.Secondly,the spectral radiant color invariant model and the I_SURF algorithm were used to match the real-time color image.Finally,the improved algorithm was combined with the update algorithm of adaptive target to match the mobile target in indoor environment.The experimental results show that the static image registration accuracy of the improved algorithm is higher than that of the SURF algorithm,and also the stability of the mobile image feature points extraction and the registration number have achieved over 97% under different rotating scales.

Key words: improved SURF, mobile target, image registration, real-time, Gaussian-Hermite moment

中图分类号: 

No Suggested Reading articles found!