Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (35): 245-248.DOI: 10.3778/j.issn.1002-8331.2010.35.071

• 工程与应用 • Previous Articles    

Research of image matching method in sea ice observation

WANG Yu-xin,LIU Yan-fei,GUO He,LIU Tian-yang,YANG Yuan-sheng   

  1. School of Electronic and Information Engineering,Dalian University of Technology,Dalian,Liaoning 116024,China
  • Received:2009-04-20 Revised:2009-06-30 Online:2010-12-11 Published:2010-12-11
  • Contact: WANG Yu-xin


王宇新,刘彦飞,郭 禾,刘天阳,杨元生   

  1. 大连理工大学 电信学院,辽宁 大连 116024
  • 通讯作者: 王宇新

Abstract: In cold areas marine project,the speed and direction of movement of the sea ice are important environmental parameters.Considering the actual situation,a new method to measure the speed of sea ice movement in the video is presented:Taking the Harris corner as the characteristic,computing the displacement according to the matching feature points of the frames intercepted from the video at different times,using the greedy Gaussian Mixture Model(GMM) model to eliminate the mismatches,and obtain the speed of the sea ice in the end.Experiments and applications show that this method has a good robustness.

Key words: Harris corner detector, sea ice speed, greedy EM algorithm, Gaussian mixture model(GMM)

摘要: 海冰流速的测定是寒区海冰观测的重要内容。结合海洋工程实际,提出一种根据视频图像的特征点提取和匹配来测算海冰流速的方法:把Harries角点作为海冰的特征点,根据不同时刻下特征点的单应得到特征点的移动距离和移动方向,通过高斯混合模型消除单应计算中的误匹配,最终得到海冰的移动方向和速度。实验结果表明,这种方法具有较好的鲁棒性。

关键词: Harris角点检测, 海冰流速, 贪心EM算法, 高斯混合模型

CLC Number: