English

应用科学学报 ›› 2010, Vol. 28 ›› Issue (4): 374-380.doi: 10.3969/j.issn.0255-8297.2010.04.008

• 信号与信息处理 • 上一篇    下一篇

多特征主成分分析与声图相结合的海底底质分类

马飞虎1;2， 孙翠羽1;3， 康永红1， 刘智敏3   

  1. 1. 华东交通大学土木建筑学院，南昌330013
    2. 精密工程与工业测量国家测绘局重点实验室，武汉430079
    3. 山东科技大学测绘科学与工程学院，山东青岛266510
  • 收稿日期:2010-03-31 修回日期:2010-05-20 出版日期:2010-07-23 发布日期:2010-07-23
  • 作者简介:马飞虎，博士，副教授，研究方向：海洋测量、声纳数据处理、GPS数据处理及应用、工程测量等，E-mail: mfh3@163.com
  • 基金资助:

    精密工程与工业测量国家测绘局重点实验室开放基金(No.PF2009-21)；国家自然科学基金(No.40704001)资助

Seabed Classification Based on Principal Component Analysis of Multiple Features Combined with Sonar Image

MA Fei-hu1;2, SUN Cui-yu1;3, KANG Yong-hong1, LIU Zhi-min3   

  1. 1. School of Civil Engineering and Architecture , East China Jiaotong University, Nanchang 330013, China
    2. Key Laboratory of Precise Engineering and Industry Surveying, State Bureau of Surveying and Mapping, Wuhan 430079, China
    3. Geomatics College, Shandong University of Science and Technology, Qingdao 266510, Shandong Province, China
  • Received:2010-03-31 Revised:2010-05-20 Online:2010-07-23 Published:2010-07-23

PDF

1537

被引次数

31

摘要：

通过提取多种回声特征构造全特征向量，并对全特征向量进行主成分分析，计算出对底质分类贡献率最大的特征组，实现海底底质的分类. 采用两种分类方法对胶州湾实测数据进行比较，可得出下列结论：应用多特征主成分分析与声图相结合的分类结果优于单纯使用声纳图像的分类结果.

关键词: 特征提取, 全特征向量, 主成分分析, 海底底质分类

Abstract:

Aimed at seabed classification, statistical characteristics are extracted from the echo, and a full feature vector is constructed. The principal component analysis (PCA) is carried out to obtain the set of characteristics that most contribute to the classification. Based on the study, seabed classification is carried out and tested with two sets of experiments. Using two types of classification methods to analyze, the data from Jiaozhou Bay and comparison is made. It is concluded that the result based on PCA of many features combined with a sonar map is better than that obtained solely from sonar image classification.

Key words: feature extraction, full feature vector, principal component analysis, seabed classification

中图分类号: