基于声纳图像纹理特征的海底底质分类方法研究

Seabed Sediment Classification Based on Texture Features of Sonar Imagery

一般不同底质的海底声纳图的纹理特征不同。重点分析了不同底质声纳图像纹理特征的差异,为底质分类的实现提供了基础。并提取了海底声纳图像的灰度共生矩阵,将灰度共生矩阵的统计特征作为特征向量。最后应用自组织迭代(ISODATA)分类算法依据对特征向量进行分类,从而实现了海底底质的划分。

Different seabed sediment hasdifferent texture features.The paper focuses on the differences in different sediment texture features which provide a foundation for the sediment classification.The GLCM is extracted from sonar imagery of the seabed sediment classification and the statistical characteristics of GLCM are used as feature vectors.ISODATA classification algorithm is used for classifying feature vectors and the division of the seabed sediment is achieved in the end.