合成孔径声纳图像纹理分类

Texture Classification in Synthetic Aperture Sonar Images

纹理是图像的重要特征。合成孔径声纳图像中,水底展示出了丰富的纹理。纹理分析在声图处理中有多种用途,具有重要意义和研究价值。本文比较了不同的纹理描述方法应用与光学图像与声图中时的性能。使用基于支持向量机的主动学习方法,降低了人工标签的工作量。

Textures are important features of image. In synthetic aperture sonar images,the seabed exhibits rich texture features. Texture analysis has many applications in sonar image processing,thus is of great significance and worth researching. This paper compares the performance of different texture descriptors with optical image and sound image. The active learning method based on support vector machine is used to reduce the workload of manual labeling.