%0 Journal Article
%T 基于SVM的海洋内波图像预选方法<br>SVM Based Pre Selection Method for Internal Wave Images
%A 陈捷
%A 于振涛
%A 李婷婷
%A 余路
%J Computer Science and Application
%P 1-6
%@ 2161-881X
%D 2024
%I Hans Publishing
%R 10.12677/CSA.2024.141001
%X <span style=\"font-size:14px;\">针对海量海洋卫星SAR数据处理和海洋内波应用急需，研究SAR海洋图像内波预选方法。根据SAR图像内波明暗条纹的周期性、延展性、独立性等特点，提取内波特征向量，通过支持向量机对这些特征向量进行训练，根据训练集开展SAR海洋图像自动内波预选。通过对典型含内波SAR图像的测试可知，本文提出的SVM与内波多特征结合的方法可有效预选含内波的海洋SAR图像区域，预选结果与人工目视结果高度一致，可极大减轻人工处理工作量，为后续内波深入处理和应用奠定基础。</span><br />
<span style=\"font-size:14px;\">The SVM based method for ocean internal wave image pre selection is urgently needed for massive ocean satellite SAR data processing and ocean internal wave applications. Therefore, research on SAR ocean image internal wave pre selection methods is needed. Based on the periodicity, extensi-bility, independence, and other characteristics of the internal wave light and dark stripes in SAR images, internal wave feature vectors are extracted. These feature vectors are trained using support vector machines, and automatic internal wave pre selection of SAR ocean images is carried out based on the training set. Through testing typical SAR images containing internal waves, it can be concluded that the method proposed in this paper, which combines SVM with multiple features of internal waves, can effectively pre select areas of marine SAR images containing internal waves. The pre selection results are highly consistent with manual visual results, which can greatly reduce the workload of manual processing and lay the foundation for subsequent indepth processing and application of internal waves.</span>
%K 支持向量机(SVM)，海洋内波，功率谱特征，条纹延展性特征<br>Support Vector Machine (SVM)
%K Ocean Internal Waves
%K Power Spectral Features
%K Stripe Elongation Features
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=79525