基于SVM的海底声纳图像底质识别  

Seafloor sonar sediment image recognition with the Support Vector Machine

作　　者：熊明宽[1] 吴自银[1] 李守军[1] 罗孝文[1] 唐秋华[2] 

机构地区：[1]国家海洋局第二海洋研究所国家海洋局海底科学重点实验室,浙江杭州310012 [2]国家海洋局第一海洋研究所,山东青岛266061

出　　处：《海洋通报》2012年第4期409-414,共6页Marine Science Bulletin

基　　金：海洋公益性行业科研专项(201105001);国家自然科学基金(40506017);国家海洋局第二海洋研究所基本科研业务费专项(JG1105)

摘　　要：对海底底质声纳图像分割,提取单元特征向量进行主成份分析,选取均值、标准差、对比度、相关系数、能量及同质性作为训练特征向量,优化支持向量机(SVM)参数。利用多分类SVM对砂、礁石、泥进行训练,有效、快速地分离出了3种底质类型,测试精度高于80%,表明该方法可有效用于海底底质声纳图像的识别和分类。With segmenting the seafloor sonar gray image and extracting characteristic vector unit with principal component analysis, the mean, standard deviation, contrast, correlation coefficients, energy and homogeneity were selected as training characteristic vectors to optimize the Support Vector Machine （SVM） parameters. Using mulfi-SVM to train the sand, rocks and mud, it could separate three kinds of sediment types effectively, fast and completely, and testing precision was all above 80 %. It proved that this method could be used effectively in the classification and recognition of seafloor sonar gray image.

关 键 词：支持向量机(SVM) 声纳灰度图像 参数优化 特征向量 

分 类 号：P737.2[天文地球—海洋地质]