基于图像的海底含油沉积物特征识别算法  

Feature Recognition Algorithm of Seafloor Oil-bearing Sediments Based on Image

作　　者：王前进 姚惠萍[1] WANG Qian-jin;YAO Hui-ping(Qingdao University of Technology School of Science,Qingdao,Shandong 266520)

机构地区：[1]青岛理工大学理学院,山东青岛266520

出　　处：《新型工业化》2020年第2期125-130,共6页The Journal of New Industrialization

基　　金：国家大学生创新训练项目(201910429100)。

摘　　要：多波束声纳数据可以被处理以获得水下声纳图像。支持向量机(SVM)和卷积神经网络(CNN)都是传统的目标分类和检测方法。关于海底含油沉积物特征识别这一问题,本文分别设计了支持向量机和卷积神经网络算法,对水下声纳图像进行目标特征的提取识别。经过两种方法试验的对比,在同一海域提取样本的情况下,支持向量机在识别含油沉积物的精度和速度上优于卷积神经网络。Multi-beam sonar data can be processed to obtain underwater sonar images.Support Vector Machine(SVM)and Convolution Neural Network(CNN)are traditional research methods for target recognition in images.In the paper,SVM and CNN are used to extract and recognize target features of underwater sonar image on the issue of feature recognition of oil-bearing seafloor sediments.By comparing the two methods,it’s clear that the accuracy and speed of the SVM for identifying oil-bearing sediments are better than the CNN when the samples are extracted from the same sea area.

关 键 词：声纳图像 海底含油沉积物 特征识别 卷积神经网络 支持向量机 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]