Study on Automatic Classification Method for Seafloor Terrain Complexity Based on BP Neural Network
JI Xue
ZHOU Xing-hua
CHEN Yi-lan
TANG Qiu-hua
ZHAO Hong-chen
Abstract：For the classification of seafloor terrain complexity ,the slope and the relief degree ,in addition to the mean water depth ,are also introduced as the classification in‐dexes for characterizing the seafloor terrain complexity and quantified .And the spatial resolution of water depth data is unified .Based on these ,a seafloor terrain complexity classification library w hich includes 18 types of typical submarine features is established and trained by using BP neural network .For testing the validity and applicability of this method ,4 experimental areas with different seafloor terrain complexity are chosen and statistics method and BP nerve network algorithm are respectively applied for the classi‐fication of seafloor terrain complexity .It is found by the comparison that by using the proposed method three types of seafloor terrain ,i .e .flat seabed ,general seabed and complex seabed ,can be identified accurately ,rapidly and automatically ,and the details of the seafloor terrain complexity in the experimental areas can also be well preserved .
Keywords：BP neural networksloperelief degreeseafloor terrainclassification index
Publication Date：2016-01-01
Online Publishing Date：2026-05-22(First online date of this platform, not the publication date of the document)
Pages：10( 32-41 )