中文摘要：

海底沙波是发育在近海陆架上的一种常见海底地貌类型,海底沙波特征与运动规律的研究具有重要的科学意义与工程应用价值,沙波脊线与谷线是表征海底沙波的最基本特征,也是精确描述沙波运动的基本参量。本文提出了一种基于复合数字水深模型的沙波特征线自动识别方法——最优方向剖面法,基于水深曲面归算得到最优剖面方向,再依据最优剖面方向求导并判定极值,自动提取沙波形态特征点,最终形成沙波脊线和谷线。以台湾浅滩复合型沙波为例进行对比实验研究,结果表明,该方法能基于不同分辨率的数字水深模型自动准确地提取海底沙波脊线与谷线,勿需设置阈值,地形自动化识别程度得到进一步提升,具有重要的实际应用价值。

英文摘要：

Sand wave is a widely occurring submarine bedform which exists on the sea floor of shallow seas due to the complex interaction between waves,tides and sediments.Knowledge of sand wave characteristics and behavior has scientific import and engineering application value.Crest lines and trough lines are basic topographic characteristics of sand wave,and the basis to describe sand wave variation.In this paper,we present a new method of automatic recognition of sand wave topographic features based on composite digital depth model（DDM）,called optimally-directional profiling method.We built an optimal direction map from DDM,and use the optimal direction to profile DDM and involve matrix calculus,by verifying extrema points to extract the topographic features.Taking composite sand waves in Taiwan Bank as an example,the comparison experiment result indicated that,the method is capable of extracting crest lines and trough lines of sand waves automatically and accurately based on multi-resolution DDM.With no requirement for setting any thresholds,the automatic degree of extracting topographic features is further increased.The method possessed important practical application value.