中文摘要：

高分辨率海底微地形可呈现更丰富的海床细节特征,对水下目标正确判读具有重要意义。除去底质因素影响,侧扫声呐图像明暗变化与海底地形起伏变化有关,挖掘侧扫声呐图像中的微地形信息,可以弥补现有测深手段分辨率上的不足。针对SFS方法从侧扫声呐图像反演出的相对地形缺少起伏尺度约束和起算基准等问题,本文通过小波分析方法从单波束数据中分别提取与侧扫声呐图像反演地形相关的高频地形及表征地形趋势变化的低频地形,以此构建侧扫声呐反演地形的尺度约束及基准改正模型。试验获得的反演地形从整体趋势到细节特征均与真实地形具有较好的一致性,统计结果表明,反演地形内、外符合均方根误差均小于15 cm,反演地形整体精度与单波束地形相当。

英文摘要：

Without considering the impact of marine sediment,the gray intensity of a side-scan sonar image mainly depends on the changing seabed micro- topography,which can be used to compensate for the low-resolution defi-ciency of traditional bathymetry methods. In order to recover the micro-topography and transform it into a valid scale and absolute reference,we used the wavelet analysis method to extract high- and low-frequency terrains from single-beam soundings,in which the high-frequency terrains correspond with topography recovered from side-scan sonar images and the low-frequency terrains represent topography change. On this basis,we constructed a size restraint and reference correction model for side-scan sonar topography recovery. Th e test results show that the final recov-ered topography is consistent with the actual terrain in overall trend and local detail. Statistical results show that the internal and external root mean square error （ RMSE ） are both less than 15 cm ,and the final recovery accuracy is the same as that for single-beam soundings.