侧扫声呐图像海底线自动提取方法研究

发布时间：2018-06-07 14:11

本文选题：侧扫声呐图像 + 复杂海洋噪声　；参考：《武汉大学学报(信息科学版)》2017年12期

【摘要】：针对现有方法在侧扫声呐水柱区图像受发射脉冲、海面回波、尾流及大面积悬浮物等干扰情况下海底线无法自动准确检测和提取,造成斜距改正后目标图像严重畸变和错位等问题,基于侧扫声呐成像机理以及图像特点,提出了海底线最后峰值检测法和基于海底变化渐进性和海底线对称性的海底线修复方法。结合Kalman滤波以及上述方法的特点和适用对象,提出了一种海底线自适应综合检测和提取的方法,并给出了完整的数据处理流程。该方法应用于烟台水域,消除了海况差、悬浮物遮挡等问题的影响,实现了复杂海洋噪声影响下海底线的自动跟踪。与外部测深数据比较,取得了均方根为±0.17m的跟踪精度。
[Abstract]:The existing methods can not detect and extract the submarine line automatically and accurately under the condition of the pulse emission, sea surface echo, wake and large area suspended matter interference in the image of the side scan sonar water column area, and the existing methods can not detect and extract the submarine line automatically and accurately. Based on the imaging mechanism of side-scan sonar and the characteristics of the image, the image distortion and dislocation of the target image after correction of oblique distance are caused. The detection method of the final peak value of the submarine line and the method of repairing the submarine line based on the evolution of the submarine variation and the symmetry of the submarine line are proposed. Combined with the characteristics of Kalman filter and the characteristics and applicable objects of the above methods, a method of adaptive comprehensive detection and extraction of submarine lines is proposed, and a complete data processing flow is given. The method is applied to Yantai water area to eliminate the influence of poor sea condition and suspended object occlusion and realize the automatic tracking of the bottom line under the influence of complex ocean noise. Compared with the external bathymetric data, the tracking accuracy of the root mean square (RMS) of 卤0.17m is obtained.
【作者单位】：武汉大学测绘学院;武汉大学动力与机械学院;
【基金】：国家自然科学基金(41576107,41376109,41176068) 海岛(礁)测绘技术国家测绘地理信息局重点实验室资助项目(2015B08)~~
【分类号】：TP391.41

【相似文献】