Analysis of underwater object pose angle estimation method based on echo prediction

摘要

The theory behind the forecast method of object pose angles has important engineering application value. In this paper, firstly, the planar elements method is used to predict near-field echoes of benchmark underwater objects, and build a characteristic sample database of object echoes; Secondly, by match-filtering measured object echoes with echo samples in the database, which are under various pose angles, parameters such as the correlation coefficient are obtained; Thirdly, a cost function is designed, the gained parameters are taken into the cost function, Finally, an estimated object pose angle is obtained. According to the time field echo structure, the pose angles are divided into several ranges: 0-40, 40-80, 80-110, 110-140, and 140-180degrees. The efficiency and robustness of the method are analyzed by statistic methodology in each pose angle range. A lake experiment was conducted, and the underwater object echo data were recorded. By comparing the estimated results with the real object pose angles, the estimation method is verified to be effective and robust. As a result of the good performance in robustness and fast computation speed, the method shows potential engineering applications.