摘要
针对传统图像处理方法不能给出声场干涉图案局部信息的局限性,介绍了一种基于图像Hessian矩阵特征值分析的多尺度线性滤波器,进行干涉图案的处理和特征提取,建立了具体条纹斜率和位置估计的有效方法,并进一步讨论了依据条纹参数估计结果进行海底地声参数反演的可行性.从数值仿真以及海上实验数据的处理结果表明,该方法可以有效地提取宽带声场的干涉结构和条纹信息,而条纹的参数信息也可进一步应用到地声参数反演中.
As the traditional image processing methods can not extract local information from the interference figures in the wideband ship noise field,a multi-scale line filter based on the analysis of the Hessian matrix eigenvalues of an image was introduced and applied in processing interference figures and extracting features.An effective method for estimating striation slope and position was proposed,and the feasibility of geoacoustic inversion was further discussed according to the estimation results of striation.Both numerical simulation and experimental data processing demonstrate that this method can effectively detect and extract the interference structure and striation feature of the wideband sound field,and the striation parameter estimation results can be used for geoacoustic inversion.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2012年第3期296-301,共6页
Journal of Harbin Engineering University
基金
国家自然科学基金资助项目(50979019)
国防科技重点实验室基金资助项目(9140C200103110C20)
关键词
宽带声场
干涉结构
波导不变量
简正波
条纹处理
地声参数反演
wideband sound field
inferference structure
waveguide invariant
normal mode
striation processing
geoacoustic inversion
