摘要
针对在舰船辐射噪声DEMON线谱信噪比低时,传统使用的基于模型或专家系统的叶片数识别方法存在错误率高且识别结果不稳定的问题,提出了一种基于模板匹配的螺旋桨叶片数识别方法。该方法建立了不依赖于样本的模板库,设计了模板匹配算法和识别结果置信度算法,较好地解决了传统的叶片数识别技术所存在的数学模型失配或规则不全、样本不完善性等问题。对实测数据的分析表明,该方法明显降低了叶片数识别的错误率,使目标类型识别结果趋于稳定。
The existing algorithm of ship propeller blade-number recognition is based on mathematical model or expert system.The result of blade-number recognition is usually instable and inaccurate, in the case of DEMON line spectrum of ship radiated noise at low signal-to-noise ratio（SNR）.This paper proposed a method for ship propeller blade-number recognition based on template matching.This method establishes a template library, which is independent of samples.Template Matching algorithm and confidence factor algorithm are designed.This method can resolve the problem of inaccurate mathematical model, incomplete rules and samples in existing algorithm.Practical experiments prove that the error rate of propeller blade-number recognition is reduced distinctly, and the result of target classification goes toward stability.
出处
《声学技术》
CSCD
2010年第2期228-231,共4页
Technical Acoustics
关键词
模板匹配
螺旋桨叶片
识别
template matching
propeller blade
recognition
