基于LMS-STFT-SPWVD的二值化融合分析方法研究  

Research on Binary Fusion Analysis Method Based on LMS-STFT-SPWVD

作　　者：张亚斌 李占龙[1] 李虹[1] 张正 武鹏 孙宝[1] 刘志奇[1] ZHANG Ya-bin;LI Zhan-long;LI Hong;ZHANG Zheng;WU Peng;SUN Bao;LIU Zhi-qi(Taiyuan University of Science and Technology,Taiyuan 030024,China)

机构地区：[1]太原科技大学,太原030024

出　　处：《太原科技大学学报》2023年第6期498-503,共6页Journal of Taiyuan University of Science and Technology

基　　金：国家自然科学基金(52272401);忻州市重点研发计划(20220234)。

摘　　要：针对二值法存在干扰信号抑制效果、中心频率能量识别效果较差且阈值参数范围难以确定的缺点,提出了LMS-STFT-SPWVD二值化融合时频分析方法。首先,采用LMS-STFT、LMS-SPWVD算法对双分量LFM含噪信号测试函数进行分析,获取LMS-STFT、LMS-SPWVD时频分布模型;结合图像处理的方法,将整个二维时频分布模型进行二值化处理,使其转化为黑白二值时频图像;将二者进行点乘运算,获取LMS-STFT-SPWVD二值化黑白融合时频图像,并与经典方法(LMS-STFT、LMS-SPWVD及STFT-SPWVD)进行了对比研究,讨论了其干扰信号抑制效果与中心频率能量识别效果。结果表明,在阈值范围内,LMS-STFT-SPWVD二值法突出了信号频率能量,具有较好的抗干扰能力。Aiming at the disadvantages of the binary method,such as the suppression effect of interference signal,the poor identification effect of central frequency energy and the difficulty in determining the range of threshold parameters,LMS-STFT-SPWVD binary fusion time-frequency analysis method was proposed.Firstly,LMS-STFT and LMS-SPWVD algorithms were used to analyze the test function of two-component LFM noisy signals,and the time-frequency distribution models of LMS-STFT and LMS-SPWVD were obtained.Then,combined with the image processing method,the whole two-dimensional time-frequency distribution model was binarized to transform it into black-and-white binary time-frequency image.Then,the two images were dot multiplied to obtain the LMS-STFT-SPWVD binary black-and-white fusion time-frequency images,and the interference signal suppression effect and center frequency energy recognition effect of LMS-STFT,LMS-SPWVD and STFT-SPWVD binary images were compared.Finally,the effect of threshold parameters and anti-noise analysis are compared and analyzed.The results show that the LMS-STFT-SPWVD binary method can highlight the signal frequency energy and has good anti-interference ability in the threshold range.

关 键 词：图像二值化 LMS-STFT SPWVD LFM含噪信号 时频分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]